Interbase And Kylix Details From Borland/Inprise Con

ghjm writes: "I'm typing this from a machine in the computer lab at the Inprise/Borland conference in San Diego, where many new details about Interbase and Kylix have been revealed."

"The conference opened with a "Matrix" themed intro, and there have been blue and red jellybeans all over the place. The major announcement at the intro was that JBuilder will be ported to Mac OS X, with full support for Aqua. The press release is here.

Paul Beach, Interbase VP of marketing and sales, gave a very informative talk. It's very clear that he has a clue. Many of his talking points were interchangeable with what Bob Young would say, asked similar questions. He even alluded to the tired old Heinz ketchup analogy to explain brand equity. Interbase's plan is to sell support contracts and box sets, just like the other big open source companies. Paul is very realistic and down-to-earth about what they expect to happen. I think they are going to do very well indeed.

The most startling revelation, though, was that Interbase 6.0 is done and ready to ship the manuals and CDs are duplicated and printed and sitting in boxes. He even brought a box of CDs to the talk. Reading between the lines, it seems like these CDs have been sitting around for some time. But Dale Fuller (Inprise CEO), who also showed up at the Interbase product address, will not permit the CDs to be distributed until all the contracts are signed and executed and appropriately lawyered. Dale promised, both at the opening keynote and at the Interbase talk, that the contracts would be done and the product would ship within two weeks. Here's hoping he holds himself accountable to that promise.

Many details about Kylix have been revealed. Kylix is installed on several machines in the lab, but unfortunately these machines are off the net, so you can't steal a copy (darnit). The IDE was demoed at one of the talks, but it wasn't in the lab. What was on the lab machines was the command-line Delphi compiler only. It seems quite solid. I played around with it, built some projects that displayed various forms and controls, that sort of thing. As expected, the compiler is blindingly fast, builds genuine native executables, and looks very similar to Delphi on Windows. The new class libraries, called CLX, are very nearly 100% source-compatible with the Delphi VCL. The CLX visual components are wrappers around QT, and will be available on Windows as well in Delphi 6, so if you code to CLX you will be cross-platform. Kylix builds fully compliant QT apps, with KDE-style theming and all the bells and whistles. As with all QT/KDE apps, you can run a Kylix-developed app under Gnome but it isn t seamless for purposes of theming or UI consistency.

Kylix will, of course, be closed-source, closed-process, proprietary, and will have a price notably divergent from zero. The initial release will be Delphi standard and professional, followed quite quickly by C++ standard and professional. The enterprise version of Kylix will be called "Kylix Studio" (or similar) and will include both compilers in a single SKU. Someone in the audience suggested that they do a combined SKU under Windows as well, which got a lot of applause. At the opening keynote, Dale said that a major goal for Kylix is to make it possible for developers to release their own projects under any license, including full-strength GPL. This is a worthy objective, but I'm not sure Borland truly understands the ramifications. Then again, I'm not sure I do myself. What would it mean to have a GPL project, but you have to use a proprietary compiler to build it? Presumably the compiler libraries will be proprietary, and you have to link with them if you want a usable binary. This bears thinking about.

For database access from Kylix, there's a new library called dbDirect. Or perhaps it's called dbExpress. And I think it might also be called DataCLX, or perhaps DataCLX is a superset including dbDirect plus other things. Anyway, whatever it's called, it's a new library that lets applications talk to databases in a consistent way just like the BDE. But unlike the BDE, it tries to be as 'thin' as possible, bringing the application code as close as feasible to the native database vendor APIs while still providing relatively good code portability. With the BDE, Borland has to write a complex driver for every new database they support. With dbDirect, it's a simple matter of wrapping a few vendor API calls. Application developers can mix and match calls to dbDirect with calls to native database APIs. dbDirect will also have a tiny footprint compared to the BDE, including the ability to statically link right into the application. DataCLX will be the only database API in Kylix, and will ship in Delphi 6 side by side with the traditional BDE setup.

Kylix will also include a web broker architecture, very similar to what's in Delphi 5 today. Where Delphi web modules can compile to traditional CGIs or ISAPIs, Kylix web modules will compile to traditional CGIs or Apache modules. The level of integration between Kylix and Apache is impressive. This part of things is called NetCLX and also includes the usual socket components and TCP-suite stuff like telnet, smtp, etc.

In addition to the Borland announcements, there are a few dozen vendors on the trade show floor. Big names include Sun, Caldera, Linuxcare and Cobalt. Corel is pointedly absent. And of course we have the usual gang of third-party component vendors people like TurboPower, Woll2Woll, Raize, Digial Metaphors, etc. Everyone seems to be planning to port their tools to Linux, promising release dates from a couple weeks to a couple months after Kylix ships.

The bottom line is, I'll be coming home empty-handed. No Kylix beta, no Interbase source. But I have a very strong sense that the Delphi community is gathering behind Kylix in a big way, and I'm very pleased to see Interbase poised on the verge of release. Just get those contracts signed, Dale!"

Comments

Cross compiler

[Spudley]

What I'd like to see is Delphi and Kylix shipped with cross compilers for each other.
In other words, I'd like to be able to build Windows apps from a Linux PC, and vice-versa.

In fact, what I'd really like to see is more platforms supported, but I guess for now I should be happy that they're supporting more than just the one...

Re:Cross compiler

[Ronin Developer]

You will. It's called Delphi 6.

Delphi 6 will include both the VCL and CLX libraries. You'll use the VCL for native Windows apps and CLX for cross platform development.

We've seen both the existing Kylix and Delphi 6. Both are not ready for shipping. When they are released, life will be good.

They seem to think Kylix will be ready for initial release this fall. D6 probably won't ship until sometime next year.

Yes, these delays were disappointing. We would have loved to come home with a copy of D6, Kylix, or even Open Sourced InterBase. Didn't happen. This is in sharp contract to the last two conferences where the stuff was near ready for shipping (or shipping). We got excited by it as we could play with it. Now, we just have to wait a little while longer. To paraphrase Dale Fuller (CEO, Inprise), "It won't ship until its ready.".

Borland is digging themselves out of a hole. No more major screwups. Rushing a product out the door might appease the stock holders, initially. But, when the problems surface, you can bet the stock price will drop. Borland doesn't want that to happen again and is taking the hit now...by not releasing prematurely.

Danger of Kylix

[jrincayc]

Kylix is one of the most dangerous pieces of software to come to Linux. One of the big differences between Windows and Linux has been that Linux has one compiler (gcc) and open libraries so that everyone can compile everything. Kylix breaks that since now Kylix owners can compile software that I cannot. This will cause a split in the devoloper community that will make QT/GTK seem like a friendly gathering. Kylix forces developers to either ignore software or develop with non-free software.

Kylix creates software that can only be compiled with non-free software.

Re:Danger of Kylix

[Spudley]

OOoooh! That's gotta be moderated as flamebait??

I completely disagree with you. Kylix and gcc have completely different markets, and will therefore not cause problems for each other.

Kylix will be sold to people who want to play with a nice friendly VisualBastic-type environment; GCC will be used by developers.
Kylix will be sold to people already using Delphi and VisualBastic, to allow existing apps to be ported to Linux which would otherwise never would have appeared; GCC is for developers.
Kylix will be used by companies looking for a short learning curve for training people; CGG is used by developers.

But the big difference is that Kylix will primarily be used to write closed-source proprietary applications. You won't get GCC people getting angry about not being able to compile Kylix code as you claim, because there won't be that much Kylix code out there for them to compile.

And no, I don't have a problem with people writing closed source apps for Linux. In some cases - especially low-volume vertical market apps - closed source is pretty much a requirement. And even where it isn't, you're not going to change a company's business policy overnight.

Re:Danger of Kylix

[novitk]

That's why they invented RPMs. End-user don't need to build software, developer do. Kylix is a developer's product and any serious developer should buy it, as it's arguable the best Compiler/Ide/Library created.

Re:Danger of Kylix

[Hard_Code]

Man, your right! We won't be able to compile Object Pascal with GCC. Shit, can we compile Ada, Eiffel or Fortran with GCC. You better look into it, because it would mean all those program written in other languages would be non-free!! Damn!!

Re:Danger of Kylix

[nd]

This isn't really a problem.

Borland intends to give away its command line compiler (dcc) for free distribution. It won't be open source, but you WILL be able to compile Delphi apps in Linux without buying it.

Re:Danger of Kylix

[GroovBird]

And there's nobody who even comes up with the fact that the Free Pascal Project has been striving for 100% compatibility with Delphi? Okay, CLX will be proprietary, but I think some of us are working on FreeCLX based on early betas already.

Re:Danger of Kylix

[MrBogus]

You are right on -- Kylix is aimed at a completely different market than gcc and the traditional GNU dev tools.

Borland is marketing Kylix at their existing Windows userbase, who primarily code internal corporate applications, and who are considering their long- and short-term dependance on Windows. I would imagine that 90% of the stuff developed with Kylix never makes it outside of the company it was developed for. I seriously doubt that Borland believes that the traditional Unix developer wants/needs their tools.

So, of course there will be some Delphi stuff showing up on Freshmeat, and some clowns might even try to market shareware to the Linux folks (har). So what -- if it's Open Source and it really bugs you, you can port it to whatever compiler you want.

Re:Danger of Kylix

[Halo1]

Since the Free Pascal Compiler can compile (most) Delphi code, you should be able to do all your design etc in Kylix and then compile the result with FPC. Unless some critical information necessary to compile is stored in the project files, that is (but I don't think that's the case).

--

Re:Danger of Kylix

[HiThere]

And since CLX (CLiX?) is basically a wrapper around QT, that should be relatively easy to emulate, once the release version is available.

Re:Danger of Kylix

[bugger]

>Borland intends to give away its command line
>compiler (dcc) for free distribution.

Has there been any announcement to this effect?

Sure, this would be a logical extension of the free Borland C++ compiler (one of the most ANSI C++-compliant compilers around) and of the free JBuilder download, recognised as the very best Java IDE around.

Re:Danger of Kylix

[Ronin Developer]

If using GCC is what makes someone a developer and not the promise to develop high quality applications in a fraction of the time, then I guess its time to seek a new profession as I must not be a real developer.

Yes, you can write outstanding code using GCC. You can write some pretty piss poor code as well. It's not the tool that makes the developer, but rather thee ability to leverage knowledge and know which tool is the right one for a particular job that defines the term developer.

How does one use a hammer to attach two pieces of wood? Why, with nails and hard blows. But, what if I wanted to use screws instead of nails? Do I use a hammer? No, I use a power screwdriver.

If I want to write handcrafted C or C++ code to extend the kernel, then GCC is one of the tools that will get me there. Similarly, if I need to develop a business application, I'll use a tool like Delphi, C++Builder, or Kylix (but not VB) simply because they are designed for the purpose of creating database-centric business and end-user applications.

I would not, under 99% of circumstance, consider using those tools for kernel work (nor Kylix). That is not their role. Even Borland makes it clear that kernel development is not the role of Kylix.

So, stop belittling those that develop using other tools. Maybe they KNOW something you don't about getting the job done in an expeditious and cost effective manner without sacrificing quality or performance.

Oh yeah..you might want to look upwards...see that bright object in the sky...It a cluster of tools like Kylix that will make you extinct.

RD

Re:Danger of Kylix

[AlisdairM]

I think you miss the worst potential though. I am a borland fan, and desperately hope Kylix is well recieved, which is why I fear scenario B.

Kylix initial market will be existing Delphi developers, trying to leverage their skills onto Linux. This may indeed produce a flurry of new GUI desktop software. The danger is that it all looks like Windows software, as that is what most Kylix early adopters are used to producing. What is the point of writing Linux sofware if it faithfully reproduces all the behaviours of an existing Windows product? Little, beyond agravating the Linux community and putting an extra check-mark on the advertising.

I hope this is not the case, but that's my doomsday scenario. Please don't let this happen...

Re:We already know it

[Spudley]

We already know it, it was posted on Slashdot 3 hours ago

Yes, but that's how long it took him to type this post...

Re:Open Source Free Pascal Compiler is available

[bugger]

Yeah, interfaces are so much click-and-drool.

None of these what you call "click-and-drool delphi converts" will ever get into contact with interfaces - because interfaces are only useful for designing.

Oh, wait. You don't know what design is about! It's about maintainability, abstraction, modelling, separation of interface (sic!) and implementation...

You want me to add it? Fine - please send over the food, pay my rent and my Internet access charges. Trust me, I could do that (and much more).

Delphi Compile Speed

[ErikTheRed]

Granted, you saw it running on a few machines that weren't on the net. It was blazingly fast. Wow. Did you see what was inside the machines? What kinds of processors? Don't we always point fingers at Microsoft for pulling the same trick, demoing software on ridiculously overpowered CPU's and shovelfuls of memory?

Unlike Microsoft's products, Delphi (upon which Kylix is based) has a solid reputation for having an extremely high-speed compiler.

Bear in mind that the Pascal language is designed to be compiled in one pass, vs C/C++ which is not. Even in some decent-sized Delphi projects, each with several dozen source files (>50,000 lines total), I don't ever recall waiting for more than 3-5 seconds to do a complete build, and that's in full GUI mode with all of the bells and whistles (code optimization, &c.) turned on. Projects of equivelant complexity written in C++ have had build times of 3-5 minutes on the same machine. Considering the quality of code generated, the compile speed of Delphi is simply unreal. Since it's basically the same language and underlying compiler technology, I'd expect Kylix to be equally fast, if not faster due to lack of Windoze overhead.

Don't get me wrong - I'm not knocking C/C++; I use C++ as my language of choice for most projects. For those of you who haven't played around with Delphi though, I'd suggest you try a few small projects in it just so you can feel your jaw come unhinged when you see how quickly it builds. Way cool stuff.

Re:Delphi Compile Speed

[Emlyn]

Delphi does compile blindingly fast! I'm currently working on a C++ Builder project; looks exactly like Delphi in the IDE, has the VCL, but it's C++. My previous project, working with the same collegue, was all Delphi.

We chose C++ Builder because we thought we needed extra efficiency (which I now think was wrong). So we built stuff in C++, using templates and all kinds of goodies.

On his PIII with the lot, it takes over half an hour to recompile. It runs fine, but a half-hour compile time for a RAD project is murder.

On our previous project, an order of magnitude more complex, it took maybe 10 seconds to compile. Really!

The slow relative compile time of C++ Builder has been a big shock to us; rue the day we chose BCB over Delphi.

People familiar with both products will know that the default in BCB is to show a "Compile progress" window, to show you where compilation is up to. In Delphi, it is turned off by default. There just isn't time to display it, before compilation is finished.

Emlyn

Re:The whole article on the front page?

[yavo]

dude you suck, if you had an article of this length on the fron page it would take up all the space use that thing you cary on your shoulders

There are very few similarities to Visual Basic.

[torpor]

In fact, there are none.

Well, maybe, Delphi and VB share the ability to drag and drop stuff to create a GUI, but then thats a similarity that Delphi shares with VC++ too.

But other than that, Delphi is more similar to VC++ than it is to Visual Basic.

Re:Conference excitement and vaporware

[KGBear]

First, my point was that I saw Kylix running on a machine I happen to know as opposed to a machine "built to order" by Borland/Inprise. My point was not that Kylix on Linux is 10 times faster than Delphi on Windows.

That said, let me comment on your posting. Since you established some credentials, let me do the same. I'm not a Delphi expert, even less a Windows expert. However, I've been working as a system administrator for 10 years and I've worked with computers at one capacity or another for over 20 years. It's from the point of view of a sys admin that I write this.

Boy does that sound like a load of bullshit, and I'm saying that as a professional Delphi 3/4/5 developer. Are you comparing this with Delphi 5? On a P-II 450 Delphi 5 (or 4 or 3) compiles all but EXTREMELY large projects in a virtual instant, so how exactly are you timing this FUD 10 times faster claim? The reality is that the demo was likely extremely simple and compiled close to instantly on both machines.

Actually I don't know which version of Delphi they used on their machine. Very unscientifically, I didn't time it. The app was a simple text editor much like MS's Notepad. It took a few minutes to compile on Windows and a few seconds to compile on Kylix. The reality was just like I described it.

Linux doesn't make the processor run any faster, and in Windows 2000 or NT 4 the processor sits at 0% when I'm not doing anything, leaving 100% for Delphi to do its thing in, so any difference between the OS' will purely be the result of the coding abilities of Borland/Inprise.

This is interesting. Considering the monitor is itself a program that must consume some CPU cycles, I think it's odd that it shows 0%. If MS came up with a way to write programs that don't consume cycles at all, that's a real achievement. More probably, though, the monitor is accounting for its own load when showing the result. I wouldn't rely on such a monitor. If it accounts for its own load, why not accounting for the OS overhead too? Heck, 'top' shows itself consuming about 1% of the CPU while Linux is doing nothing else.

The file system is unmatched in 2000 (NTFS) and the memory infrastructure is good. Having said that what would be the difference?

The file system is unmatched by whose standards? I'm yet to see an independent benchmark proving the case one way or another. It's immaterial, though. They used Win98, not 2000, for that demo.

Does Windows insert special SLOW_DOWN_OP opcodes in the instruction stream? Does Linux have special RUN_SUPER_FAST_OP opcodes? No of course it doesn't. Barring multithreaded or multiprocess issues, where 2000 is considered superior anyones, there'll be no real difference.

Again, considered superior by whom? Again, immaterial. It was 98, not 2000.

What you seem to be saying, however, is that there never should be a difference in performance between different operating systems because no OS makes the processor run slower or faster. Now that, if you excuse me for paraphrasing your opening remark, is what sounds like a load of bullshit. I won't bother to explain why.

The operating system is a facilitator, it isn't an interpreter.

Exactly. Think about what that means and you may understand what I didn't bother to explain in the last paragraph.

Cheers.

I like Becker better.

Megido

[Tal Cohen]

Anybody knows what's going on with Project Megido, which was supposed to come up with a Delphi clone for Linux? Can't seem to find anything at megido.org (could be a DNS problem).

Re:Megido

[codepunk]

The Lazarus Project is the only game in town. Megido died a long time ago.... Cliff

Re:The whole article on the front page?

[cyber-vandal]

Surely not. I've paid for a fair bit of Linux software and will carry on doing so. I hope they release a trial version first however, so I can see if I want to buy it (all software makers should do this IMNSHO)

Produce BSD executables?

[rxmd]

Seeing that QLX is based on QT, the effort of adapting the system to produce BSD executables should not be too large - it should be just a slightly different threading and memory interface that would need to be supported.
I mean, I can run Linux executables on my FreeBSD machine, but I'd like it nevertheless.
Another idea might be an option to integrate different compilers in the background (i.e. put a gcc somewhere beneath it). Of course, compiling code would be slower by several degrees of magnitude, but one might write stuff on Linux or BSD and then produce applications for any platform that QLX and/or Qt is available for... and seeing that usually Borland libraries are compilable with Borland tools, one might consider porting the QLX layer as well, thus unleashing a wave of applications not only for Linux, but for practically every OS that supports Qt. This would be really nice.

(Brazilian) Portuguese translation

[KGBear]

I have posted a Portuguese translation of the entire article over here.

Re:CNN story here (and it is shorter)

[java_sucks]

Hey...it's the Don Knotts guy!! I haven't seen you around in a while.. Welcome back !!! I know you annoyed many a slashdot reader in the past with your moronic Don Knotts links but I want to let you know that I clicked them each time and usually chuckled when that goofy picture of Don Knotts loaded on my PC.

So thanks for the laughs and long live Don Knotts, the goofiest looking guy on the planet.

Re:Conference excitement and vaporware

[Cederic]

Regarding speed; if the Kylix you could play with was command line only then I would be very surprised if it wasn't fast - I know from experience that several hundred thousand lines of Delphi code compile in seconds on a bog standard development machine, inside the IDE. So using the command line compiler should be that quick, whether on Windows or Linux.

In fact, having grown into large scale programming using (BP7 then) Delphi and then Java, being forced in my current job to use C++ and waiting ages for compilation and then linking annoys me intensely. Luckily the C++ is only a small part of the job and I use Java most of the time - code, compile, test, repeat every 2 minutes or so leads to far higher quality code far quicker, and just isn't possible if it takes more than a couple of seconds to compile.

~Cederic

The whole article on the front page?

[Stradenko]

What's the use in having a "Read more..." option?

Re:The whole article on the front page?

[cyber-vandal]

Sounds like good news to me.

Re:The whole article on the front page?

[bugger]

>I hope they release a trial version first
>however, so I can see if I want to buy it
>(all software makers should do this IMNSHO)

Look at what Borland are doing on Windows.

* Free C++ command-line compiler
* Complete C++Builder Enterprise trial edition

* Free Java IDE (recognised as the best in the industry)

* Complete Delphi Enterprise trial edition

Realise that Borland do understand the concerns of the Open Source community.

Realise that the Borland community centers around component and utility source code, not around binary don't-touch-me nightmares.

What does that tell you about the future of Delphi for Linux and C++Builder for Linux?

Re:The whole article on the front page?

[joshv]

Yeah, I loaded slashdot.org, and wondered if I had somehow mistakenly already clicked on a discussion link. Nope, that's the front page.

Anyway, Kylix sounds like a great cross platform tool. I just wonder if they can manage to sell it to a bunch of people that tend to whine quite loudly if their software isn't free.

-josh

Re: Re: That's why they invented RPMs.

[Ded Mike]

Re: "...And InstallShield doesn't? ..."

Installshield is only available for Windows and Java (at the present time).

Re: "...I will never understand the need some people have to find one little grain of sand on the ocean shore that they just can't live with..."

I am not 'picking out grains of sand,' I am objecting to a _very_serious_ danger to the community.

Recently, I had a chance to hear Linus and Maddog speak at a developers' conference here in Chicago. Linus was asked by a -somewhat- savvy newsman (she used 'fork' in context) whether the chances of a fork in the source tree represented a serious danger to the OSS community and to the concept of 'the Cathedral and the Bazaar.' Linus' response was to the effect of 'I and the FSF control the code. Anytime I want, I can prune the code tree and eliminate forks.' This was fine when we had the public compiler, gcc. But with Kylix, due to the popularity of Delphi/Pascal and the proprietary libraries, (and in accord with my understanding of the GPL), this is the road down which Microsoft (the 'embrace and extend' guys, remember?) wants us to go down (see the 'Halloween Document' for details). It is the road the commercial *nixes went down, to their doom. Should this tool be commonly used by the community, we run the risk of fragmentation, just like the commercial *nixes, because of the structure of the GPL.

Re: "...When you try to say that one whatever is better than all others of it's type you are, in essence, espousing the same position that Billy-Boy used to make the latest Great Monoply. If you have to have a winner you are missing the whole point of life.."

I _don't_ have to have a winner. I never said which distro or tools I support. I was simply pointing out, in my first post, what RPMs were originally for (initial installs of OS and necessary tools, as well as an easy way to update them, until something better comes along) and then making a point about the _true_ power and responsibilities inherent in our community.

Personally, I use an FTP install of Mandrake and strip out as much of the non-GPL-licensed stuff as I can. I will continue to do so until the GNU/FSF gives me an alternative.

Since you said that you intend to buy Kylix, please, if you would speak as a programmer, let us know the weaknesses in gcc and the tools available under the GPL. Thank you, and BTW, nice quote from 'Rush.'

Re: Re: That's why they invented RPMs.

[Ded Mike]

Re: "...How can they donate a tool that is not done yet?.."

Hmmm, you're right, they can't. I just took the gist of the original story to mean that Borland would release a tool based on proprietary, non-GPL code and libraries.

re: "...Perhaps it's time to wait for the tool to be released and then see whether it is fit for GPL, LGPL, QPL, or whatever-else-OSS licence development?.."

_Not_ if the tool will be widely used to contribute to the source trees of the kernel or the _essential_ tools.

Re: "...I have very good reason to be convinced that Borland fully understand the concerns of the Open Source community...." and "...I have very good reason to speculate that the Open Source community will be happy with what Borland delivers - *once the tool is ready for release*..."

Please share your reasons. I base my points on Borland/Inprise's past history and violation of the Java Community License and their present price points (and the inclusions in) for JBuilder.

Re:Why seperate Delphi & Kylix

[fm6]

Actually, I/B hasn't said what's separate from what. They haven't said whether you'll have to buy a separate product to compile under Linux or what.

The thing to remember here is that Kylix is a project, not a product. It's kind of a subtle distinction. Which is why even the Borland people keep forgetting it.

Re:yes, lots of Windows programmers

[WraithX]

Could you be so kind as to name some of these programs?

OTOH, it will keep Windows programmers from learning how Linux development and programming works. They'll develop in the same style that they developed on Windows, and they'll produce the same kind of software that they produce on Windows.

Surely this applies only to the Delphi users who make software you don't like to use anyway? Afterall, if most of the windows apps are made with C++ the C++ users will have no problem adjusting to linux (they just choose not to). The Delphi users are all anxious to migrate to linux. It seems to me you simply don't want Delphi users to write software for Linux.

-- In principio creavit Deus caelum et terram.

Re: Re: That's why they invented RPMs.

[Ded Mike]

Re: "...Tell me why a KDE/Qt application is closed..."

Because of the original license under which the Qt libraries were released, essentially giving them a 'taking' (legal term) without recourse for developers as applied to Windows. Also, as the poster implied, closed because it is not portable across platforms (only runs on Linux/KDE). GTK-based solutions (with some diddling) can even be made to run on Windows natively. You were _not_ allowed to port your Qt-based code to Windows because of the license structure of the Qt libraries. Same was true of MySQL.

I believe, however, that the KDE tools can now be used with GTK-based GUIs (GNOME, etc.) without Qt libs and that the licensing issues, in the main have been worked out. (Somebody please correct me if I am wrong on either of these counts.)

Re: "...this nonsense."

It is _not_ 'nonsense.' Issues like this threaten the _entire_ community and the GPL, and therefor Linux and the BSD's.

Re:Delphi developers & Open source

[WraithX]

I think the opensource VCL you are referring to might be the Jedi VCL which has a website here: http://www.delphi-jedi.org/Jedi:VCLVCL:578699194

In addition to your comments, many delphi developers release their code under the GPL - something which will occur much more frequently once Kylix is out I believe.

-- In principio creavit Deus caelum et terram.

No FP?

[hoser]

Wow. Everybody's so shocked by the whole story being on the front page there's no First Post. Good job, Hemos, you got rid of them!

Re:No FP?

[Stradenko]

My apologies - I was so caught up in the fact that I got the first post that I neglected to mention that it was the first post.

Re:Open Source Free Pascal Compiler is available

[WraithX]

Well said.

I think that many people here are confusing Delphi with Visual Basic. Most Delphi developers are not drag'n'drop idiots as you can see from the wide array of Delphi apps out there - which could not be done with drag'n'drop.

-- In principio creavit Deus caelum et terram.

Go go, I/B !!

[DeICQLady]

I'ts very interesting how brave commoners get when they suspect the Dragon will Fall:
Dell.."We want to help develop Red Hat Linux"
IBM et al.. "If you use our product-on-Linux-steroids, we will support you"

This is the Age for Kings and Princes to Fall, And the pheasant Children to Rule the Throne...
-- Nadine Edwards

Nuff Respec'

DeICQLady
7D3 CPE

Re:Go go, I/B !!

[SquadBoy]

I did not know that many birds used computers.

Re:Go go, I/B !!

[DeICQLady]

Explain yourself?
Nuff Respec'

DeICQLady
7D3 CPE

Re:Go go, I/B !!

[sgifford]

I believe he was talking about the references to "pheasants" in your original post.

Re:I just can't get excited

[absurd]

"Or maybe the Bazh-haar software-development-model isn't capabale of handling such big projects."

You are just as big stupid troll as the statement above.

FUD

[EvlG]

There's some FUD going on in here - we need to clear some things up.

Kylix is going to be a big help to a lot of Windows-only shops that are now looking to migrate their development to Linux. Imagine; spend a few days modifying your app, and now it works on Linux. Of course, that assumed that you haven't used any COM stuff, but there are a LOT of apps that don't. Besides, in the Windows community, database access is one of the major reasons to use COM anyways. Sounds like Kylix/Delphi 6 will have that covered with the new database library. Exciting stuff.

Second, some Slashdotters are concerned about being able to only compile stuff with non-free tools. While it is true that you will have to have a copy of Kylix to compile stuff for it, I don't really see why this is an objection: lots and lots of commercial vendors are now going to be shipping stuff for Linux. This is another way to accelerate that. Perhaps some Free Software vendors can get some good usage out of it too. Who knows. I don't understand the logic behind complaining about havign more than one compiler, as one Slashdotter did. We have more than one everything else (soon, more than one kernel even.) Competition is good, right?

On a side note, what is an SKU? I haven't heard that term before, and it's use in the article stumps me. What does it mean, especially in the author's context?

Re:FUD

[Emlyn]

Apparently you can do DCOM, or a passing equivalent on Linux, using EntireX from SoftwareAG. And it is free.

http://web.slashdot.org/articles/98/12/15/132422 9.shtml

From what I read on SoftwareAG's site, visual stuff was not supported (that means no activeX controls). However, if you want to port an app that talks to windows only com-objects, you could use EntireX to do that I think - you could move your delphi app to Linux and talk DCOM to the windows components.

Lots of people are going to say "why would I want to use nasty 'orrible windows stuff". Remember that heterogeneity is good for software. It protects us from viruses to some extent, and breaks monopoly power. Remember, this technology is unlikely to move Linux people to Windows ("Yay! finally I can write stuff I can port to windows!"), and more likely to move people from Windows to Linux.

--

Speaking as an experienced Delphi programmer, Kylix is very exciting. Not for porting - who cares about porting code? It's exciting because I can leverage my existing knowledge and build Linux apps. Now I am very, very likely to do so; before, I was very unlikely to.

En mass, this will mean a whole bunch of coders(quite a lot of people) suddenly looking at Linux, for serious commercial work, who would not have done so before. For Linux supporters, this would have to be a Good Thing (TM).

Also, I would councel Linux coders to have a look at Kylix when it comes out. Delphi is a beautiful, heavenly, wonderful environment for building gobsmacking applications. Before you scoff, give it a look. Those of use who liked Pascal or Modula2 at uni will find that Object Pascal is a work of art (and beats the pants off C++ as an object oriented language). Anyone who likes Java will also like the language; they are quite similar (but Delphi->Kylix makes Java development environments look just plain evil). Before I get flamed, yes, I do understand the platform independence of Java, and no, Delphi/Object Pascal does not approach that. However, if you know one, it's not such a big step to the other (I think Object Pascal is closer to Java than C++ is); worth a look, it might be a better fit to some solutions where you would have used Java.

Urm, I'm rambling. Anyway, Borland/Inprise's press says that Delphi will do wonderful things for Linux, by bringing a truly powerful RAD development environment to Linux, which they hope will lead to an explosion in apps, which might just bring Linux up to a competitive level in the world of the desktop computer. I think they might be on to something.

Emlyn

Re:FUD

[x0]

SKU == Stock Keeping Unit, ie: part number.

Re:Conference excitement and vaporware

[KGBear]

Granted, you saw it running on a few machines that weren't on the net. It was blazingly fast. Wow. Did you see what was inside the machines? What kinds of processors? Don't we always point fingers at Microsoft for pulling the same trick, demoing software on ridiculously overpowered CPU's and shovelfuls of memory?

I have seen it runnin on my machine. I own a small Linux consulting company in Brasil and was invited by Borland to talk about Linux on a conference they gave for Brazilian developers. They asked me for a linux box in which they could demo Kylix, so I brought my own box. It's a PII 400 with 64 Megs. I did no scientific bechmarks on it, but the demo apps we built on Kylix compiled at least 10 times faster than the same code on their own Windows box (which was also a PII, but that's all I know about it). As a demo, we had the app up and running under Gnome while Delphi was still compiling it on the Windows side.

Unfortunately they made me uninstall Kylix before I brought the box back. No amount of pleading, cajoling or even begging would work, which is why I can't send you a pirate copy ;)

I think they probaly have a reason for not letting anybody have it, but the reason is not that they're demoing software on ridiculously overpowered CPUs or anything like that. I wouldn't call my PII ridiculously overpowered.

Delphi developers & Open source

[sg_oneill]

Ok.. I take the comment about delphi/kylix not being opensource, and particularly the VCL. A couple of things worth noting however. In my experience as a professional delphi developer, the delphi comunity is verry open source in terms of libraries / components. Check out the Torry components site and check out the large amount of components available there, ignore the stupid comercial/shareware ones, there are tons and tons of open source freeware ones. I ain't seen nothing like that at all for vis basic.

Also, the robust object model basically means that if a VCL object is not up to the task, create a new task, inheret and modify. No probs. There are I believe projects out there to create open source VCL replacements, so even that aint a prob.

With the gnarled exception of the compiler being closed source (Free pascal is a worthy replacement tho) I think it's majik for the linux community. Wan't a RTF word processor? Drop in a rich text component. make some save/load/print buttons. run. work. joy. The linux revolution is afoot!

What happens if you eat the blue beans ?

[dingbat_hp]

The conference opened with a "Matrix" themed intro, and there have been blue and red jellybeans all over the place.

What happens if you eat the blue beans ?
Do you wake up in Redmond ?

Re:What happens if you eat the blue beans ?

[CyberPup]

I ate both beans (I was hungry).

Nothing happened -- they must have counter-acted each other.

--CP

Re:What happens if you eat the blue beans ?

[Black Perl]

What happens if you eat the blue beans ? Do you wake up in Redmond ?

If you eat the blue beans, you wake up in Armonk. (At least if you eat the big ones :-)

Re:Turbo Pascal community and free software...

[ultrabot]

Delphi was one of the most underrated software products in history. It alone saved Borland from bankrupt after a string of bad management decisions. Its amazing how many developers use it today. Maybe C++ or Java has more magazine coverage, but almost all commercial shops in Brazil use Delphi for things ranging from quick and dirty apps to full blown corporate suites. Its power and flexibility are amazing.

There's always C++ builder, which gives you benefits of Delphi (that is, VCL) without having to learn yet another language (Object Pascal) or rewrite your algorithms/background code.

Not flamebait, he has a point

[dingbat_hp]

I write with a very expensive product for the day job, so I know it well, so it's what I use in the evenings too. If a Linux developer shells out money for Kylix, then they're damn well going to use the thing!

This doesn't mean The End Of Open Source (tm), but it is a risk. There will be code written under Kylix that gets OS'ed, and this will reduce the proportion of the total that's truly freely available.

Re:Screenshots

[MadAndy]

Call me a sicko too then :)

Not wanting to learn a language is a poor excuse. Learning a language takes a couple of days at worst - learning the libraries and environment is what takes time, and you'd do that with their C++ environment anyway.

Back in university I used to hate Pascal and Modula-2 with a passion. Primarily because of the pedantic type-checking (having to cast a short int to a long int to use it in a function call - pah!). With Object Pascal, they fixed virtually all the things that irritated me (except boolean operator precedence), and did a pretty tidy job of the object stuff too.

But in fact for me there are 2 killer features:
- real string handling: no buffer overflows or mucking about with memory allocation. In addition, the reuse element means it's often faster then standard C-style handling. If you're doing any form of database programming you will sing the praises of this repeatedly.
- F9. This is the Run button. You push it and your program runs. Yeah it compiles it too, but you don't notice that, it's so fast. Even recompiling all source files from scratch only takes a few seconds.

So it's all very well calling it a language designed to 'teach programming', but for many purposes it is definitively superior to C or C++. I've been happy to bet my livelihood on it and it hasn't let me down.

If you really want to attack it, at least come up with coherent arguments, like these:
- by using Delphi you are reliant on one vendor to supply the tools.
- the fact that it's not as widespread as C/C++ means the support from third parties and community isn't as strong.
- pointer handling still sucks, Pascal style. It works but it ain't nice.

Component Libraries

[hacman]

As a delphi developer, it seems crazy to me that so many people are missing a major issue that will strike when Kylix is first released (it will be a temporary situation, but until resolved will hamstring development efforts in Kylix) That is, the lack of 3rd party components. For example, the company I work for has 3rd party components for a treeview, oracle access, serial port access, reporting, etc etc. Now a number of these make use of windows API calls, or DLL calls (eg to Oracle) that will take quite some effort to port across. So, I guess what I am saying is don't view Kylix as the holy grail to getting Win developers onto Linux - view it as a first step that may eventuate into the holy grail once the component lib's are there. Hacman

Re:Component Libraries

[wide-eyed]

As a fellow delphi developer I see this as yet another incompatability between version. Such as 2.0 to 4.0. Most of us expect this to a certain degree. One of the important things to remember is not that you have to update and change your components making native api calls. Everyone who really works delphi to its limits must call on the API's to do some fudimental work. The point is that most if not all of the standard calls are going to port without modification. That to me is a life saver if you have to port a large project.

Most of the details and time it takes to port lie in the fact that if you built components on a well rounded component structure it would lend itself to being simple to update/port. This means seperating the api to a seperate object that is used by the component. You update the lower level that actually does the work and the logic and functionality of the rest of child objects stay in tact.

All this im sure is known by most people but its nice to re-iterate it from time to time to remind ourselfs how to design correctly. Even in the light of ever pressing demands by the boss to produce not screw with objects all day.

Most people would leave porting to the birds. This is because if you don't have any intentions of porting in the first place you probibly don't really have any need to port it at all. Most desktops are not going to suddenly switch to Linux overnight. The point of Kylix is to give you a choice of platforms. Thankfully they have also given us a thin API driven database attachment. The BDE is a hulking beast that often is too slow. Speed increases using native SQL Server api have made me blink more than twice. It will be a good thing for them to start out again with the database connectivity to iron out any bloated pieces.

The bde of course is a wonderful tool. That is it makes our lives very very very simple. This does not make it the best thing. Just a good thing.

Well enuf rant for now have to save some energy for The actual release.

Re:Component Libraries

[ghjm]

This is a good point. That's one of the reasons I surveyed the big third-party components vendors. Everybody's porting. Wait a quarter after the Kylix release, and you'll have (at least) Infopower, Raize, Reportbuilder, Orpheus...

Yes, of course if your own code makes lots of Windows API calls, you'll have to learn what the Linux equivalents are. At the Delphi presentation, they said that their own rule of thumb was that a major app that took a year to write should take less than a month to port. But nobody's claiming it will be a straight recompile.

-Graham

Re:Component Libraries

[sg_oneill]

Of course that's where we come in. My normal routine with developing a Delphi app is to first go and check out torry components page and find me any o/s components that will do what I want. Run a grep to see if it contains any GNU (Important to check for licencing issues) code, black box check it (Ie is it releasing memory properly? Is it thread safe etc) and plop it in.

Most of that should be available to be ported to Kylix as source code. The syntax is the same. The API is not. None the less once the code is in place, polymorphism suggests direct X-compile of the container project.

ActiveX stuff won't be so easy, but who the feck uses that shite in Delphi anyway?

Re:These guys are good

[leftover]

It was particularly good in the context of the edit-compile-link-debug(at machine level!) status quo. Integrated environment running on a single floppy disk - remember that 10M hard drives were still gleams in eyes for most of us. Very fast on a 640k 8088.

TP1 was a major innovation in the practice of coding, but probably would have had more respect if Borland charged 10x the $49.95 price. No tool change since has made the same kind of step difference to me personally.

SKUs, UPCs, and bar codes

[DragonHawk]

SKU = Stock Keeping Unit, I think. It's that barcoded number you see on everything.

SKU does stand for Stock Keeping Unit, but the barcode you see on "everything", i.e., most consumer products, are encoded UPCs, or Universal Product Codes. UPCs are "standard", where as SKUs are usually specific to the manufacturer or retailer you're talking to.

'Tis also worth pointing out that both UPCs and SKUs are the numbers themselves, not the barcodes used to encode them.

What the...

[DragonHawk]

No, RedHat (the 'R' in RPM) 'invented' RPMs to close the gap until InstallShield Corporation or someone else came up with a solution that worked in the *nix kernels of various flavors.

Where did you get THAT from?

RPM was created to install, remove, upgrade, verify, and build software packages. Even RPP (Red Hat's first attempt at a package manager) could track what was installed.

InstallShield(TM) isn't anything like RPM. It doesn't use a standard package format, it doesn't keep track of what has been installed, it requires a executable be built for each package, and that executable and its user interface is intimately part of the package.

You're so far off base people are wondering why you left the ballpark.

Thereafter, it was assumed the user would use 'gzip,' 'bz2,' or 'tar.(gz),' to uncompress the package/product and 'make' to install it.

Huh? If that was the case, they'd just dump it in using tarballs, like Slackware does (used to?).

RPMs ... -seriously- mess up the target OS (unless it is _precisely_ the OS the RPM binary was compiled on) ...

The only time I've seen this is when the packages in question were distribution-specific things,
like the initscripts package, for example. If you have something else you're talking about, I would like to here about it...?

... strew ungodly amounts of crap throughout the file system.

First of all, it is the packager who decides where to put things, regardless of package format. Everything down to "make install" works the same way. If you don't like where the packager put things, edit the source to install where you want. Don't blame RPM.

Second, I've found most of Red Hat's RPMs adhere to the Linux File System Hierarchy Standard rather strictly.

So, again, just what is your basis for this assertion?

The _best_ thing about the Open Source movement is the ability for _everyone_ to become a developer ...

I really think there is no single "best" thing about Open Source. But, either way, 99% of the people in the world have absolutely zero interest in becoming a developer.

Now there is a non-free compiler/tool/libs that threatens to take that essential paradigm away.

Um, hello? Most software already is closed-source, non-free, locked-up-tighter-then-a-drum code. It hasn't threatened Linux yet, and I don't think it will start now. Nobody's forcing you to use anyone else's commercial compiler. If someone produces a product using a commercial compiler, and you don't like -- don't use it. This isn't that hard to figure out.

... the practical use of Kylix could be limited to in-house projects in shops that need the cross-platform compatibility or the utility of Borland's otherwise excellent optimizing compilers and customer service/community.

I think the practical use of Kylix will be to whoever wants to code in it. Be it big corporations, small businesses, or Joe Hacker. Be it custom middleware, general applications, or system code. If someone wants to use it, they will. Otherwise, they won't.

Again: Nobody is holding a gun to your head, forcing you to use Borland's (or anyone else's) products.

At the very least, sysadmins, techs, and managers of Linux and *BSD shops should _carefully_ examine the products they install for GPL-flavored libraries and modules ...

You should always carefully examine the products you install, period. Be they commercial or free, closed or open. Interestingly enough, many people in the BSD camp consider "GPL-flavored" code something to avoid, because they don't like the restrictions on closed-source use. Likewise, there are those who think the BSDL's more lax attitude gives up too many protections.

Open Source doesn't make the need for software evaluation go away.

... and watch out for the inclusion of proprietary code such as what Borland will offer, lest they get too far down a one-way highway.

This, at least, is good advice.

I really wish Borland/Inpise had donated a tool the whole community could use.

And I wish there was World Peace. But I don't think either is going to happen. Borland's in the business to make money selling their software; that is their right. They have to make their business sink or swim using their chosen strategy. If their choices aren't compatible with the Linux community, they'll either change them or leave the market.

I seen no problem with that.

MPL, not GPL

[Dacta]

Actually, most non-commercial component developers are encouraged to use the MPL, not the GPL, because (a) it gets around the issues the GPL has linking to (the non-GPL'ed) VCL and (b) many major component suites (eg, Winshoes) are MPLed, so it makes linking to them easier in commercial software.

Borland pioneered "like a book" software licensing

[BitMan]

I trust Borland. They basically pioneered high quality development tools for the PC, while also pioneering the like a book mentality to software licensing.

Understand this isn't Microsoft we're dealing with, but Borland. I trust them to license commercial software and they make high quality software at that. With their powerful Object Pascal language and Delphi system behind a number of quality Windows applications (from largely non-Microsoft partners), it will be nice to get those applications ported to Windows.

And this is coming from a guy who prefers GPL, GNU/FSF, GCC, GTK+ and Gnome itself! ;->>>

-- Bryan "TheBS" Smith

Re:Screenshots

[lscoughlin]

There is also alot of really crappy C/C++ code wandering around out there. A tool does not bad code make, only a bad programmer makes bad code.

Re:Screenshots

[hholzgra]

yes, but a fool with a tool is still a fool
and klicky-klicky tools like delphi lower the level of entry to a point where people start to think they are programmers that are obviously not knowing at all what they are doing
this is a problem we where somehow protected from till today in the ***x world

Re:Conference excitement and vaporware

[grahamsz]

Unlike Microsoft, Borland actually have a history of turning out reasonably good quality software.

Can u you think of anyone that used Windows before version 3, anyone that used visual basic before version 3, anyone that used word for windows 1 (that was word 4 i think). On the whole microsoft stuff wasn't well adopted until a few generations in.

I've used TP since TP6 (since i haven't been programming that long) and delphi since version 1 and i've been thouroughly impressed all the way. A few minor teething problems with D3 but they were traced back to some corrupt microsoft debugger left on my system.

Jbuilder was a piece of crap, but by version 3.5 it's apparently working well, and that's good :)

Borland are a company that I've felt deserve trust and respect because they have served me well for the past 8 or so years. Microsoft seem to be a little less reliable tho.

Just my $0.02 + inflation of course

I will use Kylix to write GPL programs

[Danborg]

Don't count out Kylix as being a good platform for creating Open Source code. Remember this quote from the article:

At the opening keynote, Dale said that a major goal for Kylix is to make it possible for developers to release their own projects under any license, including full-strength GPL.

I wouldn't be suprised if all the libraries and a command line version of the Delphi compiler were made "free" as in beer. That way, anyone could compile Delphi projects. However, if you want the full blown IDE and all the bells and whistles, then you have to pay. That would be fair.

Re:I will use Kylix to write GPL programs

[nd]

This is the tentative plan.

(I'm not official, of course - I just read the Borland community newsgroups)

Re:I will use Kylix to write GPL programs

[GroovBird]

The license terms for Borland products say that the IDE, the compiler AND the VCL are copyrighted software. Even though as a Delphi developer you have access to the source of the VCL, you cannot redistribute it. Big companies are nowhere near Free Software fans, because that's where they make money. And that's exactly what GNU software is: Free. Dave

Re:I will use Kylix to write GPL programs

[rowland]

One small point: GPL'd software is copyrighted software too. Without the copyright, the license has no binding power.

Brent

Re:These guys are good

[lpontiac]

The Free Country's Developer City has links to an absolute crapload of free stuff, including a heap of free Pascal and Delphi compilers. The Borland abandonware stuff is here (free registration required).

Re:Delphi's object model.

[Karellen]

Huh?

Use TObjectList as a _base_class_? Or a container to store your objects?

Whatever the method, you've still got to remember to free the TObjectList yourself. And you've got t o create one for every new scope. So for every function that uses a local class type variable, you've got to create _another_ object just to hold references to them, and then you've _still_ got to remember to free it yourself anyway.

OK, it makes things slightly better, but I don't think it really invalidates my point. It's a hack to get around a language shortcoming.

Re:Delphi's object model.

[Karellen]

Huh? On what basis _would_ you argue for a certain programming language?

(Ideally, the right tool for the right job. However, some people don't have enough time to learn enough languages, in sufficient detail to do the job to their satisfaction, to accomplish this.)

Anyway I wasn't arguing for a programming language anyway. The 'it' I was referring to was the object reference model, which, IMHO, sucks rocks. The rest of the language, for the most part, is fine.

Re:Delphi's object model.

[Karellen]

You mean COM interfaces? COM interfaces on COM objects? When Kylix isn't doing COM? Riiight

So, I have the following choices for compound types:

1: 'record' types. POD types. Can be created on stack, or I can 'allocmem' & assign pointer for heap-based records. But no member functions, and certainly no inheritance, unless I want to start contstructing function pointer tables myself, which seems a little excessive.

2: 'class' types. Delphi's own object type, which I can't create on stack. All must be on heap, and must be explicitly '.free'd at end of use. They do have member functions though, and can use inheritance.

3: 'component' types. Based on COM. True object type that is _not_ garbage collected, it is reference counted. Objects created on heap, and free themselves when ref count drops to zero. Must be careful not to make circular references, which will cause objects to stay around for ever. This is hampered by the fact that it is impossible to create weak references to objects. Also, all object must be inherited from TComponent, and all must have v-tables, as well as the ref count and the ref-counting overhead, which I don't neccessarily want for tiny little objects I might have millions of.

I'm sorry, but that's a horrible mish-mash of ways to do things. C++ is bad enough with completely extraneous 'class' keyword doing _exactly_the_same_thing_ as 'struct', except for the default visiblity of members, but aside from that, at least it's pretty consistent.

I've found the features, I just don't like the way they seem to have evolved. I think they'd have been better if they'd been _designed_ with some type of consistency in mind.

Re:Open Source Free Pascal Compiler is available

[Halo1]

Delphi gets the most speed from its internal linking. We already have an internal assembler (adapted from NASM), but no linker yet and LD is very slow. The compiling itself is quite fast.

--

Re:Why seperate Delphi & Kylix

[Darkstorm]

I think the reason for the destinction is that Delphi in know to be for windows. And as much as I hate to say it there will probably be a few people who are not really bright that might pick up "delphi for linux" and not really read that its for linux. So if its labled differently there is a destinction between the OS which is very different.

I myself have used delphi since v2 and love it. I am eagerly awaiting kylix so I can make some cool toys for linux. This will also alow me to do contracting work for linux, not to mention client server programming between winblows and linux. I imagine that being able to build the client app for winblows and a server for linux would be awesome. Stability on the server side with the user still able to use winblows.

I hope they get it done soon, been waiting since it first was announced.

Re:Sounds good

[Anonymous Coward]

Most of the Delphi developers I know (I work with a bunch of 'em!) absolutely love it, and despite its similarity to VB, it remains resptectable.

Linux has needed this for a long time. One of the barriers to Un*x (especially free ones) development is that your choice of languages are fairly limited, and they usually amount to sticking together modules of two or more languages with duct tape and chewing gum (such as the "object oriented C" in GNOME), particularly if the app is (going to end up in) a WIMP environment.

I don't think that WIMP as we know it is going to last much longer, but in the meantime, Delphi will do a nice job of bringing Linux further into the mainstream (like it or not). When a niche software developer is able to say they support Linux right out of the box (excluding Debian), it's a big endorsement. Their customers are going to start wondering why software from their bigger vendors doesn't.

Re:Conference excitement and vaporware

[Plonk]

The reason why there's no Kylix software to take home is that it's still in beta, and the beta program is closed.

This doesn't mean that betas can't be demo'd.

Incidentally, the final proceedings CD is usually never available til weeks after the conference anyway...

Delphi's compiler is already blindingly fast, so there's no reason to assume that they're using high powered machines to pull the wool over anyone's eyes.

The report looks like something that actually happened at BorCon. How does this make it a rehash of a press release?

Interbase

[mark.odonohue]

I see everyone is interested in the gui. But for me the important event is Interbase. It's a quality product that in my opinion provides a lot of feature without having to own an army of DBA's.

I was (am) concerned by the delay (it's been 6months since it was announced and the products packaged and ready to go) I've always been suspicious of the "it's in the bag - we just have a few minor details to sort out before we sign", as that usually means some sort of big problem, and a 50/50 chance of it actually being signed.

I am however encouraged that they are still talking the opensource (interbase) talk while in the spotlight, and think the attention focused on borland during this crucial period will hopefully help keep them honest.

Re:That's why they invented RPMs.

[Ded Mike]

"That's why they invented RPMs...."

No, RedHat (the 'R' in RPM) 'invented' RPMs to close the gap until InstallShield Corporation or someone else came up with a solution that worked in the *nix kernels of various flavors. It was created to distribute distros and make _initial_ installs (of the OS and ancillary packages -the 'P' in RPM) easier and more friendly. Thereafter, it was assumed the user would use 'gzip,' 'bz2,' or 'tar.(gz),' to uncompress the package/product and 'make' to install it.

RPMs have become 'the prototype that wouldn't die' since they -seriously- mess up the target OS (unless it is _precisely_ the OS the RPM binary was compiled on) and strew ungodly amounts of crap throughout the file system.

"End-user don't need to build software, developer do..."

The _best_ thing about the Open Source movement is the ability for _everyone_ to become a developer, responsible for their environment, and with the tools to exercise that responsibility. Now there is a non-free (beer and speech) compiler/tool/libs that threatens to take that essential paradigm away. The poster of the original article did not say what the retail price-point of the Enterprise version of Kylix will be, but judging from JBuilder (a _great_ product, by the way) you can expect to pay about USD$2500.00/seat (more overseas - for less!) or more for the Enterprise version of Kylix Studio.

I agree with the first poster in this thread. This is _very_ dangerous and has the potential of forking the tree in a way not envisioned in the GPL or by Linus. I wonder how this is going over in Redmond? Has Borland/Inprise bought into 'embrace and extend?'

All the above said, the practical use of Kylix could be limited to in-house projects in shops that need the cross-platform compatibility or the utility of Borland's otherwise excellent optimizing compilers and customer service/community. Otherwise, there are plenty of tools/libs for Pascal (Delphi) and C++ in the Open Source tree. These tools and libraries should be used, (rather than the proprietary tools mentioned in the article) for distros and products that _truly_ support the community. At the very least, sysadmins, techs, and managers of Linux and *BSD shops should _carefully_ examine the products they install for GPL-flavored libraries and modules, and watch out for the inclusion of proprietary code such as what Borland will offer, lest they get too far down a one-way highway.

I really wish Borland/Inpise had donated a tool the whole community could use. There is no real difference between 'Kylix Studio' and 'Visual Studio' in this context. It makes me sad that yet another company does not yet 'get it.' I only hope that this prompts the various open IDE projects (KDevelop, the GNU efforts, etc.) to move _much_faster!_

Worrying

[orblee]

This proprietary library license thing might be the failure of this package. They should LGPL all of their libraries and just close source the IDE and everything else. Then we can make open source applications on it.

As a Visual Basic Developer

[CharlieG]

I think this is GREAT - I can't wait, maybe I can get into some development on a real platform.

I'm serious!

One of the ways Microsoft got Windows off the ground was to make it EASY for Fortune 500s to write those THOUSANDS of dull, everyday apps (and many INTERESTING apps) that they need to do business (Oh no, not another timesheet app). This has been missing from Linux up to this point. Delphi has always been an "edge of radar" kind of development platform in the F500 world, but didn't offer enough to bump VB (a case of too little, to late - If they had been first....)

Anyway, the idea of cross platform, and Linux, will make Linux (and Delphi) a MUCH more viable solution for a LOT of companies

Compiling GPL source with closed compiler/OS libs

[Karellen]

I thought we'd been over this before.

If you write a GPL'd program that requires closed libs (either compiler libs or OS libs), then that's OK, because your code is a 'derived work' of those libraries, not the other way around.

If the libs are GPL'd, then you can't link proprietary code to them, which is what the FSF wants. "We'll share our libs with anyone who's willing to share back", but if you write GPL code that needs to be linked to proprietary libs, then that's your perogative as you wrote the code. No-one is 'using' your code as a base, you're using theirs.

K.

Open Source Free Pascal Compiler is available

[Synopsis]

Hello,

It is with great pleasure that the Free Pascal Development Team announces
that

Version 1.00

of the Free Pascal compiler has been officially released on 12 july 2000.

The Free Pascal Compiler features:

- A Turbo Pascal and Delphi compatible compiler for the Intel processor
family, with some extensions to the Pascal and Object Pascal dialects,
such as operator overloading.
- Full debugger support with GNU GDB.
- An OS independent Run-Time Library, equivalent to the Turbo Pascal
and Delphi Run-Time Libraries, not dependent on external libraries.
- An API allowing for OS-Independent screen, keyboard and mouse
management.
- Many units, interfacing to various API's: gtk, xforms, zlib, ncurses,
sockets, X, mysql, postgresql, Interbase, paszlib, opengl, libgdb.
- A Free Component Library, containing many base classes from the
Delphi VCL.
- An IDE resembling Turbo Pascal's IDE (currently Beta quality) with
full GDB debugger support.
- More than 800 pages of documentation in Adobe PDF format, featuring
+ User's guide
+ Programmer's guide
+ Reference guide
+ reference guide for all units in the Run-Time Library
+ More than 440 complete example programs.
(Other formats include plain text, HTML and PostScript)
- Full sources to compiler, RTL, docs, packages.

More information can be found on:

http://www.freepascal.org/

Installations are available for the following platforms:

Dos - Using the go32 extender
Windows - Native Windows binary
Linux - RPM, .deb and .tar.gz packages are available.
OS/2 - Using the EMX extender.
Amiga - Based on version 0.99.5

Select your nearest download location from:

http://www.freepascal.org/sdown.html

(Beware, not all mirrors may be up-to date yet...)

or you can go directly to the FTP site:

Dos - ftp://ftp.freepascal.org/pub/fpc/dist/Dos
Win32 - ftp://ftp.freepascal.org/pub/fpc/dist/Win32
Linux - ftp://ftp.freepascal.org/pub/fpc/dist/Linux
OS/2 - ftp://ftp.freepascal.org/pub/fpc/dist/Os2
Amiga - ftp://ftp.freepascal.org/pub/fpc/dist/Amiga

Installation instructions can be found on the web-page.

If you experience any problems with the installation, please let
us know, so we can correct them as soon as possible.

Enjoy !

Michael,
speaking for the whole Free Pascal Development Team
------------------------------------------------ -----------------------

Re:Open Source Free Pascal Compiler is available

[Junks Jerzey]

Good stuff, but the whole hook of Borland's compilers has always been the speed of compilation. Delphi is up near a million lines a minute on an average machine (300MHz). If FPC is more in line with gcc, then there's no reason to switch (for me).

Re:Open Source Free Pascal Compiler is available

[bugger]

>A Turbo Pascal and Delphi compatible compiler

Free Pascal does not support the notion of "interface"s, something that Delphi has had for more than four years now.

Sorry, but interfaces are such an important design concept that I won't use FPC - at least until they are implemented.

Re:Open Source Free Pascal Compiler is available

[Anonymous Coward]

The problem with FPC is that most of the users are click-and-drool delphi converts. When they see a missing feature, they bitch and moan, since they're too incompetent to add it themselves.

Re:Delphi's object model.

[pblanton]

I am an expert programmer in both C/C++ and Delphi, with more than 10 years of real life experience. Yes, Java's garbage collection is nice but if that's the only thing that causes you to say "Basically, it's the worst of both worlds from Java and C++." then you are indeed a simpleton.

Anybody who argues for a certain programming language solely on the basis of its technology exposes himself as an Idiot who deserves to be ignored.

- Not anonymous and not a coward-

With Regards,

Phillip H. Blanton

Re:Conference excitement and vaporware

[ghjm]

Well, yes, I agree - I'm disappointed that I'm not coming home with anything. The Interbase guys really tried. They had enough CDs with them to give one to everyone at the convention, several times over. Kylix really isn't ready (actually one of Dale's big themes at the keynote was "Kylix will ship when it's ready").

BorCon has a tradition of some sort of themed event at the opening session. In the past, it's been Indiana Jones, Star Wars, medieval knights, and what have you. Doing "The Matrix" was actually fairly predictable. I'm not sure what your issue is with this.

The machines in the lab running Kylix are battered-looking Compaq Deskpro rentals. If you're interested, I'll go back to the lab and get exact specs, but trust me, these are not high-end machines.

And I can assure you that I am writing about what I saw at the conference, first-hand; I did not read any press releases before I submitted the story, and I don't work for any of the involved companies.

-Graham

Re:yes, lots of Windows programmers

[EvlG]

You logic is flawed though.

You state that most Windows software is made with VC++, not Delphi. Then later on you say that you don't want the Windows software mindset to creep into Linux by having Delphi programmers migrate to the platform. However, The "style that they developed on Windows" that you speak of is therefore from the VC++ programmers, not the Delphi ones.

Have you ever actually used an app programmed with Delphi? I've found them to be cleaner since the OO libraries are much simpler. I've found them to be less intrusive on the system, and in general, better behaved.

Why wouldn't we want to bring this to Linux?

You've misunderstood Delphi's object model.

[CRConrad]

Karellen writes:

You mean COM interfaces? COM interfaces on COM objects? When Kylix isn't doing COM? Riiight

"Riiight" indeed -- NOT!

No, he doesn't mean COM interfaces. He means OBJECT interfaces, like in Java. (You think that uses COM, too...?)

1: 'record' types. POD types. Can be created on stack, or I can 'allocmem' & assign pointer for heap-based records. But no member functions, and certainly no inheritance, unless I want to start contstructing function pointer tables myself, which seems a little excessive.

Yes, a "little" excessive indeed...

You'd basically be reinventing OOP. That's what the 'class' syntax is FOR, already.

2: 'class' types. Delphi's own object type, which I can't create on stack. All must be on heap, and must be explicitly '.free'd at end of use. They do have member functions though, and can use inheritance.

Why are you so hung up on where in memory they reside? Who gives a shit if they're "on stack" or "on heap"?

Just create them, use them, and free them -- how stupid does one have to be, not to be able to manage that? Oh, OK, for those who are, there are a couple shortcuts:

A) Assign them to an "Owner" object, which will take care of 'Free'-ing them when it is itself destroyed; or

B) Reference them (*only*!) _through an interface_, in which case the built-in reference counter takes care of them for you.

Neither should be necessary, but if you think you are too stupid to call 'Free', they're there for you to use. You don't *have* to, though.

3: 'component' types. Based on COM. True object type that is _not_ garbage collected, it is reference counted. Objects created on heap, and free themselves when ref count drops to zero. Must be careful not to make circular references, which will cause objects to stay around for ever.

No, *no*, NOOO! Sheesh, how fucking wrong can one be...?

There are no "separate 'component' types", and they aren't "Based on COM". They're just like "2: 'class' types", because they ARE ordinary defined-as-class objects.

You just declare them as, and *reference* them through, an _interface_ type in stead of through the class' native methods and properties, if you think you are too stupid to handle them the ordinary "class object" way; then they're 'Free'-d by the RTL when they aren't referenced any more.

Yeah, so you can't indulge in circular references and shit like that... So what? The type of moron who can't be expected to clean up after himself isn't very likely to get into any "advanced" stuff like that anyway. You gotta have *some* self-discipline, somewhere -- that's the price of using a powerful language where you, the programmer, are in ultimate control.

(The type of behaviour you seem to miss in Delphi, that you apparently have in C++, seems pretty VB-ish to me -- an interesting contrast for all the fuckwits who claim that it is Pascal that is the "toy language"...)

This is hampered by the fact that it is impossible to create weak references to objects. Also, all object must be inherited from TComponent,

Dunno what you mean by "weak references to objects" (unless it is what interfaces do?), but you're certtainly wrong again: No, they don't have to be inherited from TComponent.

They only *have to* implement some interfaces; if you write the objects, then *you* get to decide what interfaces you build into them. Sure, if you want them to implement IUnknown and, uh, whatever the other two basic reference-handling ones are called, then the easiest way is to inherit from TInterfacedObject. That's basically just TObject plus the implementations of those three interfaces.

(And yes, those interfaces ARE called the same as in COM; and no, that's probably not a coincidence. But that's just so *you* can easily build COM "objects" *from* them -- it doesn't mean that *they* NEED any COM stuff at all.)

and all must have v-tables, as well as the ref count and the ref-counting overhead, which I don't neccessarily want for tiny little objects I might have millions of.

Of course they have v-tables; that's what objects are all about! If you want pure dumb data, use an array of records or something.

And if you can't build a clean enough design that your "millions of little objects" keep track of themselves, use one of the built-in container classes, bung them into a TList or something.

I'm sorry, but that's a horrible mish-mash of ways to do things.

Yeah, sure, it might seem that way.

If one hasn't a fucking clue about what one is talking about.

I've found the features, I just don't like the way they seem to have evolved. I think they'd have been better if they'd been _designed_ with some type of consistency in mind

I'm sorry, but I think you have demonstrated that you aren't quite qualified to make that judgement.

Christian R. Conrad
My ISP is the Saunalahti company, of Finland.

Re:You've misunderstood Delphi's object model.

[Karellen]

"Why are you so hung up on where in memory they reside? Who gives a shit if they're "on stack" or "on heap"?"

*sigh*

Because - in the absence of garbage collection, stack based objects are deleted as the stack is unwound.

"Component types aren't based on COM"

Uhhh...let's look at the evidence, shall we?

TComponent methods include:

function _AddRef() : integer;
function _Release() : integer;
function QueryInterface(iid : TGUID, out obj) : HRESULT;

Looks like IUnknown to me. Hence, it's a COM object. Yeah - they've been protected, so you can't call them from delphi client code. Doesn't mean they're not there though and aren't accessible to other parts of the program.

TComponents are COM objects. They're almost certainly 'classic' ActiveX controls, but I haven't bothered checking whether or not they actually implement all the interfaces required for that. (I say 'classic', as the term 'ActiveX control' has now been expanded to mean pretty much any COM object. Sigh. Don't you love M$? Anyway, I digress...) And if Kylix doens't do COM, as has been mentioned (and not challenged AFAIK), then you're SOOL there then, aren't you?

As for weak references, those are references to objects that are not reference counted.

"Of course [objects] have v-tables, that's what objects are all about"

What??? You've got to be kidding, right?

"Objects", at their most basic, are _ideally_ bits of encapsulated data, whose internal representaion is shielded from the users of said objects by access methods, so that their internal state is always legal.

That does _not_ neccessarily mean function tables, much less v-tables. C++ certainly has the ability to create objects without v-tables, if it's a nice, small, concrete class that you don't expect to be derived further, and has no virtual functions.

I don't actually know how java does OO, but I'm certain that it doesn't do it with v-tables. (I recall someone mentioning it's one of most Java implementations inefficiencies)

What about smalltalk? That's a weakly typed OO langauge and certainly doesn't do polymorphism via v-tables.

What about perl? What exactly is the point in v-tables in an interpreted OO language? You're going through the code, a line at a time, read a function name, and look for a member function with that name. As a string compare. That's not a two-pointer-jump v-table-lookup function call. It's probably going to be implemented along the lines of IDispatch's 'Invoke' method, which is another way of doing polymorphism.

And yes, I can keep track of all my objects, thanks very much. But I like it if the language gives me a hand in doing so. The less I have to worry about such things, the less mistakes I am likely to make (and don't say you've never had a single bug in anything you've ever written) and the more I can get done in a given amount of time.

Look, OO is a design philosophy. It is not tied down to any particular language's representation of it, and can be (and is) implemented in many different ways.

Re:You've misunderstood Delphi's object model.

[davidmok]

"Looks like IUnknown to me. Hence, it's a COM object."

Since when (IUnknown = COM) in Delphi ? Seriously, what is your Delphi version ?

davidmok.

Hey Spud, *your* post is kind'a flame-bait, too!

[CRConrad]

'Spudley' writes:

"Kylix will be sold [...] GCC will be used by developers."
"Kylix will be sold [...] GCC is for developers."
"Kylix will be used [...] CGG is used by developers."

Hey, WTF, over...?

Delphi (and thus, Kylix) programmers ARE developers.

Christian R. Conrad
My ISP is the Saunalahti company, of Finland.

Totally out to lunch.

[CRConrad]

'jjc' writes:

"Kylix creates software that can only be compiled with non-free software."

Yeah, so?

You buy AutoCAD, you can use CAD drawings that I can't. I buy Quattro Pro, I can use spreadsheet files you can't. That's how it goes with proprietary software.

Should we campaign to abolish AutoCAD and Quattro Pro, whining about how some people can use some files and not others?!? Not even Linus wants to, AFAIK.

You don't think buying and installing Kylix will necessarily *remove* GCC from your system, do you?

Or that the kernel developers are going to start accepting patches that won't compile under GCC?

Christian R. Conrad
My ISP is the Saunalahti company, of Finland.

Hey "drinkypoo", you're a real numbskull.

[CRConrad]

"I don't want to learn pascal. I'd far prefer they brought out the C++ version first.
They didn't, because they're sickos.
In any case, C++ will prove to be far more useful throughout my life than pascal."

How bloody nice for you.

For *me*, OTOH, a sane and reasonable language like Pascal will probably prove to be far more useful throughout my life than C++.

Lucky for me, Borland is there to provide it for me.

You can run out and buy Visual C++ or Visual C# or something.

Christian R. Conrad
My ISP is the Saunalahti company, of Finland.

Hoelzle, you don't know what you're talking about

[CRConrad]

...do you? That is, in case you missed it, the AC above just explained why your post is just plain stupid.

Bloss dass du's woisch, gel?

Christian R. Conrad
My ISP is the Saunalahti company, of Finland.

Re:Conference excitement and vaporware

[HiThere]

JBuilder is well designed, but slow. I suspect that doing JBuilder is what convinced Borland to do Kylix based around Pascal for their **REAL** crossplatform environment.

Re:Screenshots

[jmccay]

What I heard has lead me to beleive that you can also compile C and C++ with Kylix. That would the best thing about it. I know pascal, C, and C++, and it shouldn't be too difficult to pickup on the variations in Delphi--especially now that there are books on it.
Now that JBuilder will be supported on the Mac OS X, they might port the other tools (like C++Builder and Delphi) too, but that is just speculation. It would be nice to write a C++ program on Linux and have it compile (and run) on WinNT/Win9x and Mac OS X all at once. Support 3 platforms with the development of only 1.
Their venture into the UNIX world began with JBuilder on Solaris. I think this is the case because they made JBuilder 3 Enterprise, Solaris Edition available via the web on 10/19/1999. I may be wrong about some of this. I think they are heading in the right direction! I look forward to thier future.

Re:That's why they invented RPMs.

[HiThere]

But do remember that KDevelop is closed in a different manner. Linux, probably KDE (though I haven't tried), only. There are many kinds of closure. That's one of the things that makes CygWin so important. And the GTK porting project (I don't really remember if porting is a separate project, but you know what I mean). And OpenGL. And GNAT (and JGNAT). These increase the openness of the environment.

Open Source Kylix

[scorbett]

I'm having a hard time understanding all the excitement regarding Kylix, considering that it is a closed-source, proprietary development tool that will cost an arm and a leg. I've used several Borland development tools for Dos/Win and I've been impressed by them in the past, but I wouldn't consider going out and spending several hundred (or thousand? how much is Kylix anyway?) dollars on a development tool for linux when there are so many free, open source alternatives.

If there are Pascal programmers out there who want to target Linux, I suggest you check out Free Pascal. If you must have a nice, user friendly GUI rather than a straight command line compiler, then I suggest you check out Lazarus, which is an open source version of Delphi for Linux. It's still in the early stages of development, but it is coming along nicely and shows huge potential, IMO.
--

Re:Open Source Kylix

[red_crayon]

several hundred (or thousand? how much is Kylix anyway?) dollars

Kylix is still vaporware, but:
Delphi desktop costs $99- ($49 students)
Delphi professional costs ~$200 ($99 students)

More fancy versions for serious database work cost more.

Re:Open Source Kylix

[Malor]

I think Delphi Pro lists at about $500 (that's around what I paid for version 3 anyway), and there's also a Studio version for about $1800 that's aimed at professional developers.

Re:Open Source Kylix

[red_crayon]

I think Delphi Pro lists at about $500

Prices have come down. Delphi 5 professional costs $99- at the local campus bookstore (student ID req.). Pretty sure list is below $500-.

Re:Screenshots

[bugger]

>I know pascal, ... shouldn't be too difficult to
>pickup on the variations in Delphi--

Uhoh. Pascal is like the old Ford Model T. Delphi Pascal is like this year's Ferrari model.

You can't compare these two. Oh, and once you have realised how nice and enormously powerful Delphi Pascal is, you'll *rarely* want to use C++. In a sense, Delphi Pascal has the performance advantages of C and the OO features of Java - a most exquisite cross-breed.

These guys are good

[Biswa]

Anyone remembers Turbo Pascal 3.0 ? It was a full blown Pascal Compiler for DOS back in '90. The compiler was somewhere around 39K. For 2K more, you could do graphics as well. It generated pretty well-optimized native code. In those days, all my programs + DOS boot (io.sys, msdos.sys, command.com and some other stuff) used to take one 360K floppy disk. I'll bet Kylix is going to be as good.
I think Borland has made Turbo Pascal 3.0 and Turbo C 1.0 free, but can't remember the URL. Try it out if you find it and haven't used them before. You'll be impressed.

Re:That's why they invented RPMs.

[bugger]

>I really wish Borland/Inpise had donated a tool
>the whole community could use

How can they donate a tool that is not done yet?

Perhaps it's time to wait for the tool to be released and then see whether it is fit for GPL, LGPL, QPL, or whatever-else-OSS licence development?

I have very good reason to be convinced that Borland fully understand the concerns of the Open Source community.

I have very good reason to speculate that the Open Source community will be happy with what Borland delivers - *once the tool is ready for release*.

Re:Screenshots

[drinkypoo]

I don't want to learn pascal. I'd far prefer they brought out the C++ version first.

They didn't, because they're sickos.

In any case, C++ will prove to be far more useful throughout my life than pascal. The only people still pushing pascal, in fact, work for borland or inprise or whatever they're calling themselves this week. Even the Apple programmers get to use C++ these days, they're not forced in a language designed to teach programming. Of course, as far as I can tell C++ is a language designed to teach humility...

Re:Help me to get it

[bugger]

>Delphi web modules can compile to traditional
>CGIs or ISAPIs
[author forgot that Delphi will do Apache on Win32]

>Kylix web modules will compile to traditional >CGIs or Apache modules

A "CGI" is just a program called by the web server. Your Perl/stuff is a CGI (assuming that you don't use modperl), just not compiled.

An Apache module is something that bundles up intimately with Apache itself. Massive performance improvements result (modperl, modpython etc are all Apache modules)

An ISAPI [DLL] is about the same as an Apache module, just for Microsoft's IIS (which some Fortune 500 companies reportedly are still running )

Re:Conference excitement and vaporware

[Malor]

I'm not really a programmer, more of a Sunday-afternoon hack, but I bought a copy of Delphi and spent about a month tinkering with it.

It felt like being led on a tour of what programming should really be like, by about ten of the smartest people in the world. I alternated between awe and excitement; I was blown away by how elegant things were. Programming in Delphi was closer to writing poetry than the grunt programming I'm used to. Everything was thought out well, top-to-bottom, no details left out. They abstract almost everything in powerful ways; you can just ignore the details and crank out some code attached to components, or you can dig into the components themselves and rewrite them to suit your exact needs. Don't like how they abstracted something? Just go back into the hierarchy until you find a level you do like/agree with, and rebuild from there. If you get the more expensive ($500?) version, like I did, you have all the VCL source sitting in front of you to serve as an excellent example.

Just like Visual Basic, you can get in and hack together apps in minutes after you open the box. If you start digging under the pretty facade, though, you'll find that the learning curve to real mastery is a LOT steeper than Visual Basic. A lot of that is simply that you can do so much more with Delphi. The VCL is just a pretty face on a massively powerful language. The VB facade hides an ancient compiler technology and the language itself is a fundamentally weak one.

I once compared Delphi to VB this way; VB is a flat, wide road that is easy to walk down, with lots of pretty little Disneyesque buildings scattered about. But if you look inside you tend to find that the houses are actually made of paperboard and are sitting on top of landfill. And that wide, easy road abruptly dead-ends into a nearly-vertical wall. Delphi, on the other hand, is like a twisty mountain road with some really lovely scenery. You have to do a lot more work to climb (learn) it -- but once you get over the top the road it turns into a superhighway that goes *anywhere* -- and through some of the prettiest country you could imagine.

You can do almost anything with Delphi. The compiled apps absolutely scream -- they build fast, rebuild even faster, and run like greased lightning. And you have all the source to the whole component library so you can tinker and rebuild to your heart's content. I don't think there's a software project you could name that would be too big to tackle with it, or at least with the fundamental compiler and language. I'm sure there are projects (particularly Linux-flavor) that wouldn't be well-suited to the Visual Component Library (VCL), but I can't imagine a problem that you couldn't tackle using Object Pascal itself.

If I were to become a full-time programmer (I'm a sysadmin now), Delphi is the language I would want to live in. I'm jazzed about it coming to Linux, finally. I had been yelling at Borland for years that they really needed to port it. It's a shame they waited until the edge of death to get a clue.

Like the author of the article, I also wonder a bit about the GPL issues. This will be a wonderful tool, and $500 is very reasonable for a compiler of this quality, but I will feel kind of bad about releasing source that will only build on a payware compiler. Note that it should be possible to compile/tinker with most released projects with the $100 model. The biggest difference between the low-end Delphi and the high-end one is the VCL source. If developers package their code carefully, the compiler cost to newbies/young programmers will be a lot less. It's an issue that needs thinking about. The really bad part is that you *need* that VCL source if you want to program seriously in this language. You're really crippled without it.

You know, I've been kind of thinking about this in the back of my head while I wrote the rest of this comment, and I really am starting to think that Borland should release the whole app for free, and attach a 'pay-us-for-commercial-use' license like QT. It's so powerful that it would be widely accepted in the Linux community, and I think they'd end up selling more commercial-use licenses than they otherwise would. Borland's big interest is going to be grabbing a big chunk of the Linux development environment -- if you are The Standard people will just throw money at you. RedHat seems to be doing pretty well at that game.... and Delphi is good enough that while it might never be The Standard for Linux, it has an excellent chance of becoming The Other Choice. I think not having a free version will hamper its ability to grow and thrive under Linux.

I saw a preview ...

[Reality_X]

I saw a preview of Kylix and the command line compilers/interpretters at COMDEX in Australia. It was pretty impressive :-)

Long posts

[Anonymous Coward]

Holy shit...can you imagine what Jon Katz will do when he learns that it's ok to put huge posts up on the main page? Run for your lives!

I would have clicked "read more"...

[hitzroth]

You didn't have to force me.

Re:Screenshots

[drinkypoo]

Not wanting to learn a language is a poor excuse. Learning a language takes a couple of days at worst - learning the libraries and environment is what takes time, and you'd do that with their C++ environment anyway.

Uh, I'm glad you're the boy genius, or have an excellent memory, whichever it is. Now, I consider myself to be pretty bright, but my memory is for crap, and I have an attention span about half as long as the lifespan of a gnat, at best, unless the subject matter involves breasts or explosions. Exploding breasts, however, do not have twice the draw.

It takes me a lot of self-convincing to actually sit down and learn something. Most of the time, I'd rather be playing UT. This is probably the reason why the only language I've picked up thus far is perl -- It's immediately useful and you can jump in and write worthwhile code (if unoptimized) without knowing much about programming. I have no formal training, know very little about data structures, and even less about which Algorithm I should be using at a given point in time and how I would implement it even if I did know that, but I can still write fairly complex perl scripts, with the help of my new favorite ORA book, "Mastering Algorithms with Perl".

On the other hand, buckling down and picking up some strongly typed fairly strict languages is just not my cup. I wish I were better with various sorts of linguistics, but my brain couldn't even retain the fact that to spew text in C you used printf for quite a while, let alone what all the damn percent substitutions are.

Anyway, it's all well and good for you to suggest that I just learn languages as I see a need for them, but frankly, it's just not that easy for everyone. So, like, thppt. In the meantime, I'd rather learn ANSI C or C++, which will be useful in a wider range of other peoples' programming projects. In the meantime, let me get back to my network administration.

Re:Screenshots

[ralphclark]

If you never learned to program in a structured, modular *and* strongly typed language then you should never have been allowed near C let alone C++. As it is you are a danger to yourself and others.

You have to have discipline as well as understanding to use languages with low-level features properly. One needs to understand the rules, and to learn to follow them *by default* almost without thinking before one can break those rules.

Programmers who don't follow this dictum *always* write crap code, because they simply don't know any better (cf. self-taught VB programmers).

Consciousness is not what it thinks it is
Thought exists only as an abstraction

Screenshots

[RPoet]

There's some nifty screenshots here, showing Kylix being run both under GNOME and KDE.

Kylix is sure to unleash a flood of new GUI applications for Linux, as "anyone" will be able to build the GUIs, and many should be able to learn the Pascal language. Many traditional, popular Windows applications may be ported as well. All in all, Kylix will be great for Linux!
--

Re:Screenshots

[hholzgra]

i'm afraid that it is all different
well, ok, you can do nice things with delphi, but there is a _lot_ of crappy delphi-made shareware out there too...
it is like punk music: the gould thing to say about punk is that anyone can get a gueitar and play, the bad thing is that anyone does ...

Re: Re: That's why they invented RPMs.

[Ronin Developer]

For starters, Borland (yes, that's their name again...Borland...An Inprise company), is stressing Kylix as a tool for building business and end-user appliations for Linux. They are not even daring to say that it should be used for kernel or device driver development.

At the conference, one of Borland's team leaders (I can't remember which one at the moment), stress this point. They DON'T want to compete with GCC in this arena. Borland percieves a different market, one filled with developers who need to develop database-centric, cross-platform business or end-user applications.

The fear that so many are exhibiting becausing of the Kylix is absurb and simply uncalled for. Borland is aware of the concerns of the Open Source community. I suspect they have no intention on going the way of the dinosour by being bound to another vendor's OS. Hence, their tools will be cross platform.

True, tools like Delphi Enterprise or C++Builder Enterprise will run you about $2,500. But, I'd venture that the initial release of Kylix (whether it be the Delphi or C++Builder variant), will be substantially less priced than that as you can get Standard and Professional editions of their Windows counterparts at princes ranging from $99 to $400.

RD

Help me to get it

[Tony Shepps]

Where Delphi web modules can compile to traditional CGIs or ISAPIs, Kylix web modules will compile to traditional CGIs or Apache modules. The level of integration between Kylix and Apache is impressive.

I'm sorry: can someone dumb this down for me, a traditional web developer using Perl, PHP, and .shtml and very few other bells/whistles?

(I know it's not geek to admit you don't know something, but I suspect there's a lot more to this than I can grok.)
--

Re:Moderate: -1, Redundant

[HiThere]

There's no problem with using a proprietary compiler ... though that does mean that having the source doesn't automatically allow one to build the binary. If the Kylix libraries are LGPL or equivalent, that wouldn't cause a problem either. It does mean that the library source would need to be available.

However, lets consider a slightly different case. Python is GPL. It also allows one to call essentially any program on your system regardless of the license. One can build links into the program that one is writing that automatically invoke totally proprietary programs. (Anaconda must do this.) There exist GPL'd methods of translating Python code into c, where it can be compiled with gcc, and linked against core libraries. What is the License of this program? One can even ensure that all of the links are static. Assume that this starts from a script that links ... O, say, WordPerfect. Assume that there is a hook that can be called to automatically make a back up of all files in its recently used list. You write the script. You compile it. You link it. What is the license of that compiled program? Can you tell?

One doesn't need Kylix to cause this problem to become intractable.

Delphi's object model.

[Karellen]

As someone who has developed for delphi (although my language of choice is C++, and I have looked at Java + others as well), I have to say I hate the way that Delphi's object model has evolved.

Delphi is based on pascal, and retains pascal's 'record' data capabilities, which are the same as C's 'struct'. It's a compound POD (plain old data - no member functions) type. You create variables of 'record' types as you do any other variable.

So far, so good.

However, the 'object' model sucks rocks.

If you want an object, you create a 'class' object (which is derived from the base 'object' class) and create an instance of that type.

However, you can't create class objects on the stack. You can only create object references. These must set to refer to object created on 'the heap'. And they _must_ be manually deleted. There is no garbage collection, and no reference counting.

Basically, it's the worst of both worlds from Java and C++. You can't even create 'smart references' that would do auto-ref counting, or even just delete objects as the stack unwinds, as those too would be objects, which you'd have to remember to free at the end of your funciton.

Basically, you're looking at potentially worse than c-program memory leaks, as people will prefer to use class types than record types (OO-mania), and stack-based clean-up will go completely.

OK, I can write big programs in C that don't leak memory. But it takes more effort and pains to do so than in C++, which allows me to concentrate on other things more. I fear for some of the memory hogging programs that will come out of Delphi if it gets really 'big'.

K.

Re:Delphi's object model.

[bugger]

>There is no garbage collection, and no reference
>counting.

It seems as if reading up on "interfaces" would help you out. Interfaces *are* reference-counted, interfaces are garbage collected.

Hop over to the excellent Borland newsgroups on

nntp://forums.borland.com/
nntp://newsgroups.borland.com/
[you do need to connect to either of these two]

and bring that up in, say,

borland.public.delphi.objectpascal

You will be surprised what features Delphi has to offer which you haven't found yet :->

Re:Conference excitement and vaporware

[Ergo2000]

I did no scientific bechmarks on it, but the demo apps we built on Kylix compiled at least 10 times faster than the same code on their own Windows box
Boy does that sound like a load of bullshit, and I'm saying that as a professional Delphi 3/4/5 developer. Are you comparing this with Delphi 5? On a P-II 450 Delphi 5 (or 4 or 3) compiles all but EXTREMELY large projects in a virtual instant, so how exactly are you timing this FUD 10 times faster claim? The reality is that the demo was likely extremely simple and compiled close to instantly on both machines.

Linux doesn't make the processor run any faster, and in Windows 2000 or NT 4 the processor sits at 0% when I'm not doing anything, leaving 100% for Delphi to do its thing in, so any difference between the OS' will purely be the result of the coding abilities of Borland/Inprise. The file system is unmatched in 2000 (NTFS) and the memory infrastructure is good. Having said that what would be the difference? Does Windows insert special SLOW_DOWN_OP opcodes in the instruction stream? Does Linux have special RUN_SUPER_FAST_OP opcodes? No of course it doesn't. Barring multithreaded or multiprocess issues, where 2000 is considered superior anyones, there'll be no real difference. The operating system is a facilitator, it isn't an interpreter.

Something like Kylix for Linux (although it'll be astronomically more complex than a Project1/Edit1 type demos that port great) is a great step and it'll be great being able to port to Linux relatively easily, but the reality is that it brings realism to the equation : Why bother with Linux? What does it really offer (apart from the ridiculous 10x faster FUD) that you don't get (usually better) in Windows 2000? Forwarding a cult, or conversely spitting in the face of Microsoft, isn't a valid selling point for most corporations (except for Sun "you don't have any privacy so get over it" and dumpster diving Oracle).

Good to see Inprise doing this though as they are an awesome company with some incredible products. From a educated perspective I will say that the whole porting issue will be infinitely more complex than many portray it, but it'll definitely be easier than using two totally disparate development platforms.

Cheers.

Irony

[FFFish]

Oh, how ironic.

The product won't be free-as-in-beer... but is anyone whinging on about that?

Hell, no. It's exciting, it's going to be great, whoo look at the speed!

How very different from the response to other products that aren't free-as-in-beer.

Corel releases powerful, useful software... and all they get is a kick in the goolies from the Slashdot crowd. Endless bitching about it being closed-source, about it costing money, about it coming from a Windows platform.

More and more, I develop the opinion that the majority of Linux users are jackasses. They want the world to come to Linux, but they don't want Linux to be truly useful to the majority of people. In their opinion "if it's not a development tool, it's shite!"

Linux won't be used by the general population until it has easy-to-use, general-purpose tools, like word processors, checkbook managers and a few games. With this current attitude toward practical popular tools, it'll be a loooong time before it gets there.

But, hey, Kylix will help you spooge your 133T H4X0R sK1Llz!

--

Re:Irony

[drinkypoo]

Corel releases powerful, useful software... and all they get is a kick in the goolies from the Slashdot crowd. Endless bitching about it being closed-source, about it costing money, about it coming from a Windows platform.

Corel releases buggy, crashy software... and I delete it.

Corel Computer releases Netwinder systems for about $400 more than it would cost me to make an (admittedly larger) pentium-based PC with the same horsepower, memory, and storage. Make that $200 more if you want me to include video and audio input.

Corel is old and tired. Corel Draw was perhaps the buggiest commercial application over $100 ever written.

The other thing you're missing is that plenty of people have been bitching about the closed-sourceness of Kylix for a long time now. They're probably just worn out.

more info here

[red_crayon]

http://www.drbob42.com/borcon.htm
---------------------------------------
Open source projects with a free-beer compiler?
I don't see why not, though it's a bit different.

Give me a break will ya.

[Malcontent]

Jeez man how paranoid are you. It's just another programming language. Can you write GPLed code in Java?

Relax. I am looking forward to shelling out whatever it takes to buy klyx I know it's going to rock. I can't wait to be able to do cool GUI stuff for linux.

Re:That's why they invented RPMs.

[X-Nc]

No, RedHat (the 'R' in RPM) 'invented' RPMs to close the gap until InstallShield Corporation or someone else came up with a solution that worked in the *nix kernels of various flavors. It was created to distribute distros and make _initial_ installs (of the OS and ancillary packages -the 'P' in RPM) easier and more friendly. Thereafter, it was assumed the user would use 'gzip,' 'bz2,' or 'tar.(gz),' to uncompress the package/product and 'make' to install it.

RPMs have become 'the prototype that wouldn't die' since they -seriously- mess up the target OS (unless it is _precisely_ the OS the RPM binary was compiled on) and strew ungodly amounts of crap throughout the file system.

<sigh>
And InstallShield doesn't?

I will never understand the need some people have to find one little grain of sand on the ocean shore that they just can't live with.

• RPM vs DEB vs TGZ
• RH vs Slack vs Debian vs Mandrake vs etc.
• GNOME vs KDE
• GTK+ vs Qt

Hell, the whole Linux vs *BSD is just about as idiotic! (I'm not even going to go into Language Wars<tm>).

When you try to say that one whatever is better than all others of it's type you are, in essence, espousing the same position that Billy-Boy used to make the latest Great Monoply. If you have to have a winner you are missing the whole point of life.

All of these things are tools to use to get the job done. As Neil Pert once said:

What you own is your own kingdom
What you do is your own glory
What you love is your own power
What you live is your own story

To bring this rant... uh, "post", back on topic, Kylix is a Good Thing<tm>. As soon as they have the C++ version out I'm buying it. I have C++ Builder for Win (though I haven't ever used it) just in case I might have to build a Win app. You never know...

---

FreePascal, FreePascal, FreePascal

[stikves]

somebody else has already mentioned it but FreePascal has the same code syntax of Delphi, and also a free compnent library for it is under development and with a RAD tool (called lazarus).

WOW

[xianzombie]

I thought it was a Katz article juding from the main page.

Sounds good

[Jon Erikson]

This will be one of the most important events for corporate acceptance of Linux in a non-server role, since Delphi is used by a lot of companies to develop their software. Since Kylix will be pretty much the same as Windows Delphi, changes to the source code in order to port an application will be minimal (at least in terms of the usual process of porting), and may make companies think about moving over from Windows to Linux, especially those companies which provide and all-in-one of software and hardware to customers.

So we may see Linux begin to really encraoch upon the business software market, which whilst not the intention of Linux, is still a large market and a great opportunity.

---
Jon E. Erikson

wow

[Maddog_Delphi97]

That's all I can say right now... One question I have though: How much does it cost? I'm assuming the pricing is the same as previous versions of Delphi.

Why seperate Delphi & Kylix

[grahamsz]

For those of you who missed what Kylix is, here's a press release here.

Essentially it's Delphi for linux, but I have to wonder why they are keeping the linux and windows versions distint.

Sure I realise there are some very large parts of linux missing from windows, and vice versa, but surely the way forward is to abstract these details as much as possilbe from the programmer.

Sure it means that low level components have to be developed seperately for each OS, but this would make code immediately compile on both platforms.

I haven't looekd in on this project for a LONG time but you may want to check out REALbasic (just noticed they wrote IE for mac in it :)) because it lets you build Mac and Windows from the same source, despite the OS differences.

None the less, I love borland/inprise/whateverdafucktheyarecalledtoday, and their software and so long as it's not another JBuilder (cringe) i'll be happy :)

Conference excitement and vaporware

[Brento]

You really get the most from the bottom line of his message: he's not coming home with anything.

Every conference I've ever attended has revolved around building hype and excitement. The key isn't how much hype is left after an hour, but how much is left after you've gone home with (or in this case, without) a copy of the tools to actually play with and find fault with. Trust me, when you go home emptyhanded, there's a reason. Everything seems great when you take it from a jolly Matrix-themed conference with cool jellybeans, but when you're sent home without a CD, that really says something.

If Microsoft pulled a shenanigan like this and compared itself to the Matrix, we'd all be cracking jokes about it and saying that the software was an illusion. In this case, though, the guy is actually singing the company's praises! Hello? No software to take home? Doesn't that scream vaporware? We wouldn't tolerate it from Microsoft, and we shouldn't tolerate it from anyone else, either.

Granted, you saw it running on a few machines that weren't on the net. It was blazingly fast. Wow. Did you see what was inside the machines? What kinds of processors? Don't we always point fingers at Microsoft for pulling the same trick, demoing software on ridiculously overpowered CPU's and shovelfuls of memory?

Am I the only one who thinks this "story" is nothing more than a rehash of a press release?

Preview?

[cybercuzco]

(Use the Preview Button! Check those URLs! Don't forget the http://!)

Follow your own advice guys ;-)

Re:The developer community is already split

[AlisdairM]

A few points in Borland's defense. I have used their products for a decade now, and am one of the few to migrate from their Pascal products and Delphi over to their C++ line. Most migration goes the other way! I have had little trouble with version incompatibilities, and used BCB since the first version. True, no upgrade has been 100% completely hassle free, but I don't expect this when upgrading a compiler. Maybe my expectations are too low. Given that each successive version is closer to the ANSI C++ standard though, I expect to have to correct errors in my code that previously passed. Otherwise, I would be using the MS product under Windows. The idea that Delphi devlopers wanting to migrate to Linux will learn C++ is to vastly underestimate the zeal of this community. C++ as a language is public enemy number one, as it is seen as the main competitor for their tool under Windows and the chief reason it never reached the market dominance it deserved. Likewise, the C++ community often sneer at Delphi seeing it as little more that a VB-clone in pascal. This completely underestimates a powerful OO language that has been more mature than C++ until the standard was finally ratified a couple of years back. Even now standard compliance is not perfect. This down-looking attitude really aggravates the community, C++ is about as far from their shopping list as you can be, despite the languages being superficially very similar. Finally, the VCL may not be Open Source, as it is 100% owned and maintained by Borland, but it is hardly closed source either. Ever since version 1 the source code has been available for a registration fee, and shipped as standard with the 'Professional' and 'Enterprise' versions. Most Delphi libraries ship as source code, as the intermediate .DCU file is never binary compatible between product versions, something that is a goal for C++ and the .O/.OBJ files.

Re:yes, lots of Windows programmers

[jetson123]

Yes, Kylix will bring lots of Windows programmers and Windows programs to Linux. Many of the programs will be proprietary, some will be free. So, why exactly is this good?

Most mainstream, Windows programs are written in VC++, so it's not like that this will bring the few crucial apps to Linux that Linux needs for mainstream acceptance (mostly Office).

OTOH, it will keep Windows programmers from learning how Linux development and programming works. They'll develop in the same style that they developed on Windows, and they'll produce the same kind of software that they produce on Windows. That kind of software is one of the main reasons I don't use Windows, and one of the main reasons other people have fled Windows as well.

The developer community is already split

[Anonymous Coward]

Case in point: Just this week I made the decision to abandon Qt/Kde as a development environment and move my work I do at home (which fortunately hasn't been published yet under their questionable license) to a truly free C++ environment, probably WxWindows. Not quite as polished as Qt but damned good anyway.

This is heartbreaking to me personally because I love Kde and think Qt is the a wonderful class library, but there is no future for small developers in the Qt camp. Qt is strictly for large corporations and charities like Kde which can afford to be "non-commercial". Not that I write software with any intention of it being closed or even to make money from it, but that I want those who do to be able to *interface* with my projects without fear of licensing uncertainties or tolls down the road.

Borland sucks. I learned C++ mostly with Borland C++ for Windows, but Borland (I still call 'em that) kept changing the API's and changing the API's and changing them. I was forced to go back to raw Win 32 API using C, and then gravitated to the Cygnus-MingWin environment, but now have got completely out of the Windows world, thank goodness. Borland made it impossible for me to develop with their toolkits without upgrading 5 times in 5 years. (Two different versions of C++ Builder were incompatible also. I didn't even bother to waste my money on those).

I don't agree that this will result in a large infulx of new desktop apps for linux. Delphi is a good development environment, but I think you will find that most programmers coming over from Windows will opt for the free development environments and class libraries. Yes, even those who currently use Delphi. If a good pascal system licensed under GPL isn't available, they will learn C or C++ or Java, which really is the coming thing. It really is now ready for desktop applications, not just applets.

Borland will meet the same fate of other commercial entities that try to port the proprietary concept along with the proprietary software. Closed source and license fees is fine for apps for end users, but not for development libraries, *especially* not for libraries which will be primarily responsible for creating a gui which is a *major* part of any deskopt app. Didn't I just explain above what happened to me with Borland C++ and Windows?

Gtk may not be so great and I destest Gnome, but I will have to live with them. There is always Fltk which lacks some features but which is incredibly fast and streamilned, and WxWindows does encapsulate the gtk libraries in a more intuitive manner and doesn't add much overhead. Some things are even faster than naked Gtk.

Again, it is not just the cost in dollars but the certainty that you will have to pay and pay and pay again and that you have no control over the source to the libraries you are using. You will have to adjust to changes in the API's made primarily to force you to upgrade, and to buy new development tools to match.

Well, out with the old and in with the new. Don't get me wrong, for end users Kde is a great desktop system, and there is no good reason not to use it. But if you are a developer and are not getting paid to develop using Qt/Kde/Delphi/Borland, it is a complete waste of your time.

I still don't like Gnome!

Puh-leeze

[Anonymous Coward]

You've got to be kidding. For one thing, Borland's tool is a RAD package, not a mere IDE. For another, the real news is not just what it is, but what it makes possible--the porting of thousands of new programs from Windows to Linux.

This development is Very Good News for Linux and Linux users, even for those who won't use the Borland tools or programs created with them.

Turbo Pascal community and free software...

[cribeiro]

In its early times, Borland was somewhat developer friendly. Solid products and good documentation (at least judging from that times standards). At that time the GNU project was just starting, and most development was closed source. Borland was just that - a closed source company - but the Turbo Pascal community was somewhat different. There were groups like SWAG, and the ftp site at garbo.uwasa.fi (from Timo Salmi). Sharing code snippets was pretty common. Even the commercial libraries for Turbo Pascal used to cost much less than the equivalent ones for C, and almost all of them included source code. This was one of the reasons was Pascal survived so long - specially if you consider the strong pressure towards C (and later C++) in the academia and the industry.

Delphi was one of the most underrated software products in history. It alone saved Borland from bankrupt after a string of bad management decisions. Its amazing how many developers use it today. Maybe C++ or Java has more magazine coverage, but almost all commercial shops in Brazil use Delphi for things ranging from quick and dirty apps to full blown corporate suites. Its power and flexibility are amazing.

I think Kylix is going to be a very strong product, one that builds upon the strenghts of Delphi. It also can take Borland back to its initial days, and make Linux a truly viable alternative for commercial software development. The Pascal community can make a strong difference. Object Pascal structure makes easier than ever to share components, and the vast amount of quality free code for Delphi (RXLib is a wonderful example) is a tribute to this.

We already know it

[pravel]

The major announcement at the intro was that JBuilder will be ported to Mac OS X, with full support for Aqua.

We already know it, it was posted on Slashdot 3 hours ago.