Most Popular Flavor of PICK

[Thomas Rectenwald]

Hello all.  I am new to PICK and currently learning on a TigerLogic D3 PICK system.  I have heard of other types such as UniVerse, jBase, Revelation, and well, I'm sure there are many more.  I'm just curious.  Which one of these flavors in the most popular, i.e. most used and supported/actively developed?  I'm not asking which one is the best! :-)

Also, how difficult is it, in general, to pick up a different flavor?

Thanks,
Tom

[George Land]

People may argue about 'popular' but from a market presence perspective the leader is Rocket with U2 - UniVerse and UniData.  Of course they aren't Pick, nor are jBase, Revelation et al.  But they are multi-value databases that are widely used, are developed, supported and growing.

As for picking up a different 'flavour', the principles are the same but the details differ.  If you know one it won't take long to learn another though

George Land
APT Solutions Ltd
(Rocket's U2 UK Distributor)

[Anthony Youngman]

On 22/09/13 18:04, George Land wrote:
> People may argue about 'popular' but from a market presence perspective
> the leader is Rocket with U2 - UniVerse and UniData. Of course they
> aren't Pick, nor are jBase, Revelation et al. But they are multi-value
> databases that are widely used, are developed, supported and growing.
>
> As for picking up a different 'flavour', the principles are the same but
> the details differ. If you know one it won't take long to learn another
> though

Mmmmmmmmmmmm....

Just be aware that Pick systems fall into two groups, and there's quite
a culture shock going from one to the other. Pick itself is seen as the
parent version, but there are several versions of "true Pick" derived
from the original Public Domain GIRLS code. These are all pretty much of
a muchness, and have mostly been absorbed back into D3 - Pick Systems
over the years has bought them all out.

Then there's the second group - derived from the DevCom "rewrite from
scratch done right". Although very similar in philosophy, it is very
different in implementation. Most of these have likewise been absorbed,
in this case into Rocket's U2. I'm very much in this camp, and struggle
when I hit "true Pick" idioms.

That said, pretty much any modern implementation can handle either group
no problem - I've worked on a site which was originally "true Pick"
ported to UniVerse (in the DevCom camp). I guess it ported easily - I
wasn't there - but I was forever tripping up over "Pick-isms".

At the end of the day, just be aware that while all the stuff that you
learn in the land of "true Pick" will work fine in the land of DevCom,
don't forget that they speak with a somewhat different dialect there :-)

Cheers,
Wol

>
> George Land
> APT Solutions Ltd
> (Rocket's U2 UK Distributor)
>
> On Sunday, 22 September 2013 16:11:32 UTC+1, Thomas Rectenwald wrote:
>
> Hello all. I am new to PICK and currently learning on a TigerLogic
> D3 PICK system. I have heard of other types such as UniVerse,
> jBase, Revelation, and well, I'm sure there are many more. I'm just
> curious. Which one of these flavors in the most popular, i.e. most
> used and supported/actively developed? I'm not asking which one is
> the best! :-)
>
> Also, how difficult is it, in general, to pick up a different flavor?
>
> Thanks,
> 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

[Kevin Powick]

On Sunday, 22 September 2013 11:11:32 UTC-4, Thomas Rectenwald wrote:

 

 Which one of these flavors in the most popular.

Mauve.  It has the most RAM.

--

Kevin Powick 

[Thomas Rectenwald]

Thanks guys.  Very interesting stuff.  I wasn't aware of a schism as such.  Really appreciate it.

[Thomas Rectenwald]

Okay, one more question!  What is the difference between UniVerse and UniData.  Which one is more on a one on one basis with D3?  Thanks again, I'm from a strong Linux/Unix background so really find this fascinating as it is so different.

On Sunday, September 22, 2013 11:11:32 AM UTC-4, Thomas Rectenwald wrote:

[Brian Speirs]

Neither. Both of the U2 databases are based on the Pr1me Information way of seeing the multi-value world. As such, both have moderate differences from the R83 Pick databases.

Some differences are probably there just so Pr1me guys weren't accused of copying Pick. For example, the ED editor uses a C command to change something while Pick uses an R command to replace something. Likewise, the stacker on each type of system is inverted relative to the other - Pick has recent commands at the top, while Information has recent commands at the bottom.

Other differences are more structural. For example, when the query languages process multi-valued items, the Pick databases call the processing routines once for every value, whereas the Information databases process all values in one hit. That difference means that you need to structure the dictionary items a little differently.

The closest databases to D3 are probably mvBASE and UniVision (which I haven't heard anything about for ages, but the website is still up). mvBASE is now owned by Tiger Logic and is effectively a sister database to D3. They also own mvEnterprise - but I know nothing of its nature.

HTH.

Brian

[Martin Phillips]

Hi Tom,

 

Which one is most actively developed? We like to think that QM leads the pack on new innovation though I am sure that others will argue and I am, of course, biased on this question.

 

Re picking up different flavours, as stated by in another response, multivalue products essentially fall into two groups; Information style (following the model introduced by the Prime Information product) and Pick style (following the original Pick model). They are very similar in most respects but have some significant differences in others. Again, I am biased but my feeling is that the Information style products have greater capability and get away from some of the less friendly features of Pick systems (e.g. use of I-type dictionary expressions instead of F-correlatives). This is a personal opinion and I accept that Pick users may disagree.

 

QM, UniVerse and Unidata (maybe more) have option switches to enable features selectively from the two styles so you can to some extent invent your own style.

 

UniVerse and Unidata started life as rival products, both trying to capture existing multivalue users from several products as systems moved from proprietary operating systems to Unix. Given this history and purpose, it is not surprising that they are very similar – indeed, we deliver a programming course that is over 95% identical for the two environments. Since Ardent Software brought the two products together in the one company a good few years ago, they have gradually become closer though there are some significant differences that make it likely that they will continue to be two separate products.

 

 

Martin Phillips
Ladybridge Systems Ltd
17b Coldstream Lane, Hardingstone, Northampton NN4 6DB, England
+44 (0)1604-709200

--

[CDMI - Steve T]

for my 2cents, I really love the QM environment.

runs on just about anything you would want including a USB stick on Windows.

the price per user is the best in the industry.

Martin and crew give top notch support and enhancements.

I'm a lover of the QM/Linux environment, yet when I had a customer come to me with over 200 reps in the field, we simply developed QM apps on the Windows QM/USB stick with cloud server (ie: the mothership) integration when connected to the internet.

absolutely wonderful, reliable, portable, stuff!

take care,

 

Steve Trimble

Computerized Data Mgmt Inc
2705 Justin Matthews Dr
N Little Rock, AR 72116
(501) 615-8674 09:00am - 6:00pm CST

---

From: Thomas Rectenwald <ter0...@gmail.com>
To: mvd...@googlegroups.com
Sent: Sunday, September 22, 2013 10:11 AM
Subject: [mvdbms] Most Popular Flavor of PICK

[Ed Clark]

There's some useful historical information at http://www.tincat-group.com/mv/familytree.html. The pdf lists a good number of the companies and products involved in multivalue over the years, and there are some other links there to multivalue information.

The pdf is about 9 years old I think, and there have been a few changes. Most notably, Intersystem's Cache database is missing, and QM is only lists as a "see also"

The current products (afaik, someone please correct me if I'm wrong) are:
D3, mvbase, and mvEnterprise from TigerLogic (which used to be called Raining Data and before that was Pick Systems. )
Reality from Northgate (which used to be called Microdata/McDonald Douglas)
Jbase from MPower1 (jbase was an independent company, then bought by Temenos who sells a widely used financial application.)
Revelation from Revelation
universe and unidata from Rocket (these databases have changed hands a few times. They used to be owned by IBM and before that by Informix and Ardent. Before Ardent, universe had been developed by a company named vmark and unidata by a company named unidata.
Cache from Intersystems.
QM from Ladybridge. QM also has an open source version OpenQM.
OnWare from OnGroup. (I think this one is still developed and sold)

The pdf also lists uniVision, which I think is dead. By dead I mean that it isn't sold anymore, but there are still people using it and getting support from third parties.
I strongly suspect that most of the "dead" databases listed in the pdf are still in use somewhere. Multivalue systems are hardy.

There are some open source multivalue systems. OpenQM, and ScarletDB (I think that's the name) which is a branch from OpenQm. The pdf also lists Maverick, Bart, and Winter, and there are a couple of others that I don't remember the names of.

As mentioned by someone else, Pick systems licensed the database to many other companies, some of whom did creative things with it. Some of the more interesting licensees were:
ADDS (which for a while was owned by NCR which in turn was owned by ATT) who sold the Mentor database. mvBase is based on mentor.
Ultimate, which sold a database by the same name. After they stopped selling the database, another company revived it as UltiPlus.
Sequoia, which had a bulletproof, fault tolerant version of pick. MvEnterprise is based on Sequoia.
General Automation, which in the late 90's bought and consolidated a lot of the remaining pick licensees, but then had an epic fail because of bad accounting practices and ended up being merged with (bought by?) pick systems.

Look back through old group posting here and on the comp.databases.pick usenet group for more historical information and general multivalue information?

Which database is best? Well, you didn't ask so I won't answer except to say that arguably every currently available multivalue system can be used to develop great applications. Each has it's advantages depending on what you want to do. For example, most of the databases run inside of their own virtual machines (and historically, multivalue systems were both the database and the operating system on custom hardware), but jbase compiles each program to an os executable. OnWare (iirc) runs as an emulation layer on top of Oracle, so might be ideal if you mostly use Oracle. MvBase only runs on Windows, and has unique integration with that platform. Revelation provides gui tools that let you use *other* multivalue databases as the back end. Cache allows you to write programs not only in multivalue basic but also in a vb-style basic and a mumps-like language, and is object oriented. QM has an open source version and is inexpensive to license.

Which is the most popular? d3, universe, and unidata are the ones most talked about.

Which is the most actively developed and supported? I'm biased (I work for Intersystems), but I'm sure that Cache is the most actively developed, and provides the best support.

[Ed Clark]

fyi, I had nothing to do with the multivalue family tree I linked to. That was the work of Dawn Wolthuis. Look in particular for her posts here for background info about multivalue history.

[Ed Clark]

The core of all multivalue systems are the same--you have items in files delimited by attribute and value marks, an editor, proc, a query language, and a BASIC programming language which has special commands to access the database. Beyond that there are tons of little differences and bugs that might or might not make you crazy when moving from one system to another just depending on how you use multivalue.
One way to get a handle on the difference between universe and unidata would be to use them. Download the personal editions and try them. Read the manuals. In particular look at the UDT.OPTIONS manual in unidata. Every one of those options represents a difference between unidata and other platforms which you can override. Also look at the universe documentation for the basic $OPTIONS command which shows how universe emulates other platforms. The differences between the list of things you can change is representative of the differences between universe an unidata. Depending on how you use multivalue, you may or may not ever experience any of these differences.

If you play with universe or unidata, make sure to start out using universe in PICK flavor or unidata in ecltype/basictpye "P". those will be most like D3. My comments below mostly apply to working in these flavors.

Some that come to mind:
In query, universe supports A-type dictionary entries with correlatives on line 8. unidata does not. Both universe and unidata support I-type dictionary items (which d3 does not) but as Brian commented, these operate rather differently

In proc, universe always uses attribute marks for the buffer delimiter, and for the most part doesn't distinguish between PQ and PQN proc. Unidata uses spaces for PQ proc, but has no documentation for proc. (D3 only supports PQ proc and not PQN which is an advanced version).

The spoolers work differently between universe an unidata, and neither is much like D3's. There are 3rd party programs that can be used if you are trying to migrate a spooler-intensive program.

In basic:
0001 EXECUTE 'SELECT SOMEFILE WITH NAME = "SMITH"'
0002 EXECUTE 'LIST ANOTHERFILE' CAPTURING OUTPUT
On D3 and unidata, the list on line 2 will use only the items selected by the select on line 1. On universe, it will list the entire file. List passing is very different between universe and unidata.

iirc, The LOCATE statement has different semantics on unidata in basic type P than it does on universe in pick flavor.
There are lots of little differences where things like
READV VAL FROM FP,ITEMID,X THEN ... ELSE ...
will produce different results if X <= 0 or will either change or leave unchanged the value of VAL if the ELSE clause is followed. (this specific example may not be correct, but there are lots of differences like this)

When you use EXECUTE on unidata, it doesn't really create a new execute level like it does on universe, so @variables stack differently between execute levels.

Cataloging is different. On D3 an universe in pick flavor, cataloging is local by default. You don't need to recatalog every time you recompile. On unidata, you need to purposely catalog locally.

mvEnterprise is probably the database closest to D3, but I don't know if TigerLogic actively sells it to new customers. It's also somewhat moot because you're not going to get a free demo version of mvEnterprise, and if you did, afaik it still only runs on AIX. mvBase is fairly close to D3, but you're also not likely to get a free demo of that. You can get free demos or personal editions of universe, unidata, QM, Cache, and I think still jbase and reality. Maybe Revelation has demos too.

[Anthony Youngman]

On 23/09/13 09:25, Brian Speirs wrote:
> Some differences are probably there just so Pr1me guys weren't accused
> of copying Pick. For example, the ED editor uses a C command to change
> something while Pick uses an R command to replace something. Likewise,
> the stacker on each type of system is inverted relative to the other -
> Pick has recent commands at the top, while Information has recent
> commands at the bottom.

Actually no. DevCom predates all the lawsuits and, as I said, GIRLS was
public domain so copying was just fine and dandy.

The differences run a lot deeper. INFORMATION *NEVER* (well almost) had
PROCS officially - it first appeared in the very last version as a
result of the SPECTRUM unification efforts. The two dictionary styles
are very different - and INFORMATION's I-descriptors are very different
from the Pick equivalent.

PAragraphs are the INFORMATION equivalent of PROCS, and again are very
different.

Most features are now available in all modern Picks, but the fundamental
differences between the two families were much more marked in the past.

Cheers,
Wol

[Kevin Powick]

On Sunday, 22 September 2013 19:48:22 UTC-4, Kevin Powick wrote:

 

Mauve.  It has the most RAM.

In case anyone was wondering, "What?"

http://dilbert.com/strips/comic/1995-11-17/

--

Kevin Powick

[Mike Preece]

jBASE is implemented wherever Temenos T24 is implemented - which is at a lot of banks around the world. If you or your company want to implement jBASE however you will be some releases behind the latest version implemented with T24 I believe. Temenos is unique in the Pick/MV space due to being the only application vendor that owns the database vendor. jBASE International is a subsidiary that supplies jBASE to everyone other than Temenos.

[Ed Clark]

Does Ladybridge sell any applications running on QM? That would put them in the same category as Temenos/jBase, though jBase was originally developed and sold independently, and then Temenos, which had been using uniVerse, bought jBase and migrated their applications to it. Ladybridge developed QM themselves, and then open-sourced it

Some other interesting relationships:
iirc, ADP licensed Pick and maintains their own version for their applications.
Revelation has a multivalue database, and a tool named OpenInsight, but OpenInsight can also be used on other multivalue databases.
There are probably other companies who developed their own multivalue databases and run applications on them.

[Martin Phillips]

Hi Ed,

QM originally grew out of an embedded database package that we wrote to build application software for our clients. Nowadays we
concentrate on the core database package and have less direct involvement in applications though we are always interested in
development projects.

We do have one application that we sell. Originally written for St John Ambulance, a UK medical charity, our ambulance control
application is now also used by the British Red Cross and several private ambulance companies. It is designed to handle local events
such as sports venues rather than the wide area cover required for 999/112 (911 for US readers) services.


Martin

[Ross Ferris]

EDP in the UK sell & develop applications, AND the Univision database

[Rick Weiser]

Right now the most popular is Rocket as they just purchased D3 for 22 million

http://prn.to/16gpWIs

[Jan Van Schalkwyk]

Now more than ever ;-)

[Dick Thiot]

Rocket isn't a flavor of PICK.  U2 isn't a flavor of PICK.  Universe, Unidata, D3, mvBASE and mvEnterprise are flavors of PICK as well as lots of others include QM and jBASE.

Dick

--

[Rick Weiser]

WOW, Dick, a little touchy ?

[jes]

On Thursday, October 17, 2013 4:48:02 PM UTC-4, Rick Weiser wrote:

WOW, Dick, a little touchy ?

Thankfully, Dick is not a flavor. 

[Dick Thiot]

Rick, i think that you read me all wrong.  I just want to identify the flavors as database versions and the companies as companies.  I am sure my terseness came off a little too strong or condemning.  That was not my intent.  It was only commentary.

JES, agreed again!

Dick (not Pick)

--

[Ed Clark]

So, a list of vendors, as opposed to platforms/databases/flavors/products/cults, ordered roughly by descending number of seats out there that you could target a multivalue application to:

Intersystems
Rocket
Temenos
Ladybridge
Northgate
Revelation
ONGroup
EDP? (Can someone verify that you can still buy UniVision?)

Did I miss any?

[Ross Ferris]

I have to wonder about ONGroup ... is this still be actively sold in the market. Don't hear from them these days

[Dawn Wolthuis]

I beg to differ.  --dawn

--

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

--
Dawn M. Wolthuis

Take and give some delight today

[Charlie Noah]

Oooooh, you bad girl! ;-)

[Rob Allen]

I get emails from ONGroup every month or two; their new name is "Onsystex" - http://www.onsystex.com/.

And both EDP and Via still offer UniVision on their websites.

[Bob Markowitz]

How can some of you people not mention jBASE in your posts.  We have been around for about 25 years and were one of the first implementations to run natively on operating systems.

The core of jBASE development is still done in the UK at Hemel Hempstead just down the street from where it all started.  In fact several of the early key developers are still there. They are a little older but a whole lot wiser.  The jBASE development staff is probably the largest in the Pick market with over 20 people.  The intellectual property of jBASE is owned by Temenos.  They are a publicly held Swiss company marketing T24, a jBASE based international banking system. 

MPower1, jBASE International's parent organization has the rights to market, sell and fully support jBASE outside of the banking industry.  MPower1 is a UK based company.  We offer a clear alternative to Rocket's offerings. 

I noticed a comment about Northgate, Ladybridge and jBASE developments in the UK.  i guess there must still be some smart people there. 

Bob 

[Dawn Wolthuis]

Of course many of us are familiar with jBASE!  So, should Mpower1 acquire Northgate or the other way around? Which one of you will twist Martin's arm and get a nice gem of a product there?  smiles.  --dawn

--
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

[George Land]

I didn't count but last time I was in Denver there were rather more than 20 people in the U2 development team and that's only the ones working out of that location.  Of course that's if you put U2 under the banner of Pick, something that I avoid wherever possible!

[Dan McGrath]

Thanks George - I can attest that there are *many* more than 20 on my team focused exclusively on R&D for MV technology here. On top of that there is also Vinnie's team for the paid-for tooling which adds even more.

[George Land]

Northgate are kind of large!  Revenues of around $250 million last year, no disrespect to Mpower but that might be a bit too big for them to acquire.

[Dawn Wolthuis]

Did I say that with my outside voice?  Apologies.  smiles.  --dawn

[Dawn Wolthuis]

I think some folks left ONgroup http://www.ongroup.com/web/ ad fored onsystex. I do not think that ONgroup changed into Onsystex.  --dawn

--

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

--

[Rob Allen]

Wow, you're right, they do seem to be  different. I jumped to a conclusion there. 

The ONGroup website looks like it hasn't been updated in years - at the top of their "News" page is an item about Microsoft's new "SQL Server 2005" release. Maybe they really are defunct. 

[Dawn Wolthuis]

I don't think they are defunct, just perhaps fewer people there.  --dawn

[Rick Weiser]

OnSysTex is not OnWare.  They are a completely different company with a rival product, OASys.

Rick

[Dawn Wolthuis]

Does OASys have a DBMS? Do they have their own BASIC or MV Query implementations? Does DesignBAIS work with OASys? With ONware?  Thanks.  --dawn

[Ed Clark]

Has anyone looked at OnSysTex/Oasys?

If you go to OnSysTex web page at http://www.onsystex.com/ and mouse over the "products" on the menu, the bottom 2 entries in the list are DesignBais and OnWare. The Onware link is dead, but it suggests that they have support for it and DesignBais in their products.
If you go to The Leadership Team page (under "about") there are some familiar names: Gil Figueroa and Ian Sandler.

[Rick Weiser]

Dawn,

Yes, OASys has a native MV database but also all the hooks to work with relational DBs.

Yes, DesignBais works very well with OASys.  As a matter of fact, we were using OASys as our backend database for our Hands-On Developer Conference last month.

Rick

[Dawn Wolthuis]

Good info, Rick. OK, I like stories like this. I did talk to some of the OASys folks when they started, but I did not know they had their own version of MV. So where did this version come from? I would be surprised if it is a "clean room" version. What code was brought over from where/whom?

Thanks.  --dawn

[Dawn Wolthuis]

Does anyone know the answer -- where did the version of MV used by OASys come from? Is it from the ONgroup? Did they buy it? I'm thinking it has to be a fork of something -- very curious. Does anyone have some clues?

Thanks.  --dawn

[Ross Ferris]

I thought Northgate revenues were measured in the BILLIONS? though to be fair, most of this is outside of MV

[Simon Verona]

Not trying to be too pedantic, but according to http://www.northgate-is.com/download/Northgate_AR12_Web.pdf   Northgate is operating at just below the 1 billion GBP turnover (1.6 billion USD).   

Probably still too big for anybody operating in the MV space to purchase in full, but maybe in the fullness of time Reality may be sold - though I don't know how dependent on Reality Northgate are these days for their software applications..

Simon

[Dick Thiot]

While I won't say that it won't happen, Northgate selling Reality would be a little like Temenis selling jBASE. It would be a big surprise to me. 

Dick

[Ross Ferris]

Did you also notice the 1.6 Billion pounds on the order book? I would have thought 2013 results would have been published by now which will show an increase I suspect ;-)

[Rick Weiser]

OASys is built from the ground up as a new MV entry.  It has similar U2 functions but is it own entity and has nothing to do with ONgroup.

[Dawn Wolthuis]

So, I think you are telling me that OASys is a clean room version of MV that sprung up this year (?) without any chatter about its development or testing in any MV forums (or has there been)? That is very impressive. Are there live users? 

In order to have a sense of where this would fit in the MV Family Tree, which MV flavors were most influential in the versions of BASIC and MV Query (those are the questions I asked in order to draw the relationships)?

Who are the chief engineers who developed this MV clean room version?

Thanks.  --dawn

[mdsi2000]

Hello group!

I have been out of the any MV posting for about 7 years. Apparently, the same MV turf battles are still going on. lol I LOVE IT!!!! It reminded me of this commercial http://www.youtube.com/watch?v=Z19vR1GldRI I hope you get a giggle out of it.

It's great to see some of the same peeps (a FaceBook term meaning people, there was no Face Book 7 years ago) from years ago. I hope to be a little more active in this group in the comming winter months.

 

 

-Peter Gonzalez