MDRSESCO - Software for the Masses

 

The software described herein, except where otherwise noted, was designed and implemented by MARK DANA ROCKMAN Email:  usermark at mdrsesco dot biz.  Developed using C# and C++ with .Net Framework and INTEROP aka P/Invoke. Some applications run as a Windows Service.  Items are for sale.  Just send money.  Microsoft’s EULA applies to these items except where it says Microsoft substitute MARK DANA ROCKMAN.  Software development project proposals are entertained here.  Drop us a line.

 

 

Discretionary Access Control Lists (DACLs) and Ownership

 

Windows has a complex mechanism that is intended to restrict access to various operating system objects; especially files and directories.  These objects have owners and DACLs.  An end user who desires access to an object must be granted permission to have access.  Some combination of ownership and DACL act to grant or deny access.  Sometimes access is denied for specious reasons.  There is a mechanism by which the end user can prod the system to determine whether an object possesses all necessary permission for the act that the end user intends.  A file (or folder) can be read by users A and B and written by user C and deleted by user D.  The users are termed “principals” in the vernacular.  Whenever something happens in a system, a principal (which could be a process or thread acting on behalf of a user or service) acts upon objects.  Users who are not owners may be denied the ability to view the DACLs.  An end user with elevated privileges (i.e., is running under the Administrator account) has permission to take ownership of practically any file.  A complicating factor is inheritance.  The files/directories are organized in a tree structure.  The files/directories are termed “nodes.”  The topmost node is the <drive-letter>.  It is the root because the tree is upside down with respect to natural trees in a forest.  Nodes below the root can inherit permissions from the root or establish themselves as the inheritance granting authority for nodes below themselves.  The original design permitted multiple inheritance.  But that was discovered to present intractable problems.  So, there is no multiple inheritance.  Windows sometimes issues silly denial messages.  My favorite is:  “You must have permission from Everyone…”  How all of this really works remains an undocumented mystery.  Well, there exist documents.  Neutrino chirality is easier to understand than this.  Refer to this document:  https://docs.microsoft.com/en-us/archive/msdn-magazine/2008/november/access-control-understanding-windows-file-and-registry-permissions 

 

Syntax:  takeown /F <name-of-object>

 

 

 

 

For Security Reasons You Want to Know What Files Have Been Accessed and When

 

Now available is a Windows Service that records file accesses.

 

 

 

If You Are Backing Up Files

 

If your application’s purpose is to backup files then you probably want to ask Windows to permit you to have access to various files that you ordinarily wouldn’t have access to.  Our best understanding of the mechanism provided involves a certain additional permission that seems ordinarily to be granted by default.  Permission to grant the permission can be denied via a different mechanism.  This is all part of the extremely complex and not very well documented security features of Windows.  Sad but true.  Here is how to set or clear the privilege:

 

 

 

The ChangeBackupPrivilege takes an argument of 0 or 1.  When the argument is 0 then the backup privilege is turned off.  When the argument is 1 then the backup privilege is turned on.  Here is the Win32 C++ method that is called:

 

 

 

 

Volume Shadow Copy Service (VSS)

 

Microsoft defines a “shadow copy” to be the frozen image of a mass storage device in whole or in part.  The Volume Shadow Copy Service is a software component of Windows that provides an application programmer interface (API) that is of interest to programmers who are creating software that is capable of creating disk images.  The API conforms to the Component Object Model (COM).  As such, the “disk image creating software” also conforms to COM.  The software shall be written in C++ or, perhaps, Visual Basic.  Development of Microsoft VSS more or less ended long ago.

 

HRESULT = 0x80d05011

 

One of millions of desktop computers running Windows 10 is suffering a Windows Update malfunction.  The end user is notified that updates are available.  The end user requests the updates.  One of several updates downloads 5% and announces a malfunction.  This is quite reliable.

 

There are several documented steps an end user can pursue to resolve such issues.  There is even a Windows Update Troubleshooter that, one assumes, would sort out all kinds of malfunctions.  They’ve all been tried and have failed.  Microsoft’s Feedback Hub has been informed.  We await a satisfactory response.

 

 

 

So You Want To Choose the Command Prompt’s Current Working Directory

 

I searched for hours using Microsoft’s documentation platform (the point-and-click desktop, outdated textbooks for paid classes, AND Google).  (I found tons of people who want to delete C:\Windows\System32, since that appears in front of the common user as the default current working directory (CWD) of the legacy Windows shell.)  But I could not remember how I did it a while back.  And I could not find how to do it.  The solution:  Create a shortcut on the desktop that launches the legacy shell.  Then, access the Properties of the shortcut and put your chosen CWD in the Start In box.  It is allegedly possible to make a Registry change to accomplish what you want.  But design of the Registry is subject to change without notice.

 

 

 

How to Configure Windows to Bypass the Logon Screen

 

Microsoft offers, as part of Windows, a utility application called NETPLWIZ.  The purpose of NETPLWIZ, as far as I can tell, is to allow you to configure Windows so that you do not have to enter a password when you turn on the computer.  The utility formerly worked properly.  That is to say, there was a checkbox that you could tick to configure one of the user accounts as the account that you automatically get logged into when you turn on the computer.  Microsoft improved NETPLWIZ so the checkbox no longer appears.  Presumably this design decision was made for the ever-popular “security reasons.”  Actually, there is a way to configure Windows so that NETPLWIZ presents the checkbox.  Just go into the Registry editor and set the following item to zero.  By default, it is set to 2.  Reboot the computer, launch NETPLWIZ, and, voilŕ, there is the checkbox.  The advent of “Windows Hello” allows you to configure a PIN in place of a password.  You can also, with the right hardware, have the computer recognize your face or some device you plug in.  As of 2022 NETPLWIZ and Windows do not play nice unless you stay away from the advanced features.

 

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\PasswordLess\Device\DevicePasswordLessBuildVersion

 

Windows Registry Editor Version 5.00

 

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\PasswordLess\Device]

"DevicePasswordLessBuildVersion"=dword:00000000

 

Making and Recovering a Disk Image

 

In the event of a catastrophic computer failure, a disk image is the best aid to remediation.  A disk image contains all the data on a disk without regard to the file system.  It includes data that is not in the file system such as the Master Boot Record and/or its equivalent.  If you have software installed, for which you do not possess installation media, a disk image will restore that software.  If you have laboriously configured your operating system, and you do not want to spend more hours recreating that environment, a disk image will save you time and effort. 

 

In practice, the available options for creating a disk image have solved the creation part adequately.  When restoration is attempted, the solutions often present obstructions.  For example, one commercial solution has a habit of blocking access to a disk image, during restoration, that happens to be in a shared directory, or maybe on the Internet.  The solution does not distinguish between the two. For security reasons, they assume it must be the Internet.  More to its liking is that you plug in a USB drive, preferably a capacious one.  The commercial solution, for restoration, requires ancillary software that comes in two flavors:  one for Windows 10 and one for Windows 7.  Attempting recovery with the wrong ancillary software will appear to complete successfully, but the contents of the hard drive will be improperly formatted.  The end user interface is confusing and the learning curve is steep.  The vendor has added malware protection to help sales.  Malware protection is turned on by default.  It can be turned off, but only partially.  There is plenty of opportunity for the malware detector to falsely decide the software that you run is malware.  This causes end users to panic and may disrupt production.  During backup, if you are running the latest Microsoft operating system, then the commercial solution behaves flawlessly.  The open-source solution, but not the commercial solution, requires that the end user takes the production application out of service during the period in which the image is created. The open-source solution runs under Linux whereas the commercial solution runs under Windows (applications in production) during backup and under Windows Pre-boot Environment (proper version) during restoration.  The open-source solution has a horrible end-user interface with many pages and hard-to-understand wording.  During recovery, figuring out which device you want to backup to or restore from is difficult, because Linux has no concept of drive letters.  You should format the output media with a label that appears in the interface.  The label acts as your clue as to which device is the backup device.  Open-source is no-cost and there is no recourse if it fails.  (Not that there is recourse should any commercial solution fail.)  It is operating system agnostic.  It properly restores a Windows 7 hard disk. It probably works with any operating system.  Even Linux.

 

 

 

How Create a Disk Image with CloneZilla

 

WARNING:  You must be very careful when choosing a device to read from and to write to with CloneZilla.  The software has no way of knowing which hard disk drive has which drive letter.

 

1.  Gather these USB drives:

a.   CloneZilla boot drive.

b.  Formatted, empty USB drive to receive the image.  Label this drive CreatedByLinux.

2.  Unplug all USB devices from the PC.

3.  Insert the CloneZilla boot drive.

4.  Reboot the PC whose image is to be created.

5.  Communicate with BIOS or UEFI as required to boot from CloneZilla boot drive.

6.  Press ENTER on each screen until you arrive at the Start CloneZilla screen.

7.  Choose Start CloneZilla mode.  Do not choose Enter Shell mode.

8.  See CloneZilla Open Source Clone System screen.

 

The CreatedByLinux label is created by you when you format the output drive.  The label will appear, truncated, within a message issued by CloneZilla during a save or restore operation.

 

*CloneZilla is free (GPL) software, and comes with ABSOLUTELY NO WARRANTY*

///Hint!  From now on, if multiple choices are available, you have to press space key to mark your selection.  An asterisk (*) will be shown when the selection is done///

Two modes are available, you can

1.  clone/restore a disk or partition using an image

2.  disk to disk or partition to partition clone/restore

Select mode:

1.  device-image work with disks or partitions using images

2.  device-device work directly from a disk or partition to a disk or partition

OK or Cancel

 

You want device-image mode.  Press OK.

 

Mount CloneZilla image directory

 

Before cloning, you have to assign where the CloneZilla image will be saved to or read from.

We will mount that device or remote resources as /home/partimag. 

The CloneZilla image will be saved to or read from /home/partimag.

 

Select mode:

 

You want local_dev Use local device (E.g.: hard drive, USB drive)

 

(This is more than a little confusing.  The software indicates it wants you to specify the output device.)

 

OK or Cancel

 

You want local_dev.  Press OK.

 

Below the CloneZilla image directory screen appears several lines, as follows:

 

ocsroot device is local_dev

Preparing the mount point /home/partimag?

If you want to use USB device as a CloneZilla image repository, please

·        Insert USB device into this machine *now*

·        Wait for about 5 secs

·        Press Enter key

·        So that the OS can detect the USB device and later we can mount it as /home/partimag.

Press Enter to continue.

 

The software wants you plug in the USB drive that you previous formatted.  Linux will detect that you have done so if you wait about five seconds.  Then press Enter.  Notice that you have provided a place for the image to go.

 

After you press Enter several lines will appear that enumerate various partitions on the various hard disk drives.  In Linux these are named /dev/sda<n> where <n> varies and are named /dev/sdc<n> where <n> varies.

 

CloneZilla - Opensource Clone System (OCS) | Mode:

 

Now we need to mount a device as /home/partimag (CloneZilla image(s) repository) so that we can read or save the image in /home/partimag.  ///NOTE/// You should NOT mount the partition you want to backup as /home/partimag

The partition name is the device name in GNU/Linux.  The first partition in the first disk is  hda1? or  sda1?, the 2nd partition in the first disk is  hda2? or  sda2?, the first partition in the second disk is <hdb1> or <sdb1>.   If the system you want to save is MS windows, normally C: is hda1 (for PATA) or sda1 (for PATA, SATA, or SCSI), and D: could be hda2 (or sda2), hda5 (or sda5).

 

This is perfectly clear in an alternative universe.  You are specifying the output device.  Look at the various choices and select the one that contains the name of the manufacturer of the output device.  For example:  Sandisk.  To help things along, you may see CreatedByLinux, truncated, in one of the choices.  Pick that one.

 

Press OK.

 

CloneZilla - Opensource Clone System (OCS)

 

Which directory is for the CloneZilla image (only the first level of directories are shown, and the CloneZilla image (i.e. directory) itself will be excluded.  If there is a space in the directory name, it will _NOT_ be shown.

 

Various top-level directory names are shown.  I recommend that you choose the root directory (i.e. the slash).  It is shown as Top_directory_in_the_local_device.

 

Many lines will appear followed by Press <Enter> to continue.

 

CloneZilla - Opensource Clone System (OCS)

 

Choose the mode to run the following wizard about advanced parameters:

 

I recommend that you choose Beginner mode.  Do not choose Expert mode.  Press OK.

 

CloneZilla - Opensource Clone System (OCS):  Select mode

 

This software will overwrite the data on your hard drive when restoring!  It is recommended to backup important files before restoring!***

///Hint! From now on, if multiple choices are available, you have to press space key to mark your selection.  An asterisk (*) will be shown when the selection is done///

 

Choose savedisk 

Save_local_disk_as_an_image.

Do not choose saveparts

Save_local_partitions_as_an_image.

Press OK.

 

CloneZilla - Opensource Clone System (OCS):  savedisk

 

Input a name for the saved image to use

 

2022-02-23-05- img

 

Press OK.

 

This is the output file.

 

CloneZilla - Opensource Clone System (OCS):  savedisk

 

Choose local disk as source.

This disk name is the device name in GNU/Linux.

The first disk in the system is <had> or <sda>, the 2nd disk is <hdb> or <sdb>  ?Press space key to mark your selection.  An asterisk will be shown when the selection is one.

 

[*] sda 250GB_ST250LT03-9YG14-ST250LT003-9YG14C_W0456VGT

 

 

This is the input device.

Press OK.

 

Choose if you want to check and repair the file system before saving it. 

This option is only for certain file systems which are well supported by fsck on GNU/Linux, like ext/3/4, reiserfs, xfs, jfs, vfat.  Not for NTFS, HFS*.

 

Choose Skip checking/repairing source file system.

 

This is because NTFS is the most popular file system for Windows 7. 

Press OK.

 

CloneZilla advanced extra parameters | Mode: savedisk

 

After the image is saved, do you want to check if the image is restorable?  ///NOTE/// This action will only check the image is restorable, and it will not write any data to the harddrive.

 

Choose Yes, check the saved image

 

Press OK.

 

CloneZilla advanced extra parameters | Mode: savedisk

 

Do you want to encrypt the image?

If yes, eCryptfs program will be used to encrypt the image.  It uses industry-standard cryptographic ciphers, key generation, and passphrase protection mechanisms.  Without your salt/passphrase or private key, nobody will be able to retrieve your data.  //NOTE// You have to remember the passphrase, otherwise the image will _NOT_ be usable in the future.

 

Choose Not to encrypt the image

 

Be sure to put the output USB drive in the vault for safekeeping.  Press OK.

 

Are you sure you want to continue (y/n)

 

Press y.  Press Enter.

 

If your computer has multiple hard disk drives, it will probably be necessary to repeat the entire process more than once.

 

This was a long, long trip.  But finally, the disk image starts being written to the output drive.

 

At the end it becomes obvious that the work is done. You should reboot the computer after removing both USB drives.

 

 

 

 

How to Use MWE-PC-BACKUP-SERVICE

 

1.                The purpose of the MWE-PC-BACKUP-SERVICE is to maintain current backups of files on various computers.  It is configurable on a computer-by-computer basis.  Some or all of the directories on each computer can be backed up to a central computer.  The central computer communicates with the client computers via the SMB protocol.  Design and implementation of this project commenced in 2020 and ended in 2022.  Copyright Mark Dana Rockman 2022.

2.      The Server Message Block (SMB) Protocol is a network file sharing protocol, and as implemented in Microsoft Windows is known as Microsoft SMB Protocol. The set of message packets that defines a particular version of the protocol is called a dialect. The Common Internet File System (CIFS) Protocol is a dialect of SMB.

3.                The central computer exposes a single shared directory that is visible to all of the client computers.  It is expected that the single shared directory shall be located on a dedicated RAID-1 hard disk drive pair of sufficient capacity to backup several computers over a period time.  Someone will be required to monitor unallocated space and remove redundant backups as required.

4.                Each month a FULL backup is performed one time.  After the FULL backup is performed INCREMENTAL backups are performed.  A FULL backup contains copies of most of the files on the computer.  An INCREMENTAL backup contains only those files that have been added or changed since the most recent time of backup.  A file needing to be backed up is indicated by the ARCHIVE bit being set in the file’s metadata. A successful backup clears the ARCHIVE bit.

5.                A Windows service runs in the background.  It is normally invisible to a computer’s end user.  All types of backup run during the period 0000-0600.  Outside the period backups do not run to ensure that end user activity can proceed without conflict from the backup service.

6.                Installation of MWE-PC-BACKUP-SERVICE is simple.  An installation directory is provided to the system administrator. On the central computer, on the capacious drive that hosts the backups, the system administrator shall create a shared directory named

 

<drive-letter>:\HHHHHH\MyBackups

 

The system administrator launches an elevated command prompt.  Then the INSTALL.CMD script is run in the shell.  A graphical user interface appears as shown below.

 

 

 

 

 

7.                Under Drive Letter to be backed up the system administrator selects a drive letter from A to Z.  Usually drive letter C is selected.  If additional drives are to be backed up, they are listed under Additional Drive Letters without commas, spaces, or other punctuation.  For example, if drives C and F and H are to be backed up, the system administrator highlights C under Drive Letter to be backed up and puts the string FH in the text box under Additional Drive Letters.

8.                In IP Address of Computer Hosting Backup Files, the system administrator provides the IP address of the computer that hosts the backups.  Under User Name the system administrator puts the account that has sufficient privileges to support SMB.  Under Password the system administrator puts the password that protects the account.  Under Backup Type the system administrator selects the type of backup to be performed immediately.  Usually, a FULL backup is chosen after MWE-PC-BACKUP-SERVICE is first installed.

9.                Once a month a FULL backup is automatically performed.  At other times an INCREMENTAL backup is performed.

10.          The software depends on the computer being powered up between 0000-0600 as that is when the backups are performed.

11.          Should a backup be in progress, and the computer is restarted or powered off, then a recovery process must be performed before additional backups can be created.  There is a maximum of one backup per calendar day.

12.          The system administrator can monitor progress by inspecting the BACKUP-SERVICE application log.  Go to Computer Management/Windows Log/Event Viewer/Application.  Click on Filter Current Log.  Find BACKUP-SERVICE in the Event Sources dropdown list.  Highlight it. Click OK.

13.          The system administrator can remove the service from any computer, on which it is installed, by following the procedure outlined above for installation.  Run the UNINSTALL.CMD script rather than the INSTALL.CMD script.  It may be desirable to remove backup files from the backup directory after the UNINSTALL.CMD script is successfully completed.

14.          Periodic maintenance of the backup directory should be done to ensure there is backup capacity at all times.

15.          Restoration of files from the backup directory requires the system administrator to execute the GROWLER-X.EXE utility program.  The system administrator provides the name of the backup directory and the name of a directory to receive the recovered files.  Also required is the Computer Name.  The aforementioned utility program must merge INCREMENTAL with FULL backups to recover the latest revision of every file.  It is important that any computer that is running the service maintains an accurate clock.  The KEEPCLOCKUPDATED service is available to ensure the computer remains synchronized with Coordinated Universal Time.  Windows enforces a limit of about 250 characters in filenames.  Restoration of a file can be disrupted when this limit is exceeded.

16.          There is no backup provided of any files in these directories:  %windir%, Program Files, Program Files(x86).  Also, System and Hidden files are neither backed up nor restored.  Access Control Lists and metadata (except for timestamps) are not saved nor are they restored.

 

 

Windows 11

 

 

In October 2021 I adopted Windows 11 on my main production workstation.  There are two workstations and two domain controllers on my LAN.  Both workstations are joined to the domain.  How to join a Windows 11 workstation to a domain It isn’t easy.  The secret is to navigate to the About section of Settings (formerly Control Panel).  On the About page is Domain or Workgroup which you can tap or click to join a domain.

 

Next on our hit parade is a communications failure between the domain controllers and the workstations.  There may be a problem with a Norton security product which I uninstalled to debug this issue.  A ping produced an error message no matter what.  I found the Microsoft Windows Defender Firewall, Domain Section, was on full blast.  I turned it off, thereby exposing my LAN to all manner of criminal attempts to steal software and/or install ransomware.  But that configuration change caused pings to start answering. The more the market offers features and products that do things for security reasons, the more ordinary program behavior get falsely labeled as probable malware. The Internet was invented by the War Department to route electronic messages around nuked population centers.  It does not feature useful diagnostic messages when routing fails.

 

When you right-click the taskbar you are NOT given the option to close all the windows on the desktop in order to view the desktop.  You do see an extensive menu.  You now move the cursor to the extreme bottom right whereupon Show Desktop appears.  Also, where is the link to task manager?  It’s gone.  But you can put its icon on the taskbar.

 

 

 

 

 

Simplest Send Email from Web Form

 

I have not found documentation on this technique, although it probably exists somewhere.  The idea is to let the browser call itself with a specially-crafted URL that contains all of the form contents.  The browser then invokes a convenient email client with most of the text boxes filled in and all of the body filled in with form contents.

 

<form action="MAILTO:genericuser@gmail.com" method="post" enctype="text/plain">

 

<INPUT TYPE="submit" VALUE="--OK--">

 

 

Convert from MBR to GPT

 

This is an essay on putting a PC on a path to The Future.  Windows 11 is the ultra-secure, next generation release of Windows 10.  Microsoft has decided to mandate TPM 2.0 and Secure Boot to thwart various exploits that allow malware to wreak havoc on PCs.  Hard disk drives (HDD) are configured with partitions.  Partitions are individual areas of writable space.  Partitions are described in a partition table that resides inside a certain area of an HDD that also contains a bootstrap loader for an operating system.  IBM invented the Master Boot Record (MBR).  Industry found the need to redesign the MBR because disk capacity increased as technology advanced.  And there were certain other inconvenient limitations of the MBR.  The Globally Unique Identifier Partition Table (GUID Partition Table or GPT) is the redesigned Master Boot Record.  One must jump through flaming hoops to change a PC from what is called the legacy format to the new format.  A PC that sports a GPT must also use the Unified Extensible Firmware Interface (UEFI) which is the redesignated redesigned Basic Input/Output System (BIOS).  Without UEFI one cannot have Secure Boot and Trusted Computing Platform 2.0. 

 

https://docs.microsoft.com/en-us/windows/deployment/mbr-to-gpt

 

I found this document (see link above).  The end user can convert the system hard disk drive from MBR to GPT format by using a utility program calling MBR2GPT that is shipped with Windows 10 (later releases).  No file loss.  Some limitations which MBR2GPT checks for.  An example:  More than 3 partitions? That stops conversion cold A special option (allowFullOS) must be used when MBR2GPT is run on a production Windows 10 system.  If this step is successful then it becomes necessary to reboot to the PC’s native operating system.  In the Dell environment, one presses the F2 key at the Dell logo to enter the BOOT configuration screen. The PC firmware is then instructed to use Unified Extensible Firmware Interface which is happy to deal with GPT format hard disk drives. Any options that sport the word Legacy must be disabled. Once Windows 10 boots into UEFI it is possible to enable TPM 2.0 and Secure Boot.  It appears that Secure Boot can be enabled or disabled at your pleasure.  TPM 2.0 comes along for the ride.  You can check to see if TPM 2.0 is provided in the hardware.  Go to Device Manager and find Security Hardware.

 

mbr2gpt /allowFullOS /convert

 

 

 

 

 

Using PowerShell to Download (Not Play) an Audio File

 

 

[Net.ServicePointManager]::SecurityProtocol = "tls12, tls11, tls"

$source = 'https://www.virginiavoice.org/audio/1501.mp3'

$destination = 'C:\MWE\1501.mp3'

Invoke-WebRequest -Uri $source -Outfile $destination

$source = 'https://www.virginiavoice.org/audio/1502.mp3'

$destination = 'C:\MWE\1502.mp3'

Invoke-WebRequest -Uri $source -Outfile $destination

$source = 'https://www.virginiavoice.org/audio/1503.mp3'

$destination = 'C:\MWE\1503.mp3'

Invoke-WebRequest -Uri $source -Outfile $destination

$source = 'https://www.virginiavoice.org/audio/1504.mp3'

$destination = 'C:\MWE\1504.mp3'

Invoke-WebRequest -Uri $source -Outfile $destination

$source = 'https://www.virginiavoice.org/audio/1505.mp3'

$destination = 'C:\MWE\1505.mp3'

Invoke-WebRequest -Uri $source -Outfile $destination

 

The first line informs PowerShell which protocol variant to use.  We know PowerShell is playing the role of a web browser, such as Microsoft Edge.  The problem is that we are using port 443 instead of port 80.  So, the hypertext transfer protocol is secure. There are old versions of the secure variant and new versions.  PowerShell will pick the one that doesn’t work and then report failure.  It does not suggest a solution.  The first line informs PowerShell to try several variants.  Apparently, this problem has been fixed in Windows releases that include an updated PowerShell.

 

Microsoft Store advertises the availability for download of PowerShell.  It does not provide installation media.  It installs PowerShell at its option.  It may or may not do so.

 

 

Publish Software For The Masses Document

 

The PublishSoftwareForTheMassesDocument application moves this document from a directory on a workstation to the network server where you can bring it up on your browser.  The URL is www.mdrsesco.biz/.  Writing the computer program (the app or application) was challenging but worth the effort.  It complies with technical document RFC959 which is available for viewing at https://www.w3.org/Protocols/rfc959/.

 

There exists an application that transfers a Word document of the type htm and the accompanying image files such as the ones shown below.  Research has shown, as has experience, that the class system.Net.FtpWebRequest contains all the functions necessary to transfer a binary file from the local computer to a remote server.  This avoids all the complexity and odd behavior entailed in writing novel code that implements the FTP RFC from scratch.

Disappointly, Microsoft realized that the class does not work properly.  Worse, Microsoft management decided to deprecate it while leaving the documentation in place. They added the following paragraph:

 

We don't recommend that you use the FtpWebRequest class for new development. For more information and alternatives to FtpWebRequest, see WebRequest shouldn't be used on GitHub.

 

 

 

Automated Backup Service

 

 

 

 

 

 

Installation Instructions

 

 

Find <drive-letter>:\Release

Plug a USB drive into <source-PC>.

Copy the directory <drive-letter>:\Release (and its contents) to a USB drive directory <drive-letter>:\Release

Plug the USB drive into the <destination-PC>.

Copy the <drive-letter>:\Release directory to some out-of-way place on the PC.

Remove the USB device.

Then run the following script in an elevated command prompt:

 

<drive-letter>:

cd \Release

call .\INSTALL.CMD

 

 

 

 

You tell your computer operators to back up their computers on a regular basis.  They have higher priorities.  So, it doesn’t get done.  You even provide a capacious file server where their important files can be saved.  It goes unused.  PC-BACKUP-SERVICE is the solution.  Each computer that requires backing up runs this service.  The first time it runs, every file, except system and hidden files (and those files that are inaccessible), is saved to the file server.  On subsequent days, new files and changed files are saved to the file server.  If, as was the case in my work environment, a hard disk drive takes a powder and cannot be recovered, you can recover the files to a brand-new disk drive using Growler-X.  Each time the month changes (e.g., from January to February) a FULL backup is automatically done.

 

The service is simple to install.  Launch a command line prompt as Administrator.  Make the current working directory the Release directory of the Visual Studio solution.  Launch the Install.cmd script. The first task is to configure the service.  The PC-CONFIGURATOR is an application that allows you to specify which drives on this computer to back up.  The first time you run the configurator, you want to tick the FULL radio button.  You specify logon credentials for the server.  Press CREATE.  A configuration file appears in C:\ CONFIGFILE-U4syKL3SZaGc25gxe8n9. Further on in the Install script the service is installed and started.  Backing up usually commences immediately.  Remember to keep your PC powered on after installation and during the period in which the FULL backup is completed.  You can monitor progress by examining the file C:\ CONFIGFILE-U4syKL3SZaGc25gxe8n9\Configfile.txt.  When the second line changes from FULL to INCREMENTAL, or vice versa, you are free to turn your PC off. If you turn it off during the FULL backup, when the PC is turned back on, any defect in the backup will be removed.  Then FULL backup resumes from the beginning once the date changes.

 

A defect refers to damage done to the backup file when either the client or the server reboots during a backup.  There will usually be a partial file backup at the tail end of the backup file.  Also, there is a hard EOF at the end of the file.  A completed backup file has a soft end-of-file marker.  The remedy for a defective backup is to find the partial backup and remove it. This is time consuming but necessary to avoid a corruption event during file restoration.  Defect removal is automatic for backup files created today, yesterday, and a few days before.  There is a list of files for which it has been determined that there is no defect.

 

For reasons that would become apparent to users of an early version of this software, the hard work of creating backups occurs between the hours of midnight to 0600.  It is assumed that the user leaves the PC powered on at all times and that the user is not trying to do productive work in the interval 0000-0600.  Backing up and defect removal occurs at other times of the day but in such a way that productive use of the PC is not impacted.

 

A cyclic redundancy check is applied to each backed up file.  The CRC more or less guarantees that restored files differ not a single bit from what was backed up.  The file length is also checked. The three metadata dates (creation time, modify time, access time) are restored by GROWLER-X. Odd attributes such as READONLY are not restored.  Occasionally Windows will obstruct such restoration.  But file text is not impacted.  Applications in C:\Program Files and its brother are not saved or restored.  The operating system in C:\Windows is not saved or restored.  Nor the paging file.

 

Once the FULL backup has been created, except for monthly FULL backups, only INCREMENTAL backups are created.  Here is a File Explorer display that shows the result of backing up:

 

The System Administrator is required to provide ample storage capacity on the file server.  She will have to create a shared directory called HHHHHH that contains a directory called MyBackups.  That’s where the backup files go.

 

You can recover files from the MyBackup directory by using the Growler-X tool.  Just launch Growler-X, answer the questions, and wait.  Since storage capacity is not infinite, it is recommended that a system administrator periodically consolidates the backup files to remove redundant and superseded data.  Growler-X is meticulous about preserving timestamp metadata.  File attributes are not preserved.  These include attributes such as READONLY and access control lists.

 

Development of this tool took eight months.  When in the midst of backing up, it may occur that the file server goes offline.  Recovery is complicated.  There is the expected task of reconnecting to the file server.  This entails the use of a pair of Win32 methods that don’t appear in the .NET Framework.  The first method tries to delete the useless drive letter that (formerly) mapped to the shared directory on the file server where the backups are stored.  The method often reports a spurious and meaningless error.  The second method creates the  network drive? which maps a drive letter to the aforementioned shared directory.  That call is successful once the file server is back online, which generally follows a reboot.  And then there is the remnant and incomplete backup file.  The backup file contains images of many files from the PC being backed up.  Each image has a prefix and a postfix.  The prefix names the file and contains these metadata:  creation timestamp, access timestamp, modify timestamp, and byte length.  The postfix contains a cyclic redundancy check (CRC).  The CRC is used to guarantee the validity of the image.  To avoid two passes of each backed up file, the postfix follows the coded file image.  An incomplete image does not contain a postfix.  Backups are created all the time but especially in the period 0000-0600. Once a backup is completed, the service enters the idle loop awaiting a date change.  Every backup filename contains a date.  To prevent overwriting today’s backup during recovery from a file server outage, the incomplete backup defect is removed and further work is delayed until midnight. At midnight the date changes and, thus, the filename changes.

 

Windows does not do timesharing.  One application can easily monopolize the PC to the detriment of other processes.  An I/O bound process is especially prone to getting all the CPU time that it wants.  Thus, an I/O completion is followed by a tiny bit of CPU time to get the next read or write started immediately. The service inserts timed waits at appropriate points to give other processes a chance to progress.  This is especially done outside the period 0000-0600. 

?

 

MWE-PC-BACKUP-SERVICE Operating Instructions

 

1.  Assuming the software has been previously installed, the first step is to halt it.  The Release directory, which has been provided to you, contains a script XSTOP.CMD.  You should create the directory C:\MWE-PC-BACKUP-SERVICE and populate it.  You should launch a Command Line Prompt in privileged mode.  Now you can make sure MWE-PC-BACKUP-SERVICE is not running by entering CD C:\MWE-PC-BACKUP-SERVICE and then typing XSTOP.CMD.

2.  The next step is to remove a prior release of MWE-PC-BACKUP-SERVICE. Launch the Command Line Prompt in privileged mode.  Make the current working directory C:\MWE-PC-BACKUP-SERVICE.  Then type UNINSTALL.CMD. There may be some pauses in the script.  When they occur simply press Enter to continue the script.

3.  For insurance that every accessible file is backed up, you should set the Archive bit in the metadata of every file that is not System or Hidden.  You do this by launching the Command Line Prompt in privileged mode.  Then type Attributes C:\ SET.  You must do this for every drive that you intend to backup through automation.

4.  A previous configuration file may exist.  You may (your option) delete the file C:\CONFIGFILE-U4syKL3SZaGc25gxe8n9\CONFIGFILE.TXT.

5.  Next, install the software.  Launch the Command Line Prompt in privileged mode and set the current working directory as before.  Then launch the script INSTALL.CMD.  The configuration tool appears.  Fill it out and press CREATE.  The SAMPLE button will help you fill it out.  Your task is to enter the correct password.  If a previous configuration file exists (you did not delete it, for example) the form will be populated using data therein contained.

6.  There will be several pauses in the script.  Just press Enter as necessary. 

 

The software is now installed and started.  It shall run continuously.  Its first task is the massive creation of a complete backup copy of every file on the drive that you specified.  (Some files are never backed up because they are installed by Microsoft or by 3rd party software developers when software is installed.) On subsequent days only those files that are brand new or changed will be backed up.  Backing up consumes significant PC resources.  To avoid you having to get work done while competing with the backup service, the backup service runs full tilt between midnight and six a.m.  (0000-0600). At other times it waits for midnight to roll around or runs really slow.  Because an incomplete backup cannot be properly reconstructed by GROWLER-X, the service detects an incomplete backup when it starts and removes it.  GROWLER-X also removes an incomplete backup when it finds one.  The backup service requires your PC to be turned on and the file server to be turned on between midnight and six a.m.

 

 

Format of the Configuration File

 

<drive>[<drive>,<drive>, ]|<IP Address of File Server>|<username>|<password>

FULL

 

 

How to Set the Current Working Directory

 

In the command line prompt enter these commands:

 

<drive>:

CD \<name-of-directory>

 

 

How to Create A Command Line Prompt Icon and Launch A Command Line Prompt in Privileged Mode

 

If there is not already an icon on the desktop that launches the Command Line Prompt, you should create one. Right click a blank spot on the Desktop.  Click New.  Click Shortcut.  In the box that is labeled  Type the location of the item. -- enter this:  CMD.  Click Next.  Click Finish.  To launch the Command Line Prompt in privileged mode, follow these steps:  Right-click the new icon.  Click Run as Administrator.  Click Yes. 

 

 

How to Delete A File

 

erase <name-of-file>

 

 

What To Do When The Service  Gets Stuck?

 

Enter the following command in a privileged command prompt:

 

taskkill /f /pid  <Process-Identifier>

 

Example:

 

taskkill /f /pid 54068

 

You are able to obtain the <Process-Identifier> from the task manager under the services tab.

 

 

 

Attributes

 

The format of the command is:

 

Attributes <name-of-directory>? [SET | CLEAR]

 

The command sets or clears the Archive bit in all of the files contained within the stated directory.  A comprehensive report is produced.  When you do not specify SET or CLEAR, no change of attribute occurs, but a comprehensive report is produced.

 

CAMO

 

CAMO is a tool that reports, in sorted order, the names of files in a directory (datasets in a folder).  You can sort by size and by date.  There are three kinds of dates:  date the file was created, date the file was most recently accessed, and date the file was most recently written to.

 

 

 

 

 

 

 

Downdater for Windows PC

 

The longest common subsequence (LCS) problem is the problem of finding the longest subsequence common to all sequences in a set of sequences (often just two sequences). It differs from problems of finding common substrings: unlike substrings, subsequences are not required to occupy consecutive positions within the original sequences. The longest common subsequence problem is a classic computer science problem, the basis of data comparison programs such as the diff utility, and has applications in bioinformatics. It is also widely used by revision control systems such as Git for reconciling multiple changes made to a revision-controlled collection of files. [DOWN]

 

http://en.wikipedia.org/wiki/Longest_common_subsequence_problem

 

A downdater is a text file comparison program.  It produces from similar inputs (2 files) a differences file (1 file) that can be applied to the A file to produce the  B file.  Differences are denoted

 

     -N

     -N,M

 

The lines that follow N are inserted after line N in file A.  The lines that follow N,M are inserted in place of lines N through M in file  A.

 

You can compare two files with DOWN by using the following syntax:

 

     DOWN <name-of-file-A> <name-of-file-B> <differences-file>

 

You can compare files in corresponding directories with the following syntax:

 

     DOWN <name-of-directory-A> <name-of-directory-B> <differences-file>

 

When comparing directories, the various difference files are separated with a record of the form:

 

     *<name-of-file>

 

Note that only certain source code and symbolic files are eligible for comparison. A complete list is available if you send a stamped, self-addressed envelope to the author.

 

A better way to compare an old folder containing symbolic files with an updated version of the same folder involves a special call to DOWN, followed by execution of a skeleton, followed by execution of a BAT script that calls DOWN repeatedly to create several differences files rather than the single such file that otherwise is produced.  The syntax of the special call is

 

DOWN *<name-of-older-folder> <name-of-newer-folder> <name-of-new-differences-folder>

 

Notice the asterisk that comes before <name-of-older-folder.  This signifies that DOWN is to produce an SSG skeleton in the new-differences-file together with an SGS file to be used with the skeleton.  The skeleton, in conjunction with the stream generation statements, produces a CMD script that builds a DifferencesFolder inside the new-differences folder.  Only those folder pairs that exist and exhibit differences end up in the DifferencesFolder. 

 

Example:  Run DOWN *first-directory second-directory directory-to-place-skeleton.  Then CD to directory-to-place-skeleton and type RUN.BAT.  This launches a lengthy DOS script that repeatedly calls DOWN to construct a DifferencesFolder, which see.  As part of this mechanism, the tree structures of the first-named directories are compared and differences reported, if any.  This operation requires that you have installed the Symbolic Stream Generator (SSG).

 

A fourth argument on the command line is optional and may be N or R.  You can suppress the Directory Structure Examiner with the N option.  You can reverse the order of the first two arguments with the R option.

 

The option -LongRunning may appear anywhere on the command line to enable comparisons that exceed 15,000 lines.  Run times can be exceedingly long when the number of lines is exceedingly great.  Run times are O(n2) where n is the number of lines in the longest file.

 

 

SIRASM2 takes the output of DOWN and applies it to what is called a base text file to produce an updated text file.  For example, suppose the base file is

A

B

C

 

And the updated file is

A

4

C

 

Then DOWN will produce the following differences file

-2,2

4

 

The command

SIRASM2 base diff updated

reproduces the updated file given only the base file and the differences file.  [SIRASM2]

 

 

Here is a sample of real-world output.  The asterisk record signifies the following difference records apply to the named file.

Suppose you have a directory (call it Sam) that is populated with project directories.  Each project directory represents a Visual Studio solution to a different problem.  You maintain current and backup copies of Sam.  You work on improving a solution and wonder what changes you ended up making.  Knowing what changes you made can be helpful when you get around to altering the documentation to reflect the changes.  It can be helpful in detecting mistakes.  It can be helpful when you need to create two similar solutions that contain exactly the differences you want.  Here is a command file for doing a massive comparison of Sam with one of its backups:

1.  C:

2.  CD \Users\Public

3.  DELTREE .\LSA

4.  MKDIR LSA

5.  CD LSA

6.  DOWN *U:\ C:\CSOFTWARE F60

7.  CD \Users\Public\LSA\F60

8.  .\RUN.CMD

9.  CD \Users\Public\LSA\F60\DifferencesFolder

10.          ERASE .\DifferencesFolder.TXT

The command file RUN.CMD is generated by DOWN.EXE.  It invokes the Symbolic Stream Generator that sequentially calls DOWN.EXE for corresponding files in Sam and its backup.  In the end you get a directory named DifferencesFolder that contains one directory for each solution that has undergone non-trivial changes.  Crystal_DOWN is an application that produces a report from a single entry in DifferencesFolder. Such an entry corresponds to a Visual Studio solution as previously mentioned.  The OUT.CMD file produced by RUN.CMD contains one commented-out command line for each folder produced by OUT.CMD.  You can uncomment one of the lines and launch it to obtain a report that shows the source lines that the correction file deletes.  These lines are prefixed with six asterisks.  If application A contains source files B, C, and D then the correction file will contain a subset of corrections separated with star cards for B, C, and D.

 

SSG (Symbolic Stream Generator) is an app that compiles a program called a  skeleton? and simultaneously compiles a text file that is populated with Stream Generation Statements (SGSes).  The skeleton contains the logic.  The SGSes contain the data.  The program creates a text file in accordance with directives in the SGSes.  SSG is a handy tool when you want to generate, for example, a lengthy program structure (e.g. a case statement for incorporation within a C++ or C# app) based on a list of things.  I have used it many times for that purpose.  But SSG is capable of doing more complex things.  For example, a commercial software publisher uses it to generate MXML, the language of Adobe Flex, to produce web apps that allow end users to enjoy a Rich Internet Experience.  The SGSes describe what the page or pages are supposed to look like and the skeleton transforms the description into exactly the right MXML for the purpose.  You can RTFM at http://www.mdrsesco.biz/SSGManual.htm.

 

There is a tiny feature added in 2020 that allows you generate an integer with leading zeroes.  The notation [*<variable>] generates the variable length substring representation of the integer stored in <variable>.  The new feature allows you specify a fixed field width.  There will be leading zeroes to make the field the same width for all integers.  The additional syntax is [*<variable>:<field-width>]. Example: [*Z] becomes [*Z:2] which specifies that the edited field is two characters in length with a possible leading zero. [SSG]

 

                                        

 

ExFTP

 

FTP is File Transfer Protocol. ExFTP is an application that allows you to send a file to or receive a file from an FTP server.

 

 

Remote site address is the IP address of FTP server.  Local filename is the fully-qualified filename on the PC that is running ExFTP that shall send from or receive to the contents of the file.  Remote filename is the short filename on the FTP server that is the source of data or the destination for data depending on whether Receive is checked or Send is checked, respectively.  The S button is to be pressed to transfer the short filename (no drive letter, no backslashes) to Remote filename.  FSERV sets the Remote site address to the MDRSESCO local FTP server.  ADC sets the Remote site address to the MDRSESCO other local FTP server.  DiscountASP sets the Remote site address to MDRSESCO subscription service ISP file server.  Tech Road sets the Remote site address to the MDRSESCO client’s file server.  Remote site username is part of the credentials required to login to the selected FTP server.  Remote site password is part of the credentials required to login to the selected FTP server.  The GO button is pressed when all of the text boxes have been properly filled in.

 

 

LAUNCH

 

The Windows Registry is a junk drawer (a database divided into hives) of parameters that govern the behavior of the Windows operating system and various applications.  One of the parameters allows you to specify an application, script, or other thing that can be launched after an end user has entered valid credentials.  NETPLWIZ is a Microsoft application that ships with Windows that allows the end user to specify the logon credentials such that logon occurs automatically following a reboot.  Once the desktop becomes visible, after a reboot, the LAUNCH application will instruct Windows to start an application or a script.  This is done via the Registry.  LAUNCH should be executed from a non-elevated command line.  In other words, LAUNCH should be run under a user account other than Administrator. 

 

Syntax:  LAUNCH <label> <file-to-launch>

 

The label should be a unique alphanumeric string.  The file-to-launch should identify an EXE file or a CMD or BAT file.  LAUNCH verifies the existence of the file. 

 

Example:  LAUNCH  PRIME95  C:\$001\PRIME95.EXE

 

Successful completion is obvious.  At the next logon, the PRIME95.EXE application will be Windows’s responsibility to launch.  You can remove a previously entered Registry entry by providing only the label.

 

Example:  LAUNCH  PRIME95

 

From AUTOMATION to DARTH

 

From AUTOMATION to DARTH is a practical end user interface which is used at a radio reading service for the blind.  The service broadcasts constantly.  There is a daily live broadcast of the local newspaper.  There are many broadcasts pre-recorded for later broadcast.  There are broadcasts that originate in a similar service in Minneapolis-St. Paul. Twin Cities emissions are relayed from a satellite and broadcast, as usual, on a local PBS-affiliated FM radio station.  Special radio receivers are needed for blind subscribers to receive the service, which is on a sub-channel of the aforementioned FM radio station.  Sometimes a recorded radio broadcast is to be made available to clients of the DIAL-IN Service.  Blind subscribers can browse newspapers using DTMF telephones.  To move a recorded radio broadcast to the DIAL-IN service all one has to do is copy a WAV file from the SCA radio station AUTOMATION computer to the DIAL-IN Service computer.  To make this so easy that a child can do it, Mark Rockman designed and implemented this end-user interface.  Policy dictates that lengthy programs be divided into two parts.  There is a checkbox for that.

 

 

 

 

 

 

Directory Sizes (dirx)

 

From the command line, enter dirx <name-of-directory>.  Each directory in the named directory is inspected for the files that it contains and for the files in the directories that it contains, recursively.  Reported is the total size in bytes of all the files in the named directory.

 

 

 

Grid Binding for the Web and for the PC

 

For the Web:

 

using System;

using System.Collections.Generic;

using System.Linq;

using System.Web;

using System.Web.UI;

using System.Web.UI.WebControls;

using System.Data.SqlClient;

 

public partial class _Default : System.Web.UI.Page

{

? protected void Page_Load(object sender, EventArgs e)

? {

? System.Data.SqlClient.SqlConnectionStringBuilder builder = new System.Data.SqlClient.SqlConnectionStringBuilder();

? builder["Data Source"] = "(local)";

? builder["User ID"] = "sa";

? builder["Password"] = "123456";

?  system.Data.SqlClient.SqlConnection sqlConnection1 = new System.Data.SqlClient.SqlConnection(builder.ConnectionString);

? System.Data.SqlClient.SqlCommand cmd = new System.Data.SqlClient.SqlCommand();

? cmd.CommandType = System.Data.CommandType.Text;

? cmd.Connection = sqlConnection1;

? cmd.Connection.Open();

? cmd.CommandText = "USE MWEDatabase;";

? int nonqueryResult = cmd.ExecuteNonQuery();

? SqlConnection sqlConnection = sqlConnection1;

? SqlCommand sqlCommand = new SqlCommand( "SELECT * FROM Customers", sqlConnection);

? SqlDataReader reader = sqlCommand.ExecuteReader();

? GridView1.DataSource = reader;

? GridView1.DataBind();

? }

}

 

For the PC:

 

using System;

using System.Collections.Generic;

using System.ComponentModel;

using System.Data;

using System.Drawing;

using System.Linq;

using System.Text;

using System.Threading.Tasks;

using System.Windows.Forms;

using System.Data.SqlClient;

 

namespace BindQueryToGridWinForms

{

? public partial class Form1 : Form

? {

? public Form1()

? {

? InitializeComponent();

? System.Data.SqlClient.SqlConnectionStringBuilder builder = new System.Data.SqlClient.SqlConnectionStringBuilder();

? builder["Data Source"] = "(local)";

? builder["User ID"] = "sa";

? builder["Password"] = "123456";

? System.Data.SqlClient.SqlConnection sqlConnection1 = new System.Data.SqlClient.SqlConnection(builder.ConnectionString);

? System.Data.SqlClient.SqlCommand cmd = new System.Data.SqlClient.SqlCommand();

? cmd.CommandType = System.Data.CommandType.Text;

? cmd.Connection = sqlConnection1;

? cmd.Connection.Open();

? cmd.CommandText = "USE MWEDatabase;";

? int nonqueryResult = cmd.ExecuteNonQuery();

? cmd.CommandText = "SELECT * FROM Customers";

? SqlDataAdapter dataAdapter = new SqlDataAdapter(cmd.CommandText, cmd.Connection);

? SqlCommandBuilder commandBuilder = new SqlCommandBuilder(dataAdapter);

? // Populate a new data table and bind it to the BindingSource.

? DataTable table = new DataTable();

?  ?table.Locale = System.Globalization.CultureInfo.InvariantCulture;

? dataAdapter.Fill(table);

? BindingSource dbBindSource = new BindingSource();

? dbBindSource.DataSource = table;

? DataGridView dbGridView = dataGridView1;

? // Resize the DataGridView columns to fit the newly loaded content.

? dbGridView.AutoResizeColumns(DataGridViewAutoSizeColumnsMode.AllCellsExceptHeader);

? // you can make it grid readonly.

? dbGridView.ReadOnly = true;

? // finally bind the data to the grid

? dbGridView.DataSource = dbBindSource;

? }

? }

}

 

 

Creating a Word List Using the Text Editor

 

Here’s the result of combining three applications to build and maintain a list of words gleaned from newspaper articles.  The text editor tokenizes lines of text (isolates words) producing a text file. An existing text file of alphabetized words (no repeats) is combined (using cat) with the newly gleaned words.  Text editor alphabetizes the combination.  The File Comparator (DOWN) shows you the location and value of each added word.  Finally the combined list is saved.

 

The shell script:

 

C:

cd \Words-EN

ed text.txt task1.txt

C:

cd \Words-EN

cat Words.txt unique-tokens.txt merged-tokens.txt

C:

cd \Words-EN

ed merged-tokens.txt task2.txt

down Words.txt merged-tokens.txt diff.txt

type diff.txt

 

The word extractor:

 

tokenize

sort

unique

w unique-tokens.txt

 

The list combiner:

 

sort

unique

exit

 

 

Here’s a list of newly added words:

 

-220

alphabetized

alphabetizes

-1264

combination

-1266

combining

-1302

comparator

-2944

gleaned

-3624

isolates

-3995

location

-4079

maintain

-4483

newly

-4486

newspaper

-5859

saved

-6912

Tokenizes

 

Text Editor (ED)

 

There follows a description of what is nominally a text editor.  It is that and more.  [EditTechnology]

A Complete List of Rockman's Text Editor Commands

v  ace

v  add

v  annuity

v  append

v  asga

v  asgax

v  attrib

v  begin

v  bmi

v  breakup

v  bt

v  cd

v  change

v  climit

v  close

v  columnblank

v  combine

v  commands

v  copy

v  copychg

v  cpumask

v  crp

v  cwd

v  datefix

v  dcopy

v  delete

v  deleteafter

v  dhold

v  dirsize

v  dnext

v  dse

v  dup

v  easter

v  egl

v  ej

v  eject

v  elname

v  erl

v  excelhelper

v  exist

v  extract

v  f1

v  filecompare

v  find

v  flist

v  free

v  gather

v  getfilenames

v  getfiles

v  getshort

v  ghp

v  gl

v  goadmin

v  goto

v  head

v  hlbl

v  hold

v  ibefore

v  ident

v  insert

v  inuse

v  isadmin

v  isprivil

v  kill

v  last

v  lgoff

v  lizard

v  locate

v  ls

v  mdl

v  merge

v  missing

v  next

v  o

v  omit

v  opc

v  open

v  opennew

v  own

v  permutation

v  ping

v  play

v  pmask

v  print

v  ps

v  ps1

v  pswdoff

v  putquotesaround

v  pwoff

v  q

v  randgen

v  rb

v  rc

v  rdp

v  replicating

v  retype

v  reverse

v  ro

v  runsql

v  scale

v  select

v  sizes

v  sort

v  split

v  sql

v  sunrise

v  superficial

v  sz

v  tab

v  td

v  tld

v  toc

v  tokenize

v  top

v  touch

v  twilight

v  undo

v  unique

v  unsplit

v  up

v  uro

v  vw

v  wav

v  winplaces

v  wpt

v  write

v  wulog

Aliases

General Principles

The Rockman Text Editor accepts commands from the keyboard or from a file. The syntax of the shell command is

ed [name-of-file-to-be-edited [name-of-command-stream]]

You can omit both command line arguments, in which case commands can establish what is to be edited, and commands shall originate at the keyboard. You must provide name-of-file-to-be-edited when you want to provide name-of-command-stream.

Commands comprise three parts: 1) the name of the command, 2) option, 3) arguments. Example:

r,11 PAIJ

contains the r command, the option 11, and the argument PAIJ. Only the name of the command is mandatory.

The editor maintains the file that is being edited in area called the memory file.  The content of the memory file can be created from scratch or it may originate in a file in a file system on a device.  The primarily purpose of the Text Editor is to support your efforts to make changes to the memory file.   Eventually, once you are done making changes, the memory file is copied to a file in a file system on a device for long-term storage.

The editor is either in EDIT mode or in INPUT mode. You can switch between modes by entering a blank line. In EDIT mode, you can instruct the editor with commands. In INPUT mode, non-blank lines are successively added to the memory file.

The editor pulls all of the file's text into memory. In memory is where all the editing occurs. At the end of your editing session, it is your choice whether to save any changes that you have made or to discard the changes. Type EXIT to save the changes. Type OMIT to discard them.

While you are editing, keep in mind that there is always a current line, which you can see by typing PRINT.  Navigation throughout the memory file is by line number, which may be absolute (e.g. go to line 5) or relative (e.g. go 5 lines down from here).

The text editor is one of several Rockman's Tools that you may find useful.  The Symbolic Stream Generator allows you to write programs that generate text files based on simple, structured data.  The Downdater allows you to see the differences between text files presented in -n and -n,m notation.  SIRASM2 allows you to apply changes in -n and -n,m notation to a text file in order to obtain a new text file.

Ordering

Write to mrockman@acm.org.

Command Documentation

ace

The ace command displays Access Control Entries (ACEs) that are component parts of Access Control Lists (ACLs).  ACLs are instrumental in granting or denying permissions to access system resources; in this case files and directories.  Directories are also known as Folders.  The syntax of the command:  ace <name-of-file-or-name-of-directory>.  When you name a file, the command reports all associated ACEs.  When you name a directory, the command reports the ACLs associated with the directory, all the directories therein contained and all the files therein contained.

add

The add command brings in at the current location additional lines from either of two sources. If you name a text file then the full file is added at the current point in the file being edited. If you don't name a file then lines from the most recent COPY or DCOPY command are added at the same place.

annuity

The annuity command creates an amortization table following an interactive query session that requests all necessary data.  You must provide the amount borrowed, the duration of the repayment period in months, the annual percentage rate, the starting year and month.

append

The append command relocates the current edit point to after the last line in the file being edited. The editor then goes into INSERT mode.

asga

The asga command opens, when possible, the named file in such a manner that another process can also open the file.  The free command closes the file.  Only one file at a time can be open during an edit session.

asgax

The asgax command opens, when possible, the named file in such a manner that another process cannot also open the file.  The free command closes the file.  Only one file at a time can be open during an edit session.

attrib

The attrib command reports the attributes of a file.  It does not report the attributes of a directory.  The syntax:  attrib <name-of-file>.  What follows shows all the possible attributes.

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.ReadOnly) == System.IO.FileAttributes.ReadOnly)

? answer = "R" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Hidden) == System.IO.FileAttributes.Hidden)

? answer = "H" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.System) == System.IO.FileAttributes.System)

? answer = "S" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Directory) == System.IO.FileAttributes.Directory)

? answer = "D" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Archive) == System.IO.FileAttributes.Archive)

? answer = "A" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Device) == System.IO.FileAttributes.Device)

? answer = "V" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Normal) == System.IO.FileAttributes.Normal)

? answer = "N" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Temporary) == System.IO.FileAttributes.Temporary)

? answer = "T" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.SparseFile) == System.IO.FileAttributes.SparseFile)

? answer = "P" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.ReparsePoint) == System.IO.FileAttributes.ReparsePoint)

? answer = "X" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Compressed) == System.IO.FileAttributes.Compressed)

? answer = "C" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Offline) == System.IO.FileAttributes.Offline)

? answer = "O" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.NotContentIndexed) == System.IO.FileAttributes.NotContentIndexed)

? answer = "N" + answer;

? if ((System.IO.File.GetAttributes(filename) & System.IO.FileAttributes.Encrypted) == System.IO.FileAttributes.Encrypted)

? answer = "E" + answer;

begin

The begin command works with the COPY and DCOPY commands to mark the starting line of a copy operation.

bmi

The bmi command calculates Body Mass Index for an individual’s height and weight.  It accepts English and metric units values.

·         Format 1: bmi English,5,9,150

·         Format 2: bmi metric,175,68

The first form syntax is bmi English,<height-in-feet>,<height-in-inches>,<weight-in-pounds>

The second form syntax is bmi metric,<height-in-centimeters>,<weight-in-kilograms>

 

  If your BMI is below 20:

? This indicates a lean BMI, which means you have a low amount of body fat. 

? If you are an athlete, this can be desirable.

? If you are not an athlete, a lean BMI can indicate that your weight may be too low which may lower your immunity.

? If your BMI and body weight are low, you should consider gaining weight through good diet and exercise habits, to increase your muscle mass.

 

? If your BMI is between 20 and 22:

? This indicates the ideal, healthy amount of body fat, which is associated with living longest, and the lowest incidence of serious illness.

? Coincidentally, it seems this ratio is what many individuals perceive to be the most aesthetically attractive.

 

? If your BMI is between 22 and 25:

? This is still considered an acceptable range, and is associated with good health.

 

? If your BMI is between 25 and 30:

? You are considered "Hefty" and should find ways to lower your weight, through diet and exercise. 

? You are at increased risk for a variety of illnesses at your present weight.

? You should lose weight by changing your diet and exercising more.

 

?  ?If your BMI is over 30:

? This indicates an unhealthy condition, your excess "Prosperity" is putting you at risk for heart disease,

? diabetes, high blood pressure, gall bladder disease and some cancers.

? You should lose weight by changing your diet and exercising more.

 

breakup

The breakup command partitions all the lines being edited into multiple files, 20,000 lines in each file.  The new files appear in a new directory C:\Users\Public\Breakup and the files are numbered sequentially as 1.txt, 2.txt, and so forth.

 

bt

The bt command displays the operating system version, the logon server name, whether the computer is joined to a domain, the logged in user name, the operating system build number, the marketing name of the operating system version, the computer name, the time of last system reboot (also called system restart), the duration of current system uptime, and the register width in bits according to the operating system build type. The bt command also displays the various IP addresses that point to the local computer.  Additional information about the computing environment is provided in the edit file.  For example, OSArchitecture 64-bit is an environmental factoid that may interest you.

 

 

cd

The cd command changes the current working directory to the named directory.

change

The change command replaces one occurrence or several occurrences of a string with something else. The argument contains the pattern /old-string/new-string/ where the slash is some convenient delimiting character, old-string is a substring to be replaced, and new-string is what the replacement shall be. What follows is a more complete description of the syntax.

Syntax of change command:

?change[,<option>] <delimiter><first-string><delimiter><second-string><delimiter><second-option>

?Typically, <delimiter> is the slash character, like this:  /

?When <first-string> or <second-string> or both contain / then some other character that doesn't appear is used.

?The change command changes occurrences of <first-string> substituting <second-string>.  However, the meaning of the aforementioned arguments can be reversed.

?When <option> appears it is an integer and denotes the number of lines including the current line to be inspected for changes.

?For example, if we are on line 4 and <option> is 3 then lines 4, 5, and 6 will be inspected.  When <option> is * the integer is 999999.

?Ordinarily only the leftmost occurrence of <first-string> is changed.

?However, when <second-option> is A or G then all occurrences on a line are changed.

?When <second-option> is A then the scope of the change is the current line and all following lines.

?When <second-option> is G then the scope of the change is the current line and those lines within the scope specified by <option> or 1 line when <option> is not present.

?When <second-option> is R then the meaning of <first-string> and <second-string> are reversed.

?UNDO is a separate command that allows you to reverse changes you made in error since the last time the top of the file was traversed.

?The change command, once entered, can be repeated by simply entering the command by itself.  You omit everything beginning with the leftmost <delimiter>.  You may, if you choose, modify the command to include <option>.  The repeated command is possible because everything except <option> is saved by the text editor.

?The command C,S displays the saved change command status.  There must be some text in the file for you to see the status.

 

Task to Be Performed

The Necessary Command

Characterization

Change the first occurrence of ABC on the current line and the next two lines

C,3 /ABC/XYZ/

ChangeFirstOccurrenceNextSeveralLines

Change the first occurrence of ABC on the current line

C /XYZ/ABC/R

ChangeFirstOccurrenceCurrentLineOnly

Change all occurrences of ABC on the current line and the next two lines

C,3 /ABC/XYZ/G

ChangeAllOccurrencesNextSeveralLines

Change all occurrences of ABC on the current line and on all following lines

C,* /ABC/XYZ/G or C /ABC/XYZ/A?

ChangeAllOccurrencesCurrentLineAndThereAfter

Change the first occurrence of ABC on the current line and on all following lines

C,* /ABC/XYZ/

ChangeFirstOccurrenceCurrentLineAndThereAfter

Determine whether a change command has been saved.

C,S

 

climit

The climit command limits the range of columns affected by any of the forms of the change command. Specify the left and right column numbers separated by a SPACE. Example: cli 42 49

close

The close command causes the editor to write the lines in the memory file to disk and then causes the editor to terminate.

columnblank

The columnblank command operates on all the lines of the in-memory text file.  You specify which columns you want blanked and the command performs that function.  The syntax allows you specify several space-separated ranges and/or individual column numbers.  For example, the command  columnblank 42 39,80? puts a blank space in column 42 and blank spaces in columns 39 through 80, inclusive.  This is for tabular data wherein your desire is to eliminate certain columns. Your column numbers must be integers.

commands

The commands command lists the canonical names of all the editor's commands in sorted order.

copy

The copy command creates a store that contains the lines between the begin marker and the current line.

copychg

The copychg command duplicates a directory tree, including leaf nodes.  The filename extension of the leaf nodes is altered according to specification.  You name the input directory, the output directory, and the new filename extension.

cpumask

The cpumask command reports the configuration of central processors in the local computer. 

 

crp

The crp command must be run with the text editor elevated.  It creates a restore point. The System Restore button of the System Protection tab of the Advanced System Settings of the Properties page of This PC in Windows Explorer allows you to restore the settings saved in the restore point.

cwd

The cwd command displays the current working directory.

deleteafter

The deleteafter command finds a match on the current line, erases the match, and everything to the right of the match.

dcopy

The dcopy command is identical to the copy command except that the lines that are sent to the store are deleted.

delete

The delete command removes one or more lines from the file being edited starting with the current line. You can specify the number of lines to delete (e.g. DELETE 5). The default number is 1.

dhold

The dhold command saves the current line to a special store (named the "hold" buffer) and deletes the line from the file being edited. You can resurrect the line with the dup command.

dirsize

Suppose you have a folder.  In that folder are several other folders.  These are top level folders.  You want to know how much space the files in those folders occupy.  The dirsize command reports that information. The dirsize command has been superseded by the sizes command.

dnext

The dnext command deletes the current line and sets the current line to the line that follows it.

dse

The dse command performs a Directory Structure Examination.  You can detect structural differences between similar directories. A structural difference occurs when a directory or a file appears in one root directory but not in the corresponding place in the other.

dup

The dup command inserts one or more lines at the current position. The lines inserted originate in the hold buffer. You can specify the number of lines to duplicate (e.g. DUPLICATE 5). The default number is 1.

easter

The easter command displays the date upon which Easter next falls.  Based upon an algorithm published as follows:  Copyright (c) Simon Kershaw 1996. All rights reserved.

 

eject

The eject command causes the operating system to flush the buffers of a removable device and, perhaps, to physically eject the device. The format is ej drive-letter.

elname

The elname command displays the name of the file that is being edited.

erl

The erl command does a patterned extract of records from a comma-separated values file.  The idea is to get rid of the column heading records and various interstitial records that clutter up the file.  You specify, in order, the name of the input file, the name of the output file, the number of records to discard at the start of the file, and the number of records to skip following those records that are retained. The command assumes that retained records boast of a date in MM-DD-YY format in the leftmost column.  Since Excel handles that format inconsistently, the command reformats the date to Month Day, Year format.  For example, 01-02-03 is transformed to  January 2, 2003?.  If you have a text file open when you type in the erl command then the extracted lines end up in the file.  If not, not.  The command to extract retained lines could be:  erl C:\Users\Public\TESTBED.TXT "C:\Users\Public\Outlet.txt" "2" 1

excelhelper

The excelhelper command modifies a comma-separated value file such that those comma-separated fields that (apparently) contain a date are revised to a form that Excel finds unambiguous.  The Excel in Microsoft’s Office Suite 2007 is inconsistent in handling dates of the form MM/DD/YYYY, YYYY/MM/DD, MM-DD-YYYY, and YYYY-MM-DD.  So these are changed to <month-name> <day-of-month>, <year> within double quotes. The month always precedes the day-of-month unless you put an asterisk on the command.  Then the day-of-month always precedes the month as is usual is Europe.  Use  excelhelper *? when the date is 15/4/2015, for example.

exist

The exist command probes the named directory or filename for existence.  The several outcomes are: 1) the name is reported to be a directory, 2) the name is reported to be a file, 3) the system threw an exception when probing the name, 4) the name does not exist in the file system.  These outcomes are according to the .NET Framework which produces different existence results depending on Platform Target, a compilation parameter.  With Platform Target set to x86 you get one answer.  With Platform Target set to x64 you get a different answer.

exit

The exit command causes the editor to write the lines in the memory file to disk and then causes the editor to terminate.

extract

The extract command removes all those lines in the edit file that do not contain the argument pattern.  For example, the command extract InstallUtil.exe leaves the edit file with only those lines that contain InstallUtil.exe.  Matches are case-insensitive.

f1

The f1 command changes routes.rb in a Ruby on Rails application file so that the application uses named routes.

filecompare

The filecompare commands performs a byte-for-byte comparison of the two files that you name.  The answer is YES or NO.  YES: the two files are identical.  NO: the two files are different.  Suppose some program you?ve been using for years suddenly halts with an addressing exception.  Suppose you have a backup copy of the .EXE file.  Suppose you compare the backup with the production copy.  Then you can assure yourself that the problem is not due to file corruption.

find

The find command searches the lines following the current line for a matching pattern. In the pattern, space characters match any character at the same relative position on a targeted line. You can use find to locate based on columnar position. When a tab character is defined (see tab command), you can use it as part of the pattern. For example, suppose LMJ appears in columns 11, 12, and 13 respectively. With the tab character set to semi-colon and tab positions set to 11, 21, and 39, the command

find ;LMJ

matches on the first line following the current line that contains LMJ in columns 11, 12, and 13.  The command option specifies the desired number of match occurrences.  The default is 1.

flist

The flist command creates in the current file a complete listing (with headings) of the symbolic files of the named directory.

free

The free command closes the file that was previously opened with the asga or asgax command.  Just because you open a file with one of these commands doesn’t mean you are allowed to touch the file’s contents.  These commands exist solely to demonstrate how the Win32 file system file open function operates.  The number of links to the file is reported.  This relates to the concept of  hard link,  in which multiple file system name entries point to the same physical file.  It’s a kind of aliasing mechanism.  You cannot open a hard link (an alias) via asga/asgax.  You can open the file that the hard link designates.

gather

The gather command, from a specified directory tree, aggregates all the files that have the same filename extension.  Syntax:  gather <input-directory-name>,<output-directory-name>,<filename-extension>

The output directory at <output-directory-name> and all its content is removed before the command commences to locate and copy files to it.  Many files may have the same filename but only one such file (by each distinct name) ends up in the output directory.  This command was motivated by the iPhone which stores many photographs in several randomly named directories.  The desire is to aggregate all the JPG files in a single Windows directory.  Suppose there is a directory C:\A that contains directories C:\A\B and C:\A\C.  Suppose further that 1.jpg resides in C:\A\B and 2.jpg resides in C:\A\C.  The gather command will put copies of 1.jpg and 2.jpg in the output directory.  The output directory will contain no directories of its own. A special case of RASTER as <filename-extension> matches any of the following filename extensions: png bmp dib jpg jpeg jpe jfif gif tif tiff pic pict mov avi.

getfiles

The getfiles command enumerates the files in a specified directory.  If you don’t specify a directory, the default directory C:\ is used.  Enumeration depends on the .NET Framework method System.IO.Directory.GetFiles() and its GetDirectories() counterpart.  Certain files for unknown reasons cannot be enumerated this way.  Each file is listed on a separate line prefixed with its time-of-last-write.  The format of the timestamp is YYYY-MM-DD HHMMSS.  You can sort the files by timestamp.  Use the SORT command.

getshort

The getshort command uses INTEROP to access the Win32 method that converts a normal NTFS filename to its 8.3-format equivalent.  The 8.3 format originated in the FAT file system. It is retained for compatibility reasons.  The long filename is printed.  Then the short filename is printed.  The text currently being edited is erased.  A one-line file replaces it.  That line contains the short filename.  This all works splendidly unless the named file does not exist, in which case Windows returns a zero-length string to represent what it believes properly to be a? representation of the short filename.  Unless the long filename is syntactically incorrect, it would have been possible for Windows to present a valid short name version even when the file itself does not exist.  The documentation at https://msdn.microsoft.com/en-us/library/windows/desktop/aa364989%28v=vs.85%29.aspx neglects to mention that GetShortPathName (the exposed method of interest) needs the path to exist. For example, getshort C:\Documents and Settings works, yet getshort Ugly does not.

ghp

The ghp command has two forms:  ghp true and ghp false.  The command tries to obtain granted privileges with true. The command tries to revoke privileges with false.  All the privileges known at editor build time are entered into the edit file (assuming you?ve provided the name of one) along with YES when granted and NO when revoked.  Windows does not permit just anyone to obtain a granted privilege.  Some privileges are granted to mere mortals.  Some are granted only to deities.  Group Policy governs those privileges that are granted.

goadmin

The goadmin command terminates the current instance of the text editor application and restarts the application under the Administrator account.  This gives the application extraordinary privileges.  You should save any text editor work that you may have accomplished before calling goadmin.  Once the application restarts, use the old command to reload your work.  To check that you have, in fact, been granted privileges, enter the isadmin command.

goto

The goto command relocates the current position to the line whose number is given. A synonym for goto is simply the line number.

head

The head command is identical in function to goto 1.

hlbl

The hlbl command is HTML Line by Line.  It reformats the content of the file that is currently open, which should be an HTML stream, such as one might find in a file with filename extension htm or html.  The idea is to put each tag or tag-delimited content on a separate line, for ease of analysis by a human observer.  The output is intended to be semantically identical to the input.

hold

The hold command saves the content of the current line to a special place inside the editor called the "hold" buffer.  You can resurrect the line with the DUP command.

hw

The unimplemented hw command sets a horizontal window on the file being edited. The window applies to a specified range of columns. For display purposes only, columns outside the window do not appear. Suppose you want to see columns 11 through 21. You enter hw 11 21.

ibefore

The ibefore command is identical to the insert command except that the line is inserted before the current line.

insert

The insert command inserts a new line after the current line.

inuse

The inuse command probes the file you name to see if it is in use by another process.

isadmin

The isadmin command displays YES when the editor is running with elevated (i.e. Administrator) privileges and NO otherwise.

kill

The kill command terminates the process whose name you specify.  To obtain a list of extant processes, use the ps command.

last

The last command relocates the current position to the last line in the file.

lgoff

The lgoff command causes the editor to exit without saving the current file and causes the Windows logon session to terminate.

lizard

The lizard command programmatically invokes the following command:  cmd /C "dir /B <directory-name>".  The resulting lines are internally collected and separated.  Each line names a directory or a file that appears superficially in the named directory.  System.IO.File.Exists() is called for each file (directories having been filtered out) so as to seek those files that dir enumerates but that System.IO.File.Exists() claims doesn’t exist.  You must name a directory as the first and only argument to the lizard command.

locate

The locate command is similar to the find command except that there is no columnar bias in the pattern. Searching begins with the first line after the current line and is not case sensitive.

ls

The ls command lists the directories and file in the current working directory.

mdl

The mdl command maps a drive letter to a share (i.e. a shared directory (also known as a shared folder)).  The argument is of the form <drive-letter> <share>.  Example:  mdl B \\fserv\wpdocsb

next

The next command relocates the current position to the next line in the file.

o

The o command is similar to the print command except that printing starts with the line following the current line.

omit

The omit command causes the editor to exit without saving the memory file to disk.

open

The open command begin a new editing session by loading the named file's lines into memory. The named file must pre-exist.

opennew

The opennew command is similar to the open command except that the file need not pre-exist and, if it does, the content of the file is truncated.

own

The own command reports the owner of the specified directory or file.

permutation

The permutation command replaces the contents of the in-memory text file with all permutations of the string provided as an argument to the command.  Thanks to Alexander Bogomolyn or Alexander Bogomolny for the permutation algorithm.

ping

The ping command is like the command line ping command.  The syntax is:

·         ping <IPv4 address>

·         ping <symbolic-reference>

The idea is to get a rise out of the target computer.  The command indicates any of the following:  1) response received, 2) response not received, 3) unable to resolve symbolic name.  When you give a symbolic name and the name is resolvable, then the corresponding IP address is shown.

 

play

The play command uses your sound card to play an audio file that you name.  This employs Windows’s PlaySound Win32 interface via INTEROP.

pmask

The pmask command displays the processor (CPU) mask for up to 32 CPUs.

print

The print command prints the current line and the next n minus one lines that follow it, where n is the number stated in the command.  Example:

PRINT 5

means print the current line and the 4 lines that follow it.

ps

The ps command reports the names of existing processes.  Any of these could be terminated using the kill command.

putquotesaround

The putquotesaround command revises the lines of the current file so that each line is quoted (surrounded by double-quotes) and preceded by the characters TAG followed by a space.  Should a line contain a double-quote character, putting quotes around a quote character would be ambiguous.   You can provide a substitute for such characters as in PQA `.  In this example, the ` character replaces all occurrences of ".  The default substitution is \".

pwoff

The pwoff command causes the editor to exit without saving the current file and causes Windows to power off the computer.

randgen

The randgen command replaces the in-memory text file with a number of random text lines.  For example, the command  randgen 999? clears the text file and enters 999 lines of random text.

rb

The rb command causes the editor to exit without saving the current file and causes Windows to reboot (also known as "restart") the computer.

replicating

The replicating command is to be used only on Domain Controllers.  It reports the last time various roles were successfully transferred among domain controllers in a domain.  When replication is not happening, there exists a task for the System Administrator to attend to.

retype

The retype command is similar to the insert command except that the line affected is the current line.  The option of the retype command, when present, specifies the column at which the first character of the argument is to be placed.  For example,

r,11 LMJ

leaves alone the content of columns 1 through 10 and puts the argument into the line starting at column 11.

reverse

The reverse command reverses the order of the lines in the file. For example, the file 1 2 3 becomes 3 2 1.  Apply the command twice and 3 2 1 becomes 1 2 3.

ro

The ro command takes a directory argument and marks as read-only all the files in the directory (and its directories and their directories). The opposite operation is handled by the uro command.

runsql

The runsql command demonstrates how Microsoft SQL Server can be used as a method sink engine from the text editor.  This is only a demonstration.

Data Source=(local);User ID=sa;Password=123456

DROP DATABASE MWEDatabase;

NON-QUERY result is -1

CREATE DATABASE MWEDatabase;

NON-QUERY result is -1

USE MWEDatabase;

NON-QUERY result is -1

CREATE TABLE Radios(ID_NO CHAR(12) PRIMARY KEY, BRAND CHAR(12),);

NON-QUERY result is -1

INSERT INTO Radios (ID_NO, BRAND) VALUES ('WHT1521','LG');

NON-QUERY result is 1

SELECT * FROM Radios;

The number of records affected is -1.

Number of columns is 2

(WHT1521  ?) (LG? )

INSERT INTO Radios VALUES ('WHT1522','RCA');

NON-QUERY result is 1

SELECT * FROM Radios;

The number of records affected is -1.

Number of columns is 2

(WHT1521  ?) (LG? )

Number of columns is 2

(WHT1522  ?) (RCA )

UPDATE Radios SET BRAND='GE' WHERE ID_NO='WHT1521';

NON-QUERY result is 1

SELECT * FROM Radios;

The number of records affected is -1.

Number of columns is 2

(WHT1521  ?) (GE? )

Number of columns is 2

(WHT1522  ?) (RCA )

ALTER TABLE Radios DROP COLUMN BRAND;

NON-QUERY result is -1

SELECT * FROM Radios;

The number of records affected is -1.

Number of columns is 1

(WHT1521  ?)

Number of columns is 1

(WHT1522  ?)

ALTER TABLE Radios ADD? MAKER CHAR(12);

NON-QUERY result is -1

UPDATE Radios SET MAKER='GE' WHERE ID_NO='WHT1521';

NON-QUERY result is 1

UPDATE Radios SET MAKER='RCA' WHERE ID_NO='WHT1522';

NON-QUERY result is 1

SELECT * FROM Radios;

The number of records affected is -1.

Number of columns is 2

(WHT1521  ?) (GE? )

Number of columns is 2

(WHT1522  ?) (RCA )

SELECT ID_NO FROM Radios;

The number of records affected is -1.

Number of columns is 1

(WHT1521  ?)

Number of columns is 1

(WHT1522  ?)

SELECT MAKER FROM Radios;

The number of records affected is -1.

Number of columns is 1

(GE?  ?)

Number of columns is 1

(RCA  ?)

CREATE TABLE Orders(OrderID INT PRIMARY KEY, CustomerID INT, QtyPurchased INT, ItemPrice SMALLMONEY);

NON-QUERY result is -1

INSERT INTO Orders VALUES ('1','1','4','2.5000');

NON-QUERY result is 1

INSERT INTO Orders VALUES ('2','2','10','1.2500');

NON-QUERY result is 1

INSERT INTO Orders VALUES ('3','3','12','1.5000');

NON-QUERY result is 1

INSERT INTO Orders VALUES ('4','4','5','4.0000');

NON-QUERY result is 1

SELECT * FROM Orders;

The number of records affected is -1.

Number of columns is 4

(1) (1) (4) (2.5000)

Number of columns is 4

(2) (2) (10) (1.2500)

Number of columns is 4

(3) (3) (12) (1.5000)

Number of columns is 4

(4) (4) (5) (4.0000)

CREATE TABLE Customers (CustomerID INT PRIMARY KEY, FirstName CHAR(12), LastName CHAR(12));

NON-QUERY result is -1

INSERT INTO Customers VALUES ('1','William','Smith');

NON-QUERY result is 1

INSERT INTO Customers VALUES ('2','Natalie','Lopez');

NON-QUERY result is 1

INSERT INTO Customers VALUES ('3','Brenda','Harper');

NON-QUERY result is 1

INSERT INTO Customers VALUES ('4','Adam','Petrie');

NON-QUERY result is 1

SELECT * FROM Customers;

The number of records affected is -1.

Number of columns is 3

(1) (William  ?) (Smith )

Number of columns is 3

(2) (Natalie  ?) (Lopez )

Number of columns is 3

(3) (Brenda?  ?) (Harper? )

Number of columns is 3

(4) (Adam?  ?) (Petrie? )

SELECT * FROM Customers INNER JOIN Orders ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 7

(1) (William  ?) (Smith ) (1) (1) (4) (2.5000)

Number of columns is 7

(2) (Natalie  ?) (Lopez ) (2) (2) (10) (1.2500)

Number of columns is 7

(3) (Brenda?  ?) (Harper? ) (3) (3) (12) (1.5000)

Number of columns is 7

(4) (Adam?  ?) (Petrie? ) (4) (4) (5) (4.0000)

SELECT * FROM Orders INNER JOIN Customers ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 7

(1) (1) (4) (2.5000) (1) (William  ?) (Smith )

Number of columns is 7

(2) (2) (10) (1.2500) (2) (Natalie  ?) (Lopez )

Number of columns is 7

(3) (3) (12) (1.5000) (3) (Brenda?  ?) (Harper? )

Number of columns is 7

(4) (4) (5) (4.0000) (4) (Adam?  ?) (Petrie? )

SELECT *, Orders.QtyPurchased * Orders.ItemPrice AS Line_Item FROM Customers INNER JOIN Orders ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 8

(1) (William  ?) (Smith ) (1) (1) (4) (2.5000) (10.0000)

Number of columns is 8

(2) (Natalie  ?) (Lopez ) (2) (2) (10) (1.2500) (12.5000)

Number of columns is 8

(3) (Brenda?  ?) (Harper? ) (3) (3) (12) (1.5000) (18.0000)

Number of columns is 8

(4) (Adam?  ?) (Petrie? ) (4) (4) (5) (4.0000) (20.0000)

SELECT * FROM Customers LEFT JOIN Orders ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 7

(1) (William  ?) (Smith ) (1) (1) (4) (2.5000)

Number of columns is 7

(2) (Natalie  ?) (Lopez ) (2) (2) (10) (1.2500)

Number of columns is 7

(3) (Brenda?  ?) (Harper? ) (3) (3) (12) (1.5000)

Number of columns is 7

(4) (Adam?  ?) (Petrie? ) (4) (4) (5) (4.0000)

SELECT * FROM Customers RIGHT JOIN Orders ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 7

(1) (William  ?) (Smith ) (1) (1) (4) (2.5000)

Number of columns is 7

(2) (Natalie  ?) (Lopez ) (2) (2) (10) (1.2500)

Number of columns is 7

(3) (Brenda?  ?) (Harper? ) (3) (3) (12) (1.5000)

Number of columns is 7

(4) (Adam?  ?) (Petrie? ) (4) (4) (5) (4.0000)

SELECT * FROM Customers FULL JOIN Orders ON Customers.CustomerID = Orders.CustomerID;

The number of records affected is -1.

Number of columns is 7

(1) (William  ?) (Smith ) (1) (1) (4) (2.5000)

Number of columns is 7

(2) (Natalie  ?) (Lopez ) (2) (2) (10) (1.2500)

Number of columns is 7

(3) (Brenda?  ?) (Harper? ) (3) (3) (12) (1.5000)

Number of columns is 7

(4) (Adam?  ?) (Petrie? ) (4) (4) (5) (4.0000)

scale

The scale command prints a column scale extending from column 1 through column 80.

select

The select command performs a case-insensitive search of all the lines in the main memory file and removes all the lines that do not match the target value.

 

sizes

The sizes command takes a single argument specification of a directory (also known as a folder).  The command reports the number of bytes occupied by the specified directory and all of its subordinate directories. The editor initially erases the text file and then reports the byte sizes of the directly subordinate directories, if any.  Similar information is written to a comma-separated value file named C:\Users\Public\SIZES.CSV.  The size of any directory comprises the sum of the sizes of all the files in the directory and all the files in all the subordinate directories direct and indirect.

 

 

sort

The sort command reorders the lines in the memory file so that they appear alphabetically.

superficial

The superficial command works with a directory that contains several directories that individually contain files (and perhaps no directories).  Each superficial directory is listed together with the count of ordinary files that it contains.  The report replaces the in-memory text file.

tab

The tab command allows you to establish a tab character and several column stops. This is handy when entering columnar data. For example, if you enter the command

tab ; 11 21 39

then enter the command

i ;LMJ;X11,EDIT$;. Open the EDIT$ packet

you will create a new line with the various segments left-aligned at the indicated column stops.

td

The td command displays the date and time last written of the file being edited.

tld

The tld command creates a shell script for archiving top level directories.  It takes one argument that names the directory to receive a series of ZIP files produced by PKZIPC.  The data source is the top level directories of the C:\ drive. Each directory immediately under root is represented by a ZIP file.  You can individually reconstitute any of the top level directories.  It is recommended that there be no overlap between the receiving directory and the set of directories to be archived. By default the top level directories of the C:\ drive are obtained.  You can specify a different drive letter in the command’s option field.  Syntax:  tld[,<drive-letter>] [<directory-to-receive-ZIP-files>]? When you do not specify a receiving directory, the default is C:\Users\Public\Receptacle.

toc

The toc command erases the current file in memory and creates a new file in memory that enumerates the directories and files in the named directory.

tokenize

The tokenize command enters words into the in-memory text file, one word per line.  The in-memory text file is the source of the words.  Breaks between words are signified by characters that are not letters of the alphabet.  Diacritical marks on letters cause breaks in the initial release of this feature. The one word per line list replaces the text being edited. You can preserve the input by writing the list.  Use the W command.

top

The top command sets the current line to zero.

touch

The touch command updates the time-of-last-write of the argument file to match the current time. Should the file not exist, it is created with zero content. When the file has content, it remains unchanged after touch does what it does.

undo

The undo command, when possible, reverses the effects of the most recent change command(s) and/or the most recent retype command(s), as is applicable.  For example, you might mistakenly make a global change that you want to reverse.  Generally, you can undo everything since the last time you visited the top of the file.  Use the C,S command to see whether undo is possible.

unique

The unique command removes duplicate lines.  The first line in a duplicated set of lines remains in the file.  The rest of the lines are deleted.  Only unique lines remain.

up

The up command sets the current line to the line immediately above (numbered one less than) the current line.

uro

The uro command removes the read-only attribute from all the files in the argument directory (and the directories it contains and the directories they contain).

vw

The vw command sets a vertical window in which appears a subset of lines of the file being edited. If an n-line subset of the file is in the vertical window then the lines are numbered 1 to n and changes that you may apply affect only the lines in the window. You set the vertical window by stating the starting and ending line numbers as in vw 2 4 which saves the entire file to a hidden location and commences editing of the subset. You merge the subset back into the main file by entering vw without arguments.

winplaces

The winplaces command displays the folderpaths of all the system special folders. When a text file is open it is replaced with a report similar to this:

 

wpt

The wpt command waits for a process to terminate.  Just give the name of the process and wpt loops until the named process disappears from the system. This is handy when you are replacing one version of a Windows NT Service with another.  You can uninstall the service and the system reports the service has been stopped and uninstalled.  But that’s no guarantee that the process has actually halted.  Enter the wpt command that happily delays execution of the installation step until the process is gone.  When you don’t force a delay, it is quite probable the installer will complain that the uninstalled/stopped process is still hanging around and therefore your attempt to install will have terminated in error.

write

The write command transfers the lines in memory to the named disk file and terminates the editor.

 

 

 

BigDigitClock displays the current time-of-day.  This seems trivial, but at a radio station, with three analog clocks, that are supposed to be synchronized with a GPS time-source, but aren’t, this app fills a void.  Together with the KeepClockUpdated daemon, BigDigitClock let’s announcers know when to begin speaking at the start of a live broadcast. [BigDigitClock-BASIC]

 

 

 

 

 

File Integrity Check (FIT)

 

File Integrity Check compares corresponding files in two different directories (aka Folders).  The dates, lengths, and contents are compared.  A report is made.  You can detect hardware faults with FIT.  You can detect security intrusions with FIT.

 

 

Cross-Platform File Compressor (CPFC)

 

CPFC is designed to compress files on a Unisys (Sperry) ClearPath mainframe and decompress the files on a PC.  Travel in the opposite direction is also supported.  The sliding dictionary algorithm exploits redundancies, typical in symbolic data (e.g. ordinary text files), to minimize the number of bytes that need to be transmitted.  A buffer of a certain size is maintained.  Bytes are entered cyclically in order from the source file(s) and, initially, there is no compression.  Eventually a redundancy appears. Instead of literally transmitting uncompressed bytes, a reference to the repeated pattern (a redundancy) is sent.  Typical reduction in file size hovers around 50%.  The mainframe’s 36-bit word, when compared to the PC’s 32-bit word, means special consideration must be given to guarantee all bits are sent.

 

 

Infinite Precision Arithmetic

Sometimes a computer language’s data types are not up to the task.  Suppose you want to compute a very large integer exactly.  For example, what is 2 to the 1000th power Enter MPA4 (Multiple-Precision Arithmetic Version 4).  MPA4 carries out integer addition, subtraction, multiplication, and division, correct to the very last decimal place.  It computes integer powers of integers.  It factors integers.  It performs Diffie-Hellman and Mersenne prime computations.  [MPA4]

 

By the way, ?is 10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196477686542167660429831652624386837205668069376

 

 

Suppose your insurance company will dispense a 90-day supply of medication.  It takes time for a pharmacy to accept a prescription and deliver the tablets.  If you wait 90 days between doctor visits, you may have to live without your meds for a few days.  In the case of serious medical conditions, that could prove to be inconvenient.  Enter Days Between, a utility program in which you enter two dates, to reveal the number of days it will take to get you from the first date to the last date. When your doctor wants you to come back, make it fewer than 90 days from now.  In this example:  84 days from now.  You get your script and time to have it filled.  No running out of meds. [DaysBetween]

 

 

 

 

 

 

File Logger

 

Windows exposes a method whereby an application is notified in real-time when files are created, deleted, modified, and renamed.  File Logger displays these notifications in real-time.  Applications can respond appropriately beyond simply displaying the notifications.  Refer to FileSystemWatcher in Microsoft? Windows? documentation.

 

 

 

 

Sometimes you need to upload a large file.  There may be a limit on the number bytes per transmission.  Enter Split/Combine, a utility program that breaks a large file into several contiguous pieces, and later recombines them. [SpilitCombine]

 

 

MERGE finds corresponding files in two directory trees and merges them into a third directory.

Some files may appear in one directory but not the other.  These are sent to the output directory.  Those files that appear in both directories are compared to learn which one of the two is the more recent.  The more recent one is sent to the output directory. When files have identical timestamps, the longer file is copied. 

MERGE is handy when you maintain several separate but related collections of files, organized into directories and sub-directories.  It sometimes becomes convenient to combine collections through de-duplication.  For example, two people can be working on the same application.  Each has his own folder of source code.  When work is done the two efforts can be merged with Merge.

The Differences Only checkbox causes the (C) file to receive only those files that are different across the (A) and (B) files.  A file is different if 1) it appears in one directory but not in the other, or 2) it appears in both directories, but the timestamps are unequal, or when the timestamps are equal, but the file lengths are unequal.  When the file lengths are unequal and the timestamps are equal then the longer file appears in the output.  When the timestamps are unequal then the more recently modified file appears in the output.  The directory structure that is common between (A) and (B) is maintained in (C). [Merge]

 

 

 

 

CopyMaster

 

CopyMaster is an app that allows you to duplicate the latest directory structures and files so they appear on two different computers at the same time.  This redundancy has two advantages.  First, there is a current backup copy of your important files.  Second, there is convenience in being able to find all your favorite stuff on at least two different computers. [CopyMaster]

 

How You Work It

Map the A: drive to \\ComputerOne\C$ and the B: drive to \\ComputerTwo\C$.  Then launch CopyMaster.  When CopyMaster finishes doing what it does, you?ll have two computers that sport nearly identical contents (ignoring the operating system and extra software files). 

You can create a $.txt file in the C:\Users\Public\$$$DeleteMe directory that instructs CopyMaster to get rid of unwanted directories on both computers.


 

Deltree

 

Suppose you have a directory that contains a florid tree of sub-directories along with files at many levels and you want to be rid of the whole mess.

 

Windows has a rule that says you may not delete a read-only file.  This prevents you from deleting a directory that contains a read-only file.

 

Deltree: to the rescue.  Just point Deltree at the root of the mess and It’s gone in three shakes of a lamb’s tail.

 

Now try doing that with the del or erase command in the ordinary Windows shell.  Oh.  It will do something. But it won’t do what you want. [deltree]

 

 

 

Copydir

 

Sometimes you want to copy a directory tree, complete with contents from directory A to directory B.  This is, by Microsoft edict, a task for  File Explorer,  formerly called  Windows Explorer.  But what if you want to invoke the function from the command line? Enter Copydir.  The syntax is:

 

copydir A:\A B:\B

 

Everything subordinate to the A directory is copied subordinate to the B directory.  Identically-named files are copied when the file from directory A is newer than or absent from directory B.  Be aware that the files and directories inside directory A are created (when necessary) and copied as files and directories inside directory B.  The B directory will be created when it does not yet exist.  In other news, proof of function can be had by using the Ident command.  The structure of A must be identical to B or else an error message is issued.  The content of files (data NOT metadata) is physically compared.  [Copydir]

 

IDENT

 

 

Suppose, for example, you have a USB device (an external hard drive, thumb drive) that you copied to a directory on your computer’s internal hard drive.  You want to verify that everything that you copied is present and unmodified.  IDENT is a utility program to perform the function.  Just launch the application from the command prompt and name the directory on the USB device and the corresponding directory on the PC’s hard drive. The contents of each file in every directory in the directory tree shall be compared and any irregularity reported.  The two directories must be absolutely identical.  No check on various dates and times in the metadata is performed.

 

Usage:  ident <directory1> <directory2>

? The two directories must contain only those directories and files to be compared.

? The contents of corresponding files are checked to determine that they match.

 

 


WavConcatX

 

WAVConcatX, takes a collection of WAV files and concatenates them.  It optionally inserts an audio separator between the files.  The output is a single WAV (audio) file made from the contents of the WAV files in the aforementioned collection.

 

The individual WAV files can exhibit varying attributes.  For example, there is the sampling rate, one channel or two channels, and sample size (8-bit, 16-bit).  You choose the sampling rate and the number of channels in the output file.  WCX takes care of the necessary conversions?.  Sixteen-bit samples are standard in the WCX output file.  Insertion of transitional separators is possible only when the output sampling rate is 44,100 samples per second, and there are two channels. [WCX]

 

?The sample rate conversions are naively done without filtering.  Properly done, a low-pass digital filter is applied to remove frequencies above half the sampling rate to eliminate aliasing, which is audible in sampling rate conversions that are produced by WCX.  If you really want to change the sample rate, try SoX (Sound Exchange), open-source software on SourceForge.  For example, to change from 44100 samples per second (or whatever the metadata says it is) to 6000 samples per second, follow this example:

 

·         sox  r6000 input.wav output.wav

 


 

 

 

 

TFC_File_Inspector

 

 

TFC_File_Inspector came about after a store employee destroyed a file system on an Apple Macintosh computer.  The remnants of the file system comprised mainly file content without names or dates.  So the task for the human is to give names to files that now sport generic names.  Generic names are sequential numbers.  These names have zero mnemonic value. The aforementioned application allows the end user to cycle through a set of generically-named files, view their contents, and give them meaningful names.  One avoids the laborious process of launching each file and renaming in the conventional manner.  Ancillary programs named FILE_MATCHING_PROGRAM, FMP2, AND FMP3 assist in the effort to characterize the situation that TFC_File_Inspector is meant to correct.  First, hashcodes are developed from the content of each file on the recovery device.  Then, hashcodes are developed from the content of each file on an Apple MacIntosh that contains an old backup of many of the lost files.  We discover that the recovery device contains many duplicated files.  Since our task is to give names to the files on the recovery device, we can remove the duplicated files and reduce our effort.  Then the hashcodes for files on the recovery device are compared with hashcodes for files on the Mac.  With this step we identify further files that already have names and that are situated in a reasonable directory.  These need not be renamed.  Our effort is thus further reduced.

 

1.  fmp takes a directory and generates a flat file of hashcodes (full tree).

2.  fmp2 takes two flat files from fmp called P and G, searches P for identical files in G, and reports them as CORRESPONDING.

3.  fmp3 takes CORRESPONDING.TXT and deletes G files that exist in P.

 

It was discovered that certain older versions of Microsoft Office products (e.g. Word and Excel) produce what are called Compound Documents.  These objects (files, actually) were implemented as part of the Component Object Model (COM) effort.  A Compound Document is what is called a Storage Object.  A Storage Object may contain metadata that includes such juicy tidbits as Title, Author, and Last Save Date. The TFC_File_Inspector effort includes implementation of a program that extracts these metadata and changes generic names to proper names using tidbits.  This is great when all you got is file content after your file system gets corrupted.  The Compound Document was abandoned some years ago in favor of coding the same document data in very, very complex XML.  The XML does not offer the same rich lode of metadata.

 

 

 

 

IBMCHECK implements the Luhn Algorithm.  It allows you to enter a credit card number and be informed whether it is valid.  You can enter the number as it appears on your credit card, with spaces every four digits, press a button, and the spaces disappear.  The number is copied to the Windows Clipboard for pasting into a text box on some retailer’s website.  Retailer websites typically get upset when you leave the spaces in place. [IBMCHECKCS]

 

 

Corrector of Directory Dates

 

 

CORR is the Directory Date Corrector.  Ideally, Windows would ensure the timestamp applied to every directory would reflect the contents of each directory.  The file inside a directory with the latest last-date/time-of-modification would be honored to have its timestamp applied to its enclosing directory or directories.  So that is what CORR does.  Which directories (now called Folders by marketing) get what timestamps is decided by tournament rules.  CORR makes it possible to inspect the folder contents of a folder and decide which folders need further inspection because of recent updates to the files inside.  Notice that, in the latest release, you can watch counters advance whilst files and then directories are processed.  [CORR]

 

As of October, 2020, there is a New! Improved! command line version of the same application called CORR2.  It has a better algorithm that makes sure no directory is omitted from processing.  Here is the syntax:

 

     corr2 <name-of-directory>

 

You can run the application against any directory, not just the root directory of a drive letter.

 

 

 


 

Keep Clock Updated

 

 

KeepClockUpdated is a service that ensures the system clock stays accurate, with checks done for accuracy every 5 minutes.  This entails a time source query every 5 minutes.  In one commercial deployment, at a radio station, KeepClockUpdated is responsible for setting the clock on 24 different computers by comparing each computer's clock with a local time source; one that obtains its very accurate reading from GPS satellites. [This has changed because the GPS time source permanently malfunctioned.  In place of the GPS time source the service queries an Internet time source that obtains its responses based on several other Internet time sources.] (The native Windows time setting mechanism only checks the computer maybe once per day.  Its time source is a computer on the Internet.)? All the magic depends on an understanding of the Network Time Protocol (NTP) that communicates via UDP on port 123.  You can set the calendar and time-of-day to an inaccurate value and wait until some multiple of five minutes past the hour arrives on the clock.  From the time source, KeepClockUpdated obtains the date and time and compares them with the computer's own notion of date and time.  When the two are at variance, KeepClockUpdated changes the computer's clock and calendar to match.  [KeepClockUpdated]

 

Pip is an application that plays a chime on the hour and a tick on the half-hour to connote the passage of time.  In conjunction with KeepClockUpdated, one can be certain the auditory signals occur at precisely the right moment.  [PIP]

 

 

WINSVCFILCPY is service app that runs 24/7.  Once a day, at 0300, it copies key files from their usual staging area to a backup staging area. This is done in order to mitigate the risk that the usual staging area, a Microsoft software-based mirrored pair of HDDs, may become useless as a Microsoft software-based RAID-5 cluster did in September 2013. You can change the appointed time at which file copy commences by means of a named pipes client. [WINSVCFILCPY]

 

 


 

FileLogger monitors file changes.  It displays them as they occur in real-time.  It keeps a running log of changes and can copy the log to a file if you choose to do so. [FileLogger]

 

 

 


 

Password Generator generates random passwords.  They are, by default, a mixture of numerals, upper- and lower-case letters.  You specify the length you want up to 40 characters.  You can choose to have only upper-case or only lower-case letters.  A non-alphanumeric character can be randomly inserted if you choose to do so.  The generated password is put on the Windows Clipboard for ease of placement in a textbox on a web page.  These passwords are hard for hackers to guess and that makes them extraordinarily secure.  The PRODKEY button causes Password Generator to create a Microsoft-style Product Key.  [CSPASSGEN]

 

?


 

File Mover Service (FMS)

FMS is responsible for copying TRV files in folders P1 through P9 from the production computer to the backup computer.  This is in progress all the time, with throttling imposed to limit impact on an Interactive Voice Response application, except during the graveyard shift.  However, see note, below, regarding limitations imposed by an Interactive Voice Response application.  A given file is copied under the condition that the source file is newer than the destination file or when there is no destination file.  For example, suppose \\DIALIN1\P1\1547.TRV is to be copied to \\DIALIN2\P1\1547.TRV .  It follows that copying will occur only when \\DIALIN2\P1\1547.TRV does not already exist and when \\DIALIN2\P1\1547.TRV? time-of-last-write indicates that it is older than \\DIALIN1\P1\1547.TRV .

You can change the direction of copying with the application NamedPipesClient.  Simply connect, then type from-1-to-2 or from-2-to-1 in the Send Message area of the dialog box and press the Send button.  When you switch the phone lines from DIALIN1 to DIALIN2 (or the reverse), you should remember to change the direction of copying.  NamedPipesClient must be run with elevated privileges. [MWE_FMS]

 


 


 

 

 

 

Individual Computer Housekeeping Service (ICHS)

 

ICHS is responsible for performing three functions:

1.  Midnight Movers

2.  Sunday Movers

3.  Richmond Download

Midnight Movers happens at 29 minutes past midnight every night.  It prepares for the upcoming business day by clearing today’s stories out of their today’s slots and moving them to yesterday’s slots.

Sunday Movers happens at 28 minutes past one o?clock on Sunday morning.  It prepares for the upcoming business week by clearing files from their daily slots and putting them into a weekly backup location.

Richmond Download implements the File Transfer Protocol so as to acquire copies of files 1501.TRV through 1505.TRV, which are voice files from the Richmond Times-Dispatch.  This download occurs daily at five minutes past noon. [MWE_ICHS]


 

MIDNIGHT MOVERS

 

 

WASHINGTON POST

 

Original P1                Daily Copy To   New P5

1000.TRV-1099.TRV                          1000.TRV-1099.TRV

1100.TRV-1199.TRV                          1100.TRV-1199.TRV

1300.TRV-1399.TRV                          1300.TRV-1399.TRV

1500.TRV-1599.TRV                          1500.TRV-1599.TRV

1700.TRV-1799.TRV                          1700.TRV-1799.TRV

1800.TRV-1899.TRV                          ?1800.TRV-1899.TRV

2000.TRV-2099.TRV          ?               2000.TRV-2099.TRV

2100.TRV-2110.TRV                          2100.TRV-2110.TRV

3600.TRV-3620.TRV                          3600.TRV-3620.TRV

3700.TRV-3720.TRV                          3700.TRV-3720.TRV

3900.TRV-3999.TRV                          3900.TRV-3999.TRV

4000.TRV-4099.TRV                          4000.TRV-4099.TRV

4100.TRV-4199.TRV                          4100.TRV-4199.TRV

4500.TRV-4599.TRV                          4500.TRV-4599.TRV

4700.TRV-4799.TRV                          4700.TRV-4799.TRV    

 

 

USA TODAY

 

Original P6                Daily Copy To   New P6

1000.TRV-1099.TRV                          8400.TRV-8499.TRV

1100.TRV-1199.TRV                          8500.TRV-8599.TRV

3900.TRV-3999.TRV                          8600.TRV-8699.TRV

4000.TRV-4099.TRV                          8700.TRV-8799.TRV

4100.TRV-4199.TRV                          8800.TRV-8899.TRV

1300.TRV-1399.TRV                          8900.TRV-8999.TRV

1600.TRV-1699.TRV                          9000.TRV-9099.TRV

1900.TRV-1999.TRV                          9100.TRV-9199.TRV

1800.TRV-1899.TRV                          9200.TRV-9299.TRV

4500.TRV-4520.TRV                          ?9300.TRV-9320.TRV

4600.TRV-4620.TRV                          9400.TRV-9420.TRV

4700.TRV-4720.TRV                          9500.TRV-9520.TRV

5100.TRV-5120.TRV                          9600.TRV-9620.TRV

5200.TRV-5220.TRV                          9700.TRV-9720.TRV

5300.TRV-5320.TRV                          9800.TRV-9820.TRV


 

SUNDAY MOVERS

 

Per the script that formerly was launched on Sundays:

 

del from P1 4801-4899",

del from P1 9601-9699",

del from P1 9701-9799",

copy p1 2901-2999 to p1 4801-4899",sb

copy p1 2501-2599 to p1 9601-9699",sb

copy p1 2401-2499 to p1 9701-9799",sb

move p1 2901-2999 to p1 4801-4899",

move p1 2501-2599 to p1 9601-9699",

move p1 2401-2499 to p1 9701-9799",

 

The indicated files are deleted from the P1 directory, copied from the P1 directory, or relocated away from the P1 directory, in the indicated order.  The sb notation indicates that copying is directed to the SundayBackup directory.


SNMP

This picture shows how it is possible to query the Windows 2012 Operating System for its computer name (cleverly labeled sysName) using the Simple Network Management Protocol.  The bytes sent and received under the protocol are anything but simple. But, never mind.  This program untangles the bytes into a series of nested subordinate messages and picks out the system name which it displays.  The request for sysName is encoded by something called an OID.  That’s the 1.3.6.1.2.1.1.5.0. that you see in the picture.  Object Identifier?  The following is referenced without permission from http://kb.paessler.com/en/topic/653-how-do-snmp-mibs-and-oids-work.

MIB stands for Management Information Base and is a collection of information organized hierarchically. These are accessed using a protocol such as SNMP. There are two types of MIBs: scalar and tabular. Scalar objects define a single object instance whereas tabular objects define multiple related object instances grouped in MIB tables.

OIDs or Object Identifiers uniquely identify managed objects in a MIB hierarchy. This can be depicted as a tree, the levels of which are assigned by different organizations. Top level MIB object IDs (OIDs) belong to different standard organizations. Vendors define private branches including managed objects for their own products.

SNMP version 1 was the initial development of the SNMP protocol. A description can be found in Request for Comments (RFC) 1157 and it functions within the specification of the Structure of Management Information (SMI). It operates over User Datagram Protocol (UDP), Internet Protocol (IP), OSI Connectionless Network Services (CLNS), AppleTalk Datagram Delivery Protocol (DDP), and Novell Internet Packet Exchange (IPX). SNMP v1 is considered the de facto network management protocol in the Internet community.

SNMP works on the basis that network management systems send out a request and the managed devices return a response. This is implemented using one of four operations: Get, GetNext, Set, and Trap. SNMP messages consist of a header and a PDU (protocol data units). The headers consist of the SNMP version number and the community name. The community name is used as a form of security in SNMP. The PDU depends on the type of message that is being sent. The Get, GetNext, and Set, as well as the response PDU, consist of PDU type, Request ID, Error status, Error index and Object/variable fields. The Trap consists of Enterprise, Agent, Agent address, Generic trap type, Specific trap code, Timestamp and Object/Value fields.

MIBs are a collection of definitions which define the properties of the managed object within the device to be managed (such as a router, switch, etc.) Each managed device keeps a database of values for each of the definitions written in the MIB. As such, it is not actually database but implementation dependent. Each vendor of SNMP equipment has an exclusive section of the MIB tree structure under their control.

In order for all of this to be properly organized, all of the manageable features of all products (from each vendor) are arranged in this tree. Each 'branch' of this tree has a number and a name, and the complete path from the top of the tree down to the point of interest forms the name of that point. This is the OID. Nodes near the top of the tree are extremely general I nature. For example, to get to the Internet, one has to reach to the fourth tier. As one moves further down, the names get more and more specific, until one gets to the bottom, where each node represents a particular feature on a specific device (or agent).

 

 

 

Simple Backup/Restore

 

Simple Backup/Restore makes a backup file from a collection of directories.  You choose which directories to backup by specifying a root directory.  The default root directory (when you do not specify one) is the root of Drive Letter.  The backup file sports the filename extension .SB.  Each backed up file is internally associated with its name and its time of last modification, its length, and unless it is old-format, a CRC.  Data integrity is assured by a Cyclic Redundancy Check, which is applied on a per file basis.  Simple Backup/Restore was developed in response an incident where a commercial file backup app failed when the computer, which it was backing up, failed catastrophically. The partially readable backup file had to be recovered through the purchase of additional commercially available software that was capable of sifting through a complex, partially corrupted database.  The central concept of System Backup is simplicity.  A partially completed backup file remains readable.  All the work completed up to the point of failure, if any, is fully recoverable without resort to extraordinary means.  You have the option to encrypt the backup file.  Restoration of an encrypted is automatic.  You do not have to know whether the backup file is encrypted or not. [SystemBackup]

 

 

 


 

The Metropolitan Washington Ear, a radio reading service for the blind, offers audio streaming via the Internet.  They moved to their own, dedicated building and infrastructure in 2008.  Their website was not appropriately updated to cater to different audio streaming software.  MARK DANA ROCKMAN developed this page to allow listeners to hear the audio stream.  That formerly was impossible. [MWE_WEB_CLIENT_FOR_AUDIO]

 


 


 

Nutrients to Foods (NTF)

 

Anybody can look up the nutrients contained within foodstuffs.  But what if you know what nutrients you want What foods should you eat to get those nutrients The answer is provided by NTF. [NTF]

 


 

SHUTDOWN

There are two version of SHUTDOWN.  Both of them work.  The second one fully exercises the Windows API for logging off and for rebooting and powering off the computer.  [SHUTDOWNCS]

 

 

 

This app is much handier than pointing and clicking around Windows’ own user interface for doing what SHUTDOWN does.  In some contexts there is no obvious way to cause Windows to reboot.  The 2022 revision includes a combobox.  With the combobox you can choose the hour at which you want the application to perform the function.  When you do not specify an hour then the operation happens immediately. [SHUTDOWNCS]

 

FileDump

 

Sometimes you just want to see exactly what a file contains.  [FILEDUMP]

 

MWE-DIAL-IN-CRASH-MONITOR

?

 

A computer that runs 24/7 is responsible for an IVR application.  The application maintains a large set of files inside a directory structure.  There are two kinds of file:  audio and database.  Both kinds are possessed by the IVR application in that they are to be instantaneously available to the IVR application.  Interference from applications outside the IVR application is dealt with severely:  the IVR application reboots the computer.  It deals with every anomaly this way.  There is no log entry documenting the fact that the IVR application has decided to reboot the computer.  We just watch, in horror, as the machine restarts.  It would be convenient for the manager in charge of the IVR application to become cognizant of the situation immediately upon its occurrence.  Enter MWE-DIAL-IN-CRASH-MONITOR, which pings the computer and plays a tune in the event of no response.

Numerous reboots over the years have been described as being due to interference from applications that are outside the purview of the IVR application.  Somebody  locked? the file.  That is to say, some application has opened the file so it can be backed up for good and proper administrative reasons. Other potential problems revolve around file metadata such as who owns the file, whether the file is compressed, whether the file is read-only, whether the file has other strange attributes like SYSTEM and HIDDEN.  Folklore tells us the IVR application is quite strict as to what attributes a file may be blessed with.  These, we learn, are N (the attribute a file gets when it has no other attributes) and A (the attribute a file gets when it is ripe for being backed up).  All others are forbidden.  When the IVR application finds a tainted file, it reboots the computer.

Folklore isn’t necessarily an accurate reflection of reality.  In actual fact, the IVR application probably reboots the computer only in the event of a file access conflict and for reasons that have nothing whatever to do with the file system, its behavior and contents.  [MWE-DIAL-IN-CRASH-MONITOR]

 

 

TRT32 FILE REPAIR aims to change file metadata to conform to the IVR application’s strict attributes rules.  [MWE_TRT32_FILE_REPAIR]

 

UPLOADY

 

This ASP.NET application allows a web client to upload a file to the ISP that hosts www.mdrsesco.biz, the website.  The interesting thing about this application is its inability to spend the time or bandwidth on uploading large files.  Supposedly the problem is mitigated with adjustments to parameters contained within Web.config, a text file Microsoft’s web server, IIS, uses to make decisions about how a round trip is to be handled.  Yeah, those adjustments don’t do the trick.  However, for small files, this  app? is quite satisfactory.  FTP is the workaround for large files.

 

<?xml version="1.0"?>

<configuration>

        <appSettings/>

        <connectionStrings/>

        <system.web>

                <httpRuntime maxRequestLength="409600" executionTimeout="86399"/>

                <sessionState timeout="500100"/>

                <!--

? Set compilation debug="true" to insert debugging

? symbols into the compiled page. Because this

? affects performance, set this value to true only

? during development.

? -->

                <compilation debug="false" targetFramework="4.0">

                </compilation>

                <!--

? The <authentication> section enables configuration

? of the security authentication mode used by

? ASP.NET to identify an incoming user.

? -->

                <authentication mode="Windows"/>

                <!--

? The <customErrors> section enables configuration

? of what to do if/when an unhandled error occurs

? during the execution of a request. Specifically,

? it enables developers to configure html error pages

? to be displayed in place of a error stack trace.

 

? <customErrors mode="RemoteOnly" defaultRedirect="GenericErrorPage.htm">

? <error statusCode="403" redirect="NoAccess.htm" />

? <error statusCode="404" redirect="FileNotFound.htm" />

? </customErrors>

? -->

                <pages controlRenderingCompatibilityVersion="3.5" clientIDMode="AutoID"/></system.web>

        <!--

? The system.webServer section is required for running ASP.NET AJAX under Internet

? Information Services 7.0.  It is not necessary for previous version of IIS.

? -->

</configuration>

 

 

 

DMP

 

In a previous life, MARK DANA ROCKMAN worked on mainframes.  He created DMP for the Unisys (Sperry) mainframe, which is now known as Unisys ClearPath but previously was known as the Univac 1100/2200 Series.  DMP persists and Les Leist, with whom the aforementioned Rockman once worked, has kindly decided to maintain it and distribute it to those who want it.

 

http://cgibin.rcn.com/leistlc/cgi-bin/indexbld.pl

 

DMP 22R2F - Interactive, Multi-Functional Utility Processor .....[06 NOV 2013 (1405 KB) -- Downloads: 878]

DMP is a versatile program with numerous commands relating to system, MCT, and directory information; and utility functions for files and tapes. This version has been upgraded for the latest MCT, MFD, and Audit Trail structures.

*** Although DMP has commands that work with @COPY,G tapes, DMP does not support the newer @COPY,G format with more than one track per block. There is also limited LPF and LEPF support. See DMP/README. ***

 

-- Written by Mark Rockman. Enhanced by Les Leist.

(DOC no longer posted; obtain from download.)

 

How to Kill a Process

Your application craves to terminate another process.  Here is how to do it.

namespace EditTechnology

{

? public class kill

? {

? public static bool WasteHim(String whomToWaste) // true indicates the process was found and kill was called on it

? {

? string whomWeGonnaWaste = whomToWaste;

? System.Diagnostics.Process[] localAll = System.Diagnostics.Process.GetProcesses();

? foreach (System.Diagnostics.Process p in localAll)

? {

? if (p.ProcessName.Equals(whomWeGonnaWaste))

? {

? p.Kill();

?  ?return true;

? }

? }

? return false;? // process by the specified name was NOT found

? }

? }

}

 

Notice that what is going on here is enumeration of all the processes in the system.  One of those may have a name that matches the argument.  It is the one, if such exists, that gets the axe.

 

How to Launch a Process

class Launch

? {

? public bool Process(string nameOfFile)

? {

 

? System.Diagnostics.Process myProcess = new System.Diagnostics.Process();

 

? try

? {

? // Get the path that stores user documents.

? string myProgramFilesPath =

? Environment.GetFolderPath(System.Environment.SpecialFolder.System);

 

? myProcess.StartInfo.FileName = myProgramFilesPath + "\\NOTEPAD.EXE";

? myProcess.StartInfo.Arguments = nameOfFile;

?  ?myProcess.StartInfo.Verb = "Open";

? myProcess.StartInfo.CreateNoWindow = true;

? myProcess.Start();

? }

? catch

? {

? return false;

? }

? return true;

?  ?}

? }

How to Launch a Process and Gather Its STDOUT

 

This shows how to launch a console program (a program that can be found among the directories that are listed in the PATH environment variable of the logged-in user’s profile) and have its output sent to a string variable.  The lines of output are separated by carriage returns and/or line feeds.

 

 

How to Determine If a File is Already Open in Another Process

 

 

 

How to Disable the Dismiss Button

You Know.  The X in the Rectangle in the Upper Right of a Window

 

In a Windows Forms application it is possible to disable the dismiss button.  You must override an event-handling method of the form class and set a flag that tells the powers-that-be that you find it utterly inconvenient to terminate the program at this time.  Probably it would formally be better if you coded a base.OnClosing() method invocation for the case where bWorkerRunning is false.  But it seems that’s not really necessary.  Setting Cancel in eventargs to true does the trick.

 

 

Interoperability:  Hooking Managed Code to Unmanaged Code

Managed code is produced by Visual Studio for consumption at runtime by a Just-In-Time (JIT) compiler.  The JIT Compiler transforms Intermediate Language (IL) to machine code and performs runtime checks for type safeness and validity to ensure proper operation and security.  IL is code written for a machine that hasn’t been realized in hardware but is code that can readily be transformed to machine code.  Code that is written in languages such as C# and Visual Basic becomes IL before it becomes machine code.

Interoperability (INTEROP) is the mechanism that enables the methods in managed code to call the methods in unmanaged code. 

Unmanaged code is produced by Visual Studio for languages like C++.  This is direct-to-machine-code compiling.  The only validity checks are the ones that are performed during compilation.  C++ programs are ready to load and execute.  There is no JIT compiling.  There is a variation of C++ that can be compiled to IL and which uses the Common Language Runtime of the .Net Framework, the same as C#.  Why use unmanaged code It is because a C++ program is the place where you have access to the entirety of the Win32 API.  Many operating system features are virtually inaccessible from managed code. There is no support in CLR, for example, that lets a programmer instruct the operating system to reboot itself.  Named pipes There is nothing in CLR for those.  But Win32 and INTEROP provide the solution.  In the case of named pipes, there is no need for a DLL.  INTEROP takes care of the whole problem.

Consider the case where a Win32 method must be invoked in order to complete the requirements of a C# application program.  The Windows API, informally WinAPI, is Microsoft's core set of application programming interfaces (APIs) available in the Microsoft Windows operating systems. The name Windows API collectively refers to a number of different platform implementations that are often referred to by their own names (for example, Win32 API).

INTEROP entails marshaling of arguments and of a return value.  A programmer writes the Win32 method in C++, compiles it into a Dynamic Link Library with a specific release of Visual Studio, and chooses whether the DLL is targeted for a 32-bit or a 64-bit environment.  Another programmer writes C# code that calls the Win32 method.  In C# the programmer must declare the external method and decorate the declaration which names the containing DLL and its location in the file system.  When deploying the solution, the end user must be provided with the C# executable (the .EXE file), the C++ dynamic link library (the .DLL file), and instructions on which Visual C++ Redistributable Library must be installed to make a home convivial to INTEROP and Win32 executable code. 

The made-up term bitness refers to machine and operating system architecture: 32-bit or 64-bit.  The Intel 80386 and follow-on products are 32-bit platforms.  Advanced Micro Devices invented what became the industry standard 64-bit architecture, well after Intel’s introduction of its proprietary 64-bit Intel Itanium Architecture.  See http://en.wikipedia.org/wiki/IA-64. Intel now also manufactures in accord with the Advanced Micro Devices? design.  When 64-bit is referenced, the architecture that springs to mind is the AMD one, not the Intel one; although they persist and co-exist.  Windows operating system SKUs come in 32-bit and 64-bit flavors.  Either of these can be installed on 64-bit hardware platforms.  On 32-bit hardware platforms only 32-bit Windows can be installed.  Visual C++ Redistributable Library comes in 32-bit and 64-bit flavors.  (The 64-bit flavor can be used only on computers that are running a 64-bit version of Windows.)? There are several versions of Visual C++ Redistributable Library that are individually associated with specific Visual Studio releases.  The Visual Studio release is the one that is used to compile the unmanaged code.  Some combination of Visual Studio release, Visual C++ Redistributable Library, 32-bit/64-bit compilation choice, hardware platform bitness, and operating system bitness can be found that will support successful INTEROP calls.

? catch (Exception emode)? //? An attempt was made to load a program with an incorrect format. (Exception from HRESULT: 0x8007000b)

 //? On January 20, 2015, it was discovered that the ONLY problem was that the default 32-bit DLL was the "incorrect format."? The system demands a 64-bit DLL. 

For the latest revisions to Windows Server (starting with Windows Server 2012), the required runtime libraries are termed  features? of the operating system.  You should install runtime libraries via Server Manager’s Install Feature function.

 

Here is how C# is used to declare an external C++ method for calling via INTEROP.

Notice this declaration not only declares the name, arguments, and return type of the method, but it also declares the file system location of the containing Dynamic Link Library.

Here is how C# calls the unmanaged method.  The declaration makes this syntactically and semantically legal.

Finally, here is the unmanaged C++ code that could, if it wanted to, call any of the various Win32 methods.

Here is a real-world C++ method that allows a C# application to reboot the computer.

 

Factors

1.    The bitness of the hardware.

2.    The bitness of the operating system.

3.    The bitness of the compiled DLL.

4.    The version of Visual Studio that is used to compile the DLL.

5.    The version of the Visual C++ runtime redistributable library.

http://msdn.microsoft.com/en-us/library/ms235636(v=vs.100).aspx

 

https://msdn.microsoft.com/en-us/library/ms973190.aspx

 

MS973190 is a document that claims Visual Studio C#, which generates managed code, can be told to target 32-bit systems or 64-bit systems.  On a 64-bit system the Just-In-Time compiler will produce 64-bit code if that is targeted or 32-bit code if that is targeted.  Windows-on-Windows (WoW64) takes care of non-native cases.  So sometimes your managed code wants a 64-bit DLL for its unmanaged code.  Sometimes it wants the 32-bit DLL.  It’s may be (no guarantee) just a matter of the highly obscure configuration settings in your Visual Studio build.

 

 

How to Create a Dynamic Link Library in Visual Studio 2010

 

To create a new dynamic link library (DLL) project

 

 

1. On the menu bar, choose File, New, Project.

 

 

2. In the left pane of the New Project dialog box, expand Installed Templates, Visual C++, and then select Win32.

 

 

3. In the center pane, select Win32 Console Application.

 

 

4. Specify a name for the project?for example, MathFuncsDll?in the Name box. Specify a name for the solution?for example, DynamicLibrary?in the Solution Name box. Choose the OK button.

 

 

5. On the Overview page of the Win32 Application Wizard dialog box, choose the Next button.

 

 

6. On the Application Settings page, under Application type, select DLL.

 

 

7. Choose the Finish button to create the project.

 

 

The Care and Feeding of Named Pipes in Services

In computing, a named pipe (also known as a FIFO for its behavior) is an extension to the traditional pipe concept on Unix and Unix-like systems, and is one of the methods of inter-process communication (IPC). The concept is also found in OS/2 and Microsoft Windows, although the semantics differ substantially. A traditional pipe is "unnamed" because it exists anonymously and persists only for as long as the process is running. A named pipe is system-persistent and exists beyond the life of the process and can be deleted once it is no longer being used. Processes generally attach to the named pipes (usually appearing as a file) to perform inter-process communication.  http://en.wikipedia.org/wiki/Named_pipe

A named pipe is a message conduit.  It bi-directionally conducts messages between processes.  Program A and Program B can connect to a named pipe and send messages to each other without involving rotating magnetic memory devices.  All the magic happens in software and in main memory.

Named pipes are especially handy when it becomes necessary to communicate with daemons, also-known-as Windows NT Services.  The term daemon is popular in Unix and is almost unknown in Windows, but the concept is the same.  A daemon is a program that generally starts automatically after a reboot and runs constantly.  It runs independently of a shell or graphical user interface, which makes it hard to communicate with them.  Enter named pipes.  With named pipes, an application that possesses a user interface is able to send and receive messages to and from a daemon.

The daemon implements a named pipes Server that is constantly watchful for Clients that wish to connect to the Server.  Independent Client sessions are thereby established and lines of communication opened.

 

/*

?* The Server

?*

?* The purpose of The Server is to maintain and service a collection of clients who connect via a mechanism known as Named Pipes.

?* There is no .NET API for Named Pipes.  So The Server uses System.Runtime.InteropServices to call the necessary Win32 APIs.

?* Fortunately, the APIs are not too complicated nor demanding.  The Named Pipes methods are CreateNamedPipe and ConnectNamedPipe.

?* These are defined by the Windows Dynamic Link Library file kernel32.dll.  The trick is in parameterizing the calls correctly.

?* Named Pipes is the perfect mechanism for communicating with a Windows Service.  A Windows Service is a process that Windows

?* starts at boot time.  There is no conventional user interface.  For example, a Windows Service has no shell and no GUI.

?*

?* The Server maintains a non-generic List (a collection) of Client objects.  The Client object is a non-behaving structure that

?* represents an individual connected client by Named Pipes handle and Filestream, if any.

?*

?* The Server sends messages to users via an event mechanism.  Users register a method that is called when

?* The Server sends a message that it has received from a client. 

?*

?* Users of The Server are able to send messages.  When this happens, each of the connected clients is destined to receive the message.

?* Said users send messages to The Server via the SendMessage() method of The Server.

?*

?* The Start() method of The Server gets things going by forking a thread to field messages that clients send to The Server.

?* The aforementioned thread is called The Listener because it "listens" for clients to "speak."? The Listener is a

?* perpetual loop that exits only when a Named Pipes error occurs or because somebody calls the Stop() method of The Server.

?* On each pass of the loop The Server calls CreateNamedPipe to obtain what is called a Client Handle.  This call blocks

?* until a client connects or until Named Pipes (the complex of software) decides that blocking forever would be futile.

?* For example, the process may be exiting.  In that case, CreateNamedPipe returns an invalid handle, which is detected and

?* handled by The Server by terminating The Listener thread.  Normally, CreateNamedPipe returns a valid handle and this

?* represents a connecting client that usually may be found running in a separate process.  The method ConnectNamedPipe is

?* called to complete the connection between client and server.  Should this error, The Server terminates The Listener.

?* In the usual case of "no error," The Server creates a new Client object and puts the handle in it.  Then The Server

?* locks the client list and adds the Client object to it in one atomic action.  Finally, The Server forks a thread to handle

?* the arrival of messages from the client.  The Reader (as represented by the Read() method of The Server) is passed a

?* reference to the aforementioned Client object.

?*

?* The Read() method of The Server creates a FileStream object based on the handle in the Client object that The Listener passes.

?* Then The Reader enters a perpetual loop that terminates upon the occurrence of an error condition or the Stop() method is called.

?* A reference to the FileStream is put in the Client object next to the handle and the FileStream Read() method is called.

?* This call blocks until a message arrives or an error condition arises.  A zero length message is an error condition that may

?* occur after the client disconnects. In the normal case, bytes arrive and must be dispatched via an event to the user of

?* The Server. The bytes are first converted to a string.  In the error case, the handle is closed, the stream is closed,

?* and the client object is removed under lock from the client list.  The Reader thread is then terminated.

?*

?* The user of The Server is responsible to provide the name for a Named Pipe that matches the name that clients will use.

?* For example, if X is a reference to The Server object then X.PipeName can appear to the left of the assignment operator to

?* receive a string of the form \\.\pipe\<name-of-pipe>.  The user then must register a method to handle message arrival events.

?* As before, X is a reference to The Server object.  So X.MessageReceived can appear to the left of the delegate add operator.

?* The assignment is of a delegate instance that points to a method that shall be invoked by The Server when a message arrives.

?* The method receives two arguments in the following order:  1) a reference to the client object, and 2) the message as a string.

?*

?* The user of The Server is in full control of the life of The Server.  Life begins with initialization and Start() and ends

?* with Stop().

?*

*/

 

The following exhibit shows the user interface to a generic Named Pipes Client.  The dropdown list shows all the available servers.  The user first gets the app to enter privileged mode via a button that is labeled  Award Privil? and that has vanished because it has already been applied.  This causes the Client run as Administrator.  Administrator is a privileged mode in which it is permitted to open a named pipe as a stream, send and receive messages via named pipes.  The user enters a command in the Send Message area and presses the Send button.  The command is transmitted to the server indicated by the dropdown box.  The response is shown in the Rich Textbox under Received Messages.  In this example, the WINSVCFILCPY service responds with its status.  The user, having read the command response, can clear the Received Messages box and send another command.

 

 

 

Solving Mysterious System Crashes

 

 

Suppose your computer performs some vital function and it reboots during peak usage periods for unknown reasons.  You?d like to know why, wouldn’t you Enter the system dump and dump analyzer.  To enable Windows to take a dump you must configure Windows to do so.  Ideally you will collect as much data possible in order to allow the dump analyzer the greatest opportunity to spell it out for you.  Microsoft provides some guidance on how to size the page file since that file receives the memory dump during the crash.  See https://support.microsoft.com/kb/2860880

 

According to that document you should configure the page file by following to the formula A+B, where A is the amount of main memory in the computer’s hardware and B is 257,000,000 bytes.  B is space required to accommodate symbol tables and headers for the various portions of the dump.  Ideally your computer will take a full system dump when a crash occurs.  The other options are  no dump at all,  and  a minidump.

 

If your system is crashing and you cannot find a dump file (filename extension .dmp) then likely your page file is too small.  On a Windows XP machine with 2GB of main memory 4GB (the maximum allowed) of page file should be plenty.

 

Usually that which causes system crashes is third-party device drivers that contain faults and issues.  Imagine a linked list that contains an invalid address in one of its pointers.

 

Microsoft provides WinDbg, a dump analyzer.  There is a commercial product known as WhoCrashed that also provides dump analysis, perhaps in a manner accessible to mere mortals.  Visit their website at http://www.resplendence.com/whocrashed

 

Latest version: 5.03

WhoCrashed reveals the drivers responsible for crashing your computer

 

 

Whenever a computer running Windows suddenly reboots without displaying any notice or blue screen of death, the first thing that is often thought about is a hardware failure. In reality, most crashes are caused by malfunctioning device drivers and kernel modules. In case of a kernel error, most computers running Windows do not show a blue screen unless they are configured to do so. Instead these systems suddenly reboot without any notice.

 

WhoCrashed shows the drivers that have been crashing your computer with a single click. In most cases it can pinpoint the offending drivers that have been causing misery on your computer system in the past. It does post-mortem crashdump analysis and presents all gathered information in a comprehensible way.

 

Normally, debugging skills and a set of debugging tools are required to do post-mortem crash dump analysis. By using this utility you do not need any debugging skills to be able to find out what drivers are causing trouble to your computer.

 

MARK DANA ROCKMAN offers no endorsement of this product and is in no way liable for problems you may have using it, if any.

 

If you want to cause a Blue Screen of Death in order to obtain a dump, launch the following text file that I?ve named bsod.reg.  Apply this registry change.  Then reboot the computer.  Then logon.  Then hold down the rightmost CTRL key and simultaneously hold down the SCROLL LOCK key.  You can uncheck the automatic reboot box in Advanced settings under My Computer Properties to prevent a reboot following a Windows fault (i.e. BSOD).  The reboot will still happen when a program you are running decides to solve the world’s problems with a reboot.

 

Windows Registry Editor Version 5.00

 

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\kbdhid\Parameters]

"WorkNicely"=dword:00000000

"CrashOnCtrlScroll"=dword:00000001

 

 

Core Dump in Windows?

 

Back in the day, a failing computer program would leave behind pages and pages of numbers called a core dump.  It was just a memory picture which, along with register contents, could be used, occasionally, to discern why a program failed.  Such a thing is rarely used these days what with Integrated Debugging Environments (IDEs) that allow you to step through problematic code and fix it all in one go.  However, it is nice to know that Windows admits the possibility that you want to take a core dump.  Here’s how you do it.  Assuming the failed program is still in memory (it may be displaying an error message), find it in Task Manager, right click the item, click Create Dump File.  Et, voil?.  You?ve got yourself a minidump in the %TEMP% directory.  You can then launch Visual Studio and drag the minidump’s icon to Visual Studio’s window in order to visualize the dump.

 

 

WinDbg

 

WinDbg is good for many things.  Among those things is using it as a dump analyzer for when you are able to cajole Windows into producing a dump file such as MEMORY.DMP.  You get a switch to throw that picks the kind of dump to be produced.  You get a number to specify that establishes the size of the Windows Page File.  To produce a full memory dump after an unfortunate computing event (i.e. a crash), the page file will have to be larger than the number of bytes that physical memory will hold.  Naturally, a memory dump has metadata alongside a clump of bytes.  If you have 2GB of physical main memory then 4GB of page file should be more than adequate.  You can leave page size management up to Windows to perform, but the question remains whether Windows is up to the task, especially when the object of the game to is to guarantee that a panic dump is available in the event of a crash.

 

In order for dump analysis to work correctly, WinDbg requires the symbol table that is created by Microsoft when it does the RTM build of Windows.  You require a specific symbol table for the specific Windows build that produced the dump.

 

Naturally, this issue progresses to questions such as 1) Where do I get WinDbg and 2) Where do I get the symbol table Also, if I get the symbol table, where does the symbol table get put by the symbol table installer The installer makes a Federal case out of copying a file, a set of files, and/or database and neglects to mention where it is located.  _NT_SYMBOL_PATH is an environment variable (you may want to create it) that translates to a string that names the location of the symbol table if you knew where that might be.

 

It can be reliably reported that WinDbg comes with the Windows Driver Kit, which you can download at no cost to you.  And symbols are readily available at

 

?http://msdn.microsoft.com/en-us/windows/hardware/gg463028.aspx#_System_Requirements

 

Download Windows Symbol Packages

 

C:\Symbols

 

According to PROCMON, a delightful tattletale, the place where the symbols installer puts all those PDB files is C:\Symbols.  So you should go to Advanced Settings and equate _NT_SYMBOL_PATH to C:\Symbols.

 

You might also try setting _NT_SYMBOL_PATH to the following

 

srv*DownstreamStore*http://msdl.microsoft.com/download/symbols

 

This last one takes the prize cake.  It turns out that symbol tables for the exact build you are running are almost impossible to find.  To fix that little conundrum there exists a way for your favorite debugger (I love WinDbg) to satiate its craving for symbol tables by drinking from the online fountain in The Cloud.  That  DownStreamStore,  above That’s The Cloud.  So my recommendation is that you set _NT_SYMBOL_PATH to the aforementioned value.  You get far less complaint from WinDbg about undefined symbols.

 

 

Process Monitor and WireShark

 

Process Monitor, by Mark Russinovich and Bryce Cogswell.  http://technet.microsoft.com/en-us/sysinternals/bb896645.aspx

 

This splendid  app? displays and preserves a comprehensive record of actions internal to Windows.  For example, you can watch various processes open and close files, attempt to communicate on the Ethernet, and play around with The Registry. When the system is complaining about odd things, Process Monitor steps in to explain.  If you don’t find Process Monitor does enough to satisfy your need for Ethernet monitoring, I suggest you use WireShark: https://www.wireshark.org/download.html

 

Process Monitor is an advanced monitoring tool for Windows that shows real-time file system, Registry and process/thread activity. It combines the features of two legacy Sysinternals utilities, Filemon and Regmon, and adds an extensive list of enhancements including rich and non-destructive filtering, comprehensive event properties such session IDs and user names, reliable process information, full thread stacks with integrated symbol support for each operation, simultaneous logging to a file, and much more. Its uniquely powerful features will make Process Monitor a core utility in your system troubleshooting and malware hunting toolkit.

 

These tools are free of any charge.  And very, very useful.

 

 

 

 

 

Do Not Block the Message Pump

 

 

 

 

 

The message pump drives the  app. Without messages flowing, the program stops behaving.  That is to say, there is no behavior.  You, the end user, see a Not Responding message.  That means the  app? is not processing what may well be a lengthy queue of messages.  That’s because the  app? is busy doing things other than processing messages.  Solution While you think the purpose of your program is to solve a business problem, Windows knows better.  The real purpose of your program is to process messages.  So you must create a thread to handle the business problem.  The initial thread thus is dedicated to handling messages.

 

Once you have a thread dedicated to the business problem, the original thread takes care of message handling.  The two threads must coordinate to avoid strange occurrences such as happens when you click the X in the upper right-hand corner of the Windows Form.  The window closes, yet the business problem solving continues invisibly.  To avoid this happening, you can override the Windows Form method that receives control in the event that the aforementioned X is clicked.  You can set a flag to tell program management to ignore the close request for the time being.  Once the business solver finishes its task, you can reverse the close inhibit.  Here’s the code:

 

 protected override void OnClosing(CancelEventArgs eventargs)? // the X icon in the upper right corner of the dialog box is made inoperable when a worker is working

? {

? if (bWorkerRunning)  ?// if a worker thread exists

? eventargs.Cancel = true;? // then ignore the close request

?  ?else

? {

? eventargs.Cancel = false;

? base.OnClosing(eventargs);

? }

 return;

? }

 

 

 

 

 

 

 

Configure Domain Name Service (DNS)

 

From time to time, one is required to create a Domain Controller.  Many Domain Controllers are Domain Name Servers (DNS).  A DNS is responsible for transforming a symbolic Internet or LAN reference to an IP address.  The necessary steps are few and simple and yet the written material is long, complicated, and jargon-laden.  If you get it wrong, your DNS will be resolving symbolic references via the wrong servers out on the Internet.  That’s slow, inefficient, and, in egregious cases, illegal.

 

Modern Windows Server operating systems offers DNS that is integrated with Active Directory (AD).  That means when you have multiple domain controllers (in the same domain), that DNS is replicated across all of the Domain Controllers.  AD is a database that facilitates coordination between Domain Controllers and allows workstations to become  members? of the domains that Domain Controllers are responsible for protecting and serving.  A hugely important offering of Domain Controllers (DC) is Group Policy.  Group Policy, it need be pointed out, is not Policy that applies to Groups.  Rather, it is a grouping of policies.  The big deal about Group Policy is that is implements the means to set policy for member workstations from a central location.  The central location is the Domain Controller.  You insist, for example, that your workstation users employ complex passwords that expire every month Bingo!  No problem.  Just set a policy that says so.

 

https://support.microsoft.com/kb/172963/EN-US is a document that explains how you configure a DNS.  It is very good, but incomplete.  It will get you going with name resolution on your LAN.  That way, computers on your LAN can find each other.  Symbolically.  If your domain does not extend beyond your LAN, you establish an arbitrary domain name such as mdrsesco.com.  All of your member workstations will be  joined to the domain? and named <name-of-computer>.<domain-name>.  For example, P30.mdrsesco.com. 

 

The following instructions apply to Windows Server 2012 R2.  Previous incarnations of Windows Server sported a DNS configuration utility that was not integrated within Server Manager.  Nowadays one must start Server Manager and navigate its user interface to perform any administrative function.  To get started, the DNS role must be configured.  Launch Windows Server (it usually launches by itself when you restart the server).Click Manage > Add Roles and Features > Role-based or Feature-based Installation > Select a server from the pool > Server roles > DNS Server.  You just tick the checkbox next to DNS Server and follow the instructions that start the installation of the DNS role.  Once that is done, you are ready to configure DNS.  To do that, go back to Server Manger.  Click Tools > DNS.  By clicking DNS you cause to launch what is called DNS Manager.  That is your window onto the world of the Domain Name Service.

 

Notice the menu across the top of DNS Manager.  One of the menu items is Action.  Click Action > Configure a DNS Server.  This launches a  wizard? that guides you through the process of creating the DNS configuration. You choose to configure a) a Forward Lookup Zone, b) both Forward and Reverse Lookup Zones, c) Neither Forward Nor Reverse Lookup Zone.  Item c) is a bad choice because all name resolution is then done via top-level domain (TLD) name servers and those servers are not designed to nor capable of being Mister Answerman for every workstation in the Universe.  You want at least a Forward Lookup Zone and preferably Forward and Reverse.  Press the radio button that applies in your case. Do you want to create a Forward Lookup Zone now Answer yes.  Do you want a Primary Zone A Secondary Zone A Stub Zone Pick Primary.  Tick the Store the Zone in Active Directory checkbox.  Select how you want Zone Data replicated.  You want it replicated far and wide.  Throughout the domain or throughout the forest of domains. I suggest To all DNS Servers throughout the domain.  Now you get to pick your Zone Name.  In my case the Zone Name is mdrsesco.com.  Now you get to choose whether Dynamic Updates are to occur.  It’s a convenience feature.  When a new workstation appears on the LAN and points to the DC/DNS, the DNS recognizes the need to list the workstation so that other workstations on the LAN can communicate with it.  Failing to allow Dynamic Updates means you will be required to manually configure each workstation.  I suggest that you chose to have DNS perform Dynamic Updates. These can be Secure or Not Secure.  Since your DNS database is integrated within AD, you want secure updates.  Now you want to create a Reverse Lookup Zone.  Such a Zone provides the capability to take an IP address and map it to a corresponding symbolic name.  Given 192.168.1.4 you may get, for example, P30.mdrsesco.com.  Do you want a Primary Zone A Secondary Zone A Stub Zone Pick Primary.  Tick the Store the Zone in Active Directory checkbox.  To All DNS Servers throughout the domain. Next, do you want to resolve IPv6 addresses Certainly not.  Not on a LAN.  Choose IPv4.  Next, enter the Network ID that applies to your domain.  Suppose all of the computers on your LAN sport an IP address of the form 192.168.1.x, where x varies according to which computer you are considering and is unique to that computer out of all the computers on the? LAN.  The network portion of the IP address is 192.168.1.  So enter that as your Network ID.  This assumes your subnet mask is 255.255.255.0.  Do you want Dynamic Updates Yes.  Should this DNS Server forward queries Of course it should. Choose the radio button next to that option and enter the IP addresses (or the equivalent symbolic names) of the Domain Name Servers out on the Internet that your Internet Service Provider has kindly documented for you.  Congratulations!  You are done.

 

When you need to alter the list of Forwarders, go to the Menu and click Action > Properties.  See the tab-laden dialog box.  Choose the Forwarders tab, click Edit and enter the IP address or symbolic names of the forwarders. Then click OK.  See http://technet.microsoft.com/en-us/library/cc754941.aspx.

 

I find it comforting to know that DNS is successfully contacting the Forwarders, so I use a packet sniffer as I enter a bogus URL in my favorite browser.  Yes, indeed.  The DNS is contacting the Forwarders.  Done and done.

 

rem https://technet.microsoft.com/en-us/library/cc754941.aspx

rem Run this as Administrator (i.e. elevated privileges)

dnscmd /ResetForwarders /help

dnscmd fserv.mdrsesco.com /ResetForwarders 199.45.32.43 199.45.32.38 /TimeOut 5

REM

REM

REM

REM

REM In future versions of Windows, Microsoft might remove dnscmd.exe.

REM

REM If you currently use dnscmd.exe to configure and manage the DNS server,

REM Microsoft recommends that you transition to Windows PowerShell.

REM

REM To view a list of commands for DNS server management, type

REM "Get-Command -Module DnsServer" at the Windows PowerShell prompt. Additional

REM information about Windows PowerShell commands for DNS is available at

REM http://go.microsoft.com/fwlink/?LinkId=217627.

REM

REM Usage: DnsCmd <ServerName> <Command> [<Command Parameters>]

REM

REM <ServerName>:

REM IP address or host name  -- remote or local DNS server

REM  ?.  -- DNS server on local machine

REM <Command>:

REM /Info? -- Get server information

REM /Config? -- Reset server or zone configuration

REM /EnumZones -- Enumerate zones

REM /Statistics? -- Query/clear server statistics data

REM /ClearCache? -- Clear DNS server cache

REM /WriteBackFiles? -- Write back all zone or root-hint datafile(s)

REM /StartScavenging -- Initiates server scavenging

REM /IpValidate? -- Validate remote DNS servers

REM /EnumKSPs? -- Enumerate available key storage providers

REM /ResetListenAddresses? -- Set server IP address(es) to serve DNS requests

REM /ResetForwarders -- Set DNS servers to forward recursive queries to

REM /ZoneInfo? -- View zone information

REM /ZoneAdd -- Create a new zone on the DNS server

REM /ZoneDelete? -- Delete a zone from DNS server or DS

REM /ZonePause -- Pause a zone

REM /ZoneResume? -- Resume a zone

REM /ZoneReload? -- Reload zone from its database (file or DS)

REM /ZoneWriteBack -- Write back zone to file

REM /ZoneRefresh -- Force refresh of secondary zone from master

REM /ZoneUpdateFromDs? -- Update a DS integrated zone by data from DS

REM /ZonePrint -- Display all records in the zone

REM /ZoneResetType -- Change zone type

REM /ZoneResetSecondaries? -- Reset secondary\notify information for a zone

REM /ZoneResetScavengeServers? -- Reset scavenging servers for a zone

REM /ZoneResetMasters?  ?-- Reset secondary zone's master servers

REM /ZoneExport? -- Export a zone to file

REM /ZoneChangeDirectoryPartition -- Move a zone to another directory partition

REM /ZoneSeizeKeymasterRole? -- Seize the key master role for a zone

REM /ZoneTransferKeymasterRole -- Transfer the key master role for a zone

REM /ZoneEnumSKDs? -- Enumerate the signing key descriptors for a zone

REM /ZoneAddSKD? -- Create a new signing key descriptor for a zone

REM /ZoneDeleteSKD -- Delete a signing key descriptor for a zone

REM /ZoneModifySKD -- Modify a signing key descriptor for a zone

REM /ZoneValidateSigningParameters -- Validate DNSSEC online signing parameters for a zone

REM /ZoneSetSKDState -- Set Active and/or Standby keys for a signing key descriptor for a zone

REM /ZoneGetSKDState -- Retrieve dynamic state for a signing key descriptor for a zone

REM /ZonePerformKeyRollover? -- Trigger a key rollover in a signing key descriptor for a zone

REM /ZonePokeKeyRollover -- Trigger a key rollover in a signing key descriptor for a zone

REM /ZoneSign? -- Signs the zone using DNSSEC online signing parameters

REM /ZoneUnsign? -- Removes DNSSEC signatures from a signed zone

REM /ZoneResign? -- Regenerate DNSSEC signatures in a signed zone

REM /EnumRecords -- Enumerate records at a name

REM /RecordAdd -- Create a record in zone or RootHints

REM /RecordDelete? -- Delete a record from zone, RootHints or cache

REM /NodeDelete? -- Delete all records at a name

REM /AgeAllRecords -- Force aging on node(s) in zone

REM /TrustAnchorAdd? -- Create a new trust anchor zone on the DNS server

REM /TrustAnchorDelete -- Delete a trust anchor zone from DNS server or DS

REM /EnumTrustAnchors? -- Display status information for trust anchors

REM /TrustAnchorsResetType -- Change zone type for a trust anchor zone

REM /EnumDirectoryPartitions -- Enumerate directory partitions

REM /DirectoryPartitionInfo? -- Get info on a directory partition

REM /CreateDirectoryPartition? -- Create a directory partition

REM /DeleteDirectoryPartition? -- Delete a directory partition