Borland C++ For Linux

Ardax writes: "Looks like Borland is going to be releasing C++ for Linux, according to this InfoWorld article. We'll be seeing more details at LinuxWorld in NY next week. The article doesn't mention whether this will be C++ Builder for Linux, or 'just' a command line compiler. No matter what, this is a sweet thing. I wonder how it will compare to gcc? (I wonder if it will be able to compile the kernel? :-) ) If it's the whole C++ Builder shebang, I wonder if there will be an Open Edition? Borland's Community site has a blurb about this. There's no comments at the Borland community yet, but some interesting commentary might pop up there."

Comments

Market

[SonCorn]

I could see there being a market for Borland if they released a nice GUI C++ development environment; but if it was just a command line program, can someone explain to me why they would use it instead of gcc. I just see no reason to pay for it if it is a command line program. I can't believe that they would offer some feature that gcc doesn't.

Re:Market

[qurob]

can someone explain to me why they would use it instead of gcc

On many UNIX workstations, GCC makes slower/much slower code than the system vendor's compiler.

Many people argued the speed/size benefits of Watcom's DOS compilers compared to DJGPP, the (DOS port GCC)

Re:Market

[dinotrac]

It's awfully hard to know without seeing the Borland compiler first,
but...
gcc is not known for handling C++ especially well. 3.0 is supposed to support the language better, but is reported to generate larger executables and no better performance than present.
I have no idea of how the Borland compiler would get along with current linkers and whether it would be better than the current mess, which is largely responsible for the long time required to start KDE apps.
This stuff may not matter for free developers, but commercial developers may see improved performance as one of the edges they can ask people to pay for.

Re:Market

[SpinyNorman]

The same is true of PCs too. gcc doesn't come close to Intel's C++ compiler in terms of optimization or features (e.g. SSE/MMX code generation and vectorization). Unfortunately Intel's compiler is pretty expensive, but maybe Borland's will be more reasonable.

The only thing (not a bad thing, mind you) that gcc has going for it is that it's free... it's hardly the compiler of choice if you really want to optimize your code.

Re:Market

[cotcomsol]

I would happily pay for Borland C++ for my developers if it is faster than G++ (which shouldn't be hard). For example, Micrsoft's cl compiler is easily 10 times faster than g++ on moderate-to-large size c++ apps.

Re:Market

[ahde]

The gcc we all use is much slower than the gnu-pro gcc you can buy from Cygnus/Redhat too.

Re:Market

[anandrajan]

There's always been plenty of excitement on the borland.public.kylix.non-technical newsgroup regarding native RAD C++ for linux. John Kaster (from Borland) is the guy who usually confirms or denies rumors. For a while, there was a rumor that the C++ RAD version of Kylix would be called Sylix and he squashed that rumor pretty quickly.

You can access the newsgroup here (sorry for the long URL, blame google not me)
http://groups.google.com/groups?hl=en&threadm=3C 43 71D8.51AB0544%40uk.renaultf1.com&prev=/groups%3Fnu m%3D25%26hl%3Den%26group%3Dborland.public.kylix.no n-technical%26start%3D25%26group%3Dborland.public. kylix.non-technical

Also please check out an earlier message on this topic which may have already been covered on slashdot (sorry too lazy to check at 3:15AM)

http://community.borland.com/article/0,1410,2819 8, 00.html
This one gives you an email address kylixbeta@borland.com where (surprise, surprise!) you may be able to get a beta.

Re:Market

[joss]

Huh ?

I agree that g++ is a bit slow, but your figures are way off. In many cases cl is slower. If you use precompiled headers cl can be faster, but not 10x. I think gcc3.1 should have necessary performance improvements especially for heavy template useage (which is main cause of slow compile times in c++). Besides, g++ 2.95+ is a much *better* compiler than even latest VC++. Try doing some complex template stuff (eg Blitz) in VC++ - it's plain broken.

Re:Market

[tomstdenis]

"Many people argued the speed/size benefits of Watcom's DOS compilers compared to DJGPP, the (DOS port GCC)"

Many people such as? GCC optimizes just as much as Watcom [and much more than MSVC].

For example, my crypto library

http://libtomcrypt.sunsite.dk

Includes both MSVC and GCC make files. Compare the timings [ciphers] for yourself!

Re:Market

[tomstdenis]

"gcc is not known for handling C++ especially well"

Or do you mean "gcc is not known to compile invalid standard non-compliant code well".

Both borland and MSVC diverge from the standards much more than GCC ever would.

Tom

Re:Market

[Graspee_Leemoor]

Many people such as ME! Watcom c++ for dos just completely wiped the floor with djgpp (the dos port of gcc). It lost in every test made by every person ever, unless you count apps that were mostly asm anyway. Watcom's optimization was legendary. Most DOS games used Watcom. Remember the famous "DOS4GW" ? That was the default dos extender used by Watcom programs. That too helped, because it was faster than the dpmi used by djgpp.

graspee

Re:Market

[stevey]

The only thing (not a bad thing, mind you) that gcc has going for it is that it's free... it's hardly the compiler of choice if you really want to optimize your code.

IMHO the best thing that GCC has got going for it is not is price, it is the sheer portability. There exists versions of GCC that will run on anything from Solaris, Linux, BeOS, Windows, and the Macintosh.

Thats why other free compilers, such as LCC have no appeal for me - sure they're smaller, neater, and easier to understand, but they're just not portable enough.

I think that part of the reason that you can't optimize your code so much is because of the nature of the portable code generator(s) - they're written to generate code for multiple machines, and as such they don't really want to start taking advantage of features particular to one processor type, (or even on particular model of processor).

I guess something else going for GCC is the amount of existing code around that takes advantages of its "special" features...

Re:Market

[Chainsaw]

Or do you mean "gcc is not known to compile invalid standard non-compliant code well".

No, he probably means that gcc is a good C compiler but the C++ support is bad, at best. I have to say one good thing about gcc: it is probably the most standard-compliant C++ compiler available. Too bad that it generates very slow code.

Re:Market

[flipper9]

Borland has always released a C++ version soon after the release of their Pascal compiler/IDE product under Windows (first they release Delphi and then C++ Builder), so it shouldn't be any different under Linux.

What's neat is that Borland uses the same compiler for both their Pascal and C++ IDEs, so there is not a lot of reworking on their end and you can use both C++ and Pascal in the same project.

Re:Market

[-douggy]

This is even more true with big numberical simulations (that you are too lazy to port to fortan) The new Intel Compiler running on a P4 really makes the code fly even compared to MS VS6.0 (student edition)

gcc is nice in that I can run a copy to develop for my pc or use it to cross compile demos for my Dreamcast but a good IDE for C++ in linux is always going to be a good thing for those who want such a thing. There is a time and a place for two terminals and a text editor and the same is true for IDEs

Re:Market

[stevey]

I wasn't aware of that - I remember using it a long time ago, when it wasn't that portable.

I have the book "LCC a retargetable C compiler .. but I don't remember it being as prolific as GCC.

Re:Market

[SpinyNorman]

I don't think having multiple back ends is really a hindrence to implementing code optimization - many/most optimizations are abstract, and then you have architecture specific peep hole optimizations after that, that can be considered part of the particular back end.

I agree about the amount of gcc specific code out there - given that most free software uses gcc, "GNU C" has become more portable than ANSI C. As someone else pointed out the Linux kernel's impressive portability comes from the portability of the gcc compiler.

Re:Market

[SpinyNorman]

I havn't seen it available for free (it's $495 if memory serves correctly) - maybe you're talking about a free Beta version?

It's certainly worth the money given it's performance (I've paid more for compilers in the past, such as $995 for WATCOM's 32 bit DOS C compiler!), but compared to gcc's price it's hard to justify for hobby use.

Re:Market

[gewalker]

On windows, if you want VB-like development, but don't like the junk that comes with it, most developers that work with both find the Borland solution superior.

In addition to the inherent technical advantages,
one of the big reasons to use CPPB/Delphi is the active 3rd-party components environment -- nearly all available with (if not including) source code. This is a huge plus and any open source fan would agree (Even if the source is licensed)

Biggest problem in moving the apps written in Delphi today to Linux is the lack of those same components under Kylix. Many real world Delphi apps (and presumably CPPB soon) will be slow to port because of these dependencies.

Components are the biggest porting challenge (VCL to CLX) and is at least partly why the components are slow in coming. Changes in the GUI and messaging models are the biggest factors.

Of course, many existing Delphi/CPPB apps make calls to native Windows API, but usually there is either a simple equivalent under Linux, or the equivalent does not make sense under Linux. So, these don't impede the Linux port as much.

Delphi 6 does have CLX, so a Delphi 6 CLX app and a Kylix app are very compatible.

Delphi and CPPB builder are not about the compiler. It's the environment, visual programming, and component based programming that appeals to the developer.

Re:Market

[Hater's Leaving, The]

C is C, use gcc to compile C.
C++ is C++, use g++ to compile C++.
C++ is not C. C is not C++.
There exists a non-empty intersection of C and C++. That intersection can be compiled by either gcc or g++.

Geddit?

THL.

Re:Market

[Hater's Leaving, The]

"What's neat is that Borland uses the same compiler for both their Pascal and C++ IDEs, so there is not a lot of reworking on their end and you can use both C++ and Pascal in the same project."

Am I the only one who had trouble understanding that?

THL.

Re:Market

[spitzak]

Oh come on, I like gcc and all that, but it is pretty obvious that MSVC makes faster code when you turn on all the optimizations. The difference is totally obvious to an end user and does not require any timing tests. (This is with code using the fltk toolkit and huge amounts of floating point math, and not using any MFC or other MicroSoft libraries other than the ones that gcc can compile with).

(it is also true that the MSVC optimizer is full of bugs, though, forcing us to select which files we need to optimize. But this pain is easily worth it to get the speed increase).

Re:Market

[yugami]

considering c1 doesn't compile c++ compliant code half the time, especially dealing w/ nice features like template's and of course the ever stupid for loop scope bug,

who cares how fast it is since you have to change your code.

Re:Market

[ztwilight]

There would have been a huge market for Borland, if they weren't so late to the game. Call me biased, but I would much prefer CodeWarrior to Borland, even if the debugger wouldn't work when I first installed it. However there's still a huge chance that Borland will become a very popular IDE. Oh yeah, and I can give you 240 REASONS why I would use Borland instead of GCC if it were left to only that. Because those are the extra seconds I have to wait to compile a mid-sized project if I choose to use GCC. Yes, 4 and a half minutes instead of 30 seconds! And I'm never going to forget some of the 2 hour kernel compiles from back in the day with GCC, mind you, I'm no expert at compiling the kernel.

Re:Market

[tomstdenis]

"Most DOS games used Watcom"

That's because Watcom was stable before DJGPP was.

Note how Allegro and Quake [two prominent program items] started in DJGPP not in Watcom.

"faster than the dpmi used by djgpp. "

While DJGPP came with a DPMI server [cwsdpmi] you were free to use anything else [including PMODE/W and WDOSX].

The only reason Watcom was more popular than DJGPP initially [i'd argue it isn't anymore] is that it came out first.

Re:Market

[1g$man]

A more wordy (but probably still not gramatically perfect translation might go like:

What's neat is that both Borland's Pascal and C++ IDE use the same compiler back-end. Because of this, porting changes between the two products doesn't require a lot of work on Borland's part. An added benifit is the ability to use both C++ and Pascal source files in the same project.

IIRC, you can only mix C++/pascal code with the C++Builder compiler, not Delphi.

Re:Market

[1g$man]

and i spelled "benefit" wrong... *sigh*

Re:Market

[johnnyb]

The only thing (not a bad thing, mind you) that gcc has going for it is that it's free... it's hardly the compiler of choice if you really want to optimize your code.

**********

Also going for gcc are a) retargetability to a wide variety of platforms, and b) It's fairly easy to write new compiled languages which use gcc's backend. In fact, the objective-C frontend for gcc came about specifically because it was more expedient to write a free compiler on top of gcc than to write a paid-for one on its own.

Re:Market

[ComputerSlicer23]

If that is true, I will happily buy a copy and post it the patches for anybody to apply to GCC. If RedHat doesn't give me the source, I'll inform RMS personally I am sure he will want to know. It is based on a GPL'ed project, I am pretty sure your spreading non-sense here. Any facts to back this up with? They can't keep the code from escaping back to the mainline GCC, beings that they are both hardcore opensource companies I am calling bullshit on the parent post.

Re:Market

[Graspee_Leemoor]

"Note how Allegro and Quake [two prominent program items] started in DJGPP not in Watcom."

You can discount Allegro (fine library, btw) because it was open source; It only made sense to target an open source compiler- other Mr. Hargreaves would have been severely limiting the number of people who could use Allegro.

Off off topic- but I have a feeling that early versions of Allegro could be compiled with Watcom anyway- obviously the syntax of the asm bits would have to have been converted...

And I'm not sure that the reason Watcom was more popular then (note I'm in no way saying it is *now*- we're talking years ago here) was because it was more stable- if anything it was more suspicious because there were always bugs being discovered in code compiled with max opts turned on!

No, the real reason was that Watcom produced faster code. My primitive, (and extremely crappy by today's standards) 3d engine always ran about 10fps faster when compiled with Watcom. I remember using the textures out of quake in my demo scene, so we're talking quite late on in the DOS scene here too...

The two alternative dpmi extendors you mention for djgpp were fairly late arrivals on the scene too, and never really got to the same level of performance as dos4gw until very late.

Sorry if you think I'm being petty replying to your reply- I'm not- it's just that I enjoy discussing these good old days- Watcom was simply the God of compilers at that time- a true classic. Sure, they went all crappy when they got taken over, and they never really got the hang of the Windows compiler market, but man, in DOS they were so damn hot!

graspee, remembering with bliss what great days they were...

Re:Market

[puppy0341]

Can you explain why GNUPro produces faster code?
The information on redhat's site states that it's built from GCC and licencesed under GPL.
They seem to provide a stable and supported binary, that's all.
Nothing about producing faster code.

Did i miss something?

Re:Market

[DagSverre]

g++ is a front-end to gcc. It's still the gcc compiler we're talking about.

Compile the kernel?

[jjares]

Isn't the kernel written in pure c? is c++ 100% compatible with c?

Re:Compile the kernel?

[Wheaty18]

The creator(s) of C++ tried to make it as backward-compatable with C as possible. But it's not 100% backwards-compatible.

For example, in C you didn't *have* to prototype your functions; the compiler would make some assumptions (which weren't always right :p). But in C++, you *must* prototype your functions.

That's just one example, I'm sure there are more though...

Re:Compile the kernel?

[tongue]

C++ isn't compatible with C, so to speak, meaning a straight C compiler will not compile your C++ classes. But C++ is by definition a superset of C, meaning that all valid C code will compile cleanly by a C++ compiler. In theory at least.

Re:Compile the kernel?

[Score0, Overrated]

Valid C :

int main()
{
int new=3;
return new;
}

it won't compile with a C++ compiler

Re:Compile the kernel?

[AYEq]

The kernel is not written in pure C, even though it's portability would make you think so. It is writen in C with a ton of GNU extentions. So the kernel is really tied to gcc. (which actually makes it more protable because gcc runs on a ton of machines)

Re:Compile the kernel?

[Anonymous Coward]

The intel compiler uses the source file's extension to decide whether to compile with the C or C++ syntax

C++

D:\IntelC\Compiler50\ia32\bin>icl ft.cpp
Intel(R) C++ Compiler for 32-bit applications, Version 5.0.1 Build 010525Z
Copyright (C) 1985-2001 Intel Corporation. All rights reserved.
ft.cpp
ft.cpp(5): error: expected an identifier
int new=3;
^

ft.cpp(7): error: expected a type specifier
return new;
^

ft.cpp(7): warning #120: return value type does not match the function type
return new;
^

compilation aborted for ft.cpp (code 2)


C

D:\IntelC\Compiler50\ia32\bin>icl ft.c
Intel(R) C++ Compiler for 32-bit applications, Version 5.0.1 Build 010525Z
Copyright (C) 1985-2001 Intel Corporation. All rights reserved.
ft.c
Microsoft (R) Incremental Linker Version 6.00.8447
Copyright (C) Microsoft Corp 1992-1998. All rights reserved.

-out:ft.exe
ft.obj

Re:Compile the kernel?

[grammar fascist]

Any reason this is moderated as a troll? The guy's not making up the output. You can't use new as an identifier in C++, since it's a reserved word.

Re:Compile the kernel?

[igrek]

No, C++ is not 100% C compatible.
However, it's irrelevant here. The Borland C++ compiler properly supports both C and C++ languages. By default, it depends on file extension, but you can also specify the language option (C or C++) explicitly.

Re:Compile the kernel?

[dbremner]

Some code will compile under C and C++, but produce different results.
For example,

#include
int main()
{
int x = 4 //* foo */ 2;
printf("\n%d", x);
return 0;
}

// style comments were added to C99, but you get the idea.

Re:Compile the kernel?

[DarkEdgeX]

I guess the lesson here is not to use reserved words in your C code if you might compile it under C++, hm? Older code should stick out like a sore thumb as well-- the C++ compiler will knock out the specific line and column position that the error was seen at, making it simple work to rename variables that have improper names, or remove improper comments (as another example in this thread showed).

Re:Compile the kernel?

[erikdalen]

ISO C is not compatible with ISO C++.
This page lists some of the incompatibilities.

/Erik

Re:Compile the kernel?

[hatchet]

Linux kernel is written in ANSI C...

Re:Compile the kernel?

[mamba-mamba]

I don't think C++ is a superset of C. It may have been once, I doubt it is now. For example, I am pretty sure that the new C standard (C99) introduced features in C which are absent in C++.

However, I would be surprised if the Borland C++ compiler does not also includes a C compiler.

MM
--

Re:Compile the kernel?

[dbremner]

Good point. A google search on comp.lang.c would probably find a good example.

Re:Compile the kernel?

[Craig Davison]

That's a garbage example. 'a' is equally an int (the casting comes in when you assign a char to it). I can say int x = 'a', for example. Check out what the preprocessor does to your '' constants sometime.

The result of sizeof() is by definition ambiguous, because a constant is only assigned a type when used in an expression. I can't think why you would ever do that - sizeof is only useful with variables and types.

Debugger?

[Score0, Overrated]

I wonder if it will be able to compile the kernel?

And, if so, I wonder if it will have a kernel debugger.

More Mainstreaming of Linux

[$lashdot]

It's nice to see more commercial development packages coming to Linux. I would imagine that every step like this brings more developers to consider Linux developing as less and less it's own world. It's some good news to balance off the Loki happening.

They should open up their Windows products

[qurob]

It's not like anyone BUYS them...

It'd be nice to have a C++ Builder for Linux though.

Re:They should open up their Windows products

[Osty]

Quite a few people still buy Delphi. Delphi users moving to C++ often buy C++ Builder. Otherwise, most everybody else chooses Visual Studio. So, I could see a case made for opening C++ Builder, but not Delphi (which happens to fall under "their Windows products").

I got the impression

[ninth harmonic]

that the writer of the article didn't know what they were writing about. Oh well I cant wait for the Windows version of C++ Builder

IDE Me !

[beanerspace]

What an old fart I am. I remember cutting my teeth back in the 80's on Turbo C 1.0 ... and nearly did back flips when they finally got the Windows IDE right with C++ 5.0 some ten years later.

Let's hope it doesn't take them as long to give us a familiar interface. Sure, the command line would be nice, considering Borland's robust libraries and deep oop capabilities ... but an IDE would be even sweeter.

Re:IDE Me !

[An Ominous Coward]

If you're looking for a familiar interface, I don't think you'll be disappointed. Kylix, Borland's recent Delphi for Linux IDE, is for all purposes identical to Delphi 6. Kylix 1 had some stability problems, but Kylix 2 has been perfect. I'm sure that C++Builder for Linux will be stable and be immediately useful for experienced C++Builder users.

Remember Borland C 1.5...

[Lobsang]

Hmmm...

I wonder if this will be like Borland C V1.5 (or was it 1.0? I'm getting old anyways...):

main()
{
int a = 4 / 8;
printf("%d\n", a);
}

Result: 2

It's not a joke kiddos. It was a real bug, just like that.

Re:Remember Borland C 1.5...

[szomb]

Oh yeah? MS Visual C++, version 6.0, right now, today:


#include

int main()
{
const i = 1.0;
double d = 1.0;

if(i == d)
cout << "yes" << endl;
else
cout << "no" << endl;

return 0;
}


You guessed it. It prints no.

Re:Remember Borland C 1.5...

[Cryptnotic]

I think the point of this is that it shouldn't print anything at all. It shouldn't even compile. When you compile it using g++, you get this message:

test.cpp: In function `int main()':
test.cpp:3: ANSI C++ forbids declaration `i' with no type
test.cpp:3: warning: initialization to `const int' from `double'

The fact that VC++ doesn't catch things that are just plain wrong like this is pathetic.

Cryptnotic

Re:Remember Borland C 1.5...

[Andreas Rueckert]

The debugger was usually better than the compiler. I still have the BC++ Manuals for DOS standing here and remember the funny code it created for 32-bit CPUs. You could check it with the debugger, since it showsd the assembler code just fine.

Re:Remember Borland C 1.5...

[beerman2k]

Yes, well, VC++ never claims to be ANSI compliant. In fact its not. It's just close in most things.

C++ Builder more likely than not

[wrinkledshirt]

If it isn't visual, it won't be long before it is. They'll probably incorporate the same toolkit that Kylix has.

Oh man...

[pb]

Borland has always put out wonderful tools, and really worked hard on making their compilers optimized on their platforms, but I think they've missed the boat here. This is most likely for easy porting of other applications written with Borland tools to Linux, because Linux already has a solid toolchain of its own. Regardless, I hope they get back on track.

What I miss most is the old text-based Borland IDE. That was the most productive development environment ever. RHIDE is close, but wasn't stable on Linux when last I checked.

Re:Oh man...

[robbyjo]

I think that Borland is trying to make its fan "feels like home", creating the "illusion" of being "cross platform". Thus, developers seeking to embrace Linux but reluctant to lose their Windows market can easily be lured in. When Linux gets stronger, Borland already has had a real good head start.

BTW, old text-based Borland IDE can be "simulated" using Twilight scheme.

Re:Oh man...

[ipfwadm]

What, we can't use another compiler (which may or may not be better than gcc) and possibly an IDE?

Us: "We need more companies to release products for Linux!"
Borland: "OK, we'll release our C++ development environment!"
Us: "No, not you, we don't need your product, we've already got that."

Even if you never use it, it helps raise the visibility of the Linux platform when big-name companies like Borland are releasing Linux products.

Re:Oh man...

[pb]

I agree that there's nothing wrong with having more tools out there. However, I sincerely doubt that many Linux users would pay to install a third-party C++ compiler suite when they already have the standard C++ compiler for Unix (g++) free and already bundled with the system. This fact alone significantly narrows Borland's potential audience on Linux.

Factors that might change this:

(1) Borland releases it free of charge or under some open source license; this is a possibility, but isn't clearly stated in the article--perhaps the command-line compiler will be available free of charge, which would encourage many Linux enthusiasts to try it out.

(2) Borland's C++ compiler supports advanced features not in gcc, such as compatibility with Borland's existing C++ compiler for Windows, better support for templates, better optimization, you name it. I already mentioned that this product might be aimed at people porting applications already written for Borland's compiler.

However, the main problem I have with your point is your imaginary conversation; you neglected to date those statements. That first statement was made years ago! At the time, Linux was not very well known and companies were just starting to take notice of it. Borland took a survey and started work on Linux products. Then other companies actually wrote and marketed Linux versions of their products whilst Borland was busy having an identity crisis (remember Inprise?). This also caused Borland to lose credibility with some of their long-time supporters, who likely ported their applications to Linux with some other product (like g++) and forgot about Borland/Inprise. Only now are they waking up and marketing this product again!

Therefore, I sincerely hope that this is a sign that the old Borland is back, and I hope they release a wonderful product, and gain massive support on Linux, and kick the gcc/g++ development crew into high gear to keep up. But understand if I fear the worst, especially from the vague tone of that Infoworld article, where they pretend that Linux doesn't already have a decent C++ compiler. :)

Re:Oh man...

[stesch]

Hmm. The last Borland product I used was Turbo Pascal for CP/M. Running on an Apple IIe.

Re:Oh man...

[kevinank]

Oh, I don't agree. KDE developers might flock to a better C++ compiler. And users of AMD chips might be attracted to better optimization code since gcc basically sucks there (instruction schedulers in recent gcc builds are badly dis-optimized for AMD.) Precompiled headers might be nice too, and the Borland compilers have long had a reputation for compilation speed which itself is useful.

Personally I remember liking the built in debugger and editor simplifying the compile/edit/debug cycle, so that would tend to attract me as well; in fact I was considering getting a box myself, and I rarely even have time to code in C++ any more.

Re:Oh man...

[robinjo]

Why "illusion"? Kylix is very compatible with Delphi.

Re:Oh man...

[HeUnique]

Never going to happend..

Look at this: Intel released a free (for non commercial) release of their compiler (ICC)

Did you see the KDE Developers move from GCC to ICC? I didn't, and I do follow the KDE lists..

What could be is that if borland (probably) will release their command line tools free (as a beer) - then someone might build KDE binaries and will give it out as super optimized. Those will be unofficial binaries of course.

Re:Oh man...

[Chainsaw]

Intel has got a pretty good C compiler that gives quite a speedup compared to gcc. The bad news is that it crashes or generates bad code when using C++ with templates, operator overloading and other features that are clearly non-C. Even the Linux kernel isn't 100% stable with it.

Re:Oh man...

[Self-Important]

First of all, GO PACK!!

Second, I completely concur: that Borland IDE was choice. But nobody's stopping you from *still* using it, even in Windows. Works great, as always. Give it a shot if you still have a copy lying around.

Re:Oh man...

[mark_lybarger]

lots of companies purchase that SUN compiler when gcc works on that platform. i'd guess part of the reasin is that they're using libraries from other companies that are only released as sun binaries. that and corporate people really like support contracts and licenses they pay big $$ for.

Re:Oh man...

[Tassach]

You hit the nail on the head. The ignorance and knee-jerk reactionism here gets pretty damn annoying after a while.

The point that many people seem to be missing as to why this is a good thing is that this will make it easy for Borland C++ users to port their existing code base to Linux. Anything that makes it easier for people to stop using Windows is a Good Thing.

Re:Oh man...

[spitzak]

I can assure you that our company would pay *right now* for a compiler that produces faster Linux code.

It will also have to support stl and other C++ stuff as well as GCC. Because we are also doing Windoze stuff and use MSVC++ we don't use any GCC extenstions. I also would not care if the .o format and libraries was completely different and did not link with gcc output (except glibc) and did not work with the debugger.

IDE's are useless for us, due to the need to develop on Windows/NT as well. It is impossible to manage a set of source files in multiple IDE's. We use gmake (cygwin gmake on NT). I also think it would be nice if the IDE's could be seperated from the compiler. Make up some standard way to feed the compiler source and have it report errors so we can choose the IDE and compiler seperately!

Re:Oh man...

[Arandir]

...because Linux already has a solid toolchain of its own.

I see the monopolists are out in force today. A language standard like ISO Standard C++ is a Good Thing. A single compiler that becomes a standard is a Bad Thing. When there is no room in Unix for an additional compiler, the end is near.

If Borland C++ (the stand-alone compiler) won't be free then there probably won't be many users. It won't be shipped with your Redhat Subscription Service. But it will still have a place, namely with those that think choice is the first attribute of freedom. And if it is free, then expect it to be widely used.

Gcc will finally have competition. It might actualy spur GNU into action to improve their compiler. Most of you guys here are too new to remember the history of gcc. Only a few years ago gcc *sucked* at C++. The unwritten by very official stance policy of GNU was that C++ sucked so don't bother. There was little standards conformance. But someone in the GNU crowd did have a clue, and forked the compiler. Before you knew it, egcs was being used more than gcc. Eventually the two merged back together, but I hope GNU learned its lesson.

Re:Oh man...

[Billly Gates]

Linux users are not expected to pay for it. Its mainly being sold for coroporate customers which I believe would be willing to pay for it. Look at it from a corporate perspective since that is Borland's main customer base. There are many windows only programs compilied with borlands proprietary libraries. Microsoft has the advatange of VC and their MFC being heavily tied into windows and other Microsoft based products. THis means all of Microsoft apps are being integrated into one big .net application where a user selects each service he or she wants and rents it upon install. This would doom Borland and would costs corporations alot of money to run there own apps.

Borland needs to give corporate customers some more room and a non microsoft alternative. Microsoft made NT popular by selling client/server solutions which integrate directly with all the desktops. It had nothing to do with stability or security of NT. As corporations looked at this as an advatange, unix and novell lost marketshare as NT servers started apearing everywhere.

For Borland to survive they first must try to stop the microsoft only client/server model for a more open client/server model which supports more platforms. Linux is not ready as a corporate desktop yet so borland is making community releases just to wet our appetites so we can recommend them at work for servers. Borland's role in support for Solaris and Linux with jbuilder and I believe kylix(not to sure if there is a solaris port) and the upcomming cbuilder will help brake the microsoft stranglhold. This comes just as corporations are doubling there pay to microsoft to stay current with Microsoft. After this, borland will probably make client/server libraries available if they haven't already with Cbuilder which will not require NT/w2k as the backend. Corporations do not like the idea of paying microsoft every 18 months or else.... just to run there vb written client/server apps. Linux and Solaris may be a way out, if there is a real cross platform development tool to lure bussinesses.

Oh ya, which tool do you think third party applications are written in? Borland, and VC is the anwser. WIth a linux port of cbuilder, expect more windows apps being ported to linux. We can all thank Borland after this happens for making this a reality. I also hope they support the mac. Remember that apple has at least a %5 share of the client market. If they do this then we can have apps written for mac, linux and Windows. This alone would make more corporate developers more weary of VC and more likely to use Cbuilder.

Re:Oh man...

[nusuth]

I can assure you that our company would pay *right now* for a compiler that produces faster Linux code.

Use Intel compiler then, it produces amazingly fast code for P4 and very fast code for everthing else. Borland's compilers never produce very fast code, they are OK but not "fast" except when compared to gcc. The main appeal of borland tools is the whole RAD environment, its integration, ease of use, very high productivity in gui design etc. Compiler's positive contribution to this is its compilation speed.

Re:Oh man...

[Essef]

I would gladly fork out some money for a C++ Builder equivalent under Linux. I already own the Windows Version. C++ Builder is IMHO the ideal tool for developing GUI apps.

my two cent's worth :)

This Is Very Good!

[Lethyos]

Borland's IDEs (baring of licensing crap ;-) have always been exceptional. Current opensource IDEs are decent, but they are no where near the quality.

It comes down to maturity. Borland has been making powerful IDEs for a very long time. Development for opensource IDEs however is a fairly new thing (KDEvelop is good, but it is still fairly unreliable and not as featured as I'd like).

After years of tweaking, Borland's got it down pat.

Re:This Is Very Good!

[redcliffe]

The problem I find with KDevelop is the lack of integration of GUI design and coding. This is a staple in all modern IDE's, and I'd like to see it in KDevelop. An interface like Borland's would be great for KDevelop.

Competition is good

[prisoner-of-enigma]

If nothing else, it'll be nice to have an industrial strength competitor to GCC coming from a (former) heavyweight in the development community. I remember Pascal oh so fondly...

And I'll be real interested to see if it will actually compile the kernel!

Re:Competition is good

[dvdeug]

And I'll be real interested to see if it will actually compile the kernel!

It would surprise the heck out of me if it does. A compiler practically has to be bug-compatible with the version of GCC Linux used to build his kernel to build it correctly. The kernel uses a lot of gcc-specific code, including undocumented properties.

The next version of Kylix will probably have C++..

[frleong]

Kylix was supposed to be compiler independent. The current generations have only the Object Pascal compiler. IIRC, the next version of Kylix will support C++ too.

IDE from Turbo C++

[Animixer]

I hope that it comes with the same IDE as Turbo C++ 3.0....I've always liked the text-mode interface. :)

Cross-platform...

[Anonymous Coward]

Does this mean that code will be cross-platform if developed in a Builder environment?

You don't say...

[Anonymous Coward]

void main(void) {
printf("Big deal\n");
}

Re:You don't say...

[Anonymous Coward]

ahem... since we are discussing C++

int main(void)
{
printf("Big deal\n");
}

Re:You don't say...

[Anonymous Coward]

I guess this isn't my day...

int main(void)
{
printf("Big deal\n");
return 0;
}

Re:You don't say...

[mlk]

I thought C++ was that dam crappy

out "Big Deal\n";?

Re:You don't say...

[mlk]

or my day either
cout << "Big Deal\n";

Re:You don't say...

[Mulletroll]

cout "Big Deal" endl;

Re:You don't say...

[Mulletroll]

Jesus fucking christ. I guess "plain old text" doesn't mean plain old text. Nevermind.

Re:You don't say...

[mamba-mamba]

void main (void) is not valid in C. You seem to imply that it is invalid in C++ only.

Also note that you should included stdio and return a value. I.e.,

#include
int main (void)
{
printf("Big deal\n");
return 0;
}

MM
--

Re:You don't say...

[DagSverre]

The C++ standard specifically allows for not returning anything at the end of the main function, in which case it will default to "return 0". Also, who's using stdio in C++? Requiring it would be bad as everyone's using iostream these days (and I think the standard mentioning that you should include stdio is extremely unlikely though I don't know as a fact that it doesn't).

open edition? I think not.

[MathJMendl]

I wonder if there will be an Open Edition?

Yes, that would indeed be cool. Unfortunately, it's not gonna happen. If they open source it then people could recompile it for windows, and boom, there goes their whole suite for windows down the drain! But still, with GCC already out there, does it matter?

Re:open edition? I think not.

[MathJMendl]

Hmm. I didn't realize they already had open versions. I guess this makes sense from their point of views. I retract my previous statement. :P

great!

[Jucius Maximus]

Then all we need is Textpad for Linux and then all well be well in the land of CS coding...

Re:great!

[Graspee_Leemoor]

Ultraedit is written by some weird-ass Christian type who thanks god humbly for his ability to program and charge other people for it.

His editor didn't even let you search for regexps until I asked him why it didn't and he said "name me one that does and is as cheap as ultraedit!" and I said "Emacs, vi ... "

Ultraedit is also one of the most popular programs amongst cracking types, mainly because they aren't hard-enough men to handle an editor in a console window- e.g. vim for win32.

graspee

XEmacs vs Borland C++

[Mongoose]

Hhhhmmm... let's see I can use XEmacs with code generation, source templates, tags, class browser or I can use Borland's and not be able to use elisp.

I'm sure MSVC++ kiddies new to unix development in general can enjoy it however. I just hope I don't see 'project files' all over the damn place a year from now in lieu of Makefiles and autoconf.

Re:XEmacs vs Borland C++

[Mulletroll]

On the other hand, I've been using emacs for about 3 years, and I've never heard of code generation, source templates, tags, class browser, and havn't had time to learn elisp beyond the few lines I put in my .emacs file.

I enjoy emacs, the features are great, but just don't ask me how to use them. In modern IDEs, the features are right there in front of you to actually use.

Re:XEmacs vs Borland C++

[cscx]

Ok moderators, you know it's true... taking away my precious karma won't make it any farther from the truth. You know that people who want an unbaised view of the topic read at a low threshold anyway.

Re:XEmacs vs Borland C++

[mvw]

Wasn't Emacs originally written by James Gosling (of Java Fame)?

Re:XEmacs vs Borland C++

[betis70]

>>Wasn't Emacs originally written by James Gosling (of Java Fame)?

No, by James W Cooper (of VB fame).

I want this...

[Bill Henning]

Don't get me wrong; GCC is greak - but C++ Builder is an impressive way to write GUI applications quickly.

I'd immediately recompile MemTach for Linux :-)

Best Regards,

Bill

Re:I want this...

[sql*kitten]

Don't get me wrong; GCC is greak - but C++ Builder is an impressive way to write GUI applications quickly.

Great? Weak? Greek? What?! ;0)

Resume Item

[kenneth_martens]

It's not clear (at least from the sketchy information in the article) if there will be an Open Source/free version, but I hope so, and here's why: currently my university requires us to use Windows in our computer science classes, mainly because Microsoft gives us a lot of expensive software for free (if Microsoft makes it, students probably have access to it--Visual Studio 6, Visual SourceSafe, SQL Server, Windows XP Pro, ...)

That leaves people like me--who prefer to run Linux instead of Windows--at a disadvantage. I have to have a dual boot system, and I have to reboot to Windows every time I need to hack out some code for a class. Now, if Borland releases their C++ for Linux and makes it free, I know I could convince a couple of my professors to ditch the Microsoft stuff and use teach the class using Linux and Borland. That would enable me--and the rest of the university--to gain some practical experience coding on the Linux platform, and not just on Windows. Don't get me wrong, there isn't anything wrong with knowing how to code using Windows and Microsoft Visual Studio (in fact it's probably a good resume item), but I'd like to get familiar with some alternatives before I enter the workforce.

Re:Resume Item

[Cuthalion]

You know, there's nothing stopping you from using GCC and whatever editor you like (RHIDE, emacs, whatever) under linux. If you absolutely must make it work under windows too, all the better - now you're learning how to write portable code. Even neverminding that, using GCC is much more practical linux experience than using Borland C++ will ever be, in that most everyone doing linux development will not drop everything and migrate to Borland's tools immediately.

Re:Resume Item

[Pinball Wizard]

Based on Borland's past offerings, I'd say there will probably be a free C++ compiler(command line) but not C++ builder.

I'm currently learning Java using their JBuilder6 Personal Edition, (a very good Java IDE, I might add). It's available for free on both Windows and Linux. Its got everything you need to learn Java, just not the enterprise stuff. So its possible they might do the same with C++ Builder. Its just in the past(for Windows) they made the base compiler free and charged you for the IDE.

Re:Resume Item

[Peyna]

I'd hope so. Especially considering that the entire .NET SDK is free for download for all from MS, it would be a good idea for others to follow suit. If you ask me, vi/notepad makes a great IDE when you can get the rest for free =]

Re:Resume Item

[Octorian]

Hmm... At my school, all the lower-level CS courses seem to be stuck in MS-land. (well, the undergrad TA's are all MS freaks, even though the profs/grad-TAs aren't) So, back when I took those, I would code my project in Linux, then make it "VC++ Compatable and Tested" before turning it in.

Now that I'm in upper-level courses, I get to see more of what the CS department here really supports (Solaris, and some FreeBSD). So, since most of my code has to work in Solaris, I just use the Ultra 30 I got myself this summer :) (much nicer than ssh'ing into something else for testing and de-quirking) Oh, and in the rare instances that I have to use Windows... (like one class where we had to write Win32 and UNIX versions of system programs) Well, that's what the SunPCi card is for! (and I can minimize it when I want to, and it doesn't eat system resources) That Sun is really nice, as now I'm taking a class where I need to do graphical stuff on Solaris, and it's nicer to work from my dorm room, than to either go to the lab or do cross-campus remote-X.

Re:Resume Item

[Ronin Developer]

Word has it from the 2000 Borland Developer's Conference, that Kylix was going to be Delphi on Linux. Next in the queue was going to be C++Builder on Linux. During that same conference, the core C++ compiler had been built and demonstrated compiling and running natively on Linux. That was two years ago.

Since that time, we have seen Kylix and Kylix 2 released as well as Delphi 6 and a new C++Builder. The fundamental piece was the use of CLX to make the code cross platform between Windows and Linux. VCL code simply will not port. And, the VCL never worked cleanly between Delphi and Kylix. CLX was the answer to this.

I expect that the upcoming release will be C++Builder on Linux. That means it will have the same look and feel as C++ Builder on Windows. Yes...a GUI oriented development tool with all the wizbang designers that many of us have come to love.

I'd also like to point out that Borland made it clear that they were not out to replace GCC. Their aim (well, 2 years ago), was to make a development tool that enabled developers with a Windows background migrate to Linux and bring their application development skills to the Linux platform. There was alot of talk about whether the libraries would be link compatible with GCC. They didn't have an answer that most of us wanted to hear. But, like I said, that was two years ago.

Borland is position themselves to enable developers to enter a new market. While the MS folks are concerned about pushing .NET out the door, Borland developers will already be developing Web services and database applications for both Linux and Windows and beating their competitors to the punch.

Will these tools be an immediate big hit with Linux users? Doubt it. It takes a lot to effect a paradime shift of that magnitude. But, you will see Windows developers porting their code to Linux and opening up new markets. Eventually, the die hard people will see the advantages of using a tool like C++Builder or Kylix in a corporate setting. And, if we are really lucky, we'll see Borland making their .so's compatible with rest of the Linux community. The whole point of doing a C++ version is simply that the majority of Linux developers use that language. But, they ported Delphi over first (ala Kylix) because the majority of their customers are Delphi users.

Guess we'll all have to wait and see, eh?

RD

Re:Resume Item

[doubtless]

I don't think the school gets Microsoft software for free. My college used to use gcc but recently they signed a contract with Microsoft to get all their software. They have to basically pay a lumpsum and the softwares they get will appear to be 'free'.

However, I don't think that covers all of Microsoft softwares either. For example, if the school buys a new server, they still have to pay for the Windows XP Datacenter or whatever.

The story is, it's not free, it just appears to be.
--------

Re:Resume Item

[Kupek]

You don't have the choice? For programs that were auto-graded, it was always on a Windows machine running VC++ 6.0, but once we got into programs that were demoed, we always had the Linux option. (I wasn't very initiated in the world of Unix at that point, and I gotta admit VC++ is a damned fine product.)

Anyway, last semester in operating systems, we had to use Linux. If you haven't gotten into higher classes, you might find those give you the option of developing under some flavor of Unix, if it's not required.

Re:Resume Item

[psyclone]

I agree with another comment in this thread.
that's what #ifndef/ifdef...#endif are for. You simply add windows and linux specific code in those locations. It will take more than a few minutes to work compatibility issues out -- you may want to think about running VMware (or something similar so you don't have to spend time rebooting; it's probably even cheaper than some of your textbooks).

In the end, you'll have written portable code and it will be a much greater accomplishment on a resume.

Re:Resume Item

[Tim C]

Don't forget that as well as Kylix, JBuilder has been available for Linux since at least version 3. (Quitepossibly before that, too, but I came to it just as 3.5 was being released.)

True, it's almost 100% Java, so there wouldn't have been that much work invovled, but it's still very welcome - it meant that I (and about half my colleagues) could finally ditch Windows entirely at work.

Cheers,

Tim

Re:Resume Item

[ahde]

In my OOP class we started out writing an interface to Excel, moved on to animated MS Agents, Windows RPC, and GDI screensavers for fun. Which pretty much limits you to MSVC. Some projects wouldn't compile on the 5.0 I had bought the year before.

While I appreciated the instructor trying to make it more "interesting", he was a couple years behind in what was cool. Several of us had started experimenting with Linux (1998) and once you've seen the flexibility of gcc, it's tough to go back.

Not to mention how painful it is to unlearn the incompatibilities Microsoft teaches you.

Re:Resume Item

[Tim C]

Indeed, and some of my work colleagues use emacs or vim + either makefiles or ant to develop in Java, despite having plenty of JBuilder 4 Professional licences. It's what they're used to, and to be honest, JBuilder is a bit of a cow, performance wise (we only have 450MHz P3s, but that's a moan for another thread)

I had never really used an IDE until I was persuaded to give JBuilder a go (other than relatively brief uses of Visual Interdev and VC++), and it's changed the way I work. If nothing else, being able to step through the code in the debugger, inspecting any variables I choose is wonderful, and a vast improvement on peppering the code with System.out.println()s.

Of course, as I develop code in Java on a machine running Linux that's going to be deployed on a server running Linux, I don't have to worry about cross-platform issues. However, I can't imagine that there's anything stopping you from developing fully portable code using an IDE, either. You'd just have to know what you were doing, which is true in any case.

Cheers,

Tim

Re:Resume Item

[Tet]

If nothing else, being able to step through the code in the debugger, inspecting any variables I choose is wonderful

...and of course, it has nothing whatsoever to do with using an IDE. Debuggers have been around as standalone products for a very long time in the Unix world. It's only the lack of decent standalone tools under DOS/Windows that has driven people to use the all in one approach that Borland and MS are selling.

Re:Resume Item

[Tim C]

True enough, but when you can fire up the debugger just by clicking that icon there, next to the one that you click to run the code normally, there's more of an incentive to start playing with it.

If you have to go hunting around on freshmeat or google to find one in the fist place, or wade through dozens of pages of info/man pages to learn an obscure command line interface, printf()s or System.out.println()s start looking attractive. This is especially true when the bug has to be fixed now, so the client will stop shouting at the project manager, who in turn will stop breathing down your neck.

Cheers,

Tim

Re:Resume Item

[Troed]

So, did you protest then? Most of our assignments were at a level high enough for us to _choose_ what languages and environments to do them in. This was a few years ago, but I would _not_ take up classes today that were so MS specific. I'm a software engineer, working as a consultant on the Symbian OS platform - I use Devstudio (as an IDE only) but that's all the MS I see on this computer ..

Re:Resume Item

[dollargonzo]

i dont think you are going to convince the profs to teach the class in l00nix and borland. and btw, how does the release of borland make l00nix a better platform for the profs to teach and develop on.

instead, write good code, and only reboot to compile, not to write and test.

QED

Re:Resume Item

[Tet]

If you have to go hunting around on freshmeat or google to find one in the fist place, or wade through dozens of pages of info/man pages to learn an obscure command line interface, printf()s or System.out.println()s start looking attractive.

Agreed, if that was the case. Fortunately, it isn't. DDD ships with most Linux distributions, and gives you the nice GUI interface you're used to, plus some extra goodies on top (the ability to visually see the state of data structures like linked lists or binary trees is an amazing debugging tool). And since this is Unix, naturally there are other choices if you don't like DDD: Code Crusader, mxdb, mxgdb, xxgdb etc.

Re:Resume Item

[KwamiMatrix]

Well, I am a freshman Computer Engineering Student. The CS dept at my University, Rochester Institute of Technology, is all Solaris 8. Since CS is a major part of my computer engineering curriculum, I am gaining a lot of experience with Solaris 8 & Sun workstations. I must say that it is nice. We have many powerful Sun Enterprise 450 servers, and using X Window is pretty fast from my laptop in my dorm running Windows XP. My computer engineering dept. is heavy on HP-UX, but I have not had much experience at all on them. We also have other Unix systems. It is just mostly our IT dept and our Software Engineering dept that code under windows.

Re:Resume Item

[ahde]

So, did you protest, then?
<br><br>
No, we complained. It's much more effective, and doesn't take as much of your time. I wouldn't take the class again today, either, but then it was much more excusable for the instructor to be uninformed. Besides, working with the windows APIs helped me appreciated GTK+.
<br><br>
We knew what Microsoft was doing even then, and we may have opened the instructor's eyes a bit, although his kneejerk reaction was that we were just paranoid. Just to show how backwater my school was, I took a UNIX class at the same time, and the instructor talked the necessity of learning vi and sh because one day, someone might get stuck working on a legacy system.

At the time, I was doing my homework on the server for a local ISP (my dad owned it.) Just log in at the console as root. I was glad I didn't have to copy the stupid practice logs (and poems!) from the text book. I just used awk, etc. on the syslogs.

Re:Resume Item

[mazor]

And, if we are really lucky, we'll see Borland making their .so's compatible with rest of the Linux community.

Borland's .so's are already compatible with the rest of the Linux system. Kylix produces and uses Linux standard ELF .so library files. The Kylix RTL relies heavily on glibc (libc.so) and libpthread.so. Why would a C++ from Borland be any different?

-mazor

Re:who cares??

[Wheaty18]

Heh, Visual C++ is not ANSI compliant.

Borland C++ or Borland C++ Builder

[fwankypoo]

The blurb (linked to here mentions that Borland is going to announce C++Builder for Linux. Just a tasty little tidbit that needs to be adressed:P

Re:Borland C++ or Borland C++ Builder

[Wizy]

correct, this is going to be C++ Builder, the full IDE. Same as they made Delphi for linux (Kylix).

Re:Borland C++ or Borland C++ Builder

[blitzrage]

"We are taking our C++ development solution to the Linux platform. We have seen a lot of Linux developers who used to be Unix developers," said Alison Deane, a senior director of product marketing at Borland, in Scotts Valley, Calif.

She added that Borland plans to announce C++Builder for Windows next month, but declined to provide details.

No they didn't. They just said they were going to announce C++ for Linux, and C++ Builder for Windows.

Re:Borland C++ or Borland C++ Builder

[fwankypoo]

I stand corrected. Damned knee-jerk reaction.

Hey, hubris, get back in your box!

Re:Borland C++ or Borland C++ Builder

[nusuth]

Well, you are correct, but if the architecture of the platform has not changed drastically from earl builder days, it takes a certain amount of dedication to port the compiler but not the whole builder. The builder envorinment is exactly same as delphi environment except for parser and object inspector (plus a few extra tools for interoperability with other C++ environments, but that is irrelevant.) The builder parser is able to parse both C++ and object pascal, while the delphi parser can only OP. The intermediate code generator is same, the code generator is same, the linker, debugger etc. tools are identical, the code of libraries are identical except for headers etc. (and this too is non-vital, builder can use delphi units just fine, without C++ headers.) The object inspector has to be aware of C++ bindings, but that too is trivial once you have C++ parser. While normally it is rather hard to produce C++ Builder like IDE with onlt a C++ compiler in hand, with Kylix working, producing C++ Builder just takes a new parser and some fiddling. Since a C++ compiler has to have a C++ parser, borland can release a single compiler only if they specifically want not to release Builder.

Ever heard of a const int?

[Plasmic]

Const defaults to int. You're doing a direct comparison of a floating point with an integer.

I have a PhD in Visual C++. This is widely known among my graduate students.

Re:Ever heard of a const int?

[seann]

PhD in Visual C++?
What, did you get that out of a cracker jack box?

Re:Ever heard of a const int?

[spectral]

These are actually serious questions, I don't mean any disrespect or anything. 1) Is it a PhD in Visual C++, or in windows programming, or cs, or whatever? 2) Aren't the Win32 VXD's deprecated in NT? I know I had something that was made for VxD only, and it thus didn't (And never will be) workable under NT4/2k/xp.. or am I thinking of something different?

Re:Ever heard of a const int?

[WasterDave]

Const defaults to int. You're doing a direct comparison of a floating point with an integer.

It's a shitty piece of coding because it's non-intuitive and a bastard to maintain. It should never pass a code review.

I have a PhD in Visual C++.

No you don't. You just failed to pass a BS review.

Dave

Oh, by the way...I have another certification,too.

[Plasmic]

Also, I'm an MCSE.

Re:The next version of Kylix will probably have C+

[mz001b]

intel has already release C and Fortran 90 compilers for Linux that are free for non-comercial use. These are very fast compilers when used on a Pentium IV.

Re:More compilers for Linux is a GREAT thing.

[Octorian]

Mod this up!

Myself, being a bit of a platform freak, I see this all the time.

GCC is way too lenient. Heck, I've even found cases where MSVC++ catches more things.

I've also used the Sun Workshop compiler, IBM's xlc, and SGI's MIPSpro.

IBM xlc has this neat thing where error messages not only indicate line number, but also where exactly in the line it found the error.

BTW, MIPSpro is my favorate so far. It's error messages are actually quite informative. Also, if you happen to be on an SMP box, add the "-apo" compiler flag and it auto-parallelizes your program. Has anyone else seen a compiler that can make your plain single-threaded app take advantage of multiple CPUs?

Wonderful

[rice_burners_suck]

I applaud Borland for choosing to put more of their fine products on Linux. I have personally used Borland's products since version 3 of their Pascal compiler, which was a pretty long time ago. In conjunction with TurboPower's libraries, which were distributed with complete source code and no royalties, Borland's compilers, both for Pascal and C/C++, were always truly amazing products.

Now, with the increasing popularity and acceptance of Linux, I believe that Borland's products have found a new home, better than on DOS and Windows. I strongly believe that if Borland continues to implement their fine software on Linux, some great applications, brand-name commercial as well as free, will show up on Linux, making it a strong and competitive alternative to the Windows family of operating systems.

Perhaps someday, a couple of years down the road, Microsoft will begin implementing their software, such as a Microsoft Office for Linux package, just as some years ago, IBM sold native Windows versions of their OS/2 applications. Hopefully, this move by Borland will bring that a bit closer to reality.

Re:Wonderful

[ZxCv]

Pleeeease. I have been using Borland products for about 6 or 7 years now and I would definitely chime in that they are wonderful tools. But I'm also not so disillusioned to think that the reason there haven't been tons of great commercial applications on Linux is because there hasn't been a development environment good enough to create them. It's simply because there isn't enough of a market for such products. And while having a tool that makes creating applications easier is always a good thing (look what VB did for Windows), Linux needs a lot more than a better development environment for it to better compete against Windows.

Re:Wonderful

[whereiswaldo]

I don't think you're seeing the big picture here.

Think of how many C++Builder applications there are on Windows. Add to that how many MS Visual C applications exist (assuming one can fairly easily port to C++Builder - correct me if I'm wrong).

If Borland can make it easy for the developers of these applications to port over to Linux, that's an amazing influx of useful, polished software we can use.

Also, I think you might be discounting the positive effect a great IDE for Linux would have on the community. It is not easy to start coding efficiently using gcc, libglade, gtk, and what have you. Not to mention that most debuggers for Linux are not as full-featured as Borland's offerings.

C++Builder provides many, many IDE features like command completion, online help, wizards, database connectivity, CORBA, Apache support, etc.. etc.. which is all at your fingertips and well documented. I don't see how it couldn't give Linux development a good shot in the arm!

One negative point, though, is that for now this will only be for the Intel platform. I'm sure someday that will change. The core of the system, though, must always use open source software. On that note, I wouldn't be surprised to eventually see open source IDEs or conversion programs which make it possible to compile C++Builder for Linux projects using gcc and compatible toolkits.

That sucks.

[nemesisj]

Lots of my classes ask for submissions written on a particular platform (linux, windows, heck even irix sometimes). My goal has always been to write my assignments in a different platform. It's helped me learn a lot about cross platform coding and lets me experience different operating systems.

This is great news!

[red_gnom]

Maybe they have a reason to believe, that their compiler will produce better optimized binaries.

I would love to see C++Builder for Linux though. It is incredibly good product. I use it for Windows, and love it.

Yes! This is great news for Linux comunity.

Re:This is great news!

[fredrik70]

You're taking about the DOS inteface up to BC3.1? That was wicked, my first IDE... HOwever, using Delphi now for a project and it got a editor from hell (rest of Delphi OK though). Can't make it indent properly! home key takes cursor to the first pos of row rather than b4 first char on row, etc, etc. Sorry, just have got very used to VC++ editor.
Would be wonderful to see borlands c++ compiler on Linux though, with or without IDE. They got one of the best and most conforming c++ compiler around today... I definetly would pay for a licence!

Re:This is great news!

[mazor]

If you don't like the editor key bindings/behaviors in Delphi or C++Builder, write your own editor key binding module for the IDE. The source code for the emacs editor keybinding module is included in the products as demo + documentation.

-mazor

Re:This is great news!

[fredrik70]

Ah, that could be handy to know. Thanks!

Really stupid question...

[kikta]

I've done hardly any programming for Linux yet. Mostly I've just used a text editor. If Borland doesn't release a GUI for C++, what's another good one? I thought gcc was command-line only & didn't have an IDE. Is there something else? I always assumed there was but have never got around to checking into it. Thanks.

Re:Really stupid question...

[sweetooth]

kdevelop
KDE Studio Gold
or just search through
apps.kde.com

Re:Really stupid question...

[Anonymous DWord]

How about Scintilla and SciTE?

Re:Really stupid question...

[gvr]

You can also try the Codewarrior IDE.

It supports a larger subset of c++ than gcc. *Ouch*

Re:Really stupid question...

[damiam]

Anjuta is a nice IDE for GNOME.

Borland C++...

[BuffJoe]

... will probably not be able to compile the kernel. The Linux kernel is notorious for using GNU extensions to the C language.

Making amends for the license agreement SNAFU?

[jasonbrown]

I think that is a step in the right direction for Borland and I applaud them.

Perhaps after discovering the anger of the programmers over the unreasonable terms of the their misguided Borland Kylix/JBuilder license, they are perhaps trying to make peace with the growing multitude of linux programmers which would be their customers if treated with respect.

I think Borland has a pretty good track record for being a friend to the community. I do appreciate Borland showing their support for linux.

Heve a look at XWPE

[root_42]

And it even has been floating around for YEARS. Look at some screenshots of it here. I think it might be what you are looking for.

Re:Heve a look at XWPE

[micromoog]

That's the worst-looking X application I've ever seen.

How many people actually use Borland C++?

[djohnsto]

I'm sure this will marked as troll, flamebait, lame, whatever, but I'm actually curious.

How many people actually use Borland's C++ products currently? Of those that do, is this just brand loyalty from the Windows 3.1 days? I've been part of product teams (all using C/C++) developing on Windows, Mac, and various unices, and I've never seen Borland being used anywhere. The last Borland product I've seen used was Turbo Pascal for DOS back in high school.

This isn't meant to start a flame war, I've just never actually seen a Borland C++ product being used, and am curious how big their market is. It may be that the Linux version would actually outsell the Windows version due to lack of competition. And it would totally rock if they released the Borland C++ builder IDE that supported not only the Borland compiler, but gcc and icc (intel) as well.

Needless to say, if they do release the IDE, I'll be very interested to see how well it works. I've tried KDevelop, CodeWarrior (older version - 5.0?), Anjuta, a couple other gnome things, etc. And I'm sorry to say none of them allow me to be as productive as I am with VC++ (with the VisualAssist add-in). CodeWarrior was probably the worst (I hope for their sake 6.0 was better), and KDevelop the most mature. However, none of the open source efforts play nice with cross platform projects (damn it, I don't WANT the make files in the same directory as the source!!!), and are terrible when dealing with large projects. If Borland's product can deal with large cross-platform projects, I'll be a happy camper.

Re:How many people actually use Borland C++?

[banky]

>How many people actually use Borland's C++ products currently?
I can't quote actual statistics, but: we have a ton of C++ Builder boxes around the office, and we also make extensive use of JBuilder. The place where I used to work, were pretty much all Borland fanatics, and had their own NNTP server, made up of borland fans. A lot of them were "Team B", the Borland Users Group kinda thing.
>Of those that do, is this just brand loyalty from the Windows 3.1 days?
Kinda. From reading the documentation and talking to them, they tend to say the same things. 1, it was a better compiler for a LONG time when compared to Windows (arguable, of course). 2, it was more command-line friendly, if you didn't want to use the full-on IDE; this mattered because a lot of them were Unix expatriates having to get work in a PC world, and they wanted a PC compiler that acted like cc/gcc. This too is arguable; I'm just reporting, here. 3, many reported better standards compliance, and more functionality in doing things other than Windows (but on x86). For example, one person I knew did embedded x86 development and liked the Borland tools. The general consensus was Microsoft's tools were *Windows* compilers, whereas Borland's were more multifunction. Lastly, they all loved the IDE, considering it more mature and stable than VC++. The majority of them are excited about moving to Linux with BC++ and I expect a small but noticeable increase in Linux acceptance once this comes out. The glare of running cash registers may not light up the skies, but I can think of at least a dozen people off the top of my head who will be willing to spend moderately large money to get their hands on a Borland toolchain for Linux.

Re:How many people actually use Borland C++?

[fwc]

I can't vouch for the current product rev, but I quit using the Borland Compilers back about the time Windows 3.1 was out but most people still used DOS apps.

The reason I quit was because of the buggy compilers. Mind you, this was 10 years ago, but what I do recall was lots of fun with things just plain not working which should have.

I DO recall that one of the most common things to do was to remove/add whitespace or comments. E.G. something like:

a=b;
b=c;

wouldn't compile, but something like

a=b;

b=c;

would. And of course, the next time you wrote code, just the opposite would be true. I also think I remember the IDE getting confused about whether you changed something or not, and not rebuilding the file you just added 100 lines of code to.

I still avoid Borland Compilers. I also avoid Microsoft compilers as they are about as slow as cold tar running uphill mid-winter.

Re:How many people actually use Borland C++?

[MrBandersnatch]

>How many people actually use Borland's C++ products currently? From experience not too many. I used them in both my previous jobs, 1 from preference because I was technical lead and 1 because it was used in the team I was working with. The vast majority of the company used VC++ though.
I've been looking for a job that would allow me to use Builder again since I have 4 years experience with it. I've been looking for over a year which I think says it all! ( And no I'm not unemployed - I'm just not doing what I would prefer to do )

Re:How many people actually use Borland C++?

[_egg]

I used C++ Builder while I was working on game tools under win32 at LucasArts. Builder is amazing! I was a very competent programmer with NO win32 GUI experience. I needed artist-friendly game tools up *yesterday*... I flopped around with Microsoft's crappy GUI code-generating tools for a week and was tearing my hair out. Then I cracked open C++ Builder and had a working first cut of my tools in about a day!

C++ Builder is almost entirely visual... You lay out your GUI, then fill in the C++ bits for events that are important to you. You could have working stuff up in seconds, then revamp the interface several times without changing a line of your C++ code. All of this while still providing you full access to the native API should you wish to do something their GUI code doesn't do normally. Large corporations eat this up... Programmer time is best spent on functionality, not GUIs. Builder enforces Model/View/Controller programming, which IMHO is the way all GUI programming (and web programming) should be done.

Glade and friends on Linux are plainly inspired by Delphi and C++ Builder. I think Glade is great and getting better, but it doesn't compare to the polish and ease of Builder. I now provide command-line Linux tools to PS2 game developers, but would like to provide GUI versions of some. I'd like to do it in Qt so I can also support guys programming under windows, but I just don't have the time to go learn it. The GTK+ port to windows just isn't stable enough to trust, and requires cygwin underneath. But if C++ Builder existed for Linux, I'd be using it daily, as would my entire staff, and not looking back.

This is one case where a commercial development package under Linux is completely worth it because it is *that* far ahead of the open source tools that are available. Let that inspire you to improve Glade and the GTK win32 port.

Borland is making a good move because as their customers look to reduce their dependency on Microsoft, Borland makes the transition easier and tags along with them into the new space. Good luck Borland! You're doing everything right.

Re:How many people actually use Borland C++?

[scherrey]

We use the BC++ tools for commercial Windows development. We also build using g++/cygwin because your code is generally better when it builds under multiple compilers. Now we'll have that option under Linux for the first time which is great.
FWIW - we cannot use VC++ because of their lack of member templates. Our code is simply too reliant on many of the ANSI C++ features that VC++ either doesn't implement or implements wrong.

Re:How many people actually use Borland C++?

[Eythian]

While I haven't seriously used borlands C++ (I used to have a really old version around somewhere), i actually still get paid to program in Turbo Pascal 7. I work in the psychology department of a university, where TP7 is the de-facto standard, as there is a lot of already-existing code out there to work on, and a lot of people in that field who only know a little bit of programming and thats in Pascal. (For those unsure, this is TP7 with the text-based IDE writing stuff to run in DOS...IIRC the copyright date is 1991.)

Re:How many people actually use Borland C++?

[geekoid]

it was a better compiler for a LONG time when compared to Windows (arguable, of course).

Not really, if it compiles faster/smaller code then its better.

am not advocating one compiler over another, just staing that there is a way to tell which I a better compiler

Hmm

[pb]

I'm familiar with XWPE, and I tried it out a long time ago (6 years?) but I never liked it as an X application.

However, I hadn't seen WPE before, so I might try that out; thanks!

FYI, RHIDE is another TurboVision IDE clone, which is also very good, but not that stable under Linux.

"compile the kernel"?

[gfim]

What kernel? Or is everything automatically related to Linux?

Re:"compile the kernel"?

[Tim C]

From the story:

Looks like Borland is going to be releasing C++ for Linux

How much more related to Linux do you want?

Cheers,

Tim

Re:"compile the kernel"?

[Bradee-oh!]

What kernel? Or is everything automatically related to Linux?

While I agree with your point in spirit, I think it's fair to say that since we're talking about a new C++ compiler for LINUX, that the kernel in question was the LINUX kernel. :)

Re:"compile the kernel"?

[TheAwfulTruth]

And the answer to that is probably "No" since gcc has been hacked repeatedly to allow for the writing of "optimized" code in the kernel. A deplorable practice at best. c has been embraced and extended in gcc just for the (Linux) kernels sake.

BGI

[tangent3]

Hey cool, can we use BGI to do graphics in Linux too?

Re:BGI

[Graspee_Leemoor]

You evil man! I had succesfully repressed all memories of that awful slow gfx library but now you've brought them all back.

graspee

Is it just the compiler?

[jaybob20]

If so, Intel's compiler for linux is nice and makes things run alot faster. If you haven't check it out and are looking to make a program run faster it will help.
I heard libDSP now compiles with icc, but I know the kernel doesn't yet.

It's Fair

[Anonymous Coward]

The Watcom Power++ IDE was the best-designed GUI for C++ developers, but it crashed every time you scratched. Borland is way ahead of MS on ease of use, but the MS optimizations are much better than Borland's. So, it's a fair product, but not without some problems. Meanwhile, it looks like it will be a long time until we can tell if gcc is improving slowly or very slowly. It gives the patient an unequaled opportunity to display their patience. What do you expect? There's not a lot of money in compilers.

Re:Oh, by the way...I have another certification,t

[Anonymous Coward]

Must consult someone experienced.

Let the flames begin

[Vicegrip]

Exchange:
"I have a file I would like to share with you"
IIS:
"Please rape my server... please"
MSVC:
"We think we support the standards, wait, we don't... does that compile yet?"
"Oh, you want that kind of template support. That'll be another 1000$ for our next version"
Office:
"Bug fixes? Have paperclip instead."
BSOD:
"IRQL LESS THAN OR EQUAL..... Reinstall computer"
XP:
"You can use your computer how and when we say you can."
"Our new fancy QOS service now helpfully reserves 20% of network bandwith for itself. Have fun trying to find the right tool to change that"
"What do you mean you don't like our sexy new menu delays?"

Re:Let the flames begin

[cscx]

lynx:
"I have a file I would like to share with you"
wu-ftpd:
"Please rape my server... please"
gcc:
"We think we support the standards, wait, we don't... does that compile yet?"
KOffice:
"Bug fixes? If you're so smart, why don't you fix them?"

BSOD: "IRQL LESS THAN OR EQUAL..... Reinstall computer"

Quit being a retard. It should be a requirement that you know what the hell you're talking about before you can flame windows. 2 things can make that happen 1) bad hardware (e.g., bad ram) or 2) a poorly written f*cked up driver.

If you think that's NT's problem, well, NT's not stupid enough to let you type cp /dev/zero /dev/mem

Re:Let the flames begin

[Corporate Troll]

Oh...that BSOD message...I hated it: it happened to me once when I wanted to install a second SCSI card....different model of the one I had installed. Result: driver overwritten (sparrow.drv or something like that) by newer card and the boot disk was on the older SCSI card. Done....had to start from scratch: no way to fix it without installing NT4 again :-(

As for "cp /dev/zero /dev/mem"...if your permissions are set right only "root" or "wheel" group can do this...so not really a danger. (Just checked Mac OSX: crw-r----- 1 root kmem 3, 0 Jan 25 09:38 mem).

Re:Let the flames begin

[Vicegrip]

KOffice: Pull out version 1.0 of MS Office and compare... I thought so.
MSOffice:"Have you pet your paperclip lately?"

" to let you type cp /dev/zero /dev/mem "
Have you tried deleting the contents of your registry yet? NT is stupid enough to let you do that.

BSOD: "It's the vendor's fault"... how typical.. mabye NT internals just suck.

gcc: Works pretty good for me. At least it didn't cost me 1000$ to get standards support. Microsoft has been trying since verion 1.0 and msvc still isn't compliant.

Lynx never hosed my coporate network.

ftpd: Your machine got hosed? Thats too bad, maybe you should apply the free upgrade that fixes ALL the problems. No endless litany of hotfixes that may or may not fix one big gapping hole but not the other one.

Retard? I'm just playing with a Troll.. here Troll.. good Troll.

Re:Let the flames begin

[cscx]

Man, are all Linux Zealots such fucking assholes???

KOffice: Pull out version 1.0 of MS Office and compare... I thought so.

KOffice is a MS Office clone. The only reason KOffice exists is because it is _trying_ to be Microsoft Office. It will never be Microsoft Office, nor will it compare.

MSOffice:"Have you pet your paperclip lately?"

Actually, I use Rocky, not Clippit. And guess what? You can turn it off if you want to! And you don't have to use vi to edit a text file and subsequently restart Office to do so!! In OfficeXP, the default setting is a text box in the menu bar for help. The Assistant is _OPTIONAL_. Do you assholes fucking understand English? OPTIONAL - there's nothing wrong with OPTIONAL components!

to let you type cp /dev/zero /dev/mem " Have you tried deleting the contents of your registry yet? NT is stupid enough to let you do that.

Too bad NTFS doesn't mount as async as default like e2-shitty-fs... so if you reboot without unmounting you've hosed your FS. (that's what cp /dev/zero /dev/mem will do...)

BSOD: "It's the vendor's fault"... how typical.. mabye NT internals just suck.

Try pulling out an expansion card whilst your system is running. Like the BSOD? Yeah, NT interals suck, asshole....

ftpd: Your machine got hosed? Thats too bad, maybe you should apply the free upgrade that fixes ALL the problems. No endless litany of hotfixes that may or may not fix one big gapping hole but not the other one.

What the fuck? "Fixes for ftpd are free, and fixes for IIS are...uh.... free.... so consequently IIS sucks." You're a fucking idiot. I'd certainly like to see the free upgrade that fixes ALL problems. From what I recall, wu-ftpd _is_ an endless litnay of hotfixes. The solution for ALL Linux problems is called Windows (or Solaris).

Retard? I'm just playing with a Troll.. here Troll.. good Troll.

You're a fucking zealot. I'll have you know that I'm sleeping well tonight knowing that you're still running a toy OS while the rest of the world can get some serious work done.

Moderator note: The 5 instances of "fuck" and 3^H4 instances of "asshole" are necessary to get through this asshole's thick skull. Thank you.

It'll probably have something like this

[ahde]

(picture a pretty gui)

Tools->Options->Build

Compiler: [] Borland C++ []gcc |path|

Linker: [] Borland [] ld |path|

Flags:

GCC extensions

[achurch]

The kernel . . . is writen in C with a ton of GNU extentions. So the kernel is really tied to gcc. (which actually makes it more protable because gcc runs on a ton of machines)

Um, no. Using compiler-specific extensions does not make code more portable by any possible interpretation of the word. If it didn't use any extensions, then not only could GCC compile it, so could Intel's compiler, Sun's compiler, etc. That would be portable.

While we're on the subject, though, it would be nice to see at least some of the GCC extensions make it into other compilers. I try to write code without them as much as possible, but in particular I've found the typeof() construct useful, as well as the ability to initialize arbitrary members of a union (yes, I know I could just turn them into structs, but that's a waste of memory). Does anyone know alternate ways of doing things like this in standard C, or whether the GCC folks are doing anything to try and get their extensions included in the standard?

Re:GCC extensions

[AYEq]

You are completely correct. I just posted becasue I see people automatically thinking that the kernel is ANSI C. It was a design descision that I really cannot argue for or against, it is just that everybody blames the compiler if it cannot "compile the kernel"

Linux is not only a x86 OS

[chrysalis]

A really great thing with Linux (and NetBSD, and OpenBSD) is *portability* . These OS can run the same way on a variety of hardware.
Something designed on Linux x86 can run on Linux PPC with almost no change.
The master key to make it possible is GCC. Because GCC can compile (and even cross-compile) code for a lot of architectures.
Projects using Borlanc C++ specific features will work on Intel Linux. Nowhere else. This is pity. An opensource Operating System should be open to everyone.

GCC constructs that made it into the C99 standard

[Carnage4Life]

whether the GCC folks are doing anything to try and get their extensions included in the standard?

A number of GCC-isms ended up in the C99 standard. Such as support for C++-style comments, inline functions and named initializations of structs.

For more info on C99 differences from C89 try reading Are you Ready For C99? which appeared on Kuro5hin about a year ago.

Broken telephone re BCB for Linux

[StrawberryFrog]

She added that Borland plans to announce C++Builder for Windows next month, but declined to provide details.

That's got to be a journalist mangling the message. Borland has been selling C++ Builder for windows for several years already - why would they anounce it now?

This could really mean that either there is a new version of BCB for win32 to be anounced, or that BCB for linux is coming out.

Now it has been Borland's stated intent that the Kylix product line (or even the Kylix product) will support drag&drop C++ as well as Object-Pascal. I don't see why not - they share a compiler back end and a class library.

So either
- Borland has changed direction and not told anyone, and there will be no BCB for linux.
- We are getting BCB for Linux now.
- This is in fact only a commandline compiler, and is just a warmup for a later release of BCB for Linux.

Place your bets.

Re:who cares??

[psykocrime]

but I think vc++ handles templates much better than gcc, but thats just my opinion.


That's interesting, because it's widely known that VC++ has TERRIBLE support for templates.

Re:Look, Linux sucks

[ahde]

there are only about a hundred case studies where all that ultra complex exchange functionality has been recreated in a webapp with php or something in like a day.

No, it isn't built into the email client, but WTF is intuitive about launching the email program to schedule a meeting, or look up someone's phone number. A simple calendar program, connected to finger and an MTA and viola!

Re:Broken telephone re BCB for Linux soon

[StrawberryFrog]

Oops, just read further, seeing as borland says "We are taking our C++ development solution to the Linux platform." said Alison Deane

Now "development solution" doesn't sound like just a commndline compiler to me.

I expect that option 2 (BCB for linux now) is most likely, and option 1 (No BCB for linux ever) is right out.

There will likely also be a new version of bcb for win32. That's what happened on the Object-Pascal front: Kylix 1 rapidly followed by Delphi6.

Not necessarily....

[Micah]

Actually this was discussed on the borland.kylix.non-technical newsgroup a while ago. The aim is apparently to be able to compile ANY C/C++ Linux application, in fact the complete system, including the kernel. Thus it will need to emulate gcc's extensions.

We'll see if this turns out or not. That's just what I recall reading in discussions.

But if so it would be pretty sweet, assuming you're not a Free Software zealot (which I am, kind of, but I can see some coolness factor in this). If the pull it off, someone will be able to build an ENTIRE Linux distribution with their optimized compiler. Everything could run faster.

Re:Not necessarily....

[Ace Rimmer]

Everything could compile faster and run slower. Shorter compile time has always been Borland's strength but nothing has convinced me about the result binaries. They had always run slower than Watcom (which was comparable to gcc). Lets see...

Re:Not necessarily....

[BlueWonder]

The aim is apparently to be able to compile ANY C/C++ Linux application, in fact the complete system, including the kernel. Thus it will need to emulate gcc's extensions. [...] But if so it would be pretty sweet, assuming you're not a Free Software zealot.

Even though I am a Free Software zealot (i.e., I wouldn't use a proprietary compiler even if it was both technically better and available at no cost :-)), I think this would be a good thing. The more compilers support gcc's extentions, the easier it'll be to have them included in the next versions of the language standards.

Re:Not necessarily....

[briansmith]

That is what everybody says about Borland's compilers. They even had the monikers "Turbo C++" and "Turbo Pascal" for a while; presumably because Turbo Pascal could compile at an amazing speed compared to other compilers.

However, it makes me wonder how the generated code could be too bad, since I assume that Borland is compiling its own tools with its own compiler? In other words, I think the speed at which Borland's compilers execute is evidence that they can generate fast code.

Don't underestimate Borland's value...

[NerveGas]

Yes, GCC makes the kernel and all of the other parts of the OS work on lots of different platforms. But let's face it, the OS alone is not enough. Linux has fought a long, hard, uphill battle for the desktop. Why is that? Because developping GUI apps for Linux is more difficult than it is for Windows. Tools like C++ Builder and Delphi make it possible to spin out high-quality apps in a lot less time - and it also GREATLY widens the number of programmers (and programs!) that can very easily switch over to Linux!

So, this isn't going to make the kernel any more portable, and it might not make life any better for those running Linux on an Alpha - but for the great, vast majority of people who use Linux as a desktop, this has the potential to be VERY useful!

steve

This is not what Linux needs.

[Otis_INF]

What Linux needs is a kick ass developer environment like Visual Studio.NET, not another C++ compiler: the way software is written, the time it takes to cook up great reusable code, THAT's important. The compiler is just the end station of what's produced.

If you are fortunate to be able to take a look at the Visual Studio.NET releases, you know what I mean. I hope for Linux Borland will come with a toolset that gives the Linux developer the same productivity tools as Visual Studio.NET gives you. But I fear it will be only the v5.5 C++ compiler.

Re:This is not what Linux needs.

[Graspee_Leemoor]

Actually, although I am a big fan of the microsoft IDEs (oops I'll get modded to troll for expressing an unworthy opinion), even in .net Visual C++ is a misnomer. The only real RAD c++ tool for windows is C++ builder.

graspee

Borland text mode IDE clones

[MarkusQ]

What I miss most is the old text-based Borland IDE. That was the most productive development environment ever. RHIDE is close, but wasn't stable on Linux when last I checked.

Try SETEDIT; it's pretty darn close.

-- MarkusQ

Commercial compilers on Linux = Good

[Lurks]

Quite a lot of this ground has been discussed in the story I submitted Does Linux Need Another Commercial Compiler?. That being about my company's consideration of porting VectorC {PC} to Linux. It'll just be provided with the Windows version.

The upshot of that discussion was that VectorC {PC:Linux} is sheduled for release in April this year. That being based on our 2.0 engine so with C++ compatibility (currently VectorC is C only). While there was (unsurprisingly) a load of anti closed-source rhetoric in public, we did recieve a good number of serious private enquiries from people looking for a compiler such as ours on the Linux platform. It was enough to form a view that the platform is viable for us.

Ultimately I can't see any serious argument against having development tools such as these (Borland C++ and VectorC) on Linux. While it's likely of less interest to the open source/free evangelists wanting everything for free, there's definately demand where Linux is increasingly being used as an industrial platform.

Codeplay looks forward to competing with Borland on Linux.

Mat Bettinson - Codeplay Ltd.

Visual Development in Linux

[Pedrito]

Personally, I could care less about what the actual compiler is. For our product, we use MS VC++ for development and the Intel compiler for release builds (it's a better compiler).

All I ask from Linux is a similar or better IDE. I want to be able to write code, edit resources (dialogs), and do it all from one environment. I want a class wizard (again, I'm a VC++ user). Give me that, and I'll develop software for Linux in all my spare time.

Everyone talks about the advantages of Open Source. Give me this one simple thing, and I as user, will become an open source developer.

Don't make me deal with configure and manual makefiles and all that garbage. Forget it, I'm spoiled. I won't go back to that. It's like going back to the old DOS days. Give me a true IDE environment for development, and you've got me hooked. Throw on top of that a really good C++ class library for dealing with X, and you're done. So, who's doing this?

Re:Visual Development in Linux

[tree_frog]

OK, so you have a very good point about having a good IDE. But I think you are wrong about not wanting makefiles. With a makefile I can not only compile the code, I can

• rebuild the pre and post contracts for every function and method (design by contact) from the specification
  
• compile the code
  
• rebuild the documentation from embedded comments in the code (literate programming),
• regress unit tests against previous versions of the code
  

and a thousand and one other things too.

IDE's are good. They can make me (and you) very productive. But make is another fine tool that increases my productivity in another, orthogonal fashion to an IDE.

Why not use both?

regards, tree_frog

Re:Visual Development in Linux

[Pedrito]

IDE Let me ephasize "Integrated." Put it all in one place, and that works for me. Thats's what works for a number of other developers. I'm not against Linux. I want to see it grow and proliferatate. This is what I need to make it happen for me as a developer.

I know there are good tools out their for Liinux. What I want to see is MSDEV with MFC on top of it, and then you'll have me as a developer. Point me to the SINGLE download for this, and you have another Linux developer.

Re:Visual Development in Linux

[jcupitt65]

kdevelop pretty much does do everything you're asking for, check it out:

http://www.kdevelop.org/index.html?filename=feat ur es.html

If you install a recent-ish linux, check the "kdevelop" box and you won't even have to download anything.

qt is much nicer than mfc (imo).

Re:Visual Development in Linux

[mangu]

What I want to see is MSDEV with MFC on top of it

I used to think like that. Then I learned about Qt. Now I'm migrating everything I have ever done in MFC to Qt. The turning point was when I was writing a program that showed a picture in a window. In MFC I had to worry about things like creating palettes and converting image color depth. In Qt, I just do it like this:

QPixmap img("bullock02.jpg");

Another thing about MFC and VC++ was the doc/view model which was never really compatible with code generated by the project wizard. For some trickier things, the only way out seems to consult Eugene Kain's "MFC Answer Book", a great FAQ-like reference for the most frequent and difficult quirks in MFC.

Re:Visual Development in Linux

[Batou]

Yeah! And then, if someone would just port Visual Basic too ... And then maybe some ActiveX holes^H^H^H^Hlibraries! I guess someone will have to figure out how to make all that code always run as root, though ...

</sarcasm>

Seriously, I don't mean to flame, but who else here actually likes MFC, or would prefer to use it over ... well, ANYTHING else?
I can certainly appreciate the ability to churn out working code quickly, but code bloat and the necessity for distributing runtime binaries has always made it a little too close to VB for my comfort.

I do agree that VC has the IDE part down pat, and Linux sorely needs a development environment of this caliber, if for nothing else than making it more accessible to students/academia/beginners/etc. It's tough to go back to doing EVERYTHING from the command line.

Re:Visual Development in Linux

[TheAwfulTruth]

Too bad qt costs a FORTUNE to liscence even for $5 software :( TrollTech's liscencing is crap. I would LOVE to use qt for development of my shareware but not until they unfuck themselves a little bit more and stop shitting on the small developer.

Re:Visual Development in Linux

[spitzak]

Actually an "IDE" that produced a simple, readable, and editable file would probably do everything you can do with Makefiles. This is because it would be easy to write programs that could find all the source code and thus run doxygen and the other stuff you are mentioning.

Unfortunately most people are familiar with and copying VC++ and the horrid and huge "project" files it creates, that cannot be edited. I would use and IDE in a second if I could copy those files and I could edit them in a text editor to quickly add, remove, and rename source files. But just because MicroSoft software is written by morons does not mean you cannot make an IDE that is usable by real engineers.

Re:Visual Development in Linux

[Samrobb]

...who else here actually likes MFC, or would prefer to use it over ... well, ANYTHING else?

I would. Not because I think it's well-designed, or better than QT, or anything like that. I started working with MFC 8 years ago, and I've kept up to date with it over the years, and I know how it works - the little quirks, foibles, and problems that it has; it's strengths and weaknesses. Yah, it's pretty much a sucky piece of junk, but it's a sucky piece of junk that I'm very, very familiar with. Compared to working in any other GUI environment, I'm about 10x more productive when using MFC.

That may change, at some point. I've done some work with QT, and while it seems awkward in places, it's problems defintely aren't any worse than MFC's... just different. My MFC experience causes problems, because my GUI coding habits are geared towards dealing with MFC. It's just going to take some time to get to know QT well enough that I'll be as productive with it as I am with MFC.

If that reasoning sounds weird - I mean, why in the world would I want to use something technically inferior, just because I'm familiar with it? - go back and change "MFC" to "emacs" and "QT" to "vi" (or vice versa, depending on which editor you consider inferior, of course :-)

Re:Visual Development in Linux

[elflord]

I know there are good tools out their for Liinux. What I want to see is MSDEV with MFC on top of it, and then you'll have me as a developer. Point me to the SINGLE download for this, and you have another Linux developer.

You don't even need to download if your distribution is up to date -- KDevelop should come with the distribution. I suggest you invest a couple of dollars in a CD of the latest release of a decent distribution. Alternatively, you could download the ISOs

Re:Visual Development in Linux

[elflord]

Actually an "IDE" that produced a simple, readable, and editable file would probably do everything you can do with Makefiles. This is because it would be easy to write programs that could find all the source code and thus run doxygen and the other stuff you are mentioning.

KDevelop uses automake and autoconf at the back end. You can certainly edit the files, though you need to take care that you don't edit it in a way that lets it get overwritten

Re:Visual Development in Linux

[doug363]

Kylix/Delphi 6/(New versions of C++ Builder) have their class libraries (CLX) based on Qt. If you buy one of these products, you have an unlimited license to use Qt linked with programs you create which are based on CLX. The CLX widgets Qt widgets -- the CLX is a set of wrapper classes to remove multiple inheritance (Delphi doesn't support it) and add some VCL-like features. You can make direct Qt calls from your programs. (Of course, you still need to get enough money to buy Kylix/Delphi/C++ Builder in the first place...)

GREAT for us poor schmucks....

[MrBandersnatch]

who have used C++ Builder under Windows for YEARS and miss the ease and speed of putting together quick, fast and powerful C++ apps when moving to Linux. Somehow though even though Builder is my tool of choice I doubt I would use it that much if it doesnt act as a direct replacement for gcc. Also I already have problems getting Windows development jobs because I've refused to touch "Visual" C++. The last thing I'd want to find is the same situation cropping up under Linux because I happened to state that Builder was my compiler of choice.

Re:GREAT for us poor schmucks....

[JimPooley]

Also I already have problems getting Windows development jobs because I've refused to touch "Visual" C++.

And who's fault is that? That would be yours, wouldn't it now.
Maybe you should stop complaining that people won't employ you and learn Visual C++. Nobody likes a stuck up programmer.

Re:GREAT for us poor schmucks....

[MrBandersnatch]

Yup I'm stuck up!!
I wont "learn" "Visual" C++ because it is a **** POOR tool as far as I'm concerned. (Actually I probably do "know" it because of the work I've had to do with MS's embedded VC++ version - THE most unproductive time I've had as a developer due to undocumented APIs and bugs.)
Visual C++ is so good that MS have had to create a new language (C# - pronounced C hash) in order to escape the garbage that is Visual C++, MFC, COM, ATL etc.
WHY THE HELL WOULD I WANT TO WASTE MY TIME WORKING WITH THAT BAG OF ****???!!!
And did I say people wouldn't employ me? NO!
However when it comes to windows dev jobs, most of the time when I state that I have 5 years C++ experience of which less than 6 months is with VC++ I'm out the door because 90% of the C++ development work under windows IS done using Visual C++. I wont comment on what I think the reasons for that are!
Having been a faithful Borlandite for several years though..my point was..even though Linux development isn't exactly a bed of roses, I'd rather take longer to develop something on Linux that will build on X number of platforms than to get the insulation and ease of use that I've had with Borland AND suffer the marginalisation that comes with the choice to use their tools.

Re:GREAT for us poor schmucks....

[JimPooley]

Exactly. You haven't got the experience employers are looking for because you consider getting that experience to be a waste of time.
Pretty much confirms my point.

Re:GREAT for us poor schmucks....

[Kirruth]

I wont "learn" "Visual" C++ because it is a **** POOR tool as far as I'm concerned.

Bless you, Mr. Bandersnatch: you have seen the truth - that Visual C++ is a foul abomination. Really, writing code that locks people into a single operating system or set of middleware is doing them no favours, long-term. Better to keep your integrity, and put bread on the table doing things you believe in.

"property" C++ extensions?

[ahornby]

Does CLX still require the borlandc C++ extensions?

One problem with VCL is that it has "properties" which are a borland only extension which means that you can't build code that uses VCL headers with any other compiler.

Still Free software?

[PEdelman]

So, if I compile Free code with a non-Free compiler, would people regard it still Free software? This is meant serious, not as a troll.

As for me, it would not matter very much if with which compiler a program is compiled. But maybe someone has a good opinion on this.

Re:Still Free software?

[Micah]

I *think* so. The code itself is what matters. You can distribute it under any Open Source license you want.

Of course, if it depends on a proprietary compiler, it can't be part of a Totally Free(tm) system. But that doesn't mean the code itself is non-Free.

Re:Still Free software?

[An Ominous Coward]

Yes, it is still free software. According to Borland's license, any code you write, and code that is automatically generated by the compiler for your projects, belongs to you and can used / licensed however you like.

GCC 2.95 vs. Borland C++ 5.5

[Per Abrahamsen]

I get twice as fast an executable when compiling my application with GCC 2.95.2 under Cygwin, compared to Borland C++ 5.5 under C++Builder, both with full optimization.

Re:GCC 2.95 vs. Borland C++ 5.5

[Kruemelmo]

Is that just from a random single program or is it a well-founded experience based result?

Re:GCC 2.95 vs. Borland C++ 5.5

[DarkEdgeX]

That's pretty damn vague-- what, precisely, is your application? Depending upon what you're doing, Borland's compiler could be faster/better than the alternatives (VC++ or GCC).

Re:GCC 2.95 vs. Borland C++ 5.5

[Per Abrahamsen]

A numerical simulation, however I suspect the problem is with the iostream implementation. It makes no sense that BC++ 5.5 should be that bad for numeric code, especially as BC++ 5.0 produced almost as good code as GCC. Visual C++ 6.0 also give results in the same range as BC++ 5.0, slightly worse than GCC.

Re:GCC 2.95 vs. Borland C++ 5.5

[mazor]

The Borland C++ 5.5 compiler (free download from borland.com) is at least 6 years old, extracted from the original BC++ MSDOS/Win32 product. It's not the same compiler as the currently shipping Borland C++ Builder 5.0. If you're going to compare BC++ 5.5 to GCC, compare it to GCC 0.9x of the same era.

-mazor

Holy Crap , Batman!

[Evil Pete]

Geeeees!

Tears of joy ... now I can officially forget Visually Impaired C++ from my resume ... after all Borland is portable. Well maybe not ... kids still gotta eat.

I used C++ Builder on a contract recently ... absolutely wunderful fellas (and fella-ettes). So easy to do things. Not perfect but pretty damn good.

Better 'quality' kernel code?

[mccalli]

I'm using 'quality' for want of a better word, I'm not implying that the code doesn't work.

Personally, I find the number of warnings that scroll past whilst compiling kernel code to be quite worrying - particularly simple ones such as unused variables or casting errors.

It would be nice if people made the effort to write code that would compile in all environments (GCC/Borland/anyone else out there), and to try and make that code warning free. Having another compiler about might help people to check that.

Cheers,
Ian

Re:Better 'quality' kernel code?

[mccalli]

/*Damn the snobs who scorn ifdef*/

Names changed to protect the guilty, but a product I worked on in a previous company used a contractor to write some of the code. The contractor and the guy that ordered it really didn't get on, mainly because the manager was a fool who was rubbish at paying the contractor on time.

After yet another silly dictact, the line #ifdef JOHN_SMITHS_PAPAL_DECREE started creeping into the code...

Cheers,
Ian

Twice Shy

[NReitzel]

Well, Borland is going to 'support' Linux.

Whoooopie. I'm so excited I could just lift a finger in celebration. The "next article" finger to be exact.

Borland was a Godsend back in DOS days, when the Microsoft platform was unreliable and probably the least compatible C compiler ever invented. Those of us who were doing development work on MouseyDos spent our hard earned dollars on Turbo C, Borland C, release after release.

Then along came Windows and competing products from other vendors. Borland provided us with an extensible object framework better than anything that Microsoft had to offer. And then, something happened. All of a sudden, Borland was in bed with Microsoft, and those of us who worked with ObjectWindows, or (horrors) that "other" platform were abandoned like poor relatives at a party.

I personally give Borland a share of the blame for the extension of the Microsoft monopoly and the eradication of that "other" platform.

And now, they're going to bring out a product for Linux.

Well, friends, I won't spend a single cent on a Borland platform for Linux. Linux already has a mature tool chain available (more than one, actually) and in my humble opinion, Borland's products are unreliable. They are unreliable for the same reason every other proprietary product is unreliable; one never knows if tomorrow that product will even exist. If I find serious errors with the product, will Borland fix them, or will they once again decide that politics or bribery rule, and abandon their product (and me) to their own interests?

No, friend. I greet Borland's announcement of Linux support with all the enthusiasm of Borland's last six years of announcements. After spending thousands of dollars on their products only to be left standing in the rain, I will never, ever, buy another Borland product again.

Not ever.

Re:Twice Shy

[Micah]

In a way I agree and in a way I don't.

I have Kylix 1 and have been following the kylix newsgroups for a while. I'd love to be able to write some cool end user software in Linux with Kylix, but haven't had much opportunity yet (started a couple cool things though).

Problem with Kylix (and much proprietary software) is that Borland hasn't really responded acceptably to reasonable gripes. Back in AUGUST it was made fairly clear that some K1 patches would be coming out that would fix the IDE, debugger, upgrade the MySQL drivers (they shipped only supporting the old 3.22 version!) and presumably other things. Then the Borland folks teased us again in October by saying that we'd see a K1 patch before K2 shipped. Then a ***BETA*** patch came out and they say "oh yeah, we only mean beta". Now, only about a week ago did the final patch come out, and all it is is a debugger fix for the 2.4 kernel and a MySQL upgrade. That couldn't possibly have been that hard to fix. How in the name of all that's good and holy did it take them this long to put it out? Any self-respecting Free Software project would have had fixes for that kind of thing LONG ago!

So that solidifies my position that one should use only Free Software tools to develop something that is truly important or mission critical. Proprietary vendors simply can't be trusted. If something is wrong with the compiler and they won't fix it, YOU'RE SCREWED!

Having said that, however, Kylix is a great tool for the most part. It is far and away the easiest way to build end user GUI applications for Linux, especially if they involve database access. So I will probably use Kylix in that way in the future -- AS LONG AS the project isn't a super critical one where loss of support from the tool vendor would be a disaster. In that case, I'd probably use Python and wxPython, or perhaps C++ if it was performance critical.

Cool...

[PrimeNumber]

But what I *really* want to see is a Linux version of Sidekick. I can see no way of getting multiple apps running on the console at once. Besides, a "pop-up" text editor is sorely lacking on Linux. And yet people continue to slam DOS.

In all seriousness, they make super tools, and this is great news.

Re:Cool...

[WoofLu]

Use 'screen' ..

Great app, a must have.

No C++ Builder

[Morgahastu]

I don't think they will be making the C++ builder. Probably just the compiler. Why? The article in InfoWorld quotes Borland saying they want their Unix users to migrate to Linux. Borland has no unix C++ builder.

Re:Oh, by the way...I have another certification,t

[LuckyPhil]

Minesweeper Consultant Solitare Expert

is the release flawed?

[dollargonzo]

"why use borland when there is gcc?" seems to be the resounding question going around right now. this is the same as saying: "why use l00nix instead of windows?" maybe this will start another flame war, but hopefully /. crowd is too smart for that.

either way, the reason it was released is to release their OWN product on l00nix, and maybe SOMEONE will use it. and even if EVERYone is still using gcc, its competition. there is nothing competing with gcc. IF borland makes a kick ass compiler (address this later) and adds stuff that say, gcc, doesnt have, be it good ide (not for me, but w/e), or features, or good checking stuff like lint, etc....you get the point: its competition.

QED

Disclaimer

[nusuth]

Let me emphasize this before someone acuses me of being on crack: I don't know the internals of kylix, I don't know about the internals of lastest C++ Builder releaser. The above argument might not apply. It is based on structure of Delphi 3, C++ Builder 1.0 and C++ Builder 3.

Re:Disclaimer

[mazor]

Nice try, but you're still on crack. ;> The Delphi source code and unit support found in Borland C++ Builder is provided by the Delphi compiler, *not* by a C++ compiler that also has a Pascal parser. Go look in your bin directory - there is both a BCC executable and a DCC executable (dll). That's been the case in every release of C++Builder.

-mazor

Re:who cares??

[Graspee_Leemoor]

I'm glad Visual C++ isn't ANSI compliant! It means I can do dumb stuff like pass function pointers to functions requiring a void*.

Most of c++ is bondage and discipline stuff that we're better off without.

You know I'm right. You know you're right. The compiler won't let you. The compiler is your friend.

graspee

I do, but not for much longer...

[Per Abrahamsen]

We originally choose Borland C++ 5.0 for creating MS Windows binaries because Delphi was popular at our department.

Since that is getting ancient (I want to use more modern C++ features), I have been looking into an upgrade. The compiler must be cheap and easy to install, as Ph.D. students (who have never heard of Unix) will want to compile the application, and I don't want to come in a situation where I have to provide technical support for the compiler.

C++Builder is the obvious choice for a succecor, but the IDE is the worst I have ever encountered, is is slow bordering to unusable, and produce (for my application) ridiculous bad code.

Cygwin was the second choice, since I already use GCC on unix. It is also the recommended way to get CVS, and I can reuse the Makefile. However, the Unix environment is too weird for some of the users, and getting -no-cygwin to work for C++ is non-trivial. So I need a more conventional solution as well.

Visual C++ is what most people use. It has en excellent IDE, produce OK code, and mediocre C++ support. I have found work arounds for the limitations in the C++ support, so that is going to be the replacement for Borland C++.

Eclipse

[Ratbert42]

Borland is a member of the eclipse.org Consortium. Perhaps their new IDE is Eclipse. I hope so. The more I work with Eclipse, the more I like it. There's room for improvement, but those improvements are being made. Eclipse is the only open-source software my company has committed to. It's becoming a core part of our flagship product. No other open-source product can say that.

Re:More than once compiler... Good? Bad?

[SkepTech]

Having a new commonly used compiler for Linux would be a good thing. It would force GCC to better comply with the ANSI standards. As it stands, there are a number of 'minor' embrace-extend features of GCC that many coders aren't even aware of. Having to write their code to compile on multiple compilers shakes some of that kind of stuff out.

distribute binaries

[oyenstikker]

I can see how many linux users would not want to get an expensive program that might be better than their free one. But what if a group like Debian got one, and used it to compile their .deb's? Wouldn't that be able to produce better code for a whole junkload of people from one liscence?

Re:distribute binaries

[damiam]

Debian, being strictly GPL, would probably be the least likely distro to do this. Rad Hat or Mandrake might though...

Borland still sucks

[billcopc]

I know alot of pointy-headed IM/IT managers will gobble this up like hot semen, but Borland (read; Inprise) is still a big hulking piece of marketing. Their products _used_ to rock, but a few years ago (starting with BC++ 5.0) they just caved in and started producing expensive unfinished crap.

Yes, this compiler will probably produce better code than gcc and it will surely be much faster, but remember this is Borland. They're trying to 'Embrace the Linux platform', just like they tried to do with Kylix, and I have yet to see any serious Kylix-made projects in widespread use, probably because of their unfriendly license for free users.

Is BC++/Linux a piece of shit ? Not really. Is it going to be useful to most of us ? Not really.

Re:Borland still sucks

[dimsm]

>>they chose to use ObjectPascal for some strange
>>reason. Hopefully by adding C++ into the mix, >>this will change things.

Real programmers code binary.
come on there is NO difference between pascal and c++ (don't take that so verbal i talk in general)

EULA

[blkros]

Out of over 200 comments, so far, there have only been a couple of mentions of the license agreement fiasco. Why is this? Does the Slashdot community forget so soon? This was a really big thing less than 2 weeks ago, and now everyone's happy because Borland is releasing a C++ compiler for linux, hmmm. This seems hypocritical to me. Borland never really apologized for their EULA, they just excused it as a mistake made by their lawyers.See this. There enterprise agreement is not industry standard, it is ridiculous. I guess that it doesn't matter what a company does, as long as it's not Microsoft. It's hard to replace an OS, but not so hard to replace an app--I, personally, would use someting a little less polished, rather than support a company that claims to support a community, but, really, doesn't.
I'm really not trying to be a troll here, I just thought that this needed to be brought up, and, discussed.

I would disagree

[abdulla]

C++Builder 5 may look pretty, but it makes C++ in to Visual Basic, with all its bindings to VCL (written in Object Pascal!), this is a BAD thing, trying to strip all that crap from a default project takes an uneccessary amount of time, and even then you still have all the ide crap hanging around, i just want to get in to stright coding, not drawing my program, if i wanted that i'd use Visual Basic. All said and done, the compiler itself is a marvel and the default c++ and windows libraries that come with it are great (with the exculsion of all the borland specific stuff), now if only that had binary compatibility with visual c++, they'd get there market share back because then microsoft couldn't force us in to using there compiler (anyone using the directx 8.1 sdk knows what i mean)

I hope it's better than...

[ameoba]

I hope it's better than the version of Metrowerks Codewarrior for Linux. The Linux version of Metroworks' product is nothing more than an IDE wrapping the standare GNU tools.

I expect better from Borland.

Cross Platform?

[Pyrosz]

If this is BCB for Linux, I dearly hope they make it cross platform. If I can create an app in the Linux version and have it compile in the Windows version, I will BUY it the minute it comes out. Actually, whats stopping them from releasing a version that runs on both Windows and Linux? They would only have to create the IDE for Linux as they have the Windows one done and working... come to think of it, Kylix has the same interface, I wonder if they are just using that and just interfacing it for C/C++ ?? On another note, the GUI controls (VCL) are created in Delphi under the windows version of BCB so I guess they could use the Kylix versions for the Linux BCB.... All they need is a compiler for linux and they have everything else done!

Probably P4 optimized

[Self-Important]

>I wonder how it will compare to gcc?

If I were a betting man, I'd guess that it will probably be geared towards squeezing the last bit of horsepower from that floundering line of Intel processors.

/AMD fan

It's probably for the VCL

[ishmalius]

I would assume that the purpose of moving their compiler to Linux is to allow the Visual Component Library's (VCL) C++ binding to work on Linux. It's by far the nicest GUI toolkit in existence, IMHO. Borland uses it for C++/Delphi/Kylix, etc.

I seriously doubt that Borland wants to compete with gcc. (gcc is of course my favorite, as it should be for any true geek, especially you Linux/BSD guys :-)

However, it might help gcc improve itself. Gcc is far too forgiving in allowing non-ANSI constructs, and also allows mixing of C and C++ code. (Like virtual members of a C structure.) Running some typical Gnuish code such as Glib and Gtk through the Borland compiler is a good method of evaluating code quality.

BSD License

[LiquidPC]

Maybe Borland will open the source and put it under the BSD license, and we'll finally have a compiler with a FREE license. I doubt it, but it'd be nice.

More info on C++Builder for Linux

[Anonymous Coward]

I'm one of the compiler team members working on the Linux port right now, so I can offer a few more pieces of information.

What will be released is the full GUI environment, not just the command-line compiler.

bcc has not been tried on the kernel sources, and I'm doubtful it would work. We implemented enough gcc compatibility to use glibc, but some things, like inline assembly, just didn't make it on the schedule for this release. Expect code that uses lots of gcc trickery to fail.

Compared to gcc it compiles code almost twice as fast. bcc is meant to complement our GUI tool, not compete with gcc standalone. Since the GUI tool focuses on "rapid application development", having fast turnaround times is a big plus.

Probably the nicest item for command-line buffs will be our incremental linker. It should be usable in standalone projects, whether or not you use bcc (will have to do some checking to make sure..). Relinking after a single object change typically takes just a few milliseconds. First-time links are still faster than ld.

There will also be no STABS or DWARF this release. Sorry gdb users. :( The IDE has a rich debugger in it, but we would like to see a command-line alternative at some point. Some things just weren't possible for 1.0.

You know what this means!!!!

[RichiP]

The NeverWinter Nights DM Module Editor MIGHT be released for Linux after all!!!! Yahoo!!!!

They wrote the toolset in Borland C++. A release of it for Linux, they said, would depend on whether Borland would release a C++ compiler.

Three cheers for Borland! Three cheers for that Canuck company, Bioware!

Re:It won't last against KDevelop

[mikera]

Sure, it's good to have competition in the IDE space.

But last time I checked (5-6 months ago?) KDevelop didn't have anything to match Borland's VCL for RAD development. It's a good IDE, but Borland are just way ahead in terms of Visual Development.

The price is pretty negligible in TCO terms - it's well worth shelling out a few grand for even the most expensive Borland development tools if it can knock just 5% off your development time, and the speed of the compiler alone could well do that...

Don't get me wrong, I look forward to the day when open source tools are better than any commercial equivalent. My point is that we just aren't there yet.

Priced at???

[DuckWing]

If their price is anything like their Kylix software, I won't buy it. Borland is just trying to capitalize on what we already have and try to draw more people to them. It won't work with that pricing scheme.

XEmacs tricks (naze nani!)

[Mongoose]

Heh, well silly! You been using it for 3 years and haven't tried any of the nice features?

It's simple http://www.xemacs.org should be able to point you in the right direction. XEmacs/emacs has a ton to '3rd party' things like class browsers and code generators. You can also get more links to sites from google, and even install packages right from XEmacs menu. =)

Bad as in slow...

[Per Abrahamsen]

My applications works fine compiled with C++Builder (5.0, with BCC 5.5), it just run half the speed I get with GCC 2.95, Visual C++ 6.0 and even BCC 5.0.

Alternate licensing

[yerricde]

Why buy gnu-pro gcc? If it's based on gcc, their custom gcc source code must be available.

Not necessarily. The copyright holder (in this case, FSF) can choose to license the software under any license, including a "no copy, no reverse engineer, all your firstborn are belong to us" type EULA.

Re:Alternate licensing

[ejasons]

Not necessarily. The copyright holder (in this case, FSF) can choose to license the software under any license, including a "no copy, no reverse engineer, all your firstborn are belong to us" type EULA.

Are you even slightly suggesting that the FSF would ever do this? If not, then why mention it? This runs contrary to everything that the FSF stands for.

In addition, the original quote was about Cygnus, who would possibly do such a thing. Unfortunately for them, GCC is GPLed, and they don't own the copyright...

just x86?

[archen]

Sure it might work on Linux, but does it just compile for x86? I don't recall Borland ever making a compiler for other processors (although I really don't know). That in itself could be sort of annoying in that you could use borland's compiler to develop applications, but then be tied completely to x86.

Linux is not one platform

[alext]

Presumably this is Linux on i86? Yes, I'm on a PC too, right now, but if trends are to be believed Linux is going to be high-volume on ARM and other chips.

All attempts to target 'the Linux platform' for mainstream app development are therefore doomed unless the platform includes a VM (as in Java VM). Why can't Borland support Parrot or Mono rather than contributing to this proliferation of non-portable "electronic concrete"?

(Related post here).

Visibility of Borland RAD tools

[fm6]

How many people actually use Borland's C++ products currently? Of those that do, is this just brand loyalty from the Windows 3.1 days? I've been part of product teams (all using C/C++) developing on Windows, Mac, and various unices, and I've never seen Borland being used anywhere. The last Borland product I've seen used was Turbo Pascal for DOS back in high school.

Here's an irony: I work at Borland, and I'm currently working on the documentation for this product. (Do not ask me for product details. Must go through Official Channels.) The ironic part is that before I took this job, I had the same misperception. I applied for a job in the Enterprise software unit, thinking the developer tools unit must be in Legacy Support mode.

Not even close to true. A lot of developers are absolutely passionate about Delphi and C++ Builder. I think they'd riot in the streets if either product were discontinued. Not that there's any danger of this. Even in Borland's darkest days, these products were making money, and never stopped being under active development.

There are lots of reasons for this lack of visibility. The most obvious one is the reluctance of managers to commit to non-Microsoft tools, despite rabid lobbying from their engineers. And there always seem to be rumors of Borland's imminent meltdown -- even as we turn a profit and scramble to house the new hires.

One way to measure the penetration of Borland RAD products is to download and run the VCL Scanner, which uncovers installed Windows programs that use Borland libraries. First time I used it, I was shocked to discover how many of these there were. Some of them were basic system utilities that I had used for years.

What is beneficial? To Whom?

[HiThere]

Borland is entering a tough market. The Linux system already comes with C++ compilers and IDEs. Perhaps not as good, but free (GPL). This makes any sale dubious. Their products may well be better, but how large a market is there for better products against already established free products?

One place where they have a big edge is in dialog building. Kylix already builds dialogs under Linux, so they know how.

Perhaps what their goal is, is to get people using their products on Linux to be cross-platform, but to make their pile selling compilers for the Windows versions? Sounds chancy to me, but it would let them sell the Linux system at around cost, and still make SOMETHING. I'm dubious about proprietary libraries, but I believe that the GPL (NOT LGPL!) version of the library is available at sourceforge. And that Borland holds sole rights, so they can license it commercially if you pay them.

This means that you can use Borland libraries in GPL software without cost, but if you want to sell the product, Borland gets a share. (Sounds fair to me. Viable? I don't know.)

Given this evaluation, past comments, etc., and what I expect is that this C product will be the C++ companion product to Kylix. And it will probably be available on the same basis.

As to what they'll call it, companies are unpredictable, but what I think they should call it is:
Kylix C++, a C++ environment for the Kylix family.
Then they could follow it up with:
Kylix Python, a Python environment for the Kylix family.
Kylix Java, a Java environment for the Kylix family.
etc.
The kicker would be that all of the various pieces could work together in a relatively seamless way. (This takes a bit of work, but SuperCede Java did this between Java, C, and C++ on the PC side years ago. They finally got bought out by someone who raised the price to $10,000 per copy, but it worked pretty well when it was affordable.) And, of course, gcc has always worked this way.
.

The competition

[Anonymous Coward]

LCC : http://www.cs.princeton.edu/software/lcc/

GCC : http://gcc.gnu.org/

Comeau C++ : http://www.comeaucomputing.com/

Intel C++ : http://www.intel.com/software/products/compilers/c 50/linux/

KAI C++ : http://developer.intel.com/software/products/kcc/

CodeWarrior : http://www.metrowerks.com/products/linux/

SGI Pro64 : http://oss.sgi.com/projects/Pro64/

TenDRA : http://www.cse.unsw.edu.au/~patrykz/TenDRA/

And yeah, there will be a Vector C++ for Linux as well. http://www.codeplay.com

If you feel like it, you could take projects like SableCC and SUIF into account as well.

Good luck, Borland.

Borland C++

[llama_flyer]

I am a student and my CS classes all focus on the use of M$ Visual Studio. I have been trying to wing myself away from M$ but I do find that it can be difficult to leave when you get used to a particular format. I love it when Im booted in Linux I love the feel of the desktop and the look, but I and others out there still feel a need to keep a dual boot machine because a lot of classes and people are not quite as linux friendly as others. I have to boot windows everytime I have to code something for my classes, because I have to say that the Beast in Redmond has produced an enviroment that I like to program in. I tried using gcc and emacs but I had a hard time learning how to get arounf in emacs. The fact Visual Studio is pretty and it works well for my classes. Now then if Borland does go through with this...then that would be a really great thing, becuase I would be able to have the enviroment that I had using VS, but with my OS of choice...LINUX :) I love linux and I want it in the homes of more people, so I see this as "A good Thing."

Remember, the kernel started out in Borland C.

[Kaz Kylheku]

Linus used a Borland compiler to bootstrap the development of Linux, because Minix was compiled with that. (Then when the system was good enough to host GCC, and serve as its own development platform, that compiler was abandoned).

unsigned i; long j; legal?

[yerricde]

However, you're missing something. ANSI C++ forbids declaration of i with no type.

I understand that declaring a variable with a type made of only qualifiers (const, volatile, etc.) is illegal in C++, but what about unsigned i; or long j;? Are tokens such as "unsigned" and "long" considered qualifiers or part of the basic type?

That's an error, not a warning.

[yerricde]

ever tried raising the warning level? MSVC catches _much_ more errors that way.

<grammar-national-socialist> First of all, "many" not "much" to agree with plural "errors." </grammar-national-socialist>

More to the point, it's not a warning; even with a warning setting on minimum, GCC will still report errors.

test.cpp:3: ANSI C++ forbids...

That's an ERROR, not a warning. Note the lack of "warning" immediately after the line number; GCC uses this convention to distinguish errors from warnings.

Borland C++ for OS/2 - a warning!

[DrXym]

I hope Borland learns from past mistakes.

About 8 years ago I got into OS/2 in a big way and wanted to start developing for it. CSet from IBM was very expensive so I was ecstatic when Borland C++ came out for OS/2. Version 1.0 was pretty damned good - a powerful IDE, a decent set of helper classes (no GUI ones though) and a syntax highlighting editor in an age when IBM CSet++ shipped with no editor whatsoever. All was good or so I thought and I put aside concerns about the few bugs because it worked dammit and 1.01 would iron out the remaining problems.

Version 1.01 came and had some fixes for the existing problems but overall had *more* bugs than 1.0. The classes didn't work as designed, the debugger more frequently hung your machine than worked and worst of all the IDE crashed - randomly. All was getting decidely iffy but I knew that 1.5 was out so I still held out hope.

Version 1.5 and all hope flew out the window and emigrated to Australia. I had never seen a buggy piece of shit in life! The all new improved version may as well had alpha written all over it. It was unusable and I gave up after discovering threading was completely broken - a feature I had used up until then.

Soon after I switched to IBM CSet++. Now that was a compiler. It might have had a totally stinky UI (none at all), but the command line tools were excellent and it came with some pretty good classes too.

Re:Borland C++ for OS/2 - a warning!

[BlindSpot]

Actually, BC++ 2.0 for OS/2 was quite stable and worked very well for what I used it for. They seemed to have fixed all of the major problems, e.g. with compiling long functions. I used it for several DOS to OS/2 ports and also for some Presentation Manager applications.

The sad part was that shortly after it was released, Borland gave up on their OS/2 compiler. Had they given 2.0 a chance it might have cut into CSet and Watcom's share. (Though dropping it turned out to be a good move because OS/2 interest itself petered out soon after.)

Hopefully they will get it right for the Linux edition, both in terms of the product quality and the way they approach marketing it.

That is not an error

[madsdyd]

Really, the internal representations differ. No big deal. In reality, you should never use == with floats.

What!?!?

[Magnusite]

There's not a lot of money in compilers.

My friend, what have you been smoking?

How do you think Borland became so big in the first place? Because they never really had anything bigger than their Pascal and C/C++ compilers. They started out with Sidekick and Turbo Pascal. Sidekick sold really well for a year, and then sank into the swamp. They either purchased or developed Quattro Pro. That sank into the swamp. They purchased FoxPro and improved it. That sold ok until Access took off, and then it sank into the swamp. they developed their own database engine, Paradox, and sold it. That burned down, fell over and sank into the swamp. But the compilers always kept them afloat.

Also, if your company is mid-sized, you can make quite a bit of money making compilers for embedded processors. Check out Hi-Tech for a 68000 C compiler - $1150 cha-ching or an 8051 compiler - $850 cha-ching or a Z80 compiler - $850 cha-ching.

And that's only the first company that came up in my google search for "embedded c compiler". There are dozens more. Not a dozen more, dozens.

The 10 outweighs the 2

[yerricde]

I think the speed at which Borland's compilers execute is evidence that they can generate fast code.

Not necessarily. If Borland's compiler is ten times less complex than GCC, but GCC produces code that's twice as fast, Borland's compiler compiled with itself will still feel five times as fast as GCC compiled with itself.

Re:What is beneficial? To Whom?

[MikeBabcock]

Large companies wanting to port their software to Linux may well be much more comfortable buying a copy of Borland C++ (I know some of my coworkers would be) than using gcc.

You can always tell

[Vicegrip]

who has lost an argument by looking to see who is swearing first.

bye bye Troll

I used C++Builder 5.0

[Per Abrahamsen]

The compiler contained in C++Builder 5.0 identifies itself Borland C++ 5.5. I have never used the free Borland C++ compiler.

I'm surprised to hear that the free C++ 5.5 is six years old, as the Borland commercial product back then was called C++ 5.0.

The GCC of the same timeframe (six years ago) was 2.7.x. Version 0.9 (the initial beta) was from 1987, 15 years ago.

I'm never on the crack

[nusuth]

I prefer pot. And once I was smoking heavily I just renamed bcc32 to bcc32-bckp and the builder worked just fine. While I was stoned I thought that bcc32 is the command line, standalone C++ compiler while dcc32 is the compiler. Obviously I find that ridiculous now that I'm sober.

Try it, maybe your last pot hasn't weared of yet.

I misread your message, I apologize

[nusuth]

I thought you were claiming that two seperate compilers were used for compiling two languages.

I know that dcc32 for builder is dcc32 for delphi+a new parser (at least that is what charlie calvert claims, and he is supposed to know.) So it is a C++ compiler that also has a Pascal parser, even but the temporal relation is backwards (the C++ parser is addon rather pascal parser.) The question is whether bcc32 is based on dcc32, my assumption is that, it is, as this quote from borland suggests: "The Free Borland C++Builder Compiler is the foundation for Borland's fully integrated C++ development environment - Borland C++Builder 5."