gcc 3.0 Coming Soon to a Computer Near You

An anonymous submitter sent in pointers to a couple of mailing list messages about the imminent release of gcc 3.0. The official word seems to be that 3.0 might be out as early as this weekend.

Comments

Here's the funny thing...

[bconway]

Guess what? Everyone that complained about GCC 2.96 being broken (and not reading http://www.bero.org/gcc296.html) despite the fact that their code wasn't C99 complient STILL WON'T COMPILE. Now you can't complain that your code won't work because it's a developmental compiler, you'll actually have to fix it. Numerous examples of this are listed at the above URL, I'd highly suggest you try it out. I have a feeling quite a few people are gonna be red in the face over this one. ;-)

P.S. It's GCC 3.0, FYI.

Re:Here's the funny thing...

[sholden]

Guess what? Everyone that complained about GCC 2.96 being broken (and not reading http://www.bero.org/gcc296.html) despite the fact that their code wasn't C99 complient STILL WON'T COMPILE. Now you can't complain that your code won't work because it's a developmental compiler, you'll actually have to fix it. Numerous examples of this are listed at the above URL, I'd highly suggest you try it out. I have a feeling quite a few people are gonna be red in the face over this one. ;-)

That's quite humourous...

Given that the first 'correct code' example is wrong I wouldn't put to much trust in the rest...

It is 'int main' in C++ not 'main', god damn it.

I've been using the snapshots og gcc 3.0 for a while now, precisely because it is more standards compliant.

From the C perspective, what's wrong with non C99 compliant code. People still use K&R C, because it
is more portable.

Version number inflation

[ryants]

On another note, isn't 3.0 a really low numbered release?

Uh... would it make you happy if they called it version 23.0? Other vendors increase version numbers like there is no tomorrow (makes managers happy, I think), so we should "catch up" to that silliness? What difference does it make?

just seems like more proof that the Open Source development model leads to slower release cycles

What are you smoking? Free software is in continuous release. How long has it been between versions of MS VC? Almost a year and a half now...

a never ending game of catchup

I think if you compare standards compliance, gcc ranks right up there with every other vendor.

I was going to just mod you down, but there is no "-1 duh" moderation.

Ryan T. Sammartino

Re:Version number inflation

[EvlPenguin]

Version number inflation is a really stupid concept to those who know what they are doing. But since when did that mean anything? Very few people grasp the concept that just because something has a higher version number does not mean it is more dependable or bug-free (*COUGH*Mac OS X*cough*).

In particular, this reminds me of when Slackware decided to jump to version 7 from 4 just to keep up with certain other distros when Linux was becoming popular.

Idiocy never ceases to amaze and influence. Then again, the folks in Redmond are seemingly above numerical versioning after 3.11. First Windows 95, then 98, then 98B (ooooh), then ME/2K, and after that I guess they decided to switch to something other than the year (people may start noticing how much they are being fucked over if they are reminded of the present year by looking at their boot screen) and now we have XP. After that.... XP2? XIP? LMNOP? XYZ?

Ugh.
--

This is good

[Laplace]

I just tried to download KAI++ and install it on my SuSE box. Amazingly enough, Intel fucked themselved (and their customers) by releasing versions of their compiler that were dependent upon the RedHat 6.2 version of egcs, and the RedHat 7.1 version of GCC (2.9unstable). GCC 3.0 will come out, all of the distros will get up to speed, and I'll be able to use an optimizing compiler to help my fast code run faster. Slightly off topic, but not really.

Cought by a troll

[sethdelackner]

I can't believe you bit.

flame

[selectspec]

Rehat has already incorporated gcc 3.0 into a patch for 7.1 and 7.0 that they are posting as a critical update.

Re:flame

[Rick the Red]

Rehat has already incorporated gcc 3.0 into a patch for 7.1 and 7.0 that they are posting as a critical update.

Really? Where? It's not here.

Re:flame

[selectspec]

Ok, bad joke. I was kidding. RedHat took alot of heat for putting 2.96 into RH7.0. Just a little redhat humor. Obviously, it was not funny.

actually...

[StandardDeviant]

I _think_ it can be either int main() or void main() or main() (which the compiler takes as void main() ).

But then I hack perl and C, not C++. :-) I'd check but I don't have a c++ compiler handy at the moment.

--
News for geeks in Austin: www.geekaustin.org

Re:actually...

[andkaha]

You're lucky this isn't comp.lang.c++!!!

It's always

int main()

(in C you have to do "int main(void)") or

int main(int argc, char *argv[])

(as in C).

Unlike C, you don't have to "return 0;" in C++ at the end of main.

Re:actually...

[StandardDeviant]

Umm, ok. Then why does this work:

#include <iostream>

void main() {
cout << "Howdy, y'all!" << endl;
}

And that proto for main is in very widespread usage. Maybe the standard says different and it's just all the compiler makers and C++ book writers that are "wrong", heh. Maybe the compilers silently turn void return in to int return 0. Maybe C++ just sucks. ;-)

--
News for geeks in Austin: www.geekaustin.org

Re:actually...

[ksorim]

Try to compile that code with -Wall (with gcc) and you get:

test.cpp:3: return type for `main' changed to `int'

Re:actually...

[andkaha]

The correct code would read:

#include

int main()
{
std::cout "Hello all!" std::endl;
}

The main function in C++ should look the same as in C (but may have '()' instead of '(void)').

You should read the C FAQ, sections 11.12, 11.14 and 11.15 at "http://www.eskimo.com/~scs/C-faq/s11.html".

For an explanation of the 'std::' prefix, please ask comp.lang.c++ or alt.comp.learn.c-c++ about "namespaces".

With GCC, always use -Wall.

Re:actually...

[bellings]

Maybe the standard says different and it's just all the... C++ book writers that are "wrong", heh.

Yes. Many of the books you'll find in the computer section of your local bookstore are painfully incorrect, often on the most basic topics. It's almost a joke, really, how much money is charged for many books that are full of paper that is too hard and scratchy to use for toilet paper, but not useful as anything else.

The C++ FAQ recommends a few good books on C++, if you're actually interested in learning the language.

Re:It's funny...

[selectspec]

...new patches will not be accepted without thorough review...

Right, because it's his project. What is your point? Anybody can fork off GCC if they wish. So what?

...isn't 3.0 a really low numbered release? Inprise's compiler is versioned fairly high. So is Microsoft's.

Since its inception in 1987 there have been 101 released versions of GCC. Microsoft since Quick C has released 11 versions counting all Visual (Quick) C and Visual C++. Counting all of Borland/Inprise's various C and C++ compilers (Turbo, Borland, Builder, OS/2, and Vision) there are 21 versions. Of course those are complete IDE's so you can't really compare them to gcc which is only a compiler. However, I think you get the point. Neither MSVC nor C++ Builder are platform indepenedent so you can throw them in the trash (what is the point of C/C++ if you can't port to a different platform).

Re:It's funny...

[PD]

Amen, brotha! But I have to admit that I'm still using Microsoft Word for Windows version 3.0. That program does everything that I could possibly want in a MS product.

Re:What about the other compilers?

[akharon]

From what I understand, ccc (Compaq C Compiler) for the alpha is supposed to make GCC it's bitch, so to speak. Of course, if you're talking about cross-platform compilers, then, well, nevermind.

namespaces

[StandardDeviant]

well, I do know about them, I just omitted the "use namespace std;" bit because I was feeling lazy ;-)

thanks for the url, btw.
--
News for geeks in Austin: www.geekaustin.org