This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [RFC] PR6015, snprintf, glibc2.3 and all that...
- From: Paolo Carlini <pcarlini at unitus dot it>
- To: Ulrich Drepper <drepper at redhat dot com>
- Cc: libstdc++ at gcc dot gnu dot org
- Date: Thu, 04 Apr 2002 11:17:43 +0200
- Subject: Re: [RFC] PR6015, snprintf, glibc2.3 and all that...
- References: <1017911147.26729.150.camel@akkadia.org>
Ulrich Drepper wrote:
>On Thu, 2002-04-04 at 00:32, Paolo Carlini wrote:
>
>>There is a misunderstanding here... Probably due to my bad english...
>>
>I
>
>>never meant really implementing this myself! You are kind with me in
>>supposing this! I meant really /excerpting/ and /adapting/ it from
>>glibc2.2! Is this doable?
>>
>
>Not really. You'd dragged probably 150k of source code around. This
>code never is meant to be used outside glibc, it's not customizable.
>It's not impossible but it's everything but easy.
>
I see.
Naively, what I'm still finding strange is that there is no way of
having from a libc (glibc, in particular) a "C"-locale-only printf, not
sensitive, I mean, to the locale in force. Is just a subset of the whole
functionality and it would be /so/ useful in order to cleanly implement
MT-safe C++ output operations, according to the letter of the standard.
Is there a general "philosophical" point about this which you can
possibly explain to me?
Ciao, Paolo.