SD 1.02 available for testing

69 views
Skip to first unread message

D M

unread,
Apr 25, 2026, 2:38:40 AM (4 days ago) Apr 25
to Pick and MultiValue Databases
In SD 1.02 Mark has added back in the Pick compatibility and PROC features that were available in OpenQM 2.6.6.  We removed some features during development and have now added these features back in as users convinced us they were needed.

The only major things we charged from the features available in OpenQM GPL and ScarletDME at this point are ssh for remote access and how accounts are managed.

New added features include embedded Python, a replacement for the encryption features that were not included in the open source version of OpenQM and the fixing of several bugs that we discovered during our code review.  Also available as an optional feature during install is the tape backup and restore system.  

The scripts to install and remove SD have also been made more robust and now support current Debian, Ubuntu, Fedora and Arch distributions.  The system will also install on the Linux Subsystem for Windows, giving Windows users a way to access the database without having to install a full version of Linux.

If anyone would like to test this before we update the main branch, you can do:


When you run the installer, just choose the dev branch rather than main.

MAV

unread,
3:35 AM (17 hours ago) 3:35 AM
to Pick and MultiValue Databases
Hello

I think it's a truly fantastic piece of work. A great alternative to consider. Congratulations.

Have you ever thought about implementing OpenQM Collections and its excellent JSON handling? I know it's difficult, but it would make SD a very powerful alternative.

Excellent work.

Marcos Alonso Vega
INGESCO Sistemas Informáticos

D M

unread,
4:14 AM (16 hours ago) 4:14 AM
to Pick and MultiValue Databases
Thank you for your kind comments.  SD is open source, so Mark and I do this for fun, not for profit.  Our focus has been to give the MV community an easily installable more complete and tested version of the OpenQM 2.6.6 version that was released 19 years ago and converted to 64 bit several years ago by the ScarletDME developers.  I think we have succeeded in making the system more reliable and secure.  For me, every three or four months I will suddenly get the urge to work on something or suggest a new feature or do documentation.  We want the system to be useful to others, but the progress is slow.  We would need significant additional contributors to enhance the core parts of the system, to create better documentation or to have a better web site.

As an example, my latest interest is to try running various AI code evaluation tools against the C code that is a part of the base system, make the suggested changes and see if the system still works.  Will this have any practical use and would it be the best investment of my time if this was being done for profit; probably not.  But it is what interests me right now, so that is what I am doing.  

Occasionally I will suggest a feature to Mark that requires his skill set.  If he finds it interesting it gets done (and vice versa).

Since we do this in our free time, our focus has been on security first.  We got rid of remote telnet access and made SSH the only remote connection option.  Mark created a Python based compiler and then we did some other work to replace all the binary bits that still remained in ScarletDME.  Any binary files needed are created during the installation process.  The reason that we did this was to mitigate the complaints that parts of ScarletDME were not auditable.  Every part of SD can be reviewed as it is all text, no binary parts remain in the source repository.  We also added the abiltiy to encrypt files.  The GPL version of openQM did not include the encryption features found in the commercial version.  Our encryption features are fairly minimal, but they do allow for data to be encrypted.  As part of the security focus we removed the user and accounts functions and replaced them with user and group accounts all located under /home/sd.  This allowed us to enforce  security permissions more easily.

Mark also added some fun extras like embedded Python. The only features that are removed in the most current version are the menu and screen builders that were a part of openQM and ScarletDME.  We had originally removed PROCs and Pick compatibility features but have now returned them to version 1.02.

Much of the internal work on SD was done by Mark.  My work was mostly the forking process from ScarletDME including removing the ScarletDME and openQM branding as we wanted to make clear that this was a separate project and that the prior developers should not be associated with any failures on our part or our decisions to modify and/or remove features.  However we do credit them for their work in our contributors file.  I also created the original installsd and deletesd scripts to make the process of maintaining  the SD system easier than was true with either the open source version of openQM or ScarletDME.  Mark and I have continued to refine these scripts to make them more bullet proof and support a larger number of Linux distributions.  I do try to test the installer when new distribution versions are released, but it is time consuming.

Anyway, I hope that gives anyone interested an overview of our slow but steady progress in refurbishing this somewhat antique database.

John Stokka

unread,
7:23 AM (13 hours ago) 7:23 AM
to mvd...@googlegroups.com
Hey guys, are you interested in the domain mvopen.org.  I own it.

I think it's a pretty good name... :)

In Christ,

John R. Stokka


--
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/af9d7fb3-241c-438d-b347-64ac16e5a9c9n%40googlegroups.com.

Steven Martin Trimble

unread,
10:09 AM (10 hours ago) 10:09 AM
to mvd...@googlegroups.com
personally of all of the PICK flavors I have worked on over the years, my favorite remains to be QM/Linux.
thanks for ALL the work you and Mark have done.
If I can be of assistance, please let me know.

CDMI
Steven Trimble
(501) 772-3450 cell/text


On Tue, Apr 28, 2026 at 3:14 AM D M <dmon...@gmail.com> wrote:
--

D M

unread,
3:17 PM (5 hours ago) 3:17 PM
to Pick and MultiValue Databases
Steve, the testing you have done and the Raspberry Pi input you have given us has been great.  Thank you.  Since this is open source, anyone can take a look at the code and propose new features by just issuing a request for us to approve.

SD can be forked and become the basis for other visions.  Again this is a benefit of open source.  Or, people with varying skill scan choose to focus their talents on improving SD.  This is another benefit of open source.

What is SD good for.  It is great for small projects or projects focused at companies with limited users.  The language is great for hobbyists or for small business owners who have an interest in coding for their own business.  It is far less complex than the normal web application stack.  A cheap server running linux can cost as little as $100 and still provide enough computing power for as many as 50 users running Putty or another SSH client. For a graphical front end, a Gambas Basic program using the API can run well on a $100 used HP EliteDesk or a Raspberry Pi.

Much of the complexity of a modern database such as PostgreSQL or an Apache/MySQL web stack is simply not needed for many small business tasks.  And for business in niche markets where few SAAS applications are available, custom software with minimal upfront and no ongoing cost is possible.  And, with developers working with modern database tools costing upward of $100K to $200K per year, custom software is beyond the reach of most small businesses.

I am always willing to take on small projects for a much smaller fee than programmers who need to make a living from programming.  For me, being retired, a side project buys extras, like a trip to the beach, rather than necessities like food and shelter.  And, if a core feature is needed in SD, you can always offer a bounty to get it added.  If Mark or I are not interested, I am sure there are other programmers who follow this group who would be.

Steven Martin Trimble

unread,
4:09 PM (4 hours ago) 4:09 PM
to mvd...@googlegroups.com
Don
my comment about QM is also why I love SD
I agree with you about how powerful and inexpensive it is combined with a solid Linux platform.
I am so happy you guys have done this for the community.
The fact that Rocket Software owns on the MV flavors (pretty much) is a serious thorn in my side.
I am currently still supporting 5 customers who could easily use SD / Linux as their platform.
The largest user has a 30user license for QM/Linux.
So, I am quite comfortable suggesting SD as an alternative to the QM/Rocket environment.
Again, thanks for all you do.
Much respect,

CDMI
Steven Trimble
(501) 772-3450 cell/text

Reply all
Reply to author
Forward
0 new messages