International Spectrum

[Philip Ellis]

Does anyone know what's happened to https://www.intl-spectrum.com/

The site doesn't appear to be being updated, and no magazine has been published since Jan/Feb 2020.

Perhaps it's related to the pandemic?

[Nathan Rector]

International Spectrum is still around.   We are currently working revamping our training and materials, and are little but behind.

-Nathan Rector
 International Spectrum

--
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 on the web visit https://groups.google.com/d/msgid/mvdbms/7fb308d9-353c-44da-8b09-8fb629446ca8n%40googlegroups.com.

-- 
--------------------------------------------
Nathan Rector
International Spectrum, Inc
http://www.intl-spectrum.com
Phone: 720-259-1356

[RDG Information Systems]

Good to hear.  Do you want to reply? If not I can post an update.

v/r

Ray
-- 

Raymond DeGennaro
Founder & Lead Developer
RDG Information Systems, LLC
Raymond....@RDGInfoSys.com
http://RDGInfoSys.com
+1-815-308-0815

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/c1ef2713-93ea-94bc-969f-5805cf64fab3%40intl-spectrum.com.

[Tony Gravagno]

Nathan - There has always been a problem for industry companies, where marketing budgets didn't permit companies to participate in multiple events, or where single events were too far away.

I believe Gus and Monica were looking at conference virtualization as a solution to these challenges. That paradigm is now not only ever more timely, but much more possible even just within the last year as the number and scope of teleconferencing solutions have exploded.

Perhaps Spectrum could consider a series of small virtual conferences consisting of live demonstrations and presentations, focus groups, and panels. Spectrum can still charge for organizing and special functions. But the cost to everyone would be Much lower, so we might be able to assume that many more companies and individuals would participate.

I don't know how others feel about this but it seems to me that the MV industry has a huge gaping hole, like a missing heart, where Spectrum used to be. There are no other effective industry-wide channels for vendors and consumers to meet. We have no way for end-users to make their requirements known, or for developers to make their offerings known. The MV DBMS companies do their jobs to provide products, but they are unwilling or unable to function as leaders of an industry. They simply don't recognize their responsibility or ability to drive an industry as part of their own corporate plan. They serve as a buffer of isolation between demand and supply rather than conduits. Outside of those companies there is only International Spectrum.

I've been waiting for one of these companies, including Spectrum, to focus on the ecosystem, not just the monthly balance sheet. But it never seems to happen. I have full confidence that any one of these entities could become a true industry leader, not just companies providing products and services to a limited audience. Each has it within their power, within their scope of talent, and in their own financial interest to broker channel relationships which ultimately reward back in terms of brand confidence, sales, and commissions. But none of these entities seems to have the larger vision to put this all together. Each is focused on just their own space, on product features, with a wall up against competitors that also serves as a wall against the user base.

Who is going to grab that brass ring? Certainly not anyone who is reading this and still trying to figure out what it is.

Regards,

T

[eric.arm...@gmail.com]

I think this is a good idea. I miss being able to view the Spectrum conference sessions online.

--

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 on the web visit https://groups.google.com/d/msgid/mvdbms/1fe825ef-346a-48a2-8bdb-246994e20c8fn%40googlegroups.com.

---

This email has been checked for viruses by Avast antivirus software. www.avast.com

[Martyn]

Hi Tony,

Grabbing that brass ring and with the deepest of respect and I respect you highly in our MV community, I cannot agree with all of your comments.

Whilst I entirely agree that we need an independent figure head and that used to be Gus and Monica with IS, I cannot agree that it is the MV DBMS companies alone that need to take a lead.  For the longest time, I personally worked the boards inside and outside of the MV community to push MV as a whole.  I used to spend hours and hours working on a top-down approach that was purely promoting MV to help to sell OpenInsight licenses.  I spoke to whoever would listen outside of the MV world about our fabulous technology but one voice alone is like a drop of water screaming in the ocean.  Mike at Revelation has provided the MV community with data connection technologies within OpenInsight to try to help to bring the industry forward and to enable people to make the most of modern Windows and Web development.  Elkie also tried her hardest to promote MV for a good number of years.  Alas, these are all singular voices and it needs a bigger voice.

With regards to “I've been waiting for one of these companies, including Spectrum, to focus on the ecosystem, …”  I do not believe that it is for the main MV vendors alone.  In fact, several times over the years there has been an attempt to bring us all together.  Working with Revelation Software, I personally tried to build relationships with IBM and Raining Data back in the day to try to market MV as a whole but I was beaten back at every turn.  I also believe that there was an attempt, just a couple of years ago, to try to bring everyone together for the anniversary, but that also fell by the wayside.

You only have to look at the old IS conferences to see how the vendors have always played an isolated game and some (not all) looked to simply poach VARs from other vendors.  I can recall several eye-opening occasions where blatant poaching was openly being performed and it is therefore understandable that the MV companies put walls up.  For this reason alone, I have never seen any of the MV companies coming together to cohesively push the technology.  The trust is just not there and I was reminded of this during the first lockdown.  A so called independent, industry wide Zoom meeting was put together and it was quickly followed by a marketing campaign, which is only going to harm the MV community as a whole.  Not only were my customers emailed by the MV Company but I (a competitor) also got the mailings.  Why would I bring my clients to the party, if the only result is that the other MV Companies are only going to try to poach my customers?  This one event obliterated my trust in the community and I backed off promoting MV almost immediately and reverted back to simply working on promoting Revelation and OpenInsight.  Yes, I became a buffer of isolation.

What you are asking for, has been asked for many times by developers.  However, I firmly believe that it needs a cross industry group, not one or even two of the MV vendors working alone.  It needs someone from Rocket, someone from Revelation, someone from Zumasys and then several independent developers using these technologies and several VARs using these technologies to come together and formulate a plan and execute it.  It then ‘needs’ someone who is ‘independent’ to head up that group and to keep things moving forward in a coherent manner.  I had really hoped that would have been IS and I still hope that, one day, we will all get behind IS (or some other leading organisation) and help to drive things forward.

Assuming that the boat has not already sailed for the n’th time.

We, collectively, have a very strong message to tell and it just needs ‘everyone’ to come together (MV companies, VARs, developers, recruiters, etc.) and to promote MV in general.  Developers can keep hoping that the MV Companies will come together and they can keep reminding us of what needs to be done.  But, without the trust, it is just never going to happen.  Maybe the lesson here is that the MV Companies are the ‘wrong’ people to undertake what you are asking for and we (the MV Community) need to find another way of achieving what you are asking for.

Ohh and for what it is worth, I am currently working with Mike at Revelation Software, on yet another plan to try to take MV out to the wider developer community.  You could say that we have not given up yet.  However, this will be yet another attempt by just one MV DBMS company standing alone and trying (once again) to push our technology outside of the MV community.  Also, you mention about supporting developers more.  You need a place to promote your offerings (Social Media anyone??), make your needs known and to come together as a community.  For the last 30 years Revelation has been successfully doing just for their community, through their website.

So Revelation Software is trying to get the message out, what is everyone (and I mean MV Companies, VARs ‘and’ developers) doing to help to promote MV more widely???

Martyn Phillips
The other one from Revelation Software

[Tony Gravagno]

Martyn - sincere thanks for your well-considered response. We are not in disagreement of any sort. That I did not say some specific words does not mean I was exclusionary or myopic. I said I believe the MVDBMS leaders have the ability to do better. That doesn't mean I don't believe others can participate or lead, or that they should do so in independent initiatives. I said I believe Spectrum can also accept the role. That's not to say some other entity cannot. I was talking about what's happening Within the MV industry as it is now. That doesn't mean there aren't other initiatives that are being pursued or that should be pursued to expand the field.

To be specific : International Spectrum, Rocket Software, and Zumasys, each have databases with a huge number of developers and current end-users. Each of these companies use their data to direct sell via their list. None of them communicate to end-users about the wealth of products and services available in the industry which are not directly sold by themselves. Joe Programmer doesn't have a way to offer services to ACME EndUser. ACME might be in discussion to leave the platform because they see no evidence of the kind of ecosystem that binds other platforms. They have no place to go to find a platform-agnostic resource for all-things Pick.

I have created such a resource - as yet unpublished. We need more resources and ways to communication information about them to end-users. At this moment the only way to broadcast information about industry resources like this is through these few forums with limited membership. Messages we need to convey are unseen by the people who need the information. I bring up this topic in a thread about Spectrum because Spectrum is a platform-independent entity, unlike those above, Revelation, and others. I don't think Revelation has the database to do this kind of reach out, Spectrum does.

There are subtleties and caveats to initiatives like this. Industry building can be done without "giving away the store" in terms of sending end-users to competitors. It can be done in a way which respects the channel, existing application providers, and existing support providers. I've talked for years about partnerships. I've encouraged application developers to reach out to the industry to collaborate to provide solutions that will endear their end-users to the platform. That doesn't happen for the same reason that the MVDBMS companies do not reach out - it's fear of competition. People need to see the benefits of such efforts and not feel threatened by them.

The alternative, as we've been witnessing for decades, is a slow attrition of end-users from the platform, increasing DBMS prices to make up the loss, and DBMS companies that create new products because they aren't selling enough DBMS licenses - until at some point they decide to sell off the DBMS assets entirely. How's that workin for all of us?

Best,

T

[Donald Montaine]

In the SQL world, the DBMS vendors act in many ways just like the MV DBMS vendors.  However, the difference is, that there is an entry point where developers can become familiar with SQL technology without cost. The open source versions of MySQL, MariaDB and PostgreSQL (or even Express offerings such as that provided by Microsoft) provide the entrance into the world of all things SQL.  The fact that there is no equivalent entry point into the MV world means that SQL will always have a leg up.  As long as all the MV vendors refuse to provide an equivalent entry point into the technology nothing will change.

In today's world, use by small users and new idea development must be free.  Now small can be defined in many ways.

Microsoft defines it as follows:  1GB maximum memory used by the database engine, 10GB maximum database size, 1MB maximum buffer cache, CPU the lesser of one (1) socket or four (4) cores (number of SQL user connections NOT limited)

MySQL puts the free/paid barrier between GPL and non-GPL use or the need to use certain enhancements which are not available cost-free.

PostgreSQL has no barriers to use, but depends on large commercial users needing paid support and commercial add-ons.

Other organizations such as PC Soft in France provide a robust, free database for users of their development tools without any distribution overhead.

The bottom line is that the SQL world offers many different alternatives for smaller developers and users to implement ideas without an upfront cost burden. Some of these implementations will scale and need various options that are only available with an associated cost.  Not only does the MV world not offer a variety of cost-free entry points, it doesn't even offer ONE. (Note here that I am not necessarily arguing for a GPL entry point.  The SQL world offers both a cost-free commercial option - MSSQL Express, and a variety of GPL/Commercial dual-license options).

All the marketing in the world, and all the attempts to expand awareness of MV in the mainstream community will be useless until MV provides an entry point that is competitive with those provided by the SQL world (as well, it should be noted, by the other NoSQL technologies).  All the MV DBMS vendors cling to a cost and marketing model from the 1980's.  The DBMS world has changed and it is no wonder that, for most organizations implementing new ideas, MV is not even in the consideration mix.

[Tom Marracci]

D3 does allow you to install for a single user as port 0 without a license. There are some other limitations (no flash, no phantom sockets, no osfi, no shared printers) but for learning Basic and how the database works, it is free and for all intents and purposes a production environment.  The only thing Rocket has not done is make the downloads freely available. But if you can get a copy, you can install and run it.

 

Tom

 

----- Original Message -----

From: Donald Montaine

To: Pick and MultiValue Databases

Sent: Thursday, March 25, 2021 8:36 AM

Subject: Re: [mvdbms] International Spectrum

--
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 on the web visit https://groups.google.com/d/msgid/mvdbms/74c1a9ae-7071-454e-957b-0a77cd2d507an%40googlegroups.com.

[Donald Montaine]

Hi Tom - compared to what is available in the SQL world, the "free" offerings in the MV world are pathetic, if you can even find them and navigate the barriers to use.  Yes, you can get a lobotomized version for one or two users to use in learning.  But there is nothing equivalent to the ease of implementation and distribution during the start-up phase offered in the SQL world.  MSSQL; download, install at an unlimited number of clients as long as the resource requirement don't exceed the MS limitations.  PostgreSQL; download and install at an unlimited number of clients without cost.  No licensing, no hoops to jump through, etc.  

[Tom Marracci]

I agree with you that entry into SQL is easier. I have Postgresql and Mysql running in several places all for free as you noted.  I do development in a Virtual Box at home with a "free" single user D3, but as you say, getting it was not freely available.  I only have a copy because I already had one.

 

Tom

 

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/5fcc1a43-f136-4458-bfd5-6af1142aa2c1n%40googlegroups.com.

[Donald Montaine]

So, what would it take for a free MV install to be competitive  with MSSQL, for example.

1) it wouldn't have to be GPL and could be closed source

2) it should allow easy install by downloading an installer

3) It should have no limitation on redistribution or number of users

4) It should allow databases of up to 5 GB ( half of MSSQL express)

5) It should be limited to 2 cores (half of MSSQL express)

6) It should be fully functional, with all the capabilities of the base "paid" option

7) it should not require registration

8) community support should be available through an online forum

9) upgrade to a paid version should be easy

10) commercial support should be available for a fee

11) ideally it should be available to 3rd party front end developers to bundle with
       a free version of their software with similar limitations or be bundled with 

       front end developer tools from the same vendor

12) it should come with a front end management tool such as Microsoft SQL Server Management Studio

12) it should run on Windows and Linux

If any MV DBMS vendor is serious about competing for new business, rather than just acquiring existing MV users from another vendor, this is what would be required to possibly be of interest to start-ups, small developers or hobbyists.

[geneb]

On Thu, 25 Mar 2021, Donald Montaine wrote:

> In the SQL world, the DBMS vendors act in many ways just like the MV DBMS
> vendors. However, the difference is, that there is an entry point where
> developers can become familiar with SQL technology without cost. The open
> source versions of MySQL, MariaDB and PostgreSQL (or even Express offerings
> such as that provided by Microsoft) provide the entrance into the world of
> all things SQL. The fact that there is no equivalent entry point into the
> MV world means that SQL will always have a leg up. As long as all the MV
> vendors refuse to provide an equivalent entry point into the technology
> nothing will change.
>

ScarletDME is apparently a very well kept secret. :)
https://github.com/geneb/ScarletDME


g.

--
Proud owner of F-15C 80-0007
http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.
Some people collect things for a hobby. Geeks collect hobbies.

ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.
http://scarlet.deltasoft.com - Get it _today_!

[Donald Montaine]

Scarlett DME does not even come close to being competitive with what the SQL world offers as a free entry-level version.  It is a very old version of OpenQM, is GPL and limited by Zumasys's idea of what the GPL allows, is only 32 bit, requires a lot of knowledge to install, does not install easily on many current Linux distros and is only available on Linux.  It has no active user community. There are a few git commits 13 months ago, but mostly the last commits were 9 years ago.

[Will Johnson]

Donald do you not consider this forum itself to be "community support offered at a forum" ?

[geneb]

On Thu, 25 Mar 2021, Donald Montaine wrote:

> Scarlett DME does not even come close to being competitive with what the
> SQL world offers as a free entry-level version. It is a very old version
> of OpenQM, is GPL and limited by Zumasys's idea of what the GPL allows, is
> only 32 bit, requires a lot of knowledge to install, does not install
> easily on many current Linux distros and is only available on Linux. It
> has no active user community. There are a few git commits 13 months ago,
> but mostly the last commits were 9 years ago.

It installs on current releases Ubuntu, Debian, and CentOS (with the
needed 32 bit libs installed). It also works in docker containers and
MacOS. It only requires "a lot" of knowledge to install if the person
doing the installation can't read English very well.

By sheer file count, "most" of the commits were within the last 10-13
months.

You have some valid points, but those are overshadowed by what you think
you knew, or by erroneous assumptions based on outdated knowledge.

While I agree that it's got some outstanding issues, your sneering
dismissal of it just pisses me off. However, I know it's much easier for
you (and those like you) to take cheap shots from the sidelines than to
get off your assess and contribute something other than snarky
denigration.

There does exist a good open source multi-value database system. Just
because it doesn't meet your criteria doesn't invalidate that fact.

You're welcome to contribute towards making ScarletDME 64 bit (it's 99%
the way there now) or to make it more easily installed, or maybe run on a
*BSD or Windows host. Writing docs your thing? I'll happily add whatever
you contribute to the git repo. Not so hot in C, but you've got a good
grasp of the MV dialects of BASIC and want to add features that way? Send
in your code or submit a pull request. It doesn't even have to be pretty
- if it does what it says on the tin, I'll accept it. Hell, if *anyone*
reading this wants to contribute in any manner, I'll welcome you with open
arms. I'm one guy with way more on my plate than I have lifespan left to
handle, and I'm certainly not the genius that Martin is. I'll take help
with ScarletDME from anyone with a pulse (and hey, if you're undead, I'll
take you too, but keep your fingers out of my brains until *after* I'm
dead.)

[Donald Montaine]

It is not my intent to start a flame war.  If I came across as minimizing the work that went into Scarlet DME or it's usefulness in some situations, I am sorry.  I was focusing on one thing and one thing only - is there anything in the MV world that completes effectively with the current entry-into-the-technology offerings in SQL land (and other NoSQL databases land).  It is my belief that on many levels nothing in the MV world does compete considering many factors including: price, liberal licensing, ease of access, documentation, size and activity of the developer community, size and strength of the user community, organizational or corporate support, currency of release, operating system support, etc.  I strongly believe that the typical startup entrepreneur evaluating for example,  MSSQL Express, PostgreSQL, MariaDB, Couchbase, MongoDB and Scarlet DME (or Zeobase); most evaluators would not place either of those two MV alternatives near the top of their list.

[geneb]

On Fri, 26 Mar 2021, Donald Montaine wrote:

> It is not my intent to start a flame war. If I came across as minimizing
> the work that went into Scarlet DME or it's usefulness in some situations,
> I am sorry. I was focusing on one thing and one thing only - is there
> anything in the MV world that completes effectively with the current
> entry-into-the-technology offerings in SQL land (and other NoSQL databases
> land). It is my belief that on many levels nothing in the MV world does
> compete considering many factors including: price, liberal licensing, ease
> of access, documentation, size and activity of the developer community,
> size and strength of the user community, organizational or corporate
> support, currency of release, operating system support, etc. I strongly
> believe that the typical startup entrepreneur evaluating for example,
> MSSQL Express, PostgreSQL, MariaDB, Couchbase, MongoDB and Scarlet DME (or
> Zeobase); most evaluators would not place either of those two MV
> alternatives near the top of their list.

ScarletDME *could* be a good "nosql" contender in that space with just a
little more work. I'll agree that it's current state isn't "newbie"
friendly and the documentation could be loads better. The problem is that
I can't do it all myself - I just don't have the time. Someone on the
list did reach out so something may happen sooner than later. I do have
it building and running under Ubuntu 20.04 on WSL2 (Windows Subsystem for
Linux), but still 32 bit.

ScarletDME is the /closest/ thing there is to the goal you'd like to
achieve and I'd love to see that happen.

[Donald Montaine]

I would love to see a competitive entry level option in the MV world.  I do see some real barriers.  1) there is no organizational support (ie: Zumasys, Rocket, Revelation, Northgate).  All the other alternatives I mentioned either have a big company or a substantial non-profit supporting them. 2) Scarlett DME is GPL only, most of the others are dual licensed and can easily migrate from pure open source to commercial easily.  3) Two SQL products, MSSQL Express and PostgreSQL. can be used for small business commercial apps without any licensing issues and PostgreSQL obviously has extensive ability to scale as Amazon replacing Oracle with their own version of PostgreSQL can attest.  4) The other options mentioned are all current technology with regular updates.  Scarlett DME is missing 9 years of substantial improvements available in the commercial QM version.. 5) and it doesn't help that the parent of Scarlett DME is, in many ways, antagonistic (or maybe just uninterested)  about it's existence. 

It also doesn't help that the MV vendors are locked into a per seat / yearly maintenance fee licensing model for their commercial offerings and their attitude regarding marketing their products is problematic.

I do not question that having a Scarlett DME alternative in the MV world might be attractive to some current MV users (but not if they want to sell commercial apps).  However, I don't think that many from outside the MV community would be enticed, given the limitations mentioned above. (Note that I would love to be proven wrong).

[geneb]

On Sat, 27 Mar 2021, Donald Montaine wrote:

> I would love to see a competitive entry level option in the MV world. I do
> see some real barriers. 1) there is no organizational support (ie:
> Zumasys, Rocket, Revelation, Northgate). All the other alternatives I
> mentioned either have a big company or a substantial non-profit supporting
> them. 2) Scarlett DME is GPL only, most of the others are dual licensed and
> can easily migrate from pure open source to commercial easily. 3) Two SQL
> products, MSSQL Express and PostgreSQL. can be used for small business
> commercial apps without any licensing issues and PostgreSQL obviously has
> extensive ability to scale as Amazon replacing Oracle with their own
> version of PostgreSQL can attest. 4) The other options mentioned are all
> current technology with regular updates. Scarlett DME is missing 9 years
> of substantial improvements available in the commercial QM version.. 5) and
> it doesn't help that the parent of Scarlett DME is, in many ways,
> antagonistic (or maybe just uninterested) about it's existence.
>

3 and 5 could be handled if someone can convince Zumasys that their
interpretation of the GPL is incorrect with regard to the output of the
QMBASIC compiler...

#4 can be handled reasonably as well - I've got a printed list of all the
public-facing updates to OpenQM since the original GPL release.

> It also doesn't help that the MV vendors are locked into a per seat /
> yearly maintenance fee licensing model for their commercial offerings and
> their attitude regarding marketing their products is problematic.
>

This is very true.

> I do not question that having a Scarlett DME alternative in the MV world
> might be attractive to some current MV users (but not if they want to sell
> commercial apps). However, I don't think that many from outside the MV
> community would be enticed, given the limitations mentioned above. (Note
> that I would love to be proven wrong).

I would love to be able to prove you wrong, but alas, your points are very
valid and correct. Maybe Zumasys could be convinced that ScarletDME is no
"true" competitor due to the development lag and they would be interested
in either relaxing their interpretation of the GPL or granting an
additional LGPL license to it. I would love to see an uptick in new MV
users!

[Donald Montaine]

An additional sad side note is that such an entry level MVDBMS would not even need to be open source.  If it was a commercial product licensed like MSSQL for small users and app developers, it would serve the same purpose.

[geneb]

On Sun, 28 Mar 2021, Donald Montaine wrote:

> An additional sad side note is that such an entry level MVDBMS would not
> even need to be open source. If it was a commercial product licensed like
> MSSQL for small users and app developers, it would serve the same purpose.
>

The problem is prying them away from the per-seat model...

[Marcus Rhodes]

Granted, it has a few bugs, but I find the 32-bit version of ScarletDME (which means what, by the way?) offers the community most of what's being proposed.  After all, 32-bits was enough for commercial/enterprise systems just a decade or two ago; it should be enough for hobbyists now, and even for modest enterprises.  It's certainly adequate for prototyping, even if that's nothing more than ensuring the code compiles and runs.  It's what I developed AnyKey and Vignettes on before installing them to customers' (UV & jBase) systems for further development and testing, and it worked great for that.

Those bugs, though.  Gotta go.  Maybe it would be easier to just upgrade Scarlet to 3.0 (32-bit, of course).  Maybe Martin/Zumasys would even be willing to help?  Just a thought, but has anyone asked?  It is, after all, the vendors who've got to do this, and, so far, to me, at least, ScarletDME is proof that OpenQM has done the best job of this.  I mean, clearly, there's some demand there.  I know I keep getting asked to produce more.

That said, training videos seem to me to be the brass ring.  (And what I get asked about the most.)  It should be possible to send new programmers a youtube playlist to just start watching: one playlist for generic Pick, plus one more for each of the flavors and their peculiarities.  They could (and should) be improved over time, of course, based on viewer feedback, just like Pick itself has been.  In time, we could hone the training to the point that any 1st year CS student could become proficient with Pick in a weekend, and expert in a month, maybe even a master in a year.  That would be HUGE!  How long did it take you to master Pick?  I'd like to think I mastered it in just a few years, but I find that I'm still learning things.  Just learned about LIST ... FMT "L30" a few months ago.  (But, then, how many UV jockies know about LIST/COPY server:account,filename,dataportion?)  How I would have loved some good videos.  THAT is where we'll strike real gold.  Like someone said earlier, it's the training and support that's really needed.  Firms will pay well for it.  Individuals, hobbyists, tourists, whatever you want to call them, may not pay much, or anything, but, as WordPerfect proved 30 years ago, just getting your product out there builds the market.  That's why Apple and M$ started fighting each other to give schools their systems for free.  Now look at them both.  That's what we need to be doing.  Except that, now, instead of getting into schools, you have to get on youtube, which means that a free, personal edition has to be readily available.  And, right now, only ScarletDME stands in the gap.  And it exist only because of Martin/OpenQM.

But, then, that brings us to another matter, and that is the why.  Why would anyone even consider Pick over SQL?  And is that even our best target?  I've mentioned before that I use ScarletDME as a scripting language with which to manage my systems instead of, say, Python.  (Which I use quite a bit at work, but only for xlsxwriter.)  So we could position it, not only as a database, but also as a scripting language alternative.  One of the programs I wrote for myself, for example, generates playlists of videos I've downloaded in a format that VLC can run with.  I wrote another to merge directories, only replacing where the one being overwritten is either older or smaller.  (Sure help with a  problem I've seen where copied files end up with a size of 0.)  Aaannnd, I can even use Zenity to throw in some GUI pieces.  And Zenity (as well as a number of alternatives) is even available on Windows.  So, maybe Martin/Zumasys might consider packaging them, too, along with some ready-to-run applications.

Ok, ok, ok, while I'm at it ... we could really use some applications, too.  I mean, I know, all the hosts already have them, but, instead of just WED, or trying to con Pick programmers that VS will work they way they work, why not a feature-rich code editor, one that has syntax-highlighting specific to Basic, PROC, dictionaries, and data?  It could even offer hooks for any custom code you might want, like shop-specific formatting requirements, mod-logs, etc.  It could look something like Microsoft's last version of their text editor.  Wouldn't that be cool?

And what about a spreadsheet-like program that would let a user cruise any file in the system?  It could use whichever existing dictionaries the user selects, or even allow the user to create new ones on-the-fly.  'Views' could be saved as reports, and run automatically.  How about that?  Why has no vendor done that?  Think something like that would sweeten the pot?  I do.

Yeah.  There's a lot that can be done.

Martin has contributed.  Gene has contributed.  I've contributed.  Who's next?  What are you gonna do.  Anything.  Anything at all.  How about testing?  Has anyone tested AnyKey?  Have you even downloaded it?  Let's get busy!

[Donald Montaine]

Actually, I think that Revelation Technologies OpenInsight would be the best offering to get ousiders interested in the world of multi-value;  A limited version with only access to a size limited version of their native MV database (no foreign MV or SQL connectors).  It is Windows based and is the most similar to the Visual Studio Express/MSSQL offering,  It provides both front and back end development in a single package.  Not that I have a lot of hope, but if they were really serious about providing an offering truely competitive with VS/MSSQL Express, I think it would give the MV world a chance to win some hearts and minds.  But their current pricing model with a yearly cost of $1000+/yr to join their Works program + a yearly per seat cost to anyone you sell a multi-user app to is really a buzz-kill when compared with the zero cost of entry and zero distribution cost of the Microsoft offering.

[Marcus Rhodes]

And I've used OpenInsight, and liked it very much, except for what I perceived as its two major flaws: 1) OO, along with the interesting gyrations required by it in their instantiation (pardon the pun) (do they still require you to 'promote' objects?), and 2) no Linux version.  In fact, I vaguely recall Revelation openly denouncing Linux and its user community because 'they want everything for free', so it seems unlikely that Revelation will ever plug that gap.  And, yes, while most users run Windows, I'm pretty sure that any serious hobbyist, or anyone else likely to delve into Pick's depths, is going to have some proficiency with Linux, and may even prefer it to Windows, at least for server-based projects.  Besides, I think it's also less likely that a casual explorer will want to RTFM on OpenInsight just to experiment with some personal projects.  I mean, they might.  After all, when I first saw VB, I immediately bribed the guy into selling it to me, took it home, worked through the tutorials, and I was on my way.  But that's me.  AND that was the quality of the tutorials, which comes back to my comments on instructional materials.

[Will Johnson]

For entry-level, I don't think you could go wrong using the *free* single-user version of Universe.

One of it's strong points, is that it lets you play with Pick flavor, Universe flavor, Unidata flavor....

And it has one of, if not the largest, existing user-base

[antlists]

On 28/03/2021 03:11, geneb wrote:
> 3 and 5 could be handled if someone can convince Zumasys that their
> interpretation of the GPL is incorrect with regard to the output of the
> QMBASIC compiler...

On my list of projects, when I get a round tuit. A BASIC compiler. If we
replace the compiler in ScarletDME, we'll have a fully properly GPL
system. And personally, given the behaviour of BASIC and the GPL, I
personally don't see a problem with building commercial systems on a
GPL'd MV system.

Imho, it's the nature of the system that keeping the GPL'd MV layer and
the commercial application layer separate is dead easy. So in practice,
GPL, LGPL, MPL, there's practically no difference.

Cheers,
Wol

[geneb]

On Mon, 29 Mar 2021, antlists wrote:

> On 28/03/2021 03:11, geneb wrote:
>> 3 and 5 could be handled if someone can convince Zumasys that their
>> interpretation of the GPL is incorrect with regard to the output of the
>> QMBASIC compiler...
>
> On my list of projects, when I get a round tuit. A BASIC compiler. If we
> replace the compiler in ScarletDME, we'll have a fully properly GPL system.
> And personally, given the behaviour of BASIC and the GPL, I personally don't
> see a problem with building commercial systems on a GPL'd MV system.
>

The issue isn't the GPL - the issue is that they think that the QMBASIC
compiler output is creating a "derivative work" as the output, which is
basically bonkers. :(

[RDG Information Systems]

The issue isn't the GPL - the issue is that they think that the QMBASIC 
compiler output is creating a "derivative work" as the output, which is 
basically bonkers. :(

It’s actually not as bonkers as it sounds. The issue is that sufficient bits&bobs of the libraries are included that the compiled code is subject to the GPL.  It’s the same with the gcc compiler.  The difference is that gcc provides an explicit exemption for compiled code:

• https://www.gnu.org/licenses/gcc-exception-3.1.en.html
• https://www.gnu.org/licenses/gcc-exception-3.1-faq.html
• https://en.wikipedia.org/wiki/GPL_linking_exception

This is further compounded by the fact that compiled mvBASIC (with the exception of jBase) is not machine-native executable code.  It’s byte code that needs an interpreter/virtual-machine — like Java

v/r

Ray
-- 

Raymond DeGennaro
Owner & Lead Developer
RDG Information Systems, LLC
Raymond....@RDGInfoSys.com
http://RDGInfoSys.com
+1-815-308-0815

[geneb]

On Mon, 29 Mar 2021, RDG Information Systems wrote:

>> The issue isn't the GPL - the issue is that they think that the QMBASIC
>> compiler output is creating a "derivative work" as the output, which is
>> basically bonkers. :(

> It’s actually not as bonkers as it sounds. The issue is that sufficient
> bits&bobs of the libraries are included that the compiled code is
> subject to the GPL.  It’s the same with the gcc compiler.  The
> difference is that gcc provides an explicit exemption for compiled code:
>

That's just it, there ARE no libraries! It's compiled down to p-code
which is then interpreted. While I haven't dug through the compiler in
painstaking detail, I'd be astonished if it was embedding subroutine code
(like iconv() for example) in the compiler output.

[antlists]

On 29/03/2021 21:19, geneb wrote:
> On Mon, 29 Mar 2021, antlists wrote:
>
>> On 28/03/2021 03:11, geneb wrote:
>>> 3 and 5 could be handled if someone can convince Zumasys that their
>>> interpretation of the GPL is incorrect with regard to the output of
>>> the QMBASIC compiler...
>>
>> On my list of projects, when I get a round tuit. A BASIC compiler. If
>> we replace the compiler in ScarletDME, we'll have a fully properly GPL
>> system. And personally, given the behaviour of BASIC and the GPL, I
>> personally don't see a problem with building commercial systems on a
>> GPL'd MV system.
>>
>
> The issue isn't the GPL - the issue is that they think that the QMBASIC
> compiler output is creating a "derivative work" as the output, which is
> basically bonkers. :(
>

I agree. But if we swap out the compiler. what they think becomes
irrelevant.

Cheers,
Wol

[Donald Montaine]

One reason that I suggested Open Insight was that the do have connecters to QM, U2, etc - so you can use them as the front end and your favorite Linux MVDBMS as the back end once you grow past the Windows only world.  As far as OO, that is what is expected nowadays, except by us fossils.  But I do agree that, like most MV vendors, they don't seem to be willing to play today's software game.

On Monday, March 29, 2021 at 11:16:58 AM UTC-7 Marcus Rhodes wrote:

[Donald Montaine]

Sadly, what MVDBMS vendors remind me of is the last retailer in an empty mall that doesn't bother with a web store because they are just sure that customers will come to their senses and flock back to in person shopping.  

Yes, I agree that there are some options if someone wants to learn about MV technology in a limited setting.  But there is no way to start a business that is as convenient and cost free as that offered by many of the other database tech communities. They hook you and then get paid when you need a more robust database engine, or attend a conference that they sponsor, or contract for training, etc.

[joseba]

I do think that mv database providers can make more for the market. But I do not think the biggest problem is to have a "free copy" or something similar. I think it is much more about "standards" on the database side. And a very important thing, the companies that work with mv and develop software for MV. I have the feeling that many of them just want to work as 30 years ago but the database vendor "magically" converts his old apps into "new technology". There are many tools to modernize applications. I see a lot of "mv on the cloud", and it is the same old infrastructure having the machine "somewhere" but just using telnet or ssh and some "intelligent" printing services. For instance, when you talk with some people about "optimistic locking" they still see you like a "mad man". We have put in the market a product for development with many mv databases (D3, mvBase, Open QM, Reality, Universe, Unidata) in almost any language, the environment you want, etc. You can install it in Windows, Linux, macOS. You can program in .Net, Java, PHP, Python, javascript, typescript, Angular, React, Vue, easily calling mv basic subroutines, you can use VSCode for all the development (you can program in the same environment in VSCode in MVBASIC and Typescript to create an IONIC mobile app having an mv database in the backend. And the framework is open-source, so you can even modify the libraries to improve them. We have a lite version that is FREE and works perfectly with one session, and the licenses are per "session", not per user at a cost of less than $85. And people are using it, the same people that use other tools many years ago with mv. But many people think that a product like ours must be free and must be included with the database. We talk a lot about "SQL databases" but we do not talk about "develop environments".  They want to have "Evoke" or something similar, for free, so they "evolve" without the effort in time and money for learning how modern development works. This is just MHO.

[Martyn]

Hi Marcus,

Revelation did have a Linux version of OI way back when (called OIL) but it never really took off.  Many people wanted to be able to tick a checkbox on a sales proposal but it was never really used in anger and fell by the wayside.  

With regards to having to RTFM, I had reading manuals and prefer to keep those a reference resources,  I have therefore personally already published three OI tutorials on YouTube, so free for anyone to get access to.  These covered OI9.x (Desktop), O4W 9.x (Web) and more recently I updated the desktop tutorial for OI10 desktop.  I had plans for updating the O4W tutorial videos but Mike and I are currently working on something very different and much,  much better.  Watch www.revelation.com or my OpenInsight blog for more details coming in a few weeks time.

In the meantime, my door is always open to anyone and I am more than willing to spend time on TeamViewer with anyone working with the evaluation.  I am semi-technical (not a programmer) and more than willing to take questions and get back to people with answers. 

M.

[Marcus Rhodes]

Martyn!  Thanks so much for chiming in.  And thank you for that intriguing information.  ("used in anger" ... LOVE IT!)  Super busy right now myself with projects, but, on your word, I may just download the demo and try it out.  It *has* been quite a while, after all.

[Will Johnson]

You *can* start a business with the free version.  Just one user, you.

If you are so successful, that you need to hire five more people, then you can afford support at that point, can't you?

*Or* you could get five more free versions, one each, and then write programs that export and import files on the fly or something like that, with intra-maching locking schemes.

It's possible, I've done it.  It takes a couple hundred hours of coding and debugging and ......

Or just pay for support.

[Donald Montaine]

Yes, you "can" start a business, yes there a modern front end tools available.  That is not my point.  The question I raised is what is needed to be competitive in a buying decision.  I was a multi-value user and developer for 25 years.  Systems I devoloped in the state of Washington were in use by governments in every county in Washington for at least 20 yrars after their introduction and for all I know they may still be in use. 

I don't use multi-value any more, not because of the technology, but because of the licensing.  I pay over $1000 for my development tool every year. (I don't have to, I could use an older version.  But there are new features provided every year that I find valuable).  I don't mind the yearly licensing cost.  Why, because the embedded client-server database that is included is free for distribution.  It is not open source, but my developed apps do not have a vendor "tax" added on to them.

In many cases I would have preferred to use a multi-value back end, but I stopped because ALL the MV vendors use similar licensing.  By the way, my development tool would work with a multi-value back end.  I have done a proof-of-concept project to verify that.  But I have never found a situation where subjecting my clients to a 3rd party yearly fee was necessary. 

In fact if RTI had offered a similar license for Open Insight, I would have never stopped using it and they would have been the company that received my yearly payment for use of their development tool rather than the company i work with now.

In the non-MV world I have choice.  I can do full open source,  I can choose Postgresql with no limitations on use, or MSSQL with a limitation once a certain database size is reached, or I can choose a tool like I use now where I pay for the tool, but there are no ongoing costs required from my clients to a third party.  I can do SQL or no-SQL with a similar variety of licensing.  I do not have those choices available to me in the MV world.

There are many reasons why entities leave the MV fold and it isn't usually as simple as MV vs SQL.  But the bottom line is that more entities leave than join each year and the community continues to shrink.

[Will Johnson]

However, *no one* looking to start a business, would ever pick an MV database, *except* retired MV developers.

So there's that.  You have to overcome that *first* before you go looking for other reasons.

In fact, I would say 99% of all people trying to start a business, start with like... a dozen web pages, and go from there.

The database itself becomes more of a "who do I know who knows computers and what do *they* recommend"

I don't think you can overcome it.

What are they teaching undergrads about databases and which one to choose.

That's the full problem right there.

[Marcus Rhodes]

But, as I keep saying, youtube is the way.  Moreover, Pick has some big advantages over other systems, and NoSQL is just one.  I, personally, loathe OO.  I know.  I know.  I did it for years.  I even wrote TSRs.  (Remember those?)  But you'll find a lot of youtube videos critical of OO.  And I've since seen the light, too.  So, Pick has NoSQL AND NoOO going for it.  But that's not all, either.  I've also come to be a huge fan of 'everything`s a string' (except for file handles, lists, etc., but they don't count), so that gives Pick NoSQL, NoOO, and NoVariableTyping.  (Veni ad Wirt sepeliendum, non ad laudandum.)  Pick is also very tolerant of formatting curiosities (unlike, say, Python).  What can we call that?  Forgiving?  Flexible?  Format-free?  So that's four things.  And then there's Pick's excellent structuring, right down to the little-known/little-used PROGRAM keyword.  External and internal subroutines offer near C-like scope-control.  Anything else?  Oh, yeah, integrated database.  Or would that come under NoSQL?  And then, of course, the one I keep harping on: Pick makes an even better scripting language than things like Python.  Even things like xlsxwriter and panda can be used from Pick almost as if native-born.

Anyway, my point is that Pick has a lot of advantages over other systems.  We just need to beat those drums loud and long rather than falling back on 'legacy' or 'heritage' or 'enterprise' or even proven, stable, etc.  (What version of Python are we on now?)  Make the hobbyists curious about it.  The thing we're really up against is that there is some seriously big money promoting those other technologies.  I used to think that my friends who kept insisting that facebook was really conceived and built by the CIA to spy on us were just more UFO goofs, but, now that it's been all but confirmed (and not just facebook, by the way, but even IBM, Xerox, Microsoft, Google, Amazon, Oracle, ... i.e.: 'Big Tech', all had help from one or more of the TLAs) (or other 'state' powers), it casts a completely different light on things.  I'm even questioning Linux's success.  Someone's paying all those people to write all that software.  That's what we're up against.  Apple's old hammer-thrower ad may once have been legit, but, if so, they sold out long ago.  Now, it's WE who are the real underground underdogs, not them.  We are the real grassroots movement, not Mozilla.  And I, for one, will not go gentle into that good night.  Pick is worth fighting for, and watching my son do his job (web-dev for one of those TLAs) has only cemented that position for me.

[Donald Montaine]

Hi Marcus - I agree with much of what you say about the benefits of MV.  But the rest of the world is playing basketball while MV keeps trying to set up a ping-pong table in the middle of the basketball court.  Youtube may be the way, but not if a developer meets a 40 old licensing scheme that is unlike any other entry level database.  If I have a new idea that requires a database, why in the world I saddle my product with yearly per seat licensing right off the bat.  At least with  Microsoft,  my client doesn't have to face that hurdle unless they are fairly big.  And, if I choose PostgreSQL, like Amazon did, my client doesn't have to face it at all.

[Kevin King]

Donald, EXACTLY.  The licensing is killing the market.

On Wed, Mar 31, 2021 at 3:59 PM Donald Montaine <dmon...@gmail.com> wrote:

Hi Marcus - I agree with much of what you say about the benefits of MV.  But the rest of the world is playing basketball while MV keeps trying to set up a ping-pong table in the middle of the basketball court.  Youtube may be the way, but not if a developer meets a 40 old licensing scheme that is unlike any other entry level database.  If I have a new idea that requires a database, why in the world I saddle my product with yearly per seat licensing right off the bat.  At least with  Microsoft,  my client doesn't have to face that hurdle unless they are fairly big.  And, if I choose PostgreSQL, like Amazon did, my client doesn't have to face it at all.

--
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 on the web visit https://groups.google.com/d/msgid/mvdbms/1a15cfbd-29de-483c-be32-645209b60af9n%40googlegroups.com.

[Will Johnson]

Ahhhh ha ha ha ha ha ha ha

I think when I had a commercial four-process Universe license I paid something like a hundred dollars a year I can't remember

It was not at all significant in the scheme of bringing in thousand of dollars a month.

My furniture rental expense was much much higher.  Just to name one thing.

[Donald Montaine]

Currently, a four user QM license is about $500 upfront and $130 / yr.  I believe that the UX products are about 4x that amount, but I haven't checked prices in years.  It is not a large amount, agreed.  But it is extra paperwork and process that is not required with alternatives.  So I have a choice; with PostgreSQL I can sell a product to my customer and whatever I charge, they only have to deal with me.  WIth MV, the best case is that they deal with me and Revelation Technologies.  With QM, they deal with me, with Zumasys, and with, as an example,  Evoke for a front end, etc.

And the licensing for MV is not simple.  It is usually seat based, not server based.  I stopped using Microsoft servers because of their insane, complex licensing.  I stopped using MV for much the same reason (although Microsoft is MUCH worse).

My current WIndows based development environment has the ability to create both WIndows and Linux apps.  My customer deals only with me and don't ever have to be involved with my software vendor.  However, for their web development software, they use server based licensing, which I dislike, but at least it is server based and does not require a subscription unless you need newer features.

I see my function as making software as simple as possible for my client.  Licensing complexity is a part of the evaluation as to simplicity of implementation and maintenance.

[Mike Wooding]

Firstly the disclaimers, caveats, etc.

I am the wrong side of the pond to proffer what I would consider to be a definitive opinion.  So, what follows is worth no more than you're paying for it, and possibly even less than that.  It should not be considered professional legal advice nor should it be relied upon.

Right, now that's out of the way, I would respectfully point any interested parties to the following references.

The Compendium of U.S. Copyright Office Practice [1] is clear on the matter of copyright in object code as paragraph 721.8 example four demonstrates:

"Shell Games LLC submits two applications for the same computer program, one specifically for the source code and the other for the object code. Because there are no copyrightable differences between the source code and the object code, there is no basis for issuing a separate registration for each representation of the program. Moreover, if the object code was created by a computer program, there would be no human authorship in the object code, and no authorship that is distinct from the source code. The registration specialist will register the claim in the source code, and reject the claim in the object code."

In short, when object code is created by a program with no distinct human authorship between the source and object code, the source code is copyrightable, whilst the object is not.

There aren't many examples where this has been tested in court but Design Data Corporation v. Unigate Enterprise, Inc., 14-16701 [2] came very close and is a good example of the American courts using the abstraction-filtration-comparison test (See Computer Assocs Int'l v Altai [3]).

Not sure if this helps, or merely muddies the waters further but at least there are suitable references here as a basis for further research if this is deemed beneficial.

Regards

Mike Wooding

[1] https://www.copyright.gov/comp3/chap700/ch700-literary-works.pdf

[2] https://ttlfnews.wordpress.com/2016/12/03/when-is-the-output-of-a-copyright-protected-software-program-itself-protected-by-copyright/

[3] https://scholar.google.com/scholar_case?case=6976925648486076739&q=Computer+Assocs.+Int%27l+v.+Altai+&hl=en&as_sdt=40006

[Will Johnson]

Not sure what you are speaking about.

With Universe they deal with you, with Unidata they deal with you....

What are you saying?

It's *always* the case that the MV vendors, just like any front-line vendor, have licenses with software development companies and all "customers" deal with those end points, not the front points.

[Martin Phillips]

Re the comments in this thread concerning the open source version of QM, here is a statement that was on the open source description page...

The terms of the GPL require that, if you modify the modules that form the open source release in any way, you must also release those modifications in open source form under the same licence. You must not remove the copyright notice embedded in every source code module or the GPL licence message and warning re lack of support that is displayed on entry to the product.

Applications that you build using the open source version and for which you hold copyright to the source code are themselves redistributable under whatever terms you wish to apply. Software compiled on the open source version may not be executed on the commercial version or vice versa.

Since Zumasys acquisition of the commercial version of QM, the open source is no longer published on the web site. Of course, you cannot withdraw open source after its release and for anyone who wishes to develop this further, the ScarletDME fork is probably the way to go.

Martin Phillips

Zumasys Inc.

[Steve Trimble]

I really think if you are a serious developer, programmer, or small business, you can't beat the price or support with openQM / Zumasys. It is a very small price to pay for a feature rich environment.

that's MHO and I am sticking to it.

be well and safe,

Computerized Data Mgmt Inc
Steve Trimble
(501) 772-3450 cell / text

cdm...@gmail.com

--
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 on the web visit https://groups.google.com/d/msgid/mvdbms/4EE05C2F-912F-46FA-960F-72E33747F83B%40zumasys.com.

[geneb]

On Fri, 2 Apr 2021, Martin Phillips wrote:

> The terms of the GPL require that, if you modify the modules that form
> the open source release in any way, you must also release those
> modifications in open source form under the same licence. You must not
> remove the copyright notice embedded in every source code module or the
> GPL licence message and warning re lack of support that is displayed on
> entry to the product.
>

Right. I should note that I personally feel that the insistance on the
"lack of support" notice was done to discourage any use. Had I thought
that through at the time, I would have (and still should) add an additional
notice that informs the user where community support could be obtained.

Linux is under the GPL and it certainly doesn't slap you in the face with
a similar message before the login: prompt.

> Applications that you build using the open source version and for which
> you hold copyright to the source code are themselves redistributable
> under whatever terms you wish to apply. Software compiled on the open
> source version may not be executed on the commercial version or vice
> versa.
>

So maybe I've been under the wrong impression for this past decade plus
(this is not by any means outside the realm of possibility!).

I knew about the intential incompatibility between the open source and
commercial QM at tokenized binary level - however, it was my understanding
that you felt people targeting applications at the open source version
would some how be in violation of the GPL if they didn't include the
QMBASIC source code for their application, even if it didn't include any
of the original system source code.

If this is not the case (and I hope it's not!), then that changes the
entire landscape as to what usefulness ScarletDME is going to be in the
future.

tnx.

[Donald Montaine]

Regarding the open source version of OpenQM; it is great to hear the explanation from Martin..  IF anyone is interested, documentation and downloads are still available on the Rush Flat site:

https://www.rushflat.co.nz/downloads.php  The Scarlet DME web page is: https://scarlet.deltasoft.com/index.php/Main_Page.  I may turn this into my retirement hobby, even though I don't think it will do much to entice new users into the MV fold.

[Donald Montaine]

Forgot to mention, only if someone else creates a 64 bit version of the GPL OpenQM.  I am not a C coder.

[Marcus Rhodes]

Just thought I'd throw this in.

https://www.facebook.com/marketplace/item/164675062082897/

[Donald Montaine]

Dick Pick - gawd, I am getting slow.  Took me all of 3 minutes.

[Albert Kallal]

Hum, I not really convinced that "only" say some free edition of a mv database would do the trick here.

I think it comes down to what set of tools you going to work and build such systems around.  I drop in this group from time to time - (my love of the mv systems keeps me dropping by from time to time). But, it been at least 16 or perhaps a bit more since I done any mv work.

I don't think the server side licensing is the "only" issue. The issue is drop a mv system on a developers lap. Now what? So, a IDE with all the trimmings like being able to consume external libraries, intei-sense. Git-hub support? That's all required. I actually don't know even how to use the git-hub command line. But from Visual Studio? Use it all the time.

And the mv coding language? Sure, quite nice. But what IDE and system would I use for that coding environment now?

I think that mv systems live on right now because they were built up in a time frame in which the software tools in our industry were not all that great. And given the mv "get more done with less"? Then that caused a truckload of systems - very good ones to bubble up from the mv land of development.

mv-systems don't live and exist right now because the coding part was great, they live and are alive right now because of some great systems and software that started MANY years ago!

I not all that surprised we don't see much new development in mv-land.  And the reason seems rather simple:

What modern tools and IDE's do I have to jump into such development? 

Does that system let me consume and share libraries?

Does that system let me say reference your existing code libraries and work. and then extend against that code base with ease?

To be fair, even complex code I write today is not all that heavy on OO approaches. But it sure is the heck heavy on consume those code libaries as classes etc. In other words, it not a case of full on out OO coding. But consuming, referencing those code libraries and utilities and systems?

That is a huge part of the developer process and experience right now.

I don't feel that mv land offers these great tools. And if you doing web stuff, then the mv system don't matter a whole lot much anyway.

Now, it's possible that mv land has these great tools. And has these great abilities to allow say 2-10 developers to work on such projects. pure OO coding is WAY overblown. But sharing of code, libraries and consuming and referencing those existing code and applications is NOT overblown - it often quite much the bulk of work.  So consuming classes etc. from those systems? Hell yes - that is common. So that's not all full OO coding here, but it simply a means to share, consume and interface with existing systems. And it allows a developer to get up to speed with greater ease (don't have to understand all the code, or the whole code base. Such IDE's and coding approaches thus allows entry to writing code with greater ease.

There was a time when even a relative small business - say only 25 employees would develop in-house software. (no other choice). So you wind up with say a 1 man, or maybe 2 man development team that created a GREAT mv system. But today, that smaller business will not likely be doing in-house development. I written two payroll systems from scratch (on on pick). I not heard of a company doing that for 20 years now!!!

So like any industry? Well, there is consolidation. There used to be a video rental place on every corner. Then blockbuster came along, and you had fewer places - but larger places to rent those videos. Then the internet and Netflix came along, and that was that!

So, web enabling and pulling data from mv systems? Sure, but you not really using mv tools.

But that does not address the mv side and coding and development process. Are those tools better then what was 20 years ago today?

is there a great mv IDE coding system for mv systems? Could I take a developer and drop them into that environment, and they would be at home right now? Since without such environments, we not going to see much new mv software written. And without new software, you not going anywhere.

And that famous quote from Steve Balmer about 'developers, developers, developers? You don't solve that and you solve nothing for the mv market. But, it is the chicken and egg issue. Without much developer activity, then you don't need developers much, do you?

I mean, what do people use for writing mv code these days? And are/is there a great IDE to consume, manage and reference and use all that existing code built up, and written in mv land?

I mean, I really do spend a LOT of time in Visual Studio right now. And it has great tools. So for example, I need to know the difference between one installed version of SQL server. I can point the two servers at each other in VS - and it will spit out a different in the two schema's (and that includes store procs, views, and table designs). And then I can say please update that target server with all the new columns, or whatever work and changes to the structure on one server A to now update server B. This kind of LONG list of features and tools sets are simply beyond invaluable.

I have often pulled changes from one customer site. We did some updates to their SQL server - a few new columns, new views, and some store procedure code. I connected via VPN to their site. synced the changes to a local edition of SQL server (express) on my laptop. Then when I had time, I VPN to another customer site. I ran that schema compare tool - and was able to hit update - all the table changes, new columns, everything like magic just shuffles over to that other site - all from my laptop! And I did that update LIVE while they were working!

Do we have these everyday tools like this in mv land? And I use such tools all the time!

Unless you can give me that great suite of tools, a great working environment, I just don't see how I could find myself back in mv land. But, hey, if someone wanted some mv work? I'd jump right in, but I would not want some green screen and a bunch of 20+ year old legacy code base on thinking that using some source code imports is a high technology means to share some code libraries on that system.

Anyway, to be fair, I am losing my pick skills - and that seems sad - but then again I still see mv systems used by smaller business being dropped. The remains are systems with high value, and large over time investments of code - but I don't see new systems and code appearing much in the mv land. It more of a keeping what they have.

Regards,

Albert D. Kallal

Edmonton, Alberta, Canada.

On Thursday, March 25, 2021 at 10:44:50 AM UTC-6 Donald Montaine wrote:

So, what would it take for a free MV install to be competitive  with MSSQL, for example.

1) it wouldn't have to be GPL and could be closed source

2) it should allow easy install by downloading an installer

3) It should have no limitation on redistribution or number of users

4) It should allow databases of up to 5 GB ( half of MSSQL express)

5) It should be limited to 2 cores (half of MSSQL express)

6) It should be fully functional, with all the capabilities of the base "paid" option

7) it should not require registration

8) community support should be available through an online forum

9) upgrade to a paid version should be easy

10) commercial support should be available for a fee

11) ideally it should be available to 3rd party front end developers to bundle with
       a free version of their software with similar limitations or be bundled with 

       front end developer tools from the same vendor

12) it should come with a front end management tool such as Microsoft SQL Server Management Studio

12) it should run on Windows and Linux

If any MV DBMS vendor is serious about competing for new business, rather than just acquiring existing MV users from another vendor, this is what would be required to possibly be of interest to start-ups, small developers or hobbyists.

[Marcus Rhodes]

And that's another thing: What IDE do you really need with Pick?  Granted, VB was, for me, as I've said before, a huge revelation, but OpenInsight covers that.  And, sure, there's SystemBuilder, but that's swatting flies with dreadnaughts.  And I mean that because, while it practically becomes its own flavor of Pick, it's just too much for most tasks that I do at home or at work.  I would never reach for it.  But there's also AccuTerm with its GED and WED.  And I use WED constantly at work.  Can't really ask for more than that.  Although, at home, I usually just use (are you ready for this?) nano.

http://www.thinqware.com/services_training_nano.html

And there are others, too.  Nebula has one.  There's another one that seems wedded to OpenQM, whose name I forget, that I keep meaning to try.  So there's no shortage, but are they really necessary?  I watch my son deal with something like half a dozen different 'frameworks' or 'technologies', none of which he learned in school.  Except for C(++), java, and whatever other tools they used.  While in school, he worked for three different firms, and html, css, javascript, ... all the usual suspects ... were all they had in common.  And none of the other stuff was offered in any class.  All had been hand-picked by the CTO for some reason.  The point I'm making is that 'the other guys' have their own problems.  And, among those, is not only a lack of a standard IDE, etc, but a lack of schooling AND a dependence on some sort of framework and IDE.  Pick just doesn't suffer from that.  For that matter, neither does Linux.  They still fight over whether vim is better that emacs.

[Martin Phillips]

The open source version of QM was issued under version 2 of the GPL which clearly states that there is no warranty. If I remember correctly (only vaguely probable after 20 years), the no support statement was indicating that there was no support from Ladybridge Systems but did not preclude anyone else providing support.

Regarding incompatibility between the commercial and open source, this was never a deliberate act but the two versions drifted apart as we continued to develop QM. A major change to the object code format finally ended what compatibility remained.

We always saw the open source version as a sandbox for developers to try things that they felt might be useful enhancements for the commercial product. We remain receptive to suggestions as QM continues to develop.

Martin Phillips

> On 2 Apr 2021, at 22:16, geneb <ge...@deltasoft.com> wrote:

>
> On Fri, 2 Apr 2021, Martin Phillips wrote:
>
>> The terms of the GPL require that, if you modify the modules that form the open source release in any way, you must also release those modifications in open source form under the same licence. You must not remove the copyright notice embedded in every source code module or the GPL licence message and warning re lack of support that is displayed on entry to the product.
>>
> Right. I should note that I personally feel that the insistance on the "lack of support" notice was done to discourage any use. Had I thought that through at the time, I would have (and still should) add an additional notice that informs the user where community support could be obtained.
>
> Linux is under the GPL and it certainly doesn't slap you in the face with a similar message before the login: prompt.
>
>> Applications that you build using the open source version and for which you hold copyright to the source code are themselves redistributable under whatever terms you wish to apply. Software compiled on the open source version may not be executed on the commercial version or vice versa.
>>
> So maybe I've been under the wrong impression for this past decade plus (this is not by any means outside the realm of possibility!).
>
> I knew about the intential incompatibility between the open source and commercial QM at tokenized binary level - however, it was my understanding that you felt people targeting applications at the open source version would some how be in violation of the GPL if they didn't include the QMBASIC source code for their application, even if it didn't include any of the original system source code.
>
> If this is not the case (and I hope it's not!), then that changes the entire landscape as to what usefulness ScarletDME is going to be in the future.
>
> tnx.
>
> g.
>
>
> --
> Proud owner of F-15C 80-0007

> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.f15sim.com%2F&amp;data=04%7C01%7Cmartinp%40zumasys.com%7C0901acd1ed8c4058a2b208d8f61c810d%7Cbfd62fcabd384e07a0785e9943d1f9c0%7C0%7C1%7C637529949604155025%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=FixuVdfEfanC8Ryd14LGsZLoA8guDW2zyE%2BLCil5uPo%3D&amp;reserved=0 - The only one of its kind.
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.diy-cockpits.org%2Fcoll&amp;data=04%7C01%7Cmartinp%40zumasys.com%7C0901acd1ed8c4058a2b208d8f61c810d%7Cbfd62fcabd384e07a0785e9943d1f9c0%7C0%7C1%7C637529949604155025%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=uy9PZsnshCD2mz7cTxjOxQLNBXrjvHsUtuilT5SLiEM%3D&amp;reserved=0 - Go Collimated or Go Home.

> Some people collect things for a hobby. Geeks collect hobbies.
>
> ScarletDME - The red hot Data Management Environment
> A Multi-Value database for the masses, not the classes.

> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fscarlet.deltasoft.com%2F&amp;data=04%7C01%7Cmartinp%40zumasys.com%7C0901acd1ed8c4058a2b208d8f61c810d%7Cbfd62fcabd384e07a0785e9943d1f9c0%7C0%7C1%7C637529949604155025%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=sk4C8cOsoBj0xo00XbF%2Bszveu9ZA8Afhq%2F5i1HYUc4M%3D&amp;reserved=0 - Get it _today_!

>
> --
> 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 https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fgroups.google.com%2Fgroup%2Fmvdbms&amp;data=04%7C01%7Cmartinp%40zumasys.com%7C0901acd1ed8c4058a2b208d8f61c810d%7Cbfd62fcabd384e07a0785e9943d1f9c0%7C0%7C1%7C637529949604155025%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=VJWhgUip9zMdLYE40BoZLlHgo0%2FEJSDkpYUWtJ4JyW4%3D&amp;reserved=0

> --- 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 on the web visit https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fmvdbms%2Falpine.LRH.2.21.2104021345030.26660%2540sidewinder.deltasoft.com&amp;data=04%7C01%7Cmartinp%40zumasys.com%7C0901acd1ed8c4058a2b208d8f61c810d%7Cbfd62fcabd384e07a0785e9943d1f9c0%7C0%7C1%7C637529949604155025%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=eGfFNMbHs5BU1NeH4PD0A7UWZh9th%2BODiV9XcTWedPs%3D&amp;reserved=0.

[geneb]

On Sat, 3 Apr 2021, Martin Phillips wrote:

> The open source version of QM was issued under version 2 of the GPL
> which clearly states that there is no warranty. If I remember correctly
> (only vaguely probable after 20 years), the no support statement was
> indicating that there was no support from Ladybridge Systems but did not
> preclude anyone else providing support.
>

I understand that - however the GPL notice can be displayed when the
system is first started, and not be used to metaphorically slap
unsuspecting users in the face when they log in.

> Regarding incompatibility between the commercial and open source, this
> was never a deliberate act but the two versions drifted apart as we
> continued to develop QM. A major change to the object code format
> finally ended what compatibility remained.
>

There's a comment somewhere in the code indicating that the token number
values were changed in the GPL version. The comment mentions that this is
done to prevent GPL pcode binaries from running on the commercial version.

The question remains - can people build commercial applications for the
open source version without having to give up the QMBASIC source they
wrote from scratch?

g.

--
Proud owner of F-15C 80-0007

http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.

Some people collect things for a hobby. Geeks collect hobbies.

ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.

[Albert Kallal]

>> What IDE do you really need with Pick?

Ah, how about a really good one?

You not explained to me how you can work on a test copy (you don't work against production code, do you????) for a week.

Now you made 15 changes to the database schema, and a bunch of changes to the code base.

And now how do you update that customer site with the weeks worth of work you just did? Including those changes to the database schema (new columns and a few new tables). ?

Without question you need such tools.  You not outlined what tools you are using to deploy updates to those customer sites. So yes, your IDE and bag of tools better have a great deployment and update feature set.  Chopping away with a text editor? That might work if you have 5+ years of experience on the one site and all you ever do is update that one code base and system. And in that case you often as noted be updating the production code as you work. That's ok for a one man code base that you know inside-out.

Can I really work at home for a few days, make changes to the software and application, and then when back at work, or VPN in, then send all the changes to the application code base? And send all the new changes to the schema (new dict items and new tables). 

And I can do this in a automated fashion?  

I mean, sure, I can say "copy" the dicts and over write. That's fine for one table + dict. But  I still have to remember which of the 50 tables and 50 dicts I made changes to. Same goes for the code base.

mv typical will have a file. Say utiltiltySubs. Maybe another one called PO code etc. So, if I change code in these seveal files - probably cataloged programs, then how you going to update the production code base? And how you going to remember what sets of code that you changed in say those 4-5 different files that hold such code?

And the editor you note gives syntax for a given language. But they often don't do a good job of helping you consume + use the subs and functions in that code base. So do I get any type checking, and pop up parameters in code as I type? 

As I stated, if you spent 5 years (or more) working on one code base - the same code often used "live" for production? Well, you know that code like the back of your hand.  You don't think about these issues - and you don't have to!

But newer devs will not. And thus I need in that editor with GREAT ease to consume that code base - not just the mv-basic syntax - as that is minor. what you need is code helpers against that code base - so I can not only call say your subs - but also in that editor know and get at those existing subs - and not have to remember the number of parameters for that sub/functions etc. for such code.

My mv/pick experience? On customer sites, we would often say have a test account. We would make code changes in that account. Test them (and we were testing against live data - not good!!).  After we have what is working well? We would then copy those code changes from that dev account to production account - that was the total extent of the development process (and that coping of code and updates was a manual process - no automated build tools. 

That works "ok" for developing a custom code base against ONE customer site. That dev model does not scale beyond that dev model. You go beyond that one customer site, and now you in for a whole lot of pain and hurt without all these great tools I spoke about.

This suggests that the development environment is limited to one person coding, and against one customer code base. 

For the above? most pick systems can fair quite well. But beyond that one developer and one site? Then all of a sudden git-hub, and things like being able to update the database structures? Now you need tools.

Now, try supporting 5 different applications - and all of them run say on 10+ customer sites.  Just the management of the incoming requests and fixes and new feature set requires software to track that stuff  - let alone a good IDE and system to try and update that big monster mess. 

Sorry - the old days of chopping away in a text editor to maintain modern software systems? Sorry, you not in the game - you just not. You tell me how the two developers coordinate their code changes to that application? And you tell me how they move their schema and dicts and changes from the test server to the production one. This was always a manual process in for my mv work. And as noted, for a one code base one customer site? This can work quite well, but beyond that? Then no - a good set of tools is required. 

 Today, all those fancy frame works are used not because they have cool sounding names (well, we hope not!!!!).

These frameworks are used because the allow the developer to LEVERAGE other code and other libraries, and use that code. 

In fact these days you spend a LOT of time adopting and using some of these frameworks and libraries - often as much time as you will writing say custom code and you own code!

So yes, having all this stuff work seamless inside of a IDE? To do that basic day of work? 

And same goes for debugging code. Does one have a decent debugger that that lets you hover your cursor over the code and it will show the variables - including object structures? All without having to type one command?

So, say we need code to output and merge and manage some PDF files - common task, right? Ok, so that IDE better help me out! Even if say using GhostScript to do that work? Does the IDE help me consume and use and work with GhostScript? (it better!!!). 

And that software better give me some nice email integration software. It not JUST that you can send a email, or that you can say use GhostScript to create, merge and produce PDF files and then send them.

It is HOW WELL that IDE helps you, and works with say the email system, and the PDF system and using that library and code base that you now going to adopt.  (for the email, or for working with PDF files).

And does that IDE now support your web portal, and all the software that goes with that? it better!!! We not even touched the web side and issue here - and that's probably the MOST important issue of our time right now. So, does that grab bag of tools support a great web development system? (it better!!!!).

The simple matter is software has become VAST more complex today then a bunch cataloged subs in a folder, or in mv-land a file with those source code items inside.  That simple Hodge podge code base of cataloged programs is a MINOR part of the tool set that developers need, and will use. 

But worse it also not going to cut it with developers having used any of the mainstream tools to build such applications. You drop such developers into mv land without a great IDE and tool set? it going to be a VERY long and very hard road for that developer to become productive in that environment . 

One person, one code base, and years of experience on that one code base? Sure, that works. And that tend to be the mv landscape right now.

However, anything outside of that typical environment for a mv system? you not in the software game without that grab bag of great tools. 

Can your tools even build me a easy to use installer for THEIR IT and admins of that target customer site to update the software?

Can the customer have only a few - say 10 out of their 50 users run the updated software - sign off, give the ok, and then roll that software out for use by the 50 other users? 

Software and tools have become complex, and the ability's to adopt those external systems ranging from email systems, PDF systems or whatever? The developer tools and IDE have to talk to, and play nice with those systems, else that developer can't bring much value to the table.

20 years ago one would write payroll or even a simple database for customer contacts. Now, such systems exist, and it rare to see such software written in-house for one customer.  And same goes for accounting packages - there was a TON of pick/mv accounting systems back in the day. 

Now? They purchase off the self. But then again, that off the shelf system will have all the fancy direct deposit stuff. It will even have a employee web portal (they can logon - see their pay, manage when they take holidays). And now we seeing these systems being extended to allow customers to make payments to that given company. (and using their smartphones to do so). 

I can type on for a long time here. The simple matter is #1 task these days is consuming and interfacing to other code systems, be it payment systems, web systems, accounting systems - you name it. And much of the reason is all these systems can't be produced or made in house anymore - and those in-house systems can't compete with commercial systems that have large customer bases - and thus large R&D budgets.

As a result? We now have to talk to, and play with, and write software to work against all these new, but typical systems that a business needs today to stay competitive in the marketplace. That long time and great mv system? Sure, it will and can survive  - its tailored to that one business - and they are unlikely to find better software.

 But the set of tools they are using better make it easy to work with all those other systems like the customer web portal, or direct banking payments for payroll or PDF documents, or smartphones or "insert longer list of things" here.

Regards,

Albert D. Kallal

[antlists]

On 03/04/2021 20:35, Albert Kallal wrote:
> >> What IDE do you really need with Pick?
>
> Ah, how about a really good one?
>
> You not explained to me how you can work on a test copy (you don't work
> against production code, do you????) for a week.
>
> Now you made 15 changes to the database schema, and a bunch of changes
> to the code base.
>
> And now how do you update that customer site with the weeks worth of
> work you just did? Including those changes to the database schema (new
> columns and a few new tables). ?
>
> Without question you need such tools.  You not outlined what tools you
> are using to deploy updates to those customer sites. So yes, your IDE
> and bag of tools better have a great deployment and update feature set.
> Chopping away with a text editor? That might work if you have 5+ years
> of experience on the one site and all you ever do is update that one
> code base and system. And in that case you often as noted be updating
> the production code as you work. That's ok for a one man code base that
> you know inside-out.
>
> Can I really work at home for a few days, make changes to the software
> and application, and then when back at work, or VPN in, then send all
> the changes to the application code base? And send all the new changes
> to the schema (new dict items and new tables).
>
> And I can do this in a automated fashion?

Git?

What you're describing is not an IDE, it's a source code control system.

Cheers,
Wol

[Marcus Rhodes]

You keep portraying issue which, as I said, don't seem to be issues on Pick.  But, then, Pick, not being a real DBMS, kinda requires you to roll your own.  And, I'll wager everyone here has done just that for every issue they've encountered.  For example, every firm I've worked for, with the rare exception, has at least a development environment entirely separate from their production environment, in which developers can play to their hearts' content.  This, of course, does require good 'lab notes'.  But, then, you learned all about that in college, right?  Some firms I've worked for actually have different development, test, stage, and production environments, but that's mostly been the commercial software vendors and the hyper-vigilant government agencies.  But, amazingly, many, if not all of us, have had to work on production systems.  Maybe we're just that good.  Maybe Pick is just that good.  But, if you adhere to some simple and obvious practices, it works great.  Yes, someone is likely going to screw up, but just look at Microsoft's updates.  They usually blow my systems out of the water once a year.  I stopped an entire production system once.  In 40 years.  And it wasn't entirely debilitating then.  That's why so many shops have a practice of blocking M$'s updates until they've been proven, usually on someone else's system.  So, once again, Pick, and the products that run on it, aren't really any different there.

In fact, I configured one of the dev systems I worked on (because I had the authority to do that kind of thing) to always read from production files, but write to their test-environment counterparts.  Saved me a lot of time copying data from prod to dev.  And that's the kind of thing I mean.  We're programmers who haven't grown dependent on/crippled by IDEs that do too much for us.  Because Pick is that flexible.

And, I've even seen production code modified in real-time.  I've had a client or two where this was the only option because they had no dev environment.  Again, proper precautions prevent problems.  Plus the very nature of Pick makes it pretty easy to do.  Just copy the sole bit of production code to a new name, make sure you're the only user who'll hit that code, and full speed ahead.

And this database schema issue... again, not much of a Pick thing.  Most of these systems have been running for years, if not decades.  If you want to add a field, it's more likely that something's wrong with your analysis than with the schema.  That said, I've only seen a few Pick systems where adding a field wasn't trivial.  And even those systems where some genius decided to create an include to define every field of every file, that was the extent of your work.  No big deal.  Moreover, on Pick, only a maniac would decide that his latest creation just wouldn't work right without inserting a new field or two between existing fields.  Pick's very forgiving that way.  Need a new field?  Just tack another field on to the end, and carry on.  In most cases, this won't even upset a running production system.  Which makes installing new code a breeze.  Although, as I said, proper practices, not a real DBMS, etc. means that everyone here has probably also written at least one installer for some piece of code they wrote because a client didn't have much (infra)structure or procedure ensuring smooth deployment.  My current client is great about this.  They have a test environment AND TESTERS!  WHEE!

And there you go!  MODERN SYSTEMS!  Define modern!  Like I said (I'm saying that a lot, I know, but it is what it is)  I've tutored a guy in the University of Mexico through his SQL classes.  I've mentored by own son (a degreed software engineer) through his web-dev classes.  IT'S ALL STILL TEXT!  The last really 'modern' tool I used was Smalltalk, and, frankly, it sucked.  I notice that practically nothing is written using anything like that anymore.  Good riddance, too!  Yeah, you're right.  The tools have become complex.  Because the tools they work with are too complex.  Both Javier and my junior lost at least a year's life trying to cope with all the tools and features these 'modern' environments offer (not to mention Wirth's and Stroustrup's lethal ideas on software development) (Hey!  Columbia University, right?  Those who know what I'm talking about...).  Sorry.  I can't see any real benefit in VS.  I don't even see the benefit in MultiEdit, which a lot of guys I know use...  on PICK!  Wow. Now, syntax-highlighting?  I'm with ya there.  WED does that as well as I need (although I *am* working on my own editor which does it better) (kinda like I set up nano to do).  Code formatting?  WED does it.  (By the way, if anyone wants my list of WED keywords, let me know.  It's comprehensive.)  I'm the last guy on earth to defend the hammer-n-chisel ED jockies, but the fact is that they are right when they say that newer systems have just become promiscuous with their 'newer, cooler' everything.  Change for change's sake is asinine.  As one commenter put it, he's convinced that C++ was a communist plot to set our software development back a decade.  I say all these web-dev toys compound that by at least double.  I've yet to use any web interface that even matches, let alone exceeds a native application's usability.  They're sheer torture, fit only for online shopping and (what the web was originally intended for) reading articles.  (Which is why Dynix is, for all practical purposes, gone.  But who didn't see that coming?)  And I mean everything: Salesforce, Jira, even OWA.  And you want me to develop with such tools?!  You think they offer something better than WED, SystemBuilder, or OpenInsight?  Sorry.  Been there.  Done that.  Not goin' back.  This is why Columbia Ultimate's Collector is still in production today.  As far as I'm concerned, the only things that should ever be accessed with a web-browser are catalogs (Amazon, Ebay, Overstock, ...) and reference works, articles, and the like.  Yeah, ok, so your tools aren't web based, but they are mostly for web-based things, and web based anything needs to be rethought.  OpenInsight can do that now, I hear.  I really need to try it because, I'll bet, it blows the doors of everything else I've seen.

Can my tools let users ... go ahead.  Give it your best shot.  I can roll my own tool that does anything I want.  And, yes, just like I can limit experimental code to just me, I can limit it to any number of users.  Amazed yet?  But, no, 'my tools' don't do that.  I do.  I'm a programmer.  We're kinda talented like that.  On pick, it doesn't require Glitch months to develop and test such a tool.  I could probably do it in a weekend all by myself.

One customer?!  This makes no sense.  Was that supposed to be an argument in support of something?

Accounting packages?!  You do know that Ceridian was written in 'Pick', right?  (I don't know guys.  Is it still?)  By the way, Microsoft's Access is essentially Pick on Windows.  Not the best, but, if you're going to develop a custom app for a single client, you could sure do worse.  Maybe not much worse, but worse. But, again, as I said, there's always OpenInsight.

As for talking to other systems, Pick is no slouch.  I run reports through xlsxwriter in python and then zip to e-mail clients encrypted Excel spreadsheets all from within Pick.  There's even a Pick-based web-server.  I've seen UniVerse directly access on-line credit card systems.  What more do you need?  And what's that got to do with development tools?

The Pick system I use at work has a web-based bill pay facility.  What's your point?  Pick also works well with off-the-shelf commercial applications.  I've used Excel to directly access and manipulate Pick data in native Pick data structures.  Maybe you've heard of ODBC?  Not all that modern, I admit, but it works pretty well.

I once had to keep an emerging OneWorld system synchronized with a retiring, home-grown, UniData-based order entry and warehouse management system.  OneWorld choked on all the data.  I had to break it up into bite-sized chunks for it.  Then their dev team kept wanting the Pick system to do more and more of the work for them.  Maybe they were just trying to buy time, but nothing they asked of me took more that a couple of days.  One of the Pick devs called me up about 6 years (and several million dollars) later to tell me that they finally shut the old system down, and laid him off.  That firm used to be an Inc.500 cover story.  Used to be.  Not any more.  And that same story applies to Colonial Pacific Leasing, Ernst hardware, and about a dozen others I've heard of along the way.  And every one of those failures was directly attributable to their attempt to 'modernize'.

Pick isn't the problem.  The tools aren't the problem.  The platform itself is the problem.  Pick is the solution.

[Will Johnson]

Could you explain better why Pick is not a DBMS ?

[joseba]

Just now, VSCODE with no doubt about it.

[Donald Montaine]

The reality is that, whatever it's benefits, Pick-MV is dead, except at companies that have too much invested in it to move off it.  The reason most Pick-MV vendors spend all their time trying to steal each others' customers rather than attract new customers is probably that they all have come to that conclusion.   

For those who believe that the statement about Pick-MV being dead is too harsh, name one major development house without a history in MV who has chosen to develop a new product using Pick-MV technologies.  Or name one major company without a Pick-MV history who has chosen to move to Pick-MV as the basis for their business.  Say, in the last year, the last five years, ten?  Now consider how many development houses and major customers have moved in the opposite direction.  If the licensing and development tools were compelling, wouldn't the movement be toward Pick-MV not away from it?  (For a ranking of all the various DBMS see https://db-engines.com/en/ranking).  

If people were really interested in modern Pick, wouldn't Open Insight be more pervasive than it is?  It works with all the Pick-MV back end products, has a decent native MVDBMS implementation, has a pretty good IDE, has supported SQL against MV files forever, has connectors to SQL databases, etc.  Instead it does not seem to have been adopted on a large scale by either Pick or mainstream developers.  Zoominfo indicates yearly revenues for RTI at about $5M/yr and Zumasys $30M/yr.   The development tool I use is made by PCSoft, Inc, in France.  They are a fairly minor player in the SQL space, but at $30M, they are the same size as Zumasys and their proprietary embedded database is ranked ahead of all the MV databases except U2 and jBase.

What is really interesting is that RTI and Zumasys both charge yearly per seat fees to users of their technologies.  PC Soft only charges the buyers of their development products rather than the end users, except for a small per server fee for their optional web application server module.  So the PC Soft income is largely from a developer community that is willing to pay each year to upgrade their tool set (it is not a subscription.  If you don't pay you can continue to use what you have).  How much of the income for RTI and Zumasys comes from developers and how much from end users who are "lucky" enough to have legacy systems they have to pay for year after year.

[Will Johnson]

You seem very fixated on this per seat issue.

Can you explain why?

For the vast majority of customers, paying say a thousand or even ten thousand a year is one of their smallest expenses

I mean the budget for us is something like 100 million a year

[Donald Montaine]

Per seat licensing may be appropriate for larger companies . I am focused on new or smaller companies with less than 100 employees.  You hook them at the beginning with a product that has no barriers to entry.  Then if they grow bigger they will outgrow the free version and start paying per server or per user.  Then again, every database competes with PostgreSQL.  Why pay if you don't have to even if you are large?

"According to the latest US Census Bureau statistics, there were about 7.6 million employer firms (ie, businesses with employees) operating in the United States in 2017. Of those businesses, 89% have less than 20 employees"  from: https://www.businessknowhow.com/money/earn.htm.

[Will Johnson]

Donald you seem to be missing a very salient point.

Business owners do *not* choose solutions because the licensing is free or cheap.

They choose solutions based on the solution itself.

I'm opening a pet store.  I'm not going to pick a solution based on cheap or free licensing.

I'm going to pick a solution that does inventory, receivables, payables, payroll etc etc etc

And looks good

The *licensing* is rolled up in the solution providers costs.

It's not rolled up into *my* costs directly.  Only indirectly through my VAR

As it has been since 1975 or so

[Bob Markowitz]

WOW! A discussion about Pick that has evolved from where is the magazine to “…Pick-MV is dead, except at companies that have too much invested in it to move off it…and They choose solutions based on the solution itself.”

Buried in the thread is almost the whole history of who we are as a marketplace. The discussion is covering costs including the licensing model; we are discussing open-source versions; highlighting the lack of Pick evangelists to tell the world that we offer something better than they have; we are talking about the mv database vendors main marketing focus was/still is – I want your customer not growing the market; we scream about the benefits of the Pick model; some focus on the need for real training held back by the lack of manuals and documentation plus the need for 14 tons of videos – where is the present day version of Jonathon Sisk; a mention of Evoke (thank you Joseba);  and there was the reference to 20+ year old Pick applications still being used because how good they are plus we are reminded of company investment in the legacy mv applications; I love the ping-pong/basketball analogy; more!

Take out your bows and arrows boys. Here is my take. I am allowed to have an opinion. I have been around since the earliest Microdata days. I helped unpack Reality serial number 6 from its crate and I am as old as dirt!

A bit of history, Once upon a time Mr. Pick and Microdata parted ways. After a contentious law suit each party owned a version of Pick. Dick would license his to anyone that would pay. Microdata was Microdata. VARs developed business applications. The Pick system was/is a major player in auto dealer software, manufacturing, managed healthcare, retail, multilevel marketing, association membership, hotels, colleges and university, many end user companies wrote their own. What major markets have I missed.

I do not believe the demise of Pick is because of the licensing model or cost. When the market was growing it was growing primarily because of VAR software. The Pick system is/was buried in the price of application software and hardware. Some VARs are still selling amazing feature/function rich software to end-user companies, there are not enough VARs.

The successful Pick VARs are still selling product because they have changed the front-end to a modern presentation layer and kept the back-end business logic (Basic). They will tell you that they spent an obscene amount of money enhancing their software and it took many man years to create it. The smart ones are also figuring out ways to move their applications to mobile devices. They seldom mention mv in the sales process.

Some smarter VARs and end-user companies will take their back-end data and business logic and focus on creating a customer experience that locks in their customers. Customer experiences run mostly on mobile devices.

The Pick system is not disappearing because there are not enough no-charge versions that a high school kid can pick up and write a new application to inventory drug paraphernalia stored in the garage. Pick is disappearing because the presentation layer is so 1970. Accuterm is a very cool product that puts lipstick upon the proverbial pig and is so 1990.

Many end-users are looking to move their applications to the web to get that modern look and feel. Some are even buying into the marketing provided by the Pick database providers that RESTful Web Services is the solution to all of Pick’s woes. Modernizing a legacy mv system costs in time and money.

Commercial time…

Mainstream software development is moving toward lean-code/low-code/no-code. Forrester predicts that in 2021, 75% of application development shops will use low-code platforms [1]. (I wonder how many mv shops they spoke to?)

Mainstream organizations are moving to lean-code/low-code/no-code because it is significantly cheaper than long hand coding; it is significantly faster than long hand coding; it can be done by non-technical as well as professional coders; developers can create native mobile apps without learning a whole lot about the creation of native mobile apps; more. For those of you that have yet to google low-code…

Evoke is the only lean-code/low-code/no-code platform that supports Pick and Oracle and SQL and…Total cost is less to develop and maintain an app using Evoke than using RESTful Web Services. It is cheaper to modernize and maintain a legacy system than using RESTful Web Services. It is up to 10 – 20 times faster to create new and modernize legacy application than using RESTful Web Services. Evoke gives you desktop, web and mobile from a single design. Apps can be modified and enhanced in hours and days, not weeks and months; it is called “agility”. Did I say Evoke is cheap?

Oh, and for those few of you that know the strengths of Microsoft Visual Studio, Evoke is Microsoft centric and generates industry standard Visual Studio & VS-Xamarin projects that allows for the inclusion of legacy Pick code and data. Also any technology supported by VS can be incorporated into an Evoke solution. Did I also say that Evoke is cheap and you do not need to be an expert in VS. As a matter of fact, if you create a no-code solution you do not even need to enter VS. Just check a button in the Evoke developer and Evoke will generate your app without you even entering VS. And for you people worried about the cost of VS. The free community edition works just fine!

Pick will be around for the foreseeable future. We will see some end-user companies modernizing their legacy Pick apps. We will see some end users moving from one version of mv to another for whatever reason.

However, we will not see any software houses (outside of the mv marketplace) writing new apps using the Pick database. We will not see young programmers yelling to become Pick programmers. We will see the retirement of some very talented mv technicians leading to a continuing shortage of talent. We will see many companies that rely on mv stuck in maintenance mode and not creating new stuff. We will see many companies that rely on mv moving away because competitive software now offers more features and functions than their legacy system as well as modern front-ends and an enhanced customer experience. Ecommerce activity has exploded this past year because of Covid-19. There is an increased expectation for seamless digital customer experiences. MV is great for gathering, manipulating and reporting. It does not play well in ecommerce. Customer experience is driving today’s software explosion – look it up!

Whether a company is B2B or B2C Evoke makes it easy to create and enhance customer experiences based upon customer and market behaviors and that drives increased sales and profit. Evoke will not save mv but it will help to keep it around for a bit longer.

One last thought, is Tony G. influencing me on the size of my posts?

[Donald Montaine]

And how many of those solution providers (unless they are a long time MV user) choose MV today?  As a solution provider deciding to develop a new application for a vertical market I am going to do everything I can to be price competitive with the competition while remaining as profitable as possible.  And one way to do that is to make sure that my costs remain low.  How many NEW vertical market apps are being developed using MV technologies today?  Whatever the reasons, these days MV does not usually win the battle  when compared with  other embedded DBMS technologies.  My belief  is that the licensing model if is contributing factor.

[Donald Montaine]

"However, we will not see any software houses (outside of the mv marketplace) writing new apps using the Pick database. We will not see young programmers yelling to become Pick programmers. We will see the retirement of some very talented mv technicians leading to a continuing shortage of talent. We will see many companies that rely on mv stuck in maintenance mode and not creating new stuff. We will see many companies that rely on mv moving away because competitive software now offers more features and functions than their legacy system as well as modern front-ends and an enhanced customer experience. Ecommerce activity has exploded this past year because of Covid-19. There is an increased expectation for seamless digital customer experiences. MV is great for gathering, manipulating and reporting. It does not play well in ecommerce. Customer experience is driving today’s software explosion – look it up!"   I agree Bob, you said it better than I could.  That said, there are times I wish I could use MV even today, but the cost of licensing is greater than my cost of just doing it using an SQL solution.

[Philip Ellis]

Indeed ScarletDME is a well kept secret - I didn't know about it until I read this post :)

On Thursday, 25 March 2021 at 16:48:42 UTC geneb wrote:

On Thu, 25 Mar 2021, Donald Montaine wrote:

> In the SQL world, the DBMS vendors act in many ways just like the MV DBMS
> vendors. However, the difference is, that there is an entry point where
> developers can become familiar with SQL technology without cost. The open
> source versions of MySQL, MariaDB and PostgreSQL (or even Express offerings
> such as that provided by Microsoft) provide the entrance into the world of
> all things SQL. The fact that there is no equivalent entry point into the
> MV world means that SQL will always have a leg up. As long as all the MV
> vendors refuse to provide an equivalent entry point into the technology
> nothing will change.
>

ScarletDME is apparently a very well kept secret. :)
https://github.com/geneb/ScarletDME

g.

--
Proud owner of F-15C 80-0007

http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.

Some people collect things for a hobby. Geeks collect hobbies.

ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.

[Tony Gravagno]

On Tuesday, April 6, 2021 at 12:39:43 PM UTC-7 bob.m wrote:

One last thought, is Tony G. influencing me on the size of my posts?

It's not the size, Bob ...

I'd say that's another conversation but since this thread on Spectrum turned into a free-for-all, it seems almost anything goes.

 

[Will Johnson]

Red Herring

*IF* you have a great solution in Pick, customers will buy it.

*IF* you do not, or refuse to try to build one, they won't.

It's a silly argument you are making.

*ALL* of your costs contribute to the cost of making a product.  Your rent, your employees, your gas money.... the licensing is the smallest cost.

Instead of whining on here for three months about it, either go do it, or don't do it.

Build you gizmondo amazing product in SQL.

The market will not shrink because you do

[Albert Kallal]

And what IDE these days does not have rich and great GIT support built in? That was my point!

So yes, today, we see git integration in those IDE's - and it not good - it so good that you don't even have to know how to use GIT from a command line environments.   I never learned GIT well, but I use it from Visual Studio all the time. 

As for the general discussion here? I think most are missing HOW this industry works.

If a employee walks though the door, and by hiring that person, how many of those people will you hire?

Why, of course 1 million such employees, or at least as many as you can until such time that hiring more of these people will NOT make you more money then they cost.

You can ONLY add more software value by adopting NEW technologies. - it is that simple.

If I walk into a business that is running say a 30 year old payroll program. Can I really sell them on a new fancy payroll program?

Nope! - yet 9 out of 10 times, I can walk out of that office with a signed contract?

Why, and how?

Well, offering to replace their mv payroll, or sql based payroll program? That offers near ZERO return on investment.

Companies buy billions of dollars of software in a year. And the reason why? The cost of buying that software is LESS then the cost benefits they receive by buying that software!

Why do you think in the top 10 worlds largest companies are oil companies? You think it because oil is expensive? 

No, oil is no different then hiring a employee that will make you more money then they cost.

Buying software is also the same - it costs less then the benefits you receive.

So, every year, I can say buy about $4,000 dollars worth of fuel, and use to scrape off my fields $80,000 dollars worth of crops!

What product can you buy EVERY year for $4,000 that you can turn into $80,000 dollars?

I can put $20 worth of gas in my car - drive around for say near two days, and generate $1000 dollars worth of business value. 

I can deliver goods all day in that truck - the cost of that fuel in fact a BARGIN - its cheap, and I can over and over turn that cost of fuel into money - and MUCH MORE then the cost of that fuel. And even on a simple level - driving to work everyday. the cost of that fuel is SMALL compared to the benefits you receive by purchasing that fuel. That is why oil and gas companies can make so much money. The product they offer can EASY be turned onto the wealth creating process.

Software is NOT any different.

So, back to that payroll program. Offering a payroll program that does the same as before? No sale! - no benefits.

ONLY by adopting new software technologies can you make that sale. So a brand new payroll program? Not really going to help that business and the cost benefits of that new software is NOT any better then that 30 year old payroll program.

but, you see, in that typical business? The accountant or controller will take and receive about 10 phone calls per day from staff. (or those 10 stall will walk into his office). So, lets say 10 people x 5 minutes (or 5 people x 10 minutes). That is 50 minutes of time. In JUST 10 days, that is 500 minutes. And in 100 days, that is now 5000 minutes of labor!

So, what i will do is offer that client a web based staff portal. now staff at home, on the weekend? They can logon, see their pay, schedule their holiday pay and whatever. you all of a sudden now have a software box - a software box that COSTS LESS then the labor and time saved.

I mean what businessman will not buy a "box" or "thing" that costs LESS then what they make by that purchase? So be it fuel for their trucks, or hiring 1 million more employees that make them MORE then that employee costs? 

What business man is not going to buy some software that can save them 5000 minutes of labor in just 100 days? That business would be in fact STUIPD to not make that purchase.  So, I'll walk out of that office because I am using NEW technology (web) to allow that business to save MORE money then the cost of that software purchase or development. It REALLY is that simple.

because mv systems are not really offering those new technology plays in the marketplace? Then ALL they can do is canaibalize  existing business process and thus customers based on what they are offering. This explains why the only market for mv customers is grabbing other existing mv customers! 

it also explains why I can't find much work based on older technologies. It not that the technology is older, but it limited in WHAT NEW business process I can offer to that business for LESS cost then what the developer cost are. 

The above concept of new technologies being required for new business process is a critical concept. And this thus allows one to make sense of the how's and why's of the current marketplace. 

ONLY by adopting new technology changes in those business process, can one offer the benefits of that new software or technology.

For example, take a typical city. There is going to be a bunch of well run and long established companies that offer business machines. photocopiers, binding machines, postage machines etc.

Now, you try and jump into that market, open up a business machines office? Good luck!

But, WHEN a new technology play comes into the market? Then HUGE opportunities open up

So, in the mid 1980's, low cost fax machines arrived. I know of some people that sold fax machines door to door.

next thing you know, they had a office. Next thing you know, those same customers all of a sudden needed photo copiers. So, now these people based on this technology changed were able to ENTER that business in a big way. In fact after a few years, one of the larger business machines in town bought them out for over 2 million dollars (they wanted that customer base of monthly re-newable photocopier customers they had built up).

So, a change in technology? That is a inflection point in which you can NOW offer new business value. And without that change, then you can't really offer anything more then the existing marketplace. 

In fact, I know of one tour bus company that had a HUGE boom in business - all due that silly fax machine. you see, sending a tour bus full of people to the mountains for a weekend of skiing? You have to provide and phone ahead to the hotel all the bookings and people and their names. That company reached a point in which some weekends they were sending 800+ people on ski trips!! They could now do this since with a fax machine, they could produce the booking lists - and fax them to each hotel - a task NEAR impossible by using the phone.

in every city, the real estate business is competitive   (just like those office machine business - VERY hard to break into that market), and thus established players and long time players own that marketplace. And all of these successful real estate companies also tend to have a section that helps you sell business. (say a couple built up a great business - they are old, and no children to take that business over. So they go to these real estate business brokers section, and put their business up for sale).

Well, I know of a group of people that purchased some fax software and in every city you can buy those "SIC" lists of companies (a list of companies categorized by restaurants, plumbing, HVAC etc.). So what they did was setup a office for selling those business. And with automated fax software? Well, say a family is selling their dry cleaning business. So, these people will automatic fax out to ALL other dry cleaning business that a new dry cleaning business is for sale. And often in the existing marketplace, existing companies are looking to expand. Needless to say? in a VERY short time, these people had moved their offices to the nice downtown business district. And they all were driving VERY expensive cars and making piles of money.

Once again? We see that introduction of NEW technologies into that business environment is WHAT creates the opportunism.

And just like that 30 year old payroll program? No sale for a new one - but that staff web portal for each staff to self serve and get all their holidary pay, schedule holidays and all that information? yes, you near garmented a sale, since like oil or anything else? What you offering costs LESS then the savings and benefits to that business.

So, today the market is hot in mobile phones and in web based software. But NOT because it really web based, it is because like those fax machines, you can adopt and offer NEW business process with that software or technology. Without that offer, you simply not find any work. Since that 30 year old payroll program does not rust - it still work fine but ONLY if you continue using existing business process. 

In the last 30-40 years, near every business now has a good level of software and automated business process. But with the web? Well, in Canada, I believe it was Purolator courier - they save something like 8 million dollars in the FIRST year of releasing their web based package tracking. Before that, they had 4 floor in a building with operations taking calls all day long to stick a phone in their ear, take the customer call, type in that package number, and read out the results to the customer. Now the customer just goes and gets that information.

And same goes for that payroll system and staff portal.

The common theme? Well, all those systems build up over 30+ years have TONS of information locked up inside. Why should your customers phone you, you shove your phone on your shoulder, and now type away into some internal business system to read out that information? Anyone used a travel agency? (not!!!!).

why use staff to type away on that software to deal with a customer to get project status? Why not let them do that? So the web market is hot and there is work, because a web system can un-lock all that information you have from your internal systems directly to the customers - and without staff labor costs to achieve this goal.  Why have your staff start + enter some new project or even quote? Why not let the customer do that - and they can with web based systems.

And because be it a simple fax machine, or that NEW business process based on web technology? Then that business will now pay me to build such a system. And thus ZERO surprise where all the new work exists in our industry! it exists in areas that allow me to offer a system in which the benefits are GREATER then the costs of such systems.

If mv land don't give me those new tools, then there is no work in mv land over and above what the existing systems are already offering and been doing so for 30+ years.

I'll give one more example of how technology change opens new doors. A customer of mine at a business condo area? A young kid out of a few week technical school on building web sites? he dropped off a crap printed out flyer for building web sites. Just like everyone was buying fax machines in the mid 80s? Well this young kid must have picked up 40 out of those 50 business in that business condo area. Breaking into the software indutry would be challenge - but with the web play? Well, 20+ years later that person has all the work he can handle at $80 per hour - easy makes over 100k per year. But, a new grad from some web school or course? He can go to that same business condo, drop off his flyer or business card - he not get ONE customer - since they all now have those simple web sites for their business.

So breaking into the business machine market (with arrival of fax) or say IT industry with everyone wanting a simple web site? No problem.

And the same now goes for un-locking all that incredible  business value locked up in all these software systems that business built over the years. Now, customers can directly get at that information, but MORE important become part of that process. Why fire up some old mv system, and enter a bunch of parts or assemblies to create a quote for a customer? Why not let the customer do that!!!  Why fire up that software to type in that a customer just approve some project or order? They can no do that, do it 24x7 and do it without costly staff labor. And that is WHY all the work is in that area - because it can save a business more then what such software costs, and a business would be stupid thus to not make investments that cost less and produce more then what they cost!!

So buying some fuel? You can week after week or for yearly harvest on a consistent basis for that purchase produce MUCH more value then the cost of that fuel. And the same goes for software.

Only by adopting technology changes in our industry can you offer NEW value to that business. 

And if your tool bag don't have those new technologies as part of that bag? Then there's not going to be any work, since you can't offer those things that save the business more then what you cost.

R

Albert

[Will Johnson]

Sorry I'm not going to read a twenty five paragraph complaint.

I forgot to mention however that you *could* move to a lower rent district.  The *savings* would more than make up for the license costs.

[Jan Van Schalkwyk]

How the hell did this go way off topic again????

[Albert Kallal]

Well, regardless of my post that explains the why and how of our industry? Yes, any decent IDE will include hooks allow with GREAT ease to adopt source code control. I would not attempt to suggest other wise. So while SCC is a different issue then a IDE? Well, sort of - but not really!

No IDE worth it salt has left out or forgotten SCC, and they have rich support for GIT or other SCC systems. VS also has support for both Team Foundation server, also their SCC api's - and has had this integration for 20 years now. 

And now with everyone flocking to Git-hub? Well now VS has great built in git support (kind of has to - since Microsoft purchased github). . So yes, - I do stand by my claim that good IDE's have good SCC support - and that includes git support - in fact your not quite much in the software game if your IDE does not have such support.

R

Albert

[Tony Gravagno]

Will Johnson wrote:

Build you gizmondo amazing product in SQL.

The market will not shrink because you do

I disagree. The current state of the industry is proof that the market does shrink relative to the rest of the pie with each decision to base a new "gizmondo amazing product" on SQL rather than MV.

T

[geneb]

On Wed, 14 Apr 2021, Jan Van Schalkwyk wrote:

> How the hell did this go way off topic again????
>

You must be new here. ;)

[Albert Kallal]

I am 100% on board with the MANY great benefits of mv systems. A few pointed out that mv systems allow you to build little utilities, and things you need that you don't have.  I mean, I became tired at one point to figure out the last attribute value when adding a new field. So I wrote a little utility to prompt me for a new column name - and it adds it (I mean, really, go find the highest attribute number out of a dict? Not!!!!). So, a little utility to the rescue - and mv systems are REALLY nice in this area. But there is a limit of how many of these little things you should have to write. 

The real problem is that I have to invest in a set a tools. And those tools should allow me to create legacy software - desktop software, but also adopt new technology (such as web and phone based). mv systems are NOT much limited in transition to web based systems.  One can really adopt new web technologies - and they can play rather well with mv systems.

the problem? The base mv tools don't translate into those other development stacks I need to use and learn. But the set of tools and development process I can adopt for web, or phone or desktop are NOT the same technology stack I would adopt when working on a  mv system. I can't leverage that mv knowledge much if at all to say now write some web based software. At least with mainstream tools - those skills "span" the multiple technology stacks I have to work with.  What I learn say in regards to desktop software can be leveraged and that code can be used for the web stack I am using.  So learning mv systems does not open doors into web development like other tools do. So, my investment in time in those other tools gets me both desktop, internal line of business software solutions, and now of course allows me to develop web and even phone based software based on my investment into these tools and skill sets.

If the base mv tools had morphed into web tools as good as they are for existing systems? Then mv land would be in a VERY different market position right now. 

Microsoft invested huge in .net. When the web came along, that web play threatened their .net eco-system. So they leveraged that .net technology  into asp.net - it was a brilliant market move.  So, while .net for web is not the most popular today? It still a player and great choice when it comes to web development. The key concept here was Microsoft invested in .net to "transition" to the web - thus preserving their years of desktop and client-> server side software investments. 

mv vendors did not have nor make this investment to allow the base set of mv tools to become  great set of tools for web development - and this IMHO is much why mv systems are not all that popular today - (no great web stack based on a transfer of mv skills to web development). As I noted, mv systems are well able to be the server and database system to drive a web site, but I'll be using few if any mv skills for this market play.

R

Albert

[Will Johnson]

You could not be more wrong. - Big Bang Theory

The reason that mv is not as popular today as it once was has *nothing at all* to do with the Internet, Web, Web Tools.

Nothing at all.  Not one bit. :)

Are you under 40 or something?  I mean this harping on modern tools, IDE, and the web completely ignores twenty years of our history

If you want to know why mv is not as popular you need to go back to the mid 80s (at least) and explain THAT first.

[Jan Van Schalkwyk]

No geneb, I've been here (in this group) since day one.
And quite frankly I've had enough of this group

I am also an old fart but I don't need these useless discussions that lead to nothing.
Some regulars that post here need a life and start doing something useful in the mv world.

Don't bother replying as...

This is my last post

Cheers

[Albert Kallal]

> f you want to know why mv is not as popular you need to go back to the mid 80s (at least) and explain THAT first.

Sure, and I believe the reasons I outline VERY much support my view. In the 80's, I think it is VERY fair to state that the 80's were a golden age of mv systems.

With the arrival of low cost intel based boxes, then all of sudden i could offer in place of a Novell network, or even a mv pick mini computer costing 20k?

I could walk in and with a R83 system setup a 6 user system - and do it for less cost then Novel, or those mv vendors selling the VERY lucrative monthly lease contracts that housed mv systems.  We now talking less then 10k for such a system. And by the early 90's? That's when we saw the start of the min computer vendors start to get in trouble, since we could replace such systems with lower cost intel/PC based boxes. 

And back then? Well, typical developer tools? What you had with mv land was as good or even better then most other systems  Since they were are par, and in fact I would give advantages to mv land?

Well, today, we have BOTH better IDE's, better tools and we ALSO have the web play. in fact the VERY two points I make? Well, during that BEST time and growth we saw for pick/mv systems also happens to be the EXACT same period where other IDE's did not provide those advantages, were not really all that much better then mv land.

And ALSO during that time frame, there was no web play or technology in this stew pot either. So I see this era as supporting my point front and center.

So a mutli-user system? They were expensive - and mv systems were good value. No one could compete price wise for my 6 user pick systems (tour operarator systems). And the dev tools and my ability to create software on mv systems were 2nd to none.

Today? Those mv tools don't compete, and they don't offer a seamless web stack, So the way I see this?

That gold age of mv land occurred based on the two big points I made here. The web did not matter, and the IDE's were not all that great back then.

Today? The IDE's and tools are much better, and we have the web to deal with!

so, it seems to be, that during the best times and best growth for mv systems? the two points I made that are relevant today did not matter back then, and now they do.

I don't enjjoy or have a market place leverage today like I could back then with mv systems. 

Another big change was Novell  fortunes went by the way side when windows networking became PART of the windows OS package and was essentially free per client work station.

Now all of a sudden, your per seat costs favored a local network over Novell, or even a pick box.  Back then, we did have to deal with per seat data base costs, but even by the early 90's, we saw a lot of wiggle room, and reduction in per seat database costs was starting to occur.

 So, the shift to client + server systems based on networks started to gain traction.

As the same time, those other IDE's and software tools became MUCH better. So much so, that the wonderful advantages of mv systems was reduced. mv was still better, but that gap was reduced to the point that the HUGE advantages became smaller and smaller. That was MUCH due to vast improvements in relational database systems, and their tools to mange such systems.

And now since those tools are great? Well, then I need tools not only for that line of business system, but those same tools and systems also provide me with paths to the web, and mv systems don't provide that easy path.

R

albert

[geneb]

On Thu, 15 Apr 2021, Jan Van Schalkwyk wrote:

> No geneb, I've been here (in this group) since day one.
> And quite frankly I've had enough of this group
> I am also an old fart but I don't need these useless discussions that lead
> to nothing.
> Some regulars that post here need a life and start doing something useful
> in the mv world.
> Don't bother replying as...
>
> This is my last post
>

Regardles, it was a joke. You apparently take yourself entirely too
seriously.

[joseba]

You are continuously saying " I need tools not only for that line of business system, but those same tools and systems also provide me with paths to the web, and mv systems don't provide that easy path."

Have you tried to find such tools? Do you like node-js, React, Angular, Vue, asp.net .........

Take a look to:

https://kosday.com/linkar-rest-api-how-to/

https://kosday.com/by-developers-for-developers-mv-basic-linkar-vs-code-extension/

https://kosday.com/Manuals/en_WEB_LINKARFRAMEWORKNODEJS/Main/

https://kosday.com/Manuals/en_WEB_LINKARFRAMEWORKJAVALIB/Main/

https://github.com/kosday/Linkar-with-node.JS

[Marcus Rhodes]

[Albert Kallal]

Sure, yes there are some better tools and plays one can make with vm-land. While you can get say VS to edit that mv-basic code? You still really don't get all that great of a library manager and the means to consume other code libraries with that setup. But , I sure wish I did have that when I worked in mv land.

And I repeated stated that mv systems can make the jump to web based, but in doing so the set of tools I will be using are not tools that come from mv land, nor are those set of tools and technologies that I used in mv land transfer as a skill set to those other tools. 

But the tools I am using now MOST certainly do leverage and transfer that skill set.

I mean, some are suggesting this is some kind of binary argument like 0, or 1. It not (never) that simple in regards to the mv market.

Some suggested that better licensing options - sure, that's a factor.

Some suggested some better free mv database editions with greater ease to get in the hands of developers? - sure that's a factor.

And some suggested (me) that the sets of tools and things from mv land don't translate well into those other platforms - and I also think that is a significant factor.

All those tools and links you suggest? Sure, they can be used, but the skill sets for using those tools don't come from mv land. That's not all a bad thing. but those tools don't really promote a mv development first environment. 

But, the tools i am using for SQL server, Visual Studio, and the web development I am doing? All those SAME tools are ALSO leveraged for desktop and projects that I work on. Those projects span desktop development, span web development, span database development (sql server). And those skills also get me Android development which I am also doing.

All those tools I use apply equal well to all of these environments.

So, if I have a choice between a set of tools and systems that gets me in the door for all of the above? What tools then do you think I going to adopt for development? Why of course those tools that open all those doors.

So yes, I stand by, and maintain that the IDE and sets of tools in mv-land are lacking. 

There are choices, but even then project builds, deployment? Well, that goes FAR beyond JUST being able to edit mv code in VS - but I will say that is a great option, and one I wish I had in my mv days.

So most of these tool plays for mv land that are kind of "me too"? 

they are still really built and for adopting the eco systems built around those tools - and they are not really mv centric. So they tend to lose even more benefits of what mv land offered in the first place.

At then end of the day? What is going to drive the market is me being able to take a set of tools and deliver software to that customer.  As I pointed out, hands down in the 80's, mv systems gave me that better option. But now it does not. 

I can state, in fact without ZERO doubts:

If a typical mv coder could just write some reports, and write code that produced nice ready viable web interfaces?

Then EVERY single long time mv coder in this group without ANY doubt would have spent the last 10 years building up all kinds of web interfaces for their customers.

Customer (or your boss) needs a report? You would create that - put it up on the internal web site, and you would be done.

Customer needs to see some project status - see some numbers and then approve that project, order, production or whatever? You would have chopped up some mv code, build a little screen and web page, and had the customer go to that page. Buckets and buckets of small little web sites and web interface systems would have spring up in mv land in the last 10 years. have ZERO doubt.

but, the base set of mv tools did not cross that bridge into web land. As I stated, .net was faced with this SAME problem. If Microsoft had not cooked up a web play and a means to LEVERAGE your base use of their tools and .net? Then no one or VERY few would be using .net right now. 

mv land and vendors had to endorse the web market and endorse it as PART of the deal and PART of the great and easy development model that mv offers.

And that means having provided a path, and a LEVERAGE of the existing tools, and existing skills that typical mv developers have. And that I dare say is a LARGE portion of the readers here (folks with great long time mv experience and tools - but VERY few have jumped on web development. And the reason was you entering a whole new world of hurt, and one that has and had little to do with mv land.

I mean, really, what developer going to dive deep into a set of tools that don't translate into web land development?

I recall the big change from green screen to GUI.  I had a better system then what Windows + GUI systems were offering. However, it fast became impossible to over come those market forces. My opinion on this matter did not matter!  What mattered was who was going to open their wallets, and why - and I can't change that.

So, I jumped from mv/pick to windows land. And my licensing costs went from around $300 per seat to $0!

But, it really was not the licensing costs that hurt or forced this change. It was customers wanting a GUI system - like it or not.

But, the web play while all cool and shiny? it not really whats driving the web play. What driving it is as I noted in my other long posts. it is the ability for a business to interact with customers in a new way, and save money by doing so. So that's where they spend money right now, and thus that's where all the work is, and thus that drives what tools you going to use.

In any software system?

I have to piece together bits and parts that allow one not only to MAKE software, but ALSO make software they are willing to buy and spend money on. As noted, there is no booming business in writing a payroll system. (but back in the 80's? Yup - wrote one for pick!).  

So, no real money in those payroll systems - they don't rust - everyone has one, back then? Wow - what a big deal and what a huge time saving investment for a company.

LOTS of companies were looking for, and buying payroll systems - or opening their wallets to build such systems. And that included mv systems!

So, a combination of the right hardware, the right fads (color screens, or GUI anyone?). And the right price, and the right cost of development. 

All these factors result in a soup pot mix that allows me to exist in that marketplace.  You miss one or some of these factors?

Then the reason to adopt and jump on mv systems - or any other system goes oout the window.

I still am of the view that if mv systems created and adopted a set of standards that leveraged the typical tool set and mv land into web land? We would be talking a VERY different tune right now.

All those mv folks using mv systems would have thus been faster to market with providing companies using such systems a web play. 

As it stands now? Those mv tools did not really facilitate that transition to the web market. There were some options - but the base tool set really did not help this transition to new technologist. And I do remember "version 1" of the pick systems attempt to provide a GUI for the desktop - it was lame, and too little too late. 

I mean, it became REALLY hard to sell Foxpro for DOS - so they re-vamped, and allowed those Foxpro/dbase developers to now build apps with a GUI. 

So, mv land didn't do all that great during the GUI transition in our industry, let alone the transition to web based. That was a ship that sailed and one that mv land failed to jump on.

So, I have no intention of claiming that my answer is the ONLY answer, but it is a factor in the mv market that struggles right now. 

And the other points/posts here? They are also valid points. 

You have to give me a platform, software and choices - and those set of choices have to be the BEST choice to turn that time + effort into money for me. At one time? With out a doubt, the mv systems I had and worked on ? 2nd to none - and were the best choice - both in terms of cost per seat, cost of development, and even cost of hardware.

Put all of the key ingredients into a pot - and if using that pot is compelling for me to use and make money (which means the users of such systems are making money), then sign me up. 

As I stated, hiring an employee that cost less then the benefits? Of course!

 Same as software - if a company is spending money on that software, they are spending money that provides more benefits then the cost of that purchase. Failure to offer me this deal? 

Then I am not using your tools until such time that soup pot makes me money, which means in turn the business consuming that choice is ALSO making money as a result of that choice.

R

Albert

[joseba]

I honestly think you are mixing a lot of concepts. I think it is a common mistake in the MV world to think that EVERYTHING must be in the database, everything must be able to be done within the database, and that manufacturers must include everything in the database. LINKAR for example is an MV PRODUCT, created by Kosday an MV company. You talk about libraries. LINKAR offers a set of CRUD libraries (and more options) for practically any current programming IDE (.NET, JAVA, JS, TS, NodeJS, C, PHP, Python ...) on Windows, Linux, and macOS. Allows you to send and receive information in MV, JSON, or XML format. You can use IDES like React, Angular, Vue for any web development using the Linkar framework libraries. You can develop mobile applications in IONIC, Xamarin ... that work on Android or IOS. You can use the C libraries from Android. And you can do it all from an IDE like VS Code with the addition of programming in the same IDE in BASIC and TypeScript, as well as being able to create dictionaries, document, even open a TELNET or SSH session from the same IDE (VS Code). You can do it with practically all MV databases (D3, jBASE, mvBase, Open QM, Reality, Universe, Unidata). All Linkar development frameworks are OPEN SOURCE and you can download the libraries for Linux or Windows or you can download their sources from GitHub to expand, reduce them, or do whatever you want. That is, you can use any of the development environments you speak about for SQL with the difference that behind it there is an MV database.

One thing is the development of the Client part and another the database. I think you mix it up.

You also find MV products, from MV companies that allow you a RAPID ZERO CODE development such as EVOKE (you're welcome Bob, you know I've always believed that we are on the same boat and on the same team) from another MV company (Bluefinity)

And neither Linkar nor EVOKE are “me too”. They are products with great technology. Linkar allows you to use “universal” development tools. Because for example, Visual Studio is a Microsoft development tool. POINT. It is not a development tool for SQL Server. The same could be said of Angular, VUE, ECLIPSE, BLUEJ, JCreator, NetBeans …… These IDEs then receive and send data from any source, SQL Server, Oracle, MySQL, Mongo DB, Cassandra… and D3, jBase, mvBase, OpenQM, Reality, Universe, Unidata.

And it's not about “building a payroll app,” it's about using the power and business rules of the payroll app you already have and adding value to it. And adding value does not mean sending the employee their payroll to the mobile (which also), it means the ability to record hours from a mobile when the employee travels, it means knowing the miles by car they travel, it means integrating with other products (for example, SaleForces)

You say “the sets of tools and things from mv land don't translate well into those other platforms”. Do you really believe it? Do you think that for example JSON is understood well with SQL and badly with MV?

You say “but the skillsets for using those tools don't come from mv land”. It doesn't care. If, as you say in your example, I want a form and a report on the web, I will give the developer the JSON interface that he needs to obtain data and the JSON interface that he needs to present it. And that programmer does not care if behind the database a SELECT is executed in MV, a BASIC SUBROUTINE is executed in MV, an SQL statement is executed in ORACLE, or a noSQL query is executed.

For example, if you want to get data from Mongo DB it will send something similar to:

db.customer.find ({name: "John"}

And you will receive something similar to:

{

"_id": ObjectId ("600c1806289947de938c68ea"),

"name": "John",

"age": 32,

"gender": "male",

"amount": 32

}

If you want to get data from a database, mv will send something similar to:

{FILE_NAME: 'employers',

      SELECT_CLAUSE: 'name = john',

      SORT_CLAUSE: 'BY name',

      DICT_CLAUSE: 'name age gender amount',

And you will receive something similar to:

{

"id": ‘101’),

"name": "John",

"age": 32,

"gender": "male",

"amount": 32

}

Ohh, it is EXACTLY the same for the developer!

From there we can talk about why use MV and not MongoDB.

But it will not be due to the limitations of the MV environment.

[Will Johnson]

So you respond by completely ignoring my point.

The internet, IDE, web, etc did *not* *exist* in the mid 80s

WHY of why did mv not sweep the world at THAT time

[Tony Gravagno]

Will Johnson wrote:

WHY of why did mv not sweep the world at THAT time 

Blame International Spectrum.

There. We're back on topic.

With that, I hope our colleagues will take anything not directly related to Spectrum to a new thread. Or rather to email, because we're not covering any new ground here.

Thank you.

T

[Nathan Rector]

Gee... thanks for the blame. :)

-Nathan

--
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 on the web visit https://groups.google.com/d/msgid/mvdbms/7cc8468b-e8b8-41e6-91d6-349096635119n%40googlegroups.com.

-- 
--------------------------------------------
Nathan Rector
International Spectrum, Inc
http://www.intl-spectrum.com
Phone: 720-259-1356

[Will Johnson]

Yes International Spectrum always seemed fluffy, they are probably to blame

I did enjoy the old Spectrum Tech editions that were discontinued many decades ago

[Eric]

I appreciate International Spectrum. They have been a big help to me.

 

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/4be82729-88ee-014c-9363-137d8da9b74e%40intl-spectrum.com.

[Albert Kallal]

Been busy:

>> WHY of why did mv not sweep the world at THAT time

Well, why did not Beta win over VHS if it was better? There are lots of reasons and factors. But in that time frame mv/pick systems WITHOUT question experienced a real growth, and without question it was a hey day of pick and mv systems. During that time - I went to spectrum in mid 80's - the vendors, the money, the sales -  "action"? it was incredible. And while in Europe, I went to C-BIT - the Comdex equivalent in Europe in the 80's. At c-bit I was able to find multiple pick vendors - so sure, they did not take over the market anymore then did Beta over VHS - but sure was a rocker and rolling time in mv land.  So even Europe vendors were doing well. Even Ultimate had offices in France.  Even IBM office in France was selling Blue Pick systems based on their RT platform. 

Throw a pick system at a developer today - now get them to build and make a web play with that choice? Not at all fun - nor does the mv system help much to achieve that goal. As I stated, there are many factors, but the mv vendors needed to leverage the mv systems, and the skill set one obtained in mv land to allow that same group of mv developers with ease to jump to the web. That's what saved .net, and such a move would have saved the mv land in a very big way. Not the only reason that mv struggles today, but it would have helped huge. 

R

Albert

[Albert Kallal]

>> I will give the developer the JSON interface that he needs to obtain data and the JSON interface that he needs to present it. 

And why do you need that web developer? How come the mv developer not able to do that? That is quite much my whole point. 

They needed to make a web play for the mv market - and mv land simply did not provide the tools and ability to leverage your mv skills. That skill set should have morphed into easy web development. As I pointed out, that's what occurred with .net - it was dead on arrival if they failed to add the web bits and parts to the base .net framework. All those TONS of developers were as a result saved, held on to, and were able to continue to use their .net skills and jump into web development based on that base set of skills. And this is the ONLY real reason why .net is still valid in the marketplace - it gave that community (the .net developers) a seamless bridge into web land based on using their previous sets of skills that they spent years learning.

No such bridge or path was provided to mv developers - and as a result few if any mv developers do web work. But I can tell ya darn near all .net developers did make the jump to some web development - and they were able to do so since they could use their current .net skills that took years to learn and use that same base skill set now for web development. 

As I stated, I have zero doubts that if mv developers had a easy path to web development based on their mv skill set, then EVERY single long time mv developer here would have spent the last 10 years making all kinds of web things for their customers. But now? Nope - you have to do as you suggest - hand that off to some web developer.  You should not have to do that.

It was and is a question of how big and hard of a bridge it was and is to jump on the web stuff. The easier that bridge and transition based on a given set of developer know how? Then the more web sites and web plays that would have occurred in mv land.  I mean, to be fair, SQL server, or oracle? That's just the database - and even mv ones can work well with web sites. The problem is what tool set, and how close and how familiar are the current tools, and what kind of bridge to web land do those tools offer. 

mv systems are a database, but that platform also represents a set of tools and means to create software - and it don't give me the tools for web land, and mv systems should have. That ship has sailed long ago - but it is what I would have done if I was the captain of the mv ship. And again, this is not the ONLY thing that would have helped mv in the marketplace, but it would have helped huge. 

R

Albert

On Friday, April 16, 2021 at 12:38:47 PM UTC-6 Will Johnson wrote:

[Will Johnson]

Name the factors.  I mean you're shooting at one goose in a flock of elephants.

If Pick was the end-all be-all in the mid80s.... What Happened Next.

Not thirty years later, five years later....

[Albert Kallal]

Sure, I think we all living what occurred.  As I stated, this not just some 0 or 1 binary type of position I have. I just think in the mid 80's, mv systems had a great run and such platforms were a fantastic choice - they were! And I not stating that ONLY due to say improving the dev stack and adopting a good web play for mv systems is the ONLY thing. As noted, I see this much like the Beta vs VHS debate. There are "many" reasons.

But what did happen 5 years later? Sure, by the early 90's the rise of low cost PC based networked systems occurred.  So in place of a 30k pick solution for 5 users or 15 users? I could adopt a windows based solution without any licensing costs on my side. So the "many" vendors that could and would adopt say a pick system for 15 users, or 20 users? They were starting to get eaten alive in the marketplace. With the rise of great pc based software tools like say FoxPro etc?

Well, all of a sudden, I would see point of sale systems and all other kinds of market plays now based on low cost PC hardware, and no central pick box was required. Those low cost computer network options? Well, in place of seeing say a 2-3 user Pharmacy system based on mv/pick? All of a sudden, such systems were being replaced with low cost Pc hardware - tools like FoxPro and several others were offering astounding performance and price points. And their user interfaces were also taking the market by storm. 

I had a great mv system - but the market was getting their pants all wet with GUI systems - and that's what business wanted and were buying. 

Toss in that the tools for development for SQL server and other database vendors were STILL moving forward, offering GUI and nice interfaces and GREAT TOOLS? Then the advantages that mv systems in 80s had were all but gone by the mid 90's.

So mv systems already missed the better GUI offerings. And by the late 90's, then the web was starting to appear. 

And now here we are!!! - so sure, the progressing of the industry? mv land missed most of the GUI revolution - that hurt it in the marketplace. And now we 20 years into the web revolution. that is  two big industry changes that mv systems were slow to embrace. 

I know I could not sell my mv system anymore - and it was the transition to GUI that was quite much what killed that market for me. Customers simply did not want to consider text based UI choices. But then again, I jumped over to Access based solutions. And toss in SQL server and I had not only a winning combo, but one that was without licensing costs to boot. And I had quite much zero license costs - even up to 20 or 30 users.  

I mean the transition to GUI also killed a truckload of PC based tools. dbase, FoxPro (DOS) and Paradox - the list goes on. 

And like any industry that  matures? used to be a video rental store near every corner. (and 5 computer shops on that same street!!!). Now block buster came along - all the little video stores gone. And then the web came along - blockbuster gone!

Seems to me these "big" changes such as the big gold rush that occurred when GUI became the hot ticket in the market place, and now the web is that hot ticket. In fact, looking at this general progression? Most text only dev tools simply fell out of use and out of viability in the market place.  And this relegated mv systems to being ONLY a database, not a set of tools to provide that solution to customers based around a GUI solution. 

In fact, this progression sealed the fate of most text only systems - and it don't just apply to mv developer tools - but any sets of tools that did not embrace these BIG industry changes form text to GUI. And in fact any sets of developer tools (even non mv ones) in the marketplace that failed to make the jump to GUI and then failed to make the jump to web? They barely exist in the marketplace, and they are insignificant players in the marketplace right now.

I find it VERY interesting that my above narrative not only applies to mv systems, but quite much all systems from that era that failed to adopt these major industry shifts. 

Software tools that did not support the GUI and then move on to the web are simply not all that relevant today - and that applies to the mv land in terms of those developer tools. 

Without adopting and providing developers with the new tools of the trade, then those vendors become stuck in their respective time periods of computer history - and quite much stopped moving forward the VERY same instance that such systems and tools did not embrace the new trends - that being GUI and now the web.

R

Albert

[Will Johnson]

You have failed to name perhaps the one most crucial factor of all

Not licensing, not cost, not visual perception (GUI), not peripherals (the rise of the mouse)....

You are missing one

[Albert Kallal]

I am open to other issues. As i stated, I don't think the GUI or now the web was the ONLY issue or factor.  I mean there are also some VERY good things that mv vendors did. The best move? They attached their wings to Linux - that was by far the best move. Pick systems was in fact the first commercial vendor at Linux world with a database system.  At that time Pick systems was working hard to "hide" the Linux side. But with time, and having to support peripherals  etc? Then they embraced Linux - and that was a real nice market play. All of a sudden that Linux server could not only support the database system, mv software, but it also make a great general purpose business server for files, and all kinds of other uses. A great time for mv systems that period was. 

I don't think the missing the boat on some of these major industry transitions is the only reason - but it was one of them - and without developers and vendors seeing mv systems as that first choice for business solutions - then it was going to be a hard up-hill - despite having that great mv system. 

I don't have or even think there is "one" issue alone here - but several of them piled up that much resulted in mv vendors now have a tough go in the marketplace. 

And I still think the web play for many of these systems can breath a lot of new life into such systems. You wind up skipping the GUI desktop revolution and skip right on to the web revolution - and mv systems can play in that field - you just not be using mv tools much for this web play.

But, this thread and discussion has truckloads of ideas  and suggestions - 100% open to that "big" issue that you feel was critical as for or against the mv market.

R

Albert

[Albert Kallal]

>. I think it is a common mistake in the MV world to think that EVERYTHING must be in the database

 

Not suggesting it should be. However, that vendor DOES have a responsibility to integrate these systems.

 

I mean, the LAMP stack is so common – that it quite much a “noun” or thing in our industry.

 

LAMP (Linux, Apache, MySqL, PHP)

 

But I don’t see mv in that term – do you?

 

So sure, some vendors have done a great job – and they package up and add and extend all those cool systems. But it done on a case by case bases.

 

As a result, each system is quite much a custom system – not a system based say on the above common stack.

 

Oracle always had high licensing costs. But they ALSO put into their platforms a standard development stack for their web play.

 

A common set of API’s, even based on Apache would have been fine and great for pick. Vendors MOST certainly have done this, but it not all that standard from one system to the next.

 

And those developer stacks are in most cases not mv centric.

 

As I just posted – I think the web area and web play holds the best area in which mv systems can play into the market right now.

 

But, the pool of developers is not using that LAMP stack much with mv systems, are they?

 

And as I stated? If IBM did not jump on to providing web servers as their product lines, then they too would have struggled.

 

And the same goes for .net. They morphed  their .net systems and products to embrace and WORK with the web - and made sure their web stack dev tools was to become part of that web play in the marketplace.

 

Joel on Software calls this the “user mental model”. And it is a important concept. If you take a given platform, and give those existing developers a set of tools that “conforms” to their existing known mental model for development? Then you can save that group of developers and companies that built their existing systems around that development model. And thus you capture years of experience and marketplace value – and you thus keep that platform alive and relevant in the marketplace. Oracle did this, Microsoft did this, and so did IBM.

 

And well Linux and “LAMP” – it rose up in that new period after those long-established vendors existed. Regardless? Those vendors did move and make plays into these new systems and embraced them as par of the course for their basic offerings. And that included extending their developer tools and practices.

 

I make no such claim that the database system needs to do everything. But the sets of tools that the vendors in general offer and WERE using MOST certainly do have to provide these stacks and provide TRANSITION technologies. In all of the above cases? We see the vendors of their systems and products having adopted web technologies as part of there overall offerings. And they ALL extended their development tools to play nice with those new web offerings.

 

You can rather easy fire up a web server and connect it to a mv system. But then you outside of the base vendors offerings. And if they do tack on say the LAMP stack? The base mv tools are not what you going to be using to achieve that end goal.

 

The core mv code systems really did not change from what they were  in the 80’s. That’s fine, but mv systems needed to vast increase that base set of tools to embrace the web. Since they did not? Then you now relegated to tacking on that LAMP stack – and as a result, then your core group of developers and vendors with their mv experience did not experience a easy transition to the web. The tools were and are there – but they not much mv centric.

 

So, things like the output of selects etc. should have obtained web output and web friendly features. They did not, so now mv quite much can only behave as the database engine. The server side language is great! – but again – its mv centric.

 

So, ZERO surprise that mv systems now are being pushed more and more towards being ONLY that database system. Since the base system and mv language never received extensions to work with web stuff – then we find that mv systems are not all that much of use other then being the database!

 

So, no surprise at all things are they way they are now.

 

Horrible that too many mv solutions are now relegated to connection technology that wind up flattening out the mv model when in fact the WHOLE web is moving in the direction of XML and now these days more even JSON – which really is a mv model!!!!

 

 

> POINT. It is not a development tool for SQL Server

 

Ah, but VS sure as the heck plays nice with SQL server. And there are BUCKETS of tools in VS that play nice with SQL server. You can quite much do everything in VS – and not have to jump to say SSMS (SQL server management studio) in a whole day of work.

 

So I respectively  disagree here. VS is a beyond rich tool and environment to develop with and for and against SQL server. This includes schema mannagment, schema compare tools, and even tools to script out changes from one database to another – in fact those tools don’t even exist in SQL studio – but they do in VS. The database server explorer in VS is really  nice – I could go on, but suffice to say? VS has a lot to offer in regards to playing nice with SQL server.

 

I only spend good time in SSMS as opposed to VS since I used SSMS for a long time. But, VS? It has HUGE leverage and features that plays right into the hands of SQL server. VS plays VERY but VERY nice with SQL server – FAR better then any other database server when working in VS – it not even close.

 

Again? My whole point:

The base developer stack has to play nice with the other tools and systems – including the web play.

 

You can tack on all the web stuff in the worlds for mv systems (you really can!!!). But as I pointed out, the instant you do so? That’s also the instant you are now moving away rather fast from using mv centric tools. And that’s quite much been my point all along here.

 

And then you wonder why the pool of mv developers is shirking?

 

It not that you can do these other things. It that the instant you do this, then you not looking at the mv development tools anymore. And from what I can see, that quite much describes the state of the mv development market right now as a result.

 

R

Albert

 

 

 

 

From: mvd...@googlegroups.com <mvd...@googlegroups.com> On Behalf Of joseba
Sent: April 16, 2021 2:07 AM
To: Pick and MultiValue Databases <mvd...@googlegroups.com>
Subject: Re: [mvdbms] International Spectrum

 

I honestly think you are mixing a lot of concepts. I think it is a common mistake in the MV world to think that EVERYTHING must be in the database, everything must be able to be done within the database, and that manufacturers must include everything in the database. LINKAR for example is an MV PRODUCT, created by Kosday an MV company. You talk about libraries. LINKAR offers a set of CRUD libraries (and more options) for practically any current programming IDE (.NET, JAVA, JS, TS, NodeJS, C, PHP, Python ...) on Windows, Linux, and macOS. Allows you to send and receive information in MV, JSON, or XML format. You can use IDES like React, Angular, Vue for any web development using the Linkar framework libraries. You can develop mobile applications in IONIC, Xamarin ... that work on Android or IOS. You can use the C libraries from Android. And you can do it all from an IDE like VS Code with the addition of programming in the same IDE in BASIC and TypeScript, as well as being able to create dictionaries, document, even open a TELNET or SSH session from the same IDE (VS Code). You can do it with practically all MV databases (D3, jBASE, mvBase, Open QM, Reality, Universe, Unidata). All Linkar development frameworks are OPEN SOURCE and you can download the libraries for Linux or Windows or you can download their sources from GitHub to expand, reduce them, or do whatever you want. That is, you can use any of the development environments you speak about for SQL with the difference that behind it there is an MV database.

--

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 a topic in the Google Groups "Pick and MultiValue Databases" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mvdbms/hitHNvfogo0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mvdbms+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/1b0aa7a0-0597-4c4c-9026-14a0918014f1n%40googlegroups.com.

[joseba]

--> "  I mean, the LAMP stack is so common " ...

--> " You can rather easy fire up a web server and connect it to a mv system"

--> " And if they do tack on say the LAMP stack? The base mv tools are not what you going to be using to achieve that end goal."

I think that you see the MV market as something where the players are only the database vendors. Anyone can use a MV server using standard tools. We have a lot of people that use Linkar and have no idea what a mv database is. And Linkar has a very very easy way to create web services and Linkar includes a web server, completely configured. So you only need somebody that knows how to use a development environment. You can develop with React, Angular, Vue, IONIC, or any tool. You can create Wordpress plugins in PHP. You can "forget" the webserver included in Linkar and use Apache, IIS..... ¿Is Linkar a database vendor product? No, it is a MV company product. But it is a MV product in the market that was created for all these things you are talking about. 

--> "VS is a beyond rich tool and environment to develop with and for and against SQL server. This includes schema management, schema compare tools, and even tools to script out changes from one database to another –"

I will really appreciate it that you have some time to take a look at the Linkar VSCode extension. If you or anybody want to take a look at it, we will love to do a webinar about Linkar VS Code Extension, Linkar Rest API,

 and Linkar Framework.