Have the source codes of the "Pick Database" ever been published?

[Ivan Ivanov]

Hi,

Have the source codes of the "Pick Database" ever been published?
For example, for some old versions? How was it for "OpenQM (ScarletDME)" version 2.6?

Currently, I have found only two open source "Pick Database"

1. ScarletDME
2. Maverick

Have the sources for the very first versions of the "Pick Database" ever been published?

[Dawn Wolthuis]

No, I don’t think so. There were a couple of other starts at open source or at least indie efforts: Bart and Winter. Others on the list might be aware of the history or status of these. 

None of the companies owning the IP for D3 or U2 made early versions, by various names, open source, as best I know (and I have taken a reasonably deep dive).

—Dawn

On Dec 16, 2024, at 5:25 PM, Ivan Ivanov <gangst...@gmail.com> wrote:

Hi,

--
You received this message because you are subscribed to
the "Pick and MultiValue Databases" group.
To post, email to: mvd...@googlegroups.com
To unsubscribe, email to: mvdbms+un...@googlegroups.com
For more options, visit http://groups.google.com/group/mvdbms
---
You received this message because you are subscribed to the Google Groups "Pick and MultiValue Databases" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mvdbms+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/mvdbms/34972a82-7942-449d-aabe-b6bd3cb49e3en%40googlegroups.com.

[Will Johnson]

The very first versions were all machine code for a specific chip.  You would not be able to read it today imho.

[Jay LaBonte]

Is MaVerick strill around? Are they still actively developing it? I haven’t heard anything about it since early 2003.

 

Jay LaBonte

--

You received this message because you are subscribed to
the "Pick and MultiValue Databases" group.
To post, email to: mvd...@googlegroups.com
To unsubscribe, email to: mvdbms+un...@googlegroups.com
For more options, visit http://groups.google.com/group/mvdbms
---
You received this message because you are subscribed to the Google Groups "Pick and MultiValue Databases" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mvdbms+un...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/c37fb042-e7d1-4c62-9a3a-3bf39f2f15fbn%40googlegroups.com.

[Dawn Wolthuis]

I don’t think so.

—Dawn

On Dec 17, 2024, at 10:35 AM, 'Jay LaBonte' via Pick and MultiValue Databases <mvd...@googlegroups.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/mvdbms/00f701db5099%242f99d270%248ecd7750%24%40paradigm-systems.us.

[Christopher Jeune]

I remember BART that’s going back a ways…

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/5B3B36FC-D29A-4EFC-AD08-A2CA507BBB4C%40gmail.com.

[allen bell]

Maverick was the work of Robert Colquhiun

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/CACHAbR2-q3qgCM9%2BM-tt1rDdE5hRXZUwLkrFGVCBSuQO05NYSg%40mail.gmail.com.

[allen bell]

Sorry, hit the button too quick! Robert Colquhoun was developing Maverick for his PhD thesis. It was never truly completed. I was one of the drones working under his lead on it for a year or two. I thought it was truly genius. It could output either java source code or have byte code. It got as far as the Basic compiler being mostly done, the file system and the printer interface. There was some work in English/Access/Recall, but that was nowhere near complete. Robert is one seriously brilliant guy, but he moved on

[Wols Lists]

On 17/12/2024 17:14, allen bell wrote:
> Sorry, hit the button too quick! Robert Colquhoun was developing
> Maverick for his PhD thesis. It was never truly completed. I was one of
> the drones working under his lead on it for a year or two. I thought it
> was truly genius. It could output either java source code or have byte
> code. It got as far as the Basic compiler being mostly done, the file
> system and the printer interface. There was some work in English/Access/
> Recall, but that was nowhere near complete. Robert is one seriously
> brilliant guy, but he moved on
>
>
> On Tue, Dec 17, 2024, 11:08 AM allen bell <apbe...@gmail.com
> <mailto:apbe...@gmail.com>> wrote:
>
> Maverick was the work of Robert Colquhiun
>

And Robert didn't start MaVerick, I did :-)

But he put far more work in, and effectively made it his own project.
I'm a firm believer in one of the Open Source mantras - he who does the
work makes the rules - so seeing as he did so much more than me ...

Unfortunately, he made a load of changes from my initial vision. Not
saying my vision was better than his ... but that was one of the things
that took the project away from me.

Cheers,
Wol

[Will Johnson]

In the beginning, the Reality machine from Microdata.  The code was written for that chip.

An upgrade meant, a technician coming to your site and replacing the chip and the code in firmware

I'm not even sure there was a way to view the code from a terminal.

Ultimate followed suit.

Adds was the first implementation, iirc, where you could actually dump the ABS frames and view this arcane machine code

You could write your own ABS frame as well, at that time, for specific needs, but the code book itself was not public.

[Nivethan T]

https://github.com/Krowemoh/R83

The r83 floppies are available here. A decompiler and some AI magic may get you somewhere.

It would be interesting to be able to build from source.

[Joe Goldthwaite]

Hopefully Pete Schellenbach can correct me here but the story I was told is that his father had a Microdata computer for his practice (Dental?) and Pete was writing software for it.

Pete learned the internals by writing a disassembler for the microcode. He used that knowledge to write AccuPlot.

I wouldn't be surprised if he still had that code somewhere but it could be on 9 track tape. That code is probably still copyrighted and owned by Microdata's prodigy.

The legality of disassembly is a lot more complex now so I would bet that there's no way to legally publish any of it without the express permission of whoever currently holds the copyright.

--

You received this message because you are subscribed to
the "Pick and MultiValue Databases" group.
To post, email to: mvd...@googlegroups.com
To unsubscribe, email to: mvdbms+un...@googlegroups.com
For more options, visit http://groups.google.com/group/mvdbms
---
You received this message because you are subscribed to the Google Groups "Pick and MultiValue Databases" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mvdbms+un...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/2ba8fe19-885a-417a-9df1-c6c53f84d44en%40googlegroups.com.

[Toby Kraft]

It's actually a bit more interesting than that.  Don Nelson dreamed up a virtual machine designed to manipulate strings that ran in a virtual memory environment.  The Pick database environment was written in the 'assembly' language of that virtual machine (before Reality existed).  Microdata had a micro-programmable computer that had no assembly language of its own, you had to write microcode for it to do anything.  This made it ideal for Pick as it provided a platform on which to actually run the virtual machine.  So the early Pick team (Dick, Chandru Murthi, Phil Earl, ? not sure who all) wrote microcode to execute the virtual object code for the virtual assembly language along with implementing the virtual memory management needed.  Thus the Pick OS was born running its own environment on its own dedicated hardware which became Reality, then Pick R77, R80, R83.

What was cool was that at the end of the early system release tapes, there was an account-save of the assembler account with source code for many of the system utilities.  That's how I got my start doing system level programming on Pick.  FWIW, John Timmons developed a disassembler and managed to disassemble most if not all of the system code (not the microcode of course).

To read more on the early days -> https://www.facebook.com/legacy/notes/145445760681669/

[Will Johnson]

Well AccuPlot was actually loaded to floppy disks as well in the 1980s

[Will Johnson]

Naming the early versions is problematic until R81

I'm sure Ultimate would pitch a fit to have all of their own work called Pick R anything

Same with Adds.

There was a lot of heavy lifting going on directly inside those companies before R81 came along

[Steven Martin Trimble]

I think Sir Robin with ADDS believed the OS Software was more important than the hardware it ran on. My favorite PICK flavor all the way through MVBase. Now, for quite some time, I am and have been an openQM follower. The Linux/QM combination is very hard to beat. Thank you Sir Martin. 😃

I hope some youngsters come along and keep all of this alive.

CDMI

Steven Trimble

(501) 772-3450 cell/text

cdm...@gmail.com

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/f8290830-6abf-4869-ba50-1c49886e03c0n%40googlegroups.com.

[Dawn Wolthuis]

I agree that OpenQM rocks. Unless Rocket were willing to have OpenQM compete, as it did before, with U2 or were willing to sell it (why would they?), it looks like it might have reached its height at this point. 

ScarletDME seems to have gained some interest. I am curious whether there are any live, profitable software applications running on ScarletDME. Does anyone know if it is in production anywhere?

—Dawn

On Dec 19, 2024, at 9:34 AM, Steven Martin Trimble <cdm...@gmail.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/mvdbms/CAJDLGm7iasDGbBhpeJUiK74xtWWV1OXR2nt1v1Ko9TqyfZYGVQ%40mail.gmail.com.

[Steven Martin Trimble]

I have moved 'accounts' from Linux/openQM to Linux/ScarletDME.

One main difference is the naming convention of dynamic files.

ScarletDME dynamic file:

file named PRODUCT looks like this on Linux:

/home/qm/PRODUCT/PRODUCT/~0

/home/qm/PRODUCT/PRODUCT/~1

commercial openQM dynamic file:

file named PRODUCT looks like this on Linux:

/home/qm/PRODUCT/PRODUCT/%0

/home/qm/PRODUCT/PRODUCT/%1

Other obvious differences is:

ScarletDME is QM version 2.6.6

I have the ScarletDME docs here: https://scarl.cdmiweb.com/

openQM is version 4.0-5 (or greater)

I have the openQM docs here: https://qmhelp.cdmiweb.com/

I would think that ScarletDME would have no trouble running in a production environment.

One really cool thing is there is no 'MAX Users' license involved.

I have to give kudos to Donald Montaine credit for the installers, etc.

He is also working on what he calls 'StringDB'. Not sure if it's going to be GPL or not.

Maybe he is watching this thread and can comment on this.

CDMI

Steven Trimble

(501) 772-3450 cell/text

cdm...@gmail.com

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/73831463-8F9F-44CF-809A-ADC6CDA0577C%40gmail.com.

[QRY Ltd]

Mark B and I have been working on SD (The String Database).  It is GPL3 licensed and is located at   https://github.com/stringdatabase/sdb64.  It is based on 64 bit ScarletDME with several legacy features removed and some features added, such as python integration.  Also much of the C and Basic code has been reviewed and some issues found and resolved.  In addition, it has been integrated more tightly with the underlying OS, an example being allowing the micro or nano editors to be used as alternatives to SED.  It comes with improved installer scripts.  I recently moved (ie: real-life intervened) but I am planning to get back to the project in 2025.

[Jay La Bonte]

I think all the work people are doing on ScarletDME is great. However, that said, I think if we as a community are going to make ScarletDME a viable database for production environments, we need to get organized and formalize the development/enhancement of the ScarletDME.

I think with the proper support, ScarletDME could be an viable option for businesses.

I’m open to idea as to what everyone thinks about this?

Jay LaBonte

[Joe Goldthwaite]

On my first programming job, one of our clients had a Microdata computer. It seems like it had 32 KB of ram and a 2 MB hard drive but it might have been bigger. The hard drive was divided into 512 byte frames. The entire OS fit in the first 512 frames. That's 262,144 KB of disk. I don't remember how many terminals they had attached to it but it was running their business. That 262,144 KB held the virtual memory manager, the database manager, the code for English and the Basic compiler and debugger.

It would be an interesting project to port something like that to the embedded platforms like the ESP32. I built a musical clock using the ESP32-S3. It has 8 MB of built in flash memory, 512 KB of SRAM and a dual core 240 Mhz processor. It would be funny to run a business on a Pick based system that costs $12.50.

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/e493cfc8-6ce6-4a41-a064-c9a9f6fd528bn%40googlegroups.com.

[Steven Martin Trimble]

I think if we (as a community) got behind anything, the StringDB from Donald Montaine might be a better option. They both require Linux. Not sure if ONLY a certain LINUX distro is required. I had to jump through some hoops to get it to run on Rocky Linux 9.

I had a very early version, so I am not quite sure what all has changed since then. Lots, I believe.

CDMI

Steven Trimble

(501) 772-3450 cell/text

cdm...@gmail.com

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/f0d2fd6a-cd90-49cd-bed3-afea8a4d3881n%40googlegroups.com.

[QRY Ltd]

Regarding requirements for SD; there are install scripts provided for Fedora and Debian derived distros.  The problem with Redhat, and clones, is that they are so far behind Fedora that the GCC compiler is not new enough to correctly compile SD, which is 64 bit only.  That is also true for the 64 bit version of ScarletDME.  Ubuntu server and Fedora server work if you want a server distribution.  Debian 12 and later work (and derived distros such as Ubuntu and Mint), Fedora 39 and later work.  I will be checking to see if CentOS works, as it is now a preview of the next version of Redhat Enterprise Linux.

The last commit to the ScarletDME repository was 2 years ago.  The last commit to the SD repository was 4 months ago and Mark B has been doing continuing work while I moved.  I will be getting involved again soon.  While we have removed some features like procs, we have improved many features and added some like embedded python and the integration of operating system utilities like micro, nano and less. We have also corrected some errors in the C and Basic code and have removed all the QM branding, including  in procedure and variable names.  We welcome any additional help and will be glad to provide whatever information is needed.  
--

[Will Johnson]

The first several Microdata Reality's running pick I believe had 16 ports

[Joe Goldthwaite]

Good point. You'd probably have to go with one of the boards that have Ethernet and put an ssh library on it. That's going to drive the hardware price up to the $20 range.

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/828f871a-6be2-4299-8665-9f895ffbdaa9n%40googlegroups.com.

[Donald Montaine]

A couple of other points about SD.  There are no "binary bits" in the code, nothing is pre-compiled, everything is compiled during the install.  Hashed files are created during the install.  The VOC files are directory files rather than hashed files.  This  means that VOC items can be edited with an operating system editor such as micro.  Hashed DICT files are created, populated and items compiled if necessary during the install.  This removes the problem that existed with ScarletDME, where some users complained about not being able to verify the safety of all elements prior to install.  Security is improved in that ssh is required for remote connections.  One recent change is that Debian 12 is not supported as the current version 12.8 broke the previously working install.  Ubuntu 24.04, Mint 22 and Fedora 40 & 41 are the tested distros. Other derived distros of Ubuntu and Fedora, your mileage may vary.

[Steven Martin Trimble]

hey thanks for the update Donald.

glad u r back in the saddle.

Happy Holidays

CDMI

Steven Trimble

(501) 772-3450 cell/text

cdm...@gmail.com

To view this discussion visit https://groups.google.com/d/msgid/mvdbms/8aabe010-3a38-4c40-b09f-294e69c7fa53n%40googlegroups.com.

[James A]

Where would this 'StringDB' be found? I've googled and searched github etc and can't find it.

[Brian Speirs]

The URL was posted earlier in this thread:

https://github.com/stringdatabase/sdb64

Cheers,

Brian