Re: Warning when compiling GNUstep

[Alexander Malmberg]

"Chris B. Vetter" wrote:
[snip]
>   <alexm_> cbv: I'll assume glibc is correct; get them to change, and
>            I'll change too :)
> 
> I interpret Alex' answer as
> 
>         Because it's in glibc it's therefor 'correct' behaviour....
>         (aka: another Linux'ism)

Then you interpreted it incorrectly. :)

It's not correct because glibc does it; glibc does it because it's
correct.

As I noted, this is consistent with the standard (which used to be
inconsistent, but was changed to remove the const where it occurred):

http://www.opengroup.org/onlinepubs/007904975/functions/iconv.html

and with the fact that the glibc guys have looked at this, and the
standards, and decided that non-const is correct (iow, they are aware of
the issue, and the lack of const is not an oversight).

See eg.:

http://www.opengroup.org/austin/mailarchives/austin-group-l/msg00269.html
http://gcc.gnu.org/ml/libstdc++/2000-11/msg00078.html
http://www.geocrawler.com/mail/msg.php3?msg_id=5232650&list=360

I agree that this is stupid and that the standard should say "const",
but it doesn't, so I'll trust the glibc team on this. If you can get
them to change, by changing the standard, then I'll likely change with
them.

- Alexander Malmberg