Hi,
On 24/09/14 18:11, Lancelot SIX wrote:
> The most notable one would be to use new / delete instead of malloc /
> free [1]:
>> s = new char[sb.st_size + 1];
I would recommend to use a std::vector instead, which avoids
manually deallocating the buffer. Maybe some QT4 functions do raise
excepti
On 16/05/14 16:32, Svante Signell wrote:
>> If [L_tmpnam] is currently not used, it should probably be dropped to avoid a
>> later reintroduction of PATH_MAX.
>
> OK, we will se what the debian maintainers/upstream says.
Agreed.
The updated patch looks good to me now.
Thanks,
Philipp
--
T
Svante,
On 15/05/14 22:39, Svante Signell wrote:
> On Thu, 2014-05-15 at 13:38 +0200, Philipp A. Hartmann wrote:
>
> Yes, you are right. However, setting len to zero would have solved that.
I'm not sure if passing len=0 to snprintf is portable in all cases. If
you insist on repor
Svante,
I have some additional comments. Please see below.
On 15/05/14 08:43, Svante Signell wrote:
>
> static void
> -mkfifofn(char *restrict buf, size_t bsz, const char *key, unsigned int tid)
> +mkfifofn(char **buf, const char *key, unsigned int tid)
> {
> - snprintf(buf, bsz, "%s out
On 28/08/12 18:20, Svante Signell wrote:
> On Tue, 2012-08-28 at 18:14 +0200, Samuel Thibault wrote:
>> Svante Signell, le Tue 28 Aug 2012 17:55:07 +0200, a écrit :
>>> --- a/libetpan-config.h.in 2008-05-23 00:39:44.0 +0200
>>> +++ b/libetpan-config.h.in 2012-08-28 17:36:00.00
Pino,
On 18/06/12 13:04, Pino Toscano wrote:
>
> Alle lunedì 18 giugno 2012, Philipp A. Hartmann ha scritto:
>
>> Alternative sketched below, untested of course.
>>
>> On 18/06/12 09:50, Samuel Thibault wrote:
>>> Barry deFreese, le Mon 18 Jun 2012 00
Samuel, Barry,
it's C++, therefore we should probably use appropriate C++ features
instead of platform-specific things like strdupa().
Alternative sketched below, untested of course.
On 18/06/12 09:50, Samuel Thibault wrote:
> Barry deFreese, le Mon 18 Jun 2012 00:20:17 -0400, a écrit :
>> +#ifd
7 matches
Mail list logo
