At 10:17 AM 07-09-00 +1000, Robert Schieck wrote:

>I seem to keep inferring from the messages about the test control system/suite
>that if Ann had got a hold of InterBase then she would have released the
>entire test suite, which as I have been trying to point out is not correct.
>
>With InterBase OpenSourced, there are only three things of value left.
>
>1) The InterBase Name
>2) The InterBase Web Site
>3) The full test control system
>
>As Ann outlined, the Test Control Suite is used to make certified builds which
>can be sold......
>
>Helen: Feel free to ask for the full test suite, but I doubt that anyone will
>give it to you.

Robert, thank you for raising this issue of testing tools.

I think you will admit that the whole question of testing tools in general
and the TCS in particular has been totally confused. The rage occurred
when the only spokesperson for Inprise was Ted Shelton. He was unprepared
to discuss the issue as anything other than "to release or not to release
the testing tools". Beyond that, there was silence on the issue. In
private communication with me, it was very clear that Ted didn't understand
what was being asked for.

>Asking for the test control systems and the 24 hour vector test is not an
>unreasonable request.

These are the pieces that open source coders want. It would be a stretch
to expect that individuals communicating by email would have the equipment
to set up and run full battery testing with the integrated suites. This
whole issue appears to me as something which could have been easily
resolved by Inprise R & D if the executive had allowed the right person to
communicate with us. Shaunak Mistry was trying hard to get permission to
provide testing tools for the InterClient team and we expressed our
appreciation for that. We still don't understand why we heard no more of
it, even a month later.

>As Ann outlined, their plan was to clean them up before
>they are released from which I infer that, Ann didn't have time to do it.

I'm sure you are right, given the amount of core code work that was
necessary to take the beta kits (in the case of Win32) and sub-alpha kits
(in the case of Linux) through to release with sub-minimal R & D staff who
were being shared with the development and support effort for Inprise's 5.x
and 4.x customers. There is no "blame" to apportion here. ISC had
promised to release testing tools so we can assume that it would have been
high on ISC priorities to do so, in order to get the OS coders working
productively in their chosen specialities.

Inprise could not release what it did not have to release. The issue has
arisen because, instead of explaining to the coders what it was intending
to do about it and how long it would take, Inprise said "You can't have any
of it. It belongs to us and we intend to use it to make money." This
delivers an undeniable message "We expect you to develop for us but we are
not going to help you to do a good-quality job unless we can make money out
of it."

>If Borland plans to do the same, I doubt that they have had enough time to
>clean them up as the InterBase web site hasn't changed in four or five weeks
>and it, from a marketing perspective, is more important.

Perhaps it is, perhaps it is not. It's a side-track. They have not done
anything about the web site or about making test tools available. As far
as the public knows, they are just letting InterBase lapse into
obscurity. They can hardly expect either the Open Source community or
their customers (that's most of us) to be encouraged.

>It appears to me that both Ann and Borland expected that the new InterBase
>Company would happen and both were caught off guard by the outcome. I suspect
>Ann would have made different decisions (fiscal amongst others) and Borland
>would have had a management team and plans in place to handle InterBase if
>they all new that the New ISC wasn't going to happen. But that is water under
>the bridge and everyone is scrambling to pick up their respective pieces.

Dale Fuller knew.

>The community needs to make a decision as to whether they want to cut Borland
>some slack and see what they plan or continue to bash. Large companies move
>slowly and bashing drives existing and potential customers away.

From where everyone else but Rob Schieck stands, nothing is even slightly
clear. The people who are bashing them are their customers, as
individuals, not the IBDI, not IBPhoenix. Many of those doing the bashing
**are** Inprise's large customers. If they are moving away, it's because
of months of unsettling delay followed by a sudden and total clampdown on
information. Support customers are unable to get any response to requests
about their contracts. We have to rely on hearsay to find out
anything. People like yourself, who are in contact with Inprise, are under
NDA and can't tell us anything. What is coming through from you, Jeff
Overcash and John Kaster is smear and fud, and people know it it.

Ask Inprise to open up and stop being defensive. Ask your colleagues to
research the facts before hitting the lists with

>I hope the
>community chooses carefully and regulates their decision.

Rob, Jeff and John, I hope that you guys at least can realise how important
it was, is and will continue to be, to understand that the people who have
been cut off by Inprise are, first and foremost, their
customers. Developers who buy and use Inprise's tools are also the guts of
Inprise's sales force for those tools. Developers make tools choices for
their companies. Developers are influenced by other developers, never by
salespeople.

More responses in the next message.

Helen

>The community also has to decide how they are going to handle the InterBase
>R&D people when the show up on this list. You can continue to "jump" them when
>the surface on this list and drive them away or welcome them to the community
>and use their guidance. After all, the most knowledgable person on the current
>releases of InterBase isn't Ann Harrison nor Jim Starkey, it is Charlie Cairo
>and he is with Borland/InterBase R&D.
>
>The community also had to decide how they are going to handle InterBase Tech
>Support when they surface on this list, and they will surface here. I hope it
>is with a warm welcome not a "where is my test suite".
>
>Since I am on a roll here, let me continue....
>
>I put up the original InterBase CVS archive with the expectations that there
>would be three of them eventually. Mine, IBPheonix/NewCo/FireBird, and
>Borlands. I also expected that the only one left standing in the end would be
>Borlands.
>
>You see (please excuse the plagiarism here from one of my favourite movies),
> >From a source perspective, InterBase isn't easy. You have to want it bad.
>It's
>not easy to build and even tougher to modify if you don't have a lot of
>experience with it.
>
>What I had hoped would happen with the CVS archives is:
>
>1) someone figures out how to build it.
>2) someone create scripts to simplify the build process so everyone can play
>the game (many hands make light work)
>3) the builds are checked against a known build for accuracy so we have a
>known base to work from (there was a plan for this)
>4) cosmetic clean ups of the code are done, changes to correct compile
>warnings are done and builds are checked against the base from 3
>5) when 4 is completed, then bug fixes etc are done.
>
>Unfortunately this is not what happened and you can see the results. On
>FireBird, changes are put in and then taken out as they break things. The side
>effects of changes are difficult to understand without a lot of experience
>with the product.
>
>I should point out that at InterBase R&D, you compile it on windows and
>compile on a Unix box to see if you have messed anything up before you put
>change back to the archives which is why item 2 was listed above.
>
>Of the three CVS archives, the InterBase one is the best. The InterBase CVS
>has a complete set of backed up databases fully integrated in with their build
>scripts. I ran over them a couple of time for a windows build and got it down
>to a two script build. I was working on getting rid of the korn shell
>requirement but I haven't got back to it in a month or so. From their
>archives, it was possible to do a complete build including the codes program
>with what they presented without a whole lot of work.
>
>I am a committer on the InterBase CVS, I haven't committed anything. All my
>changes (very simple stuff) were sent to InterBase R&D and they made the
>changes after checking out my requests. As it had been said, Open source needs
>a benevolent dictator and InterBase R&D fills the role nicely as they have the
>experience with InterBase to assess the side effects of changes.
>
>As an added bonus, the likely hood of getting a build certified off of the
>InterBase CVS is significantly higher than from any other site including mine.
>
>Enough stuff, I leave it for the community to make their choices and hope they
>choose wisely
>
>
>Rob
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>Robert Schieck
>MER Systems Inc.
>Inprise/Borland/InterBase Search Engine http://www.mers.com/searchsite.html
>
>
>
>
>_______________________________________________
>Interbase mailing list
>Interbase@...
>http://mers.com/mailman/listinfo/interbase
>
>
>Community email addresses:
> Post message: IBDI@onelist.com
> Subscribe: IBDI-subscribe@onelist.com
> Unsubscribe: IBDI-unsubscribe@onelist.com
> List owner: IBDI-owner@onelist.com
>
>Shortcut URL to this page:
> http://www.onelist.com/community/IBDI

http://www.interbase2000.org
___________________________________________________
All for Open and Open for All
___________________________________________________
Team JEDI Member