Slashdot Mirror
Inprise Considering Open Sourcing InterBase
Keith Russell writes "Caught this news blurb on ZDNet. Apparently, attrition has taken its toll at on Interbase's top levels, and Inprise is seriously considering open source as an alternative to pulling the plug. A likely possibility, given their recent enthusiasm for Linux. This could be a Good Thing. I'd rather see "end of life" software opened than hoarded. "
Eeek!
*Please* don't send email to these addresses.
I guarantee you, the "other contacts" won't care: with one exception they're all drop-boxes designed for a specific purpose; mailing them would be as annoying as mass-mailing "postmaster@slashdot.org" about wanting a job working for slashdot.
The few addresses up there that lead to real people are not people who will have the power to influence upper management in any event: again, you'll merely succeed in irritating innocent bystanders.
Post here, or otherwise communicate through normal channels; don't send mail to random Borland e-mail addresses.
--Robert West
Delphi R&D
Linux isn't exactly without open source databases. Postgres is the one that comes to my mind immediately, but there are others. However, I suspect that we (Linux users/programmers) would be thrilled to see an open source solution that was palatable to the corporate world as well. An open source database engine, with supporting language interfaces, available for multiple platforms, with a company selling support contracts for it could be just the ticket. Inprise would get the benefit of a lot of open source development effort. They are still the best positioned company to provide the support. Linux would get a mature, commercially accepted database engine. Done right, this could be a big win.
The net will not be what we demand, but what we make it. Build it well.
Well, I've used Interbase on and off for a good few years, and v5.0 intensively over the last year. It's always been a doddle to administrate, requires almost no resources, and has _never_ lost any data. It dies maybe once every 3 months, but always with a full recovery. :)
Open sourcing InterBase is effectively giving it away. InterBase requires no "after sales support" for Inprise (I agree with the previous poster about the name - Borland _is_ better
So, although it might not make any money _yet_ with a little patience (and with the targeting of Linux as a new OS platform to develop for), Borland may well kill Access with it. Add a pretty front end; there you have a fully featured DB back end to put any GUI on. And I find it a crying shame that neither Delphi or C++ Builder haven't penetrated the markets yet. Delphi is a piece of piss to use, (surely, ALL programmers play with Pascal at some point), and compiles code with an execution speed that makes VB programmers cream themselves. If you've ever used C++ Builder and MS VC++, well, you already know what I'm talking about... DRAG AND DROP, Bill!!! And what the fuck is all that message pump nonsense about? Sheesh. In a test for a pretty basic GUI and simple database (10 tables), here's what I got:
Delphi / Interbase - 8.5 hours.
C++ Builder / Interbase - 10 hours (I got distracted and wrote a whole load of thread classes so all DB activity is done in the background. Then included it into the Delphi version. Try doing that between VC++ and Visual Basic...)
MSVC++ / SQL Server - 16 hours
And I think I have relatively even spread of skills with all the above.
Anyway, don't hold your breath for source code for InterBase.
Strong data typing is for those with weak minds.
First, let's face it: InterBase has never been a wildly successful product. From what I can tell (as a Inprise outsider) the only people who bought it were people who were already firmly entrenched in the Borland development environments.
It seems to me that Borland's real cash cow has always been their development tools. Open sourcing Interbase could lead to it being more widely used, which could sell a /lot/ more development tools. Especially in the Linux environment.
Second, this would firmly entrench Borland in the Linux world. I think that the past few years and the (ridiculous) success enjoyed by Visual Basic have made it clear that it is going to very difficult for Borland to compete on Microsoft operating systems.
There hope of radical success (as opposed to the kind of mediocre success they've been enjoying for a while) is to become the premier provider of Linux tools. And I think they know it.
All in all, this could be very cool.
Also, let me comment that, from what I've seen, Interbase is a very cool product that has never gotten the recognition it deserves. The big advantage over MySQL is that it has full transaction support. Sorry guys: but there are some applications where you just need transactions.
-- Slashdot sucks.
If a company considers that a software package is no longer "profitable", opensourcing should be considered a way of generating advertising at the very least.
Look at all the companies recently who have benefitted greatly in terms of good PR by opensourcing their stuff.
However, this does not necessarily mean that opensourcing software is *always* going to be good PR.
Look at the recent Quake1 debacle.
What should have been (and Is still, in my opinion) a fantastic move has been perverted by whiners complaining that the release of the code has resulted in "iffy clients" or that the Graphix were not released with it.
I think opensourcing should be done in conjunction with the open source community:
Ask if anyone will support it.
Explain exactly what it is you are intending to do.
Don't expect adulation and worship.
The secret of success is honesty and fair dealing. If you can fake those, you've got it made. (Marx)
This looks good... It's encouraging to see companies that traditionally have had nothing to do with the community/movement considering open source alternatives.
Inprise used to be Borland, which was a favorite of MS-DOS/Windows programmers everywhere. As we all know, companies that devote themselves almost exclusively to MS-based products have traditionally been far from the open source mindset, as open source has traditionally been associated with Linux et al.
I like it when those traditions are abandoned. The software industry is teeming with neophobes.
Russia used this militarily to destroy both French and German armies; they performed strategic retreats when "outgunned," destroying crops and other infrastructure so that when the Russian winter set in, opponents were overextended, and despite "winning the battle," wound up losing the war.
This obviously came at the cost of considerable Russian destruction, and with Inprise, the cost is that of not getting revenues from license sales, whilst the immediate benefit is that this may injure sales of competing DB vendors.
The open question is of how this affects already-free DBs like MySQL and PostgreSQL.
Effects on them are severalfold, and some are dependent on what license Inprise comes up with:
- Regardless of the license, it should be useful to have source code access as this can allow folks to see an implementation of transaction locking, stored procedures, SQL-CLI, ODBC, as well as the data storage mechanis, which may be quite useful when they try to add such functionality to other DBMS systems even if there is no reuse of code.
- If the license is sufficiently compatible, it may prove possible to integrate code one way or another either into "OpenInterBase" or into one of the other DBMSes.
Note that the folks likely to get particularly injured by this are the second/third tier companies selling licenses to things like:- Altera
- SOLID SQL Server
- OpenIngres
- Mimer
- Faircom
- Raima
- Yard
- Empress
whilst people will still likely be prepared to "pay the bucks" to move up to "Tier 1" DBMSes like Sybase/Oracle/Informix/DB2If you're not part of the solution, you're part of the precipitate.
Another, similar route to open source software is through research projects that, for one reason or another, aren't commercialized; the research code is released and often becomes an important open source/free software system.
We should be happy about that: much (if not most) open source and free software started out that way.
Because so much free software starts out as commercial or research projects that, I think it's important to think about how to encourage development and research organizations to build it in such a way that the transition to free software will be easy. That means that such organizations should find it easy to use existing free software libraries, build on open APIs with free implementations, and should not feel the need to rely on proprietary libraries (which would make freeing the software later much harder).
One thing that I think is very important is to use licenses like LGPL or BSD (as opposed to GPL or QPL) for important libraries. Research and development organizations will not use software if that means making a strong commitment early on to open sourcing their software later or face uncertain expenses later. Both GPL and QPL, unfortunately, impose such uncertainties and limit options. If there is no unencumbered free or open source software, they will pick the best and most affordable proprietary libraries to build on.
The LGPL and BSD licenses, on the other hand, allow development and research organizations to keep their options open for what to do with their code. When infrastructure libraries (standard libraries, networking, gui, etc.) are released under those licenses, research and development organizations can use them, and when they decide to release their software as "free software", it will be so much more useful to the free software community than if it had been based on proprietary libraries or APIs.
For similar considerations, I think it's also important to get as much free software infrastructure on Windows. If companies start programming to free software APIs on Windows (and they have to cover the Windows market), when they go open source, their software will be much more useful to the free software community. So, the more unencumbered networking, database, and GUI libraries we can get onto Windows, the better.
So, keep that in mind when thinking about policies and licenses. While the idea that all free software is created by altruistic volunteers is appealing (and a significant amount of free software is), the reality is that a lot of free software is created by companies and donated if the software turned out not to be a winner in the market or is otherwise not commercializable. Making the life of those companies easier and allowing them to develop code that interoperates well with other free software is a win for everybody.
I'd love it if companies were legally required to release the source of any product that's been discontinued--including in the event of the demise of the company.
It's hard to imagine how you would make this stick legally: it's easy for a company to claim "no really, it's not discontinued, we still have 1/10th of a developer working on this, it'll be done in 5 or 6 years." Then there's all the issues with intellectual property rights entangled in the source. I wonder, though, if you could craft a law along the lines of the Freedom of Information Act, requiring source code to be released at a certain point in time. (It'd be the Freedom of Source Code Act--FOSCA) It would give interested parties a basis to sue the holder of the code for release. This would work in particular in situations where a company gets acquired and one of its products is squelched by the new owners.
Alternatively, it might be possible to accomplish this with a culture shift: if developers regularly required as part of their employment contracts a clause that the product code be released under certain circumstances. We can all think of tons of cases where developers labored for years on supercool a product, only to have the startup go under and the code disappear into a legal black hole.
All of the stuff that Borland has completly done away with, such as Turbo Prolog, Turbo Basic, and Turbo Modula (for z80-based CP/M systems) reverted back to the original authors. Philippe Kahn, Borland's founder, made his fortune and his company by licensing and distibuting other's people's software. So, when things went bad, he simply dropped them.
Turbo Basic became PowerBasic from PowerBasic
Turbo Prolog became Visual Prolog from PDC (www.pdc.dk)
Turbo Modula became JPI Modula. The x86 port became TopSpeed Modula-2. TopSpeed was later sold to Clarion, and now their compilers are only available with the Clarion 4GL products.
Of course, dBase became visual with Visual dBase 5.5, which was a good idea, but incredibly buggy and unstable. And the native compiler cost extra. Supposedly 7.0 is good, but it's pretty much lost the end user DB wars to Access.
I for one would like to see and OSS Turbo Pascal for Unix. That thing had the fastest compiler I've ever seen! Add a nifty IDE and good online help, and you've got a winner. (Yeah, I know there's Free Pascal and GPC, but still...)
~~~~~~~~~
auntfloyd
According to the company history:
Borland might be an evidence against the common contention that "Microsoft is the company that never produces anything, but merely buys out products from other companies that are creative," as many of Borland's products were not natively produced, but rather resold on behalf of other componies.
By the way, that was Ashton Tate that used to own the dBase trademark...
As for integration with DBM variants, I see little importance to that. InterBase is a relational database (or at least, as relational as they come), as opposed to merely being a data store. The value would be in sharing code between InterBase and PostgreSQL or MySQL, or maybe using InterBase as a "data store" for persistent data in KDE or GNOME.
If you're not part of the solution, you're part of the precipitate.
It has nothing to do with price. For large enterprise databases, price isn't an issue. For smaller applications where Interbase would be used, Oracle would be a huge overkill.
A pickup truck can't haul as much as an 18 wheeler, but if you just have a desk to haul, an 18 wheeler isn't needed. Likewise, if you do have enough stuff to haul to need an 18 wheeler, the cost of the truck probably isn't going to be a problem.
-BrentYes, but when the Interbase code was still young, "Open Source" wasn't something that was heard of. You want them to not open source now, just because the concept wasn't around when the product began?
Open sourcing older programs is a good idea. Companies have no experience in Open Source development. By starting with a product that you can't "lose" with, it gives companies a feel for how to best work with the Open Source community with their flagship products.
-BrentLots of paradoxical effects are possible; the fact that PostgreSQL uses a BSD-like license means that it may be easier to do code integration between it and Interbase, as compared to GPLed MySQL, where the somewhat "infectious" nature of the license may discourage attempts to integrate code.
I certainly agree that YAPOPL (Yet Another Pseudo-Open Proprietary License) would discourage development efforts, but suggest that the relative merits/implications of GPL versus BSDL are quite nonintuitive and nonobvious.
If you're not part of the solution, you're part of the precipitate.
Okay... Who is going to work on it ? eh... when Mozilla went open source we all rejoiced.. "Yeah.. wer'e saved".. but how many of us were willing to wade through all of that code, then devote any of our precious spare time to working on it.. ?? About a handfull. (That's not to say that mozilla isn't going to work, just not at the speed that Linux does)
I love the whole open source development model, but it doesn't always work. You can't always save a product is in trouble by going open source.
I just thank God I dumped all of my Inprise stock. What a trip that was...
A genius writes code an idiot can understand, while an idiot writes code the compiler can't understand.
Those that want 24x7 support contracts and the likes for big SMP boxes with big RAID arrays aren't going to be much more attracted to InterBase because it becomes free than they were before when it wasn't.
In contrast, those that were price-sensitive, and went with SOLID/ Altera/ ... may seriously consider trying out Interbase as an even less expensive alternative that doesn't tie them to proprietary licensing.
The company most likely to lose from this is SOLID. They got pretty seriously "flamed" a while back due to a change in licensing strategy; they used to sell individual licenses for around $200-$300, but have moved to selling groups of licenses so that the minimum price granularity is rather higher, more like $10K.
That is a market rather more vulnerable to Interbase's "scorched earth" strategy...
If you're not part of the solution, you're part of the precipitate.
Just to make it clear this rumor does have some basis in fact:
Anders Ohlsson (Borland) posted this message to a BBS on Wednesday evening:
From The "Save Interbase" Website
Anders Ohlsson is a farily well known Borland Developer.
Exactly... If you need Oracle, you NEED Oracle... If your application doesn't require it, you're just throwing away your cash. They compete in COMPLETELY different areas of the market.
:)
Just like somewhere once i saw a checklist comparison of MySQL vs. Oracle. The most pointless waste of time i'd ever stumbled across. Yeah you can say MySQL supports this and that that Oracle may not support, or maybe it's faster when dealing with 25MB files, but you'd better not be buying Oracle to deal with said file.
I liked your analogy there... But let's make that desk into a 13" TV...
If anybody remembers who they were, and what the story was, I'd appreciate it if they could post it.
I always thought this product was an ideal candidate for open sourcing. If we could find them maybe we could convince them.
--SolidGold
--SolidGold
Everything you know is wrong. Or more accurately, inaccurate.
In most applications, the quality of database design and code manipulating the data have more to do with the robustness of the application that the dbms does.
-- Slashdot sucks.
They may also be looking to save money. If Inprise can save money by getting open source developers to take over some or all of the load for new development and support for InterBase, it may be a win for them. A penny saved is a penny earned.
... just say "it's open source, you can fix it yourself and get lots of free on-line help" and problem solved.
...
That was my first thought. Just think, now they can avoid all those Y2K-hungry lawyers looking to get bug fixes
And, as you also point out, this will cause more of us to think of buying Borland/Inprise development tools first, before we consider other ones. After all, even if it's proprietary today, it will be GPL tomorrow
Will in Seattle
Shouldn't that be:
Thanks for contributing the requisite uniformed Windows/NT user's perspective!
Will in Seattle
What are "most" applications? Most of your applications? Most of my applications? You are right, Oracle isn't worth the price for most applications. So why would you use it? Oracle doesn't *want* you to use their database for most applications. That's why other databases like MySQL and Interbase exist.
For use as a database to track orders for a a company that does $1 million dollars a year in sales, Oracle would be a poor choice. But for a company that does $100 billion dollars a year in sales, Oracle is probably a very good price.
Don't think that 18 wheelers are a wasted product just because all your move around is a 13" TV. Be content with a pickup if that meets your needs.
-Brentagreed, go here instead,
http://www.interbase.com/cgi/contact.cgi
peterrenshaw ~ Another Scrappy Startup
Just think, now they can avoid all those Y2K-hungry lawyers looking to get bug fixes
:-) But in all reality, it is probably too late to avoid the Y2K vulture lawyers if InterBase has any Y2K problems. At any rate, if it isn't too late, they had better hurry up, because Y2K is only TWO days away! :-)
Well, lawyers don't want bug fixes, they want big legal fees...
I've used interbase, and it is an excelent product.
It and it's predecessor (DEC RDB) in my opinion are better than most of the big names by a mile.
I was really shocked when got to try other relational databases. Interbase developed the blob stuff and has had transactions and row level locking in it from the start. The other notables, Oracle Ingress, Sybase and Informix were all latecomers to this (Im not sure if Ingress actually has it yet (and it was renamed OpenRoad/OpenIngress or something by Computer Associates)
Interbase was also beautifly simple to use, and virtually required no DBA, even for rather large data sets (couple of million rows). The constant tinkering and adjusting required by a full time DBA for the other big names, DBA time is a real hidden killer that is often the real cause for DB down time.
Oracle now owns DEC RDB (now called Oracle RDB) and are porting it to other platforms (either NT or Unix) although it is seperate from their other Oracle SQL product.
A good strong "commercial strength" relational database with an open source philosophy would be a nice accompanyment to a linux system.
Postgres and MySQL (although MySQL is not transactional) are pretty damn good and are filling this role fairly quickly.
I think that Borland could give a real kickstart to interbase by making it opensource and closely associating it with linux.
And I for one would relish the oportunity to work with it again.
Cheers
Nice to see there is someone else besides myself seeing through the crap. Interbase is profitable, they may drop it on certain platforms, but not WinNT or Linux. Too many developers are using/deploying it. Personally I hope the ex-employees get sued for breach of contract. My email box was overflowing with various FUD/Interbase crap. Had to setup a custom filter in Outlook to send it to the trash.
As someone who's just beginning SQL programming on Linux I'd like to know what's wrong with postgresql. It has transactions...its' open source, and it has a JDBC driver. I'll probably test using EJB's with it..but even if I don't go the EJB route I'll use it from servlets...It's even included in the Linux Mandrake release I'm using. Dorwin
Do you really think that spamming a Y2K address or the HR department (resume) or the FTP admin will have any endearing effect?
/. effect now the spam.now effect?
Should we call the
Don't you think that select, well-reasoned emails to a TARGETED address will have a much more positive effect than spamming. Is spamming for an - allegedly - good cause (OSS) *not* spamming? Hey, I'll spam you tomorrow and ask you to donate a quid to your favourite charity. Feel better now?
Really, a *brilliant* idea.
Sure there are applications where you need Oracle or Informix (FWIW, Informix generally kicks Oracle's butt IMNSHO). However, as you conceed they are few and far between.
Way too many people have a blind, knee-jerk loyalty to a particular RDBMS that leads them to waste a lot of money. Consider, for example, Medic Computer Systems (www.medcmp.com), a company I am intimately familiar with.
One of their products uses Informix -- and requires 10 times the hardware resources it would need if it used Interbase.
-- Slashdot sucks.
What will happen: whoever looks at it will find out the code is a huge pile of crud, it will take an year or so to start looking like something, during this time the decision of opening up will be questioned a lot of times, yadda yadda. Of course, if the code has any value at all, in 2 or 3 years we will have a real killer database. (sigh) Don't people realize the difference between Opening up vs. developing openly?
I thought it was under some semi-free Licence where you had to pay licence-fees for non-Linux platforms.
Has this changed? I seem to remember something about GPLing an old version of MySQL. Is this what happened?
Interbase is a good database. It is small, highly portable and quick.
The Client/Server version is roughly comparable to SQL Server 7 in performace and features - maybe it won't quite outscale Oracle, but for anything smaller than that it is pretty nice. (IMHO)
There is a "Local" version, too. I know quite a lot of people who have sold apps using that (Maybe I hang out with too many Delphi people, but still)
The list of platforms it comes for isn't bad, either:
It's not missing any features, either - unlike MySQL.
He was asking on the Mers mailing list about what people would think.
(David Intersimone is Borland's developer relations manager, or something like that)
John Kaster sure hasn't ruled it out on the News Groups, either.
I think I was...
One of their products uses Informix -- and requires 10 times the hardware resources it would need if it used Interbase.Didn't I say, "Why use an 18 wheeler to move a 13" TV?" I'm not justifying using a product that wasn't designed for what you are using it for. And Oracle certainly wasn't designed for workgroup type applications.
-BrentHave you looked into PostgreSQL? How does it compare to Interbase?
-- Don't Tase me, bro!
And on another point... Inprise is presumably looking to make money.
They may also be looking to save money. If Inprise can save money by getting open source developers to take over some or all of the load for new development and support for InterBase, it may be a win for them. A penny saved is a penny earned.
Your vote doesn't earn them dollars...
Perhaps not directly, but it does get them publicity, developer and user mindshare, and perhaps sales of other commercial products. Perhaps even sales of commercial or shrinkwrap boxed versions into some sites.
a pledge to purchase their commercial development tools would help, maybe.
If people are using their database, presumably they will be more likely to buy their development tools, especially if they offer features in their development tools that make it easier to develop for InterBase.