Re: Problem compiling sparse-xdiv.cc (typename)

[Jaroslav Hajek]

On Sun, May 10, 2009 at 10:40 AM, Michael Goffioul
<address@hidden> wrote:
> On Sun, May 10, 2009 at 8:20 AM, Jaroslav Hajek <address@hidden> wrote:
>>> I find this a little unfortunate as well , as many (not so) older systems 
>>> are
>>> still running on gcc-3.
>>
>> Finding something unfortunate is not constructive. I think there are
>> more places that gcc 3 chokes on and they aren't probably that easy to
>> workaround. But feel free to suggest a constructive solution.
>
> Wow! This wasn't really meant to be "constructive", but to express
> the opinion from someone who contributed signifcantly to octave in
> the past. I didn't know the policy was now "contribute ot shut up".

Forget about "policy". There's no such thing. Sorry about that word. I
don't understand where you took that "shut up" from. I agree that the
fact that gcc3 does not compile Octave is "unfortunate". So what? I
have a list of such "unfortunate" properties Octave, and the gcc3
incompatibility does not even make the top ten. And it is the kind of
feature whose importance fades over time.

> That being said, I don't use gcc and I'm not the one who broke
> compilation for gcc3, so I don't see how I could be constructive on
> that point, except than saying that the responsible for the breaking
> might want to take a look.

I don't use gcc 3 either. Someone reported the breakage (the one in
oct-inttypes.cc) a couple of weeks after the code was checked in (and
a number of fixes afterwards). (surprisingly) I *did* take a look. I
verified the code was correct C++ and the bug was in gcc 3, and I
offered several times creating a workaround patch. None of the
reporters wanted it, mostly the answer was they'd like to upgrade gcc
anyway. So, what else you suggest I do?

cheers

-- 
RNDr. Jaroslav Hajek
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz