Makert Share D3 and Universe

[euobeto]

Hy... i always question myself why the share of MV of U2 developers and market is bigger than D3...  is there any obvious reason for this? or is there another reason?

--

Alberto  Leal
T.I  Campo Grande
LPI ID: LPI000191272
E-mail: alb...@tecwebcg.com
Gmail:  ees....@gmail.com
          

[Wols Lists]

On 22/03/2023 12:49, euobeto wrote:
> Hy... i always question myself why the share of MV of U2 developers and
> market is bigger than D3...  is there any obvious reason for this? or is
> there another reason?
>

Because U2 is a completely different product to D3? (In the same way as
MS-SQL-Server is a completely different product to Oracle.)

There's a bunch of versions (D3, Reality, a few more) all based on the
original GIRLS code. These have been pretty much swallowed up into D3,
and were hampered by Dick Pick's penchant for "marketing by lawyer".

Devcom cloned Pick, so to some extent were immune to Dick's lawyerly
antics (and were based in Oz), although this became Prime Information.
Plus - because it was a clone - they ditched a lot of the baggage from
Pick and the original design was very nice and well thought out.

Information was then cloned into UniData, UniVerse, and QM. Universe (or
rather its owners) mopped up the others, and Rocket has mopped up pretty
much the lot.

But the reality (sic) is that the U2 line used marketeers not lawyers to
gain market share, and so did a much better job of it. Not that even
they did that well, though.

Cheers,
Wol

> --
> Alberto  Leal
> T.I  Campo Grande
> LPI ID: LPI000191272

> E-mail: alb...@tecwebcg.com <mailto:alb...@tecwebcg.com>
> Gmail: ees....@gmail.com <mailto:ees....@gmail.com>

[Will Johnson]

D3 never had a great marketing team

And then of course when Unidata and Universe got out of the gate they could run on UNIX !!!

D3 only ran on Windows.  WINDOWS!  I mean WHO uses Windows as an ERP in 1990 ?

Nobody probably

Meanwhile everyone wanted Unix

[Bill H]

If I'm not mistaken D3 first ran on Unix then was later ported to Windows.  Instead of running within a directory with a shell on Unix I believe they ran in a kind-of virtual machine, where they carved out part of the disk and ran within that environment.  On Windows they created the same sort of structure with a kind-of virtual machine (I think it was called the VME) then allowed an external file system called the FSI (File System Interface).

Yes, U2 is different but D3 was originally ported to run on Unix and was later ported to Windows (I'm pretty sure).

[James A]

Maybe because U2 is a better product? That's always been my impression; though I haven't used D3 myself; my understanding is that it's more of a 'monolith'; i.e. one giant Linux/Unix file; and controls it's own processes (at least it was years ago; has this changed?); which leads to all kinds of issues for sys admins etc.

We currently have 15 'primary' UV servers; not including their backups and 'dr' (in another state) backups; heavily using rsync on selected files; a complex home-grown replication system that uses scp on specific files; etc.

- James Adrig

DocMagic, Inc

[Wol]

On 22/03/2023 19:27, Bill H wrote:
> If I'm not mistaken D3 first ran on Unix then was later ported to
> Windows.  Instead of running within a directory with a shell on Unix I
> believe they ran in a kind-of virtual machine, where they carved out
> part of the disk and ran within that environment.  On Windows they
> created the same sort of structure with a kind-of virtual machine (I
> think it was called the VME) then allowed an external file system called
> the FSI (File System Interface).
>
> Yes, U2 is different but D3 was originally ported to run on Unix and was
> later ported to Windows (I'm pretty sure).

I think it was D3 ... was the FIRST commercial database to be released
and marketed on Linux.

Cheers,
Wol

[geneb]

On Wed, 22 Mar 2023, 'Will Johnson' via Pick and MultiValue Databases
wrote:

> D3 only ran on Windows. WINDOWS! I mean WHO uses Windows as an ERP in
> 1990 ?

D3 never ran on Windows 3.0 or 3.1 :)

When I first ran into D3 in 1998, it had been running on SCO Unix for
quite a while. That's the host I was administering it on.

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_!

[geneb]

On Wed, 22 Mar 2023, Bill H wrote:

> If I'm not mistaken D3 first ran on Unix then was later ported to Windows.
> Instead of running within a directory with a shell on Unix I believe they
> ran in a kind-of virtual machine, where they carved out part of the disk
> and ran within that environment. On Windows they created the same sort of
> structure with a kind-of virtual machine (I think it was called the VME)
> then allowed an external file system called the FSI (File System Interface).
>

D3 on Unix/Linux uses a raw partition that isn't managed by the host OS.

D3 is pretty cool until you run into the fact that Rocket wants something
like $700/seat for it. When they bought OpenQM, they jacked the per seat
price up by 20%. It's obvious that Rocket has zero interest in increasing
their market share. They're going to ride the whole MV DB industry into
the ground and then congratulate themselves on how well they were able to
maintain their coke & hooker budget througout the whole ride.

[Jim Idle]

It’s statements like this that remind me to go back in time and have a word with my 24 year old self. 

--
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/211dd636-a32f-25f5-54d8-3509bb7b9515%40youngman.org.uk.

[Marcus Rhodes]

Yeah, that first Linux release was back in 1998, wasn't it?  But I also once installed D3 ... well, ok, AP ... natively on a standard PC (with a tape cartridge drive).  Did any of the others ever do that?

[David Knight]

Hi Alberto [& others reading this thread],

To throw in my 2c worth and summarise/correct what has been said by others so we can get a clear[er] picture:

I cannot confirm the genesis of Prime/Unidata/Universe/QA streams; but it sounds about right insofar as it was a separate beast not 'aligned' with the 'PICK' flavours.

'PICK' on the other hand was always a Virtual Machine [VM] system which had a 'monitor' layer that sat between the physical hardware and the VM. Pretty advanced stuff seeing as this was the 70's. Sooooo... that lead to a bunch of 'versions' of PICK which could work on 'other' hardware platforms, because only the monitor had to change and be platform-specific. Some licensed out to those same manufacturers with a 'name change' [think  'Reality' for example, McDonnell Douglas], and some under the 'PICK' name, so it became available on various sets of hardware. But 'natively'. That is, no underlying o/s. In effect for these 'versions' PICK was the o/s.

But then there was the rise of 'real' [for want of a better term] o/s such as SCO Unix, Linux, MS-DOS, early Windows [which was really MS-DOS anyway]. So the 'PICK' incarnations simply needed a new 'monitor' layer to sit on these o/s and away you went. Just like the various h/w platforms. Therefore, contrary to what some have said here; some early versions of 'PICK' were available on SCO Unix, Linux. Who delivered first, PICK or Prime/Unidata/Universe? I really do not know and cannot remember back that far; but PICK was certainly available on Unix quite early.

The comments of a political/marketing nature parallel these timescales are [sadly] also true. In fact I remember there being marketplace arguments about the 'war' between SCO Unix and PICK as to which was going to win the "o/s" war. While that was going on, Microsoft stole the show anyway! I guess you could argue that the o/s systems all won, not just SCO since the marketplace all moved to a relatively small collection of o/s while the database systems figured out their new role after the regime change. [At that time there were a few bastard-child PICKs such as an AP version which bundled an early Linux so that one "didn't require an o/s".]

Up until this time, conceptually the PICK database sat in a giant BLOB on the HDU containing executables AND data, so the comments about being 'monolithic' were true at that time. Since then 'PICK' had morphed into 'd3' as we pretty much know it now; and maintained the VM concept BUT; the BLOB split into two: The 'VME' [Virtual Machine Environment] which is kinda like the VM of old, containing SOME internal data and essentially resides in a BLOB; and the 'FSI' [File System Interface], which is hosted by the o/s using that system's filesystem.

Now the VME/FSI concept greatly modernised and opened up d3 [PICK?] since the data was no longer buried within a BLOB and enabled all sorts of other goodies such as the 'OSFI' [Open Systems File Interface] which is basically a method by which a d3 application can EASILY see data from pretty much anywhere, including other d3 and non-d3 databases. And vice-versa. But I am starting to ramble.

But NONE of this really answers your question, apart from providing some historical context.

I suggest the likely reason is 'marketing': Prime/Unidata/Universe etc did not allow itself to get bogged down; while sadly PICK under Mr Richard Pick's guidance did. And I say that with no disrespect to the man; whom I had the privilege of meeting and whose influence on the world of computing is largely ignored as a footnote when his ideas were way before their time. I say it because that is just the way it was.

That said, for me, the marketing guys stole the march and lead us to the current undeniable facts concerning the roughly binary split of the mv marketplace now controlled by Rocket: UniData/Universe [U2] is the favoured son; with 'd3' unfairly receiving the 'runt of the litter' award.

But if we ask which is the 'better' mv system from a technological/performance/power/flexibility standpoint, for me again it is no contest. d3 wins hands down.

I expect I've pissed off a number of U2 fans out there, and I am certainly in favour of a lot of the concepts I've seen in U2 which I wish d3 had. We can still be friends.

As for comments made about the price of a database system, I shall not engage: I will come back to 'marketing' as the answer since the world is full of more expensive and less costly 'things' ALL of which are purchased by people making conscious decisions.

All of which leads me in this essay to the crux of your question Alberto,  which I think really is: "Why does U2 have a greater market share when d3 is better?" The answer is mainly marketing and history.

So the real question we should be asking therefore is: What about tomorrow?

What do those who control these systems plan to do and are we all participants in that future?

Will the mistakes of the past be corrected, or will inferior tech rule the day?

I for one would prefer to see the best of U2 taken and merged into the underlying power and tech of d3 so that the combined STRENGTHs of each fit into a new replacement which allows us all to move forward.

Let's call it UD8, U2 TAKEN TO THE POWER OF D3.

Now THAT would be a thing.

David Knight

[Wols Lists]

On 23/03/2023 15:25, David Knight wrote:
> But if we ask which is the 'better' mv system from a
> technological/performance/power/flexibility standpoint, for me again it
> is no contest. d3 wins hands down.
>
> I expect I've pissed off a number of U2 fans out there, and I am
> certainly in favour of a lot of the concepts I've seen in U2 which I
> wish d3 had. We can still be friends.

As I die-hard Pr1mate, you can include me in that list of people you've
pissed off :-)

By the way, it was, I believe, Don Nelson who was ahead of his time, not
Dick Pick.

But imho the DevCom rewrite was true to the spirit of Pick, while
completely redesigning the implementation, and it shows. Okay, all these
"flavours" have muddied the water, but Pr1me Information was a very
clean, logical, well-thought-out system.

Cheers,
Wol

[David Knight]

Ahh yes, Mr Nelson. Point well made.

Perhaps an even stronger case for the shame of being but a footnote. He doesn't even get a mention.

[Jon Wells]

Hi Alberto,

I began working with MV using R83 Pick, which was running on a Monolith PC with a 386 processor.  We upgraded to a 486 processor along with a SCSI hard drive and tape backup system.  That allowed me to create a full backup every night.  I loved Pick's control of the backups as well as with restoring data.  If a record was damaged, I could load the last tape, and restore the one record, and everyone could keep on working.  

Of course, I did not care for Group Format Errors (GFE).  When we moved to the SCSI hard drive, along with a really good UPS / line conditioner, those became very rare.

I started to support a telemarketing unit which had a predictive dialer.  Data for the system had to be in dBase2 compatible tables.  We added PickLan as well as ViaDuct (with a PC running Windows 3.1).  This combination made it extremely easy to convert data from customers that arrived using all kinds of databases  We even added a 9 track tape machine since we had customers sending data that way.  Pick's data tape controls were very easy to use.  

The database I started with was designed by someone that worked in the world of manufacturing, and did not really understand the database needs of our non-profit.  So, the database design had a lot of limitations.  I was able to get around a lot the problems because of the power of Pick, however, we were growing beyond this.  The accounting package was a manufacturing oriented product that we had turned into a fund oriented system.  I wanted to move to a system that would have run on Advanced Pick, and was designed for non-profits.  The company's owner separated from  Author Anderson, and created his business that specialized in serving the needs of non-profits.  I think we would have paid about $35K for everything, including a lot of customizations.  [This was in the mid 90's.]   Then, a "free" consultant came along who clearly had an anti-MV chip on his shoulder.  He talked our CEO into putting out a RFP.  This RFP was crafted in such a way that he clearing was wanting us to go with Blackbaud's Raiser's Edge.  When Blackbaud decided we would need to go with Oracle to handle our data.  I remember telling my boss that there was no way we were going to be able to afford this.  Sure enough, it would have cost about $175K to make the move, along with hefty maintenance fees.  After dropping out from the Blackbaud deal, our CEO and our board were talked into using a system based on Visual FoxPro that I REALLY did not like.  So - I left.

After this, I entered the world of higher education, working with Datatel's Unidata based Colleague / Benefactor system.  There were a lot of things I liked about Unidata, however, I really missed the tape backup system from R83.  I also liked Pick's "sysprog" account, which handled a lot of things at a system wide level.  I created a Unidata "sysprog account" which started to be where I put utility programs, program templates, etc.  

After reading David's post, I suspect that I would really like working with D3.

Cheers,

Jonathan Wells

--
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/2dc27716-cf46-4f7d-9f43-8a7741c9a785n%40googlegroups.com

.

[Joe Goldthwaite]

I don't think any of the MVDB companies did much marketing at all to end users. It was the software companies that were writing applications on top of top of the database that were marketing it to the end users. I think word-of-mouth from those software companies were the only way anyone heard of Pick or MVDB. Most of the end users didn't know or care.

I've been working for many years at a good sized company. They started with a Honeywell Ultimate computer running Pick and later upgraded to a Data General running Universe on DGUX.

Do you know what all the users call it? It's the AccuTerm system. They call it that because AccuTerm is the Windows program they run to access it.

 

I remember talk that Dick Pick wanted to blow their entire marketing budget on a superbowl commercial!

There was talk in the Pick community about the Pick, Unix war or as Dick called it, the Pick eunuchs war. Ask people outside of the Pick community about it and they'd say "What's Pick?".

I think that the reason some were more successful than others is that they had better software houses writing and marketing applications for them.

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

--
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/4f0570e6-653a-4de7-ab6a-627ab14f63adn%40googlegroups.com.

[Joe Goldthwaite]

Just a note.  When I worked at Interactive Systems (CompuSheet+), we got some floppies with the Pick OS for the original 4.77 mhz IBM PC XT.  It worked just like any other native Pick system. There was no MSDos, Unix, Linux or anything else. It was slow but it worked.

I don't remember the year but there was no PC AT and I don't think the clone market had happed yet.

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

--
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/2dc27716-cf46-4f7d-9f43-8a7741c9a785n%40googlegroups.com.

[Ed Clark]

I think that would have been the infamous R83 ?

[David Knight]

That it would have been!

For me, that was the very first PICK I ever saw, running on the very first IBM PC - dawn of the PC era.

This was one of the 'monitor' incarnations I spoke of.

[euobeto]

hahah i'm baby dinosaur ... when I started working with MV, it was with D3 7.5

--
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/3e00e3da-e998-4a28-ab6f-7cf13df8aae5n%40googlegroups.com.

[geneb]

On Mon, 27 Mar 2023, David Knight wrote:

> That it would have been!
>
> For me, that was the very first PICK I ever saw, running on the very first
> IBM PC - dawn of the PC era.
>

My first exposure to Pick was using R83 on an IBM PC/AT via a Wyse 50
terminal.

[tomma...@aircraftspruce.com]

We had an Evolution and an ADDS that we timeshared. GFE city on the Evolution when the temp went over 65 degrees.

Tom

--
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/alpine.LRH.2.21.2303270938560.2511%40sidewinder.deltasoft.com.

[Ed Clark]

It’s probably a combination of marketing, name recognition, and history:

U2 has a mostly solid history. UniVerse and Unidata based themselves on Pr1me Information and had some solid advantages over Pick. They also both took some efforts to make it somewhat less painful to move to them from Pick. The 2 merged into Ardent in the mid 90’s, and at that time there was some concern that they appeared more interested in developing the O2 database than U2 (which I’m believe was named ironically in opposition to O2). Then they were bought by Informix, then IBM—both big brands, which reduced the pressure of users to switch platforms. Then Rocket acquired them. Basically, anyone who developed a U2-oriented application over the last 30 years had no reason to move to a new platform unless they had a compelling reason.

On the other hand, Pick Systems licensed their database to dozens of hardware vendors, so a good portion of pick application developers never had a relationship with Pick systems: they dealt with Adds, Ultimate, Sanyo, General Automation, Sequoia, Stratus, etc. Each of those licensed systems had unique advantages. For the ones who dropped support for their Pick databases, their customers had to choose a new platform—maybe not Pick-oriented. For those that were bought by other companies—GA consolidated several in the mid-late 90’s—customers were encouraged to migrate to the company’s flagship product, and might not have liked it. Eventually GA committed suicide-by-financial-error and was bought by Pick systems, but then Pick merged with Omins and changed their name to “Raining Data”, which was more interested in selling Omnis Studio than Pick. Then they changed their name to “Tiger Logic”, and their marketing made it feel like they weren’t interested in databases anymore at all. So people on the Pick/R83/AP/D3 track had many points in time when they might have moved away.

Lots of people moved from Pick to U2. I doubt many moved from U2 to D3.

Just my 2 cents of course. My perspective is that of an end-user developer, not a VAR, and both VARs and people who worked inside of Pick Systems might have better answers.. I have moved through Adds Mentor, Stratus, uniVerse. GA’s Power95, MVEnterprise, and jBase (which I would have said was the best 20 years ago, but haven’t seen it in a while), several of them at the same time serving different function, and with side-steps through Ultimate, Mentor PC/OS, Unidata, and pcVerse. It’s often commented on here that you can probably do anything you need to with *any* of the currently available multivalue platforms.

--
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/CAPZXEvPxe4yUdGH3kcb2p1UBAg3S_BEKnrKBcFpG_UmWp5CWZw%40mail.gmail.com.

[Will Johnson]

I have to say I do not believe Universe was based on Prime.

If anything I would say it developed from mvBase

Or maybe just R83

R83 by the way wasn't just PCs.  The development of R83 is what launced the massive explosion of Pick-alikes

Suddenly there were about fifteen hardware manufacturers who ported R83 onto their systems.

I think ADDS was the first and they actually called it R81 at first, because it came out in 1981

[Ed Clark]

Gosh no, universe long predates mvBase. I’m reasonable sure they were looking at Pr1me Information, because among their 6 emulation “flavors”, their default “ideal” is closest to their INFORMATION and PI/Open emulations.

I think there might have been even more than 15 licensees. IIRC the Pick Pocket guide has a list. I think you can find that on the JES website.

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/5d85b560-383f-41a5-9f44-326e26c76e67n%40googlegroups.com.

[Ed Clark]

Correcting myself, it’s “Exploring the pick operating system” at https://www.dropbox.com/s/f3eopjbsfolxlmi/Exploring%20the%20PICK%20Operating%20System%202nd%20Edition%20by%20Jonathan%20E%20Sisk%20and%20Steve%20VanArsdale.pdf?dl=0

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/B114FB05-6CCE-442B-A920-44F40DECD566%40aol.com.

[Joseba Real de Asua]

Some years ago Zumasys and the Tincat Group created an MV Family Tree poster where you can see the evolution. I cannot find the latest version, the last one is this, from 2015:

http://www.tincat-group.com/mv/MVFamilyTree2015.pdf

joseba

[Will Johnson]

Considering that mvBase is the Adds Mentor system, which began in 1981 I don't think you can say that Universe predates it.

ADDS was the very first R81 - R83 emulator system.

[Will Johnson]

Dawn has Cosmos Revelation leading directly with a solid line to Vmark's Universe

I find that problematic.  I worked on one of the first Vmark systems and I didn't personally see any relationship to Revelation

I wonder if the answer can be dug up with more actual sources than this graph Dawn created

[Wols Lists]

1981? That sounds a little late to the party.

I always thought UniVerse was an INFORMATION clone, and while I don't
know what version PI was in 1981, I first met PI in 1985 as version
(iirc) 5.3.

So UV could well have been around in 1981.

Cheers,
Wol

[Rob Allen]

My recollection is that uniVerse was based on a project called uPix, whose creator was later found to have had Pick source code - which is why the owners of uniVerse eventually had to pay royalties to Pick. A group of Pr1me resellers formed vMark, took over uPix, renamed it uniVerse, and added the features they needed most.

These are decades-old memories so accuracy is not guaranteed.

[jes]

[Will Johnson]

Well https://en.wikipedia.org/wiki/Rocket_U2

Someone should update this if there is more information on this history

I should say that Pick-"approved" versions were the Reality, next Ultimate Corp in 1979 and then in 1981 Adds

Once Adds created a software only upgrade approach (no chip swapping), suddenly you have an explosion of look alikes like Scan Optics and GA Zebra and Fujitsu and MicroVax etc etc etc

I mean in 1983 I think there were perhaps fifteen look alikes.  I certainly worked on at least eight.

I can't speak for what DevCom and Prime were doing behind closed doors, working with source code.....

I'm not sure the full story of that has ever been published.  Some one should get the lawyers to speak

[Wol]

On 29/03/2023 19:42, 'Will Johnson' via Pick and MultiValue Databases wrote:
> I can't speak for what DevCom and Prime were doing behind closed doors,
> working with source code.....
> I'm not sure the full story of that has ever been published.  Some one
> should get the lawyers to speak

Well, speaking as someone who worked with an early(ish) version of PI,
"True Pick" has always seemed very strange.

PI "never" supported PROCS or assembler (PROCS only appeared in the very
last version - v8.1).
"True Pick" didn't have PAragraphs or sit over a file system
PI didn't have Q-pointers.
PI had "D" and "I" dictionary items, and didn't have RPN in its
dictionary items.

"True Pick", on the one occasion I met it, felt completely alien in that
it seemed familiar on the surface but - seeing as I had been set a task
in a job interview - nothing worked the way I expected it.

The claim is DevCom rewrote it from scratch, and given all the stuff
that (a) got left out, and (b) new stuff that got added in, that
statement seems self-evidently true.

Cheers,
Wol

[Donald Montaine]

If I remember right, the five "founding mothers" of Cosmos had worked on a PC implementation of Prime Information for the PC.  They decided to create their own product, rented a house in the small logging town of Morton, Washington, and eventually produced the first version of Revelation.  It ran only on the PC.  The product took off and they opened an office in SeaTac, Washington a few blocks from the airport.  I visited there several times.  Revelation was very Information-like, Advanced Revelation had the same underpinnings but added a TUI IDE.

[Scott Ballinger]

It is my recollection (as an R83/AP/D3 developer since 1983) that:

• Prime Information was created by DEVCOM (Bellevue WA, maybe a Microdata dealer?) in the late 70s under contract by Prime Computer to develop a commercial database product for PrimeOS (written in FORTRAN).
• Universe was created by a group of renegade Prime dealers in the early 80s as a Unix alternative to Information that would run on non-Prime hardware.
  
• Pick Systems started directly marketing R83 on the PC (as an OS) sometime around 1987 or later. At that time Pick also had licensing deals with a lot of hardware vendors, including Honeywell/Ultimate, Intertechnique (later Siemens/Nixdorf), ADDS/NCR (Mentor,) General Automation (Zebra), Sanyo/Icon (Up Board), Sequoia, CDI (Seattle, IBM Series/1), Seattle Lab (Pick Blue IBM RT & RS6000), Stratus, Fujitsu, DEC, WICAT, Wyse, Alpha Micro, Pyramid, Altos, Rexon, Encore, and god knows how many others. (Dick Pick was a very promiscuous licensor!)
• Eventually Pick Systems morphed R83 into AP (1989?) and directly marketed it as AP/DOS which ran natively on PCs as the OS, and on Unix (AP/SCO and AP/AIX), and then AP/Pro which ran on freeware Red Hat but tried to hide the Linux part and eventually AP/Linux which put Red Hat out front.
• By 2000 Pick Systems had become Raining Data after Dick died, which became Tiger Logic and AP became D3, with versions that ran on WindowsNT, various flavors of Unix including SCO, ATT, HP, IBM AIX, and more?, and Red Hat Linux.
• Eventually D3 consolidated onto NT (later Windows Server), Red Hat Linux, and IBM AIX.
• Later, Rocket Software bought everything.

/Scott Ballinger

[Felipe Gordillo]

Sorry, but I think AP/DOS didn't run natively on PC as OS. It is a program that runs under MS/DOS, just like AP/SCO and AP/AIX, which ran under SCO and AIX respectively.

The only AP that ran natively as S.O. it was AP/Pro.

Greetings.

Felipe Gordillo

To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/5866878d-d40c-4ddd-a173-e9f15158f82dn%40googlegroups.com.

[David Knight]

I think that is what Scott meant: 

R83 installed and ran 'natively' onto an XT and later the AT early pc's of the day. [See my essay on the concept of the 'monitor layer']. In that case PICK provided it's 'own' OS and monitor to suit. Then along came AP which as others have said suddenly ran on a bunch of OS's including MS-DOS [& thus technically early Windows] simply by supplying a different 'monitor'. So AP-DOS ran on MS-DOS; AP-SCO ran on SCO Unix etc etc. To 'keep' the idea of being able to take over the whole machine as in the R83 days; one incarnation bundled a version of Linux [as others have confirmed] so it 'behaved' more like the PICK of old. I never personally used that; but have started with R83 which then went through the AP stage where I used AP-DOS, AP-SCO; and finally onto 'd3'.

A minor side-step involved Fujitsu and the IBM RISC mini computers which for their days were awesome!

[geneb]

On Tue, 28 Mar 2023, Donald Montaine wrote:

> If I remember right, the five "founding mothers" of Cosmos had worked on a
> PC implementation of Prime Information for the PC. They decided to create
> their own product, rented a house in the small logging town of Morton,
> Washington, and eventually produced the first version of Revelation. It
> ran only on the PC. The product took off and they opened an office in
> SeaTac, Washington a few blocks from the airport. I visited there several
> times. Revelation was very Information-like, Advanced Revelation had the
> same underpinnings but added a TUI IDE.
>

I've got a pile of Cosmos, Rev G and Advanced Revelation stuff. I was
heavy into it due to a friend that worked there in the late 80's early
90's. I wrote an extension that would allow access to EMS from R/BASIC,
but it never went anywhere. :(

[Scott Ballinger]

Felipe, you are right. An AP/DOS machine booted DOS (or Windows?), then invoked the Pick environment with the "pick" command.

I think AP/DOS was limited to three users (monitor + 2 serial ports?), whereas the R83 version of Pick available at the same time of 286/386 machines could support up to 17 users (monitor + 2x 8-port serial card).

[Will Johnson]

It's certainly possible that Pick Systems did not  market R83 until 1987, and then only on PCs.

But R83, as it's name suggest was created in 1983 and already on the Adds in 1983

The Adds had an earlier version they called R81 in 1981 when they started selling it.

I never knew or heard that Universe was created by renegade Prime developers to run on non-Prime hardware

So I learned something new today.

[chandru murthi]

>By the way, it was, I believe, Don Nelson who was ahead of his time, not

Dick Pick

No need to set up a rivalry there. Nelson indeed had the original ideas behind the Pick file system, GIRLS, etc., and also did develop detailed flow charts from which the first GIM system was implemented on the IBM360. A tremendous accomplishment, but he was a thinker, not a doer.

Dick, however, was way the driving force behind the sophistication of the Pick system and its marketing. He had the genius to recognize the power of microcding the Microdata 800/1600 boxes, without which there would have been no way an 8-bit 1mHz machine could've supported 8-12 interactive terminals with  max 65K memory. His marketing "genius" developed severe hiccups as he accreted more systems and licensees, too bad.

Chandru Murthi

On Thursday, March 23, 2023 at 10:37:10 AM UTC-7 Wols Lists wrote:

On 23/03/2023 15:25, David Knight wrote:
> But if we ask which is the 'better' mv system from a
> technological/performance/power/flexibility standpoint, for me again it
> is no contest. d3 wins hands down.
>
> I expect I've pissed off a number of U2 fans out there, and I am
> certainly in favour of a lot of the concepts I've seen in U2 which I
> wish d3 had. We can still be friends.

As I die-hard Pr1mate, you can include me in that list of people you've
pissed off :-)

By the way, it was, I believe, Don Nelson who was ahead of his time, not
Dick Pick.

But imho the DevCom rewrite was true to the spirit of Pick, while
completely redesigning the implementation, and it shows. Okay, all these
"flavours" have muddied the water, but Pr1me Information was a very
clean, logical, well-thought-out system.

Cheers,
Wol