| commit | da60914c2634242f38b5461f59dd5c34b63f49bf | |
| author | Petr Písař <petr.pisar@atlas.cz> | |
| Wed, 16 Dec 2015 12:37:10 +0000 (16 13:37 +0100) | ||
| committer | Petr Písař <petr.pisar@atlas.cz> | |
| Wed, 16 Dec 2015 12:37:30 +0000 (16 13:37 +0100) | ||
| tree | 1eb2f0dbd622500bfe93a181eadd1df5f64070b3 | treesnapshot (tar.gz zip) |
| parent | 0b29210bcd9557ab4b837170d17abe972ba95c40 | commitdiff |
Format tv_usec as int32_t in timeval2timestring()
Windows' PRIdMAX is not compatible with MinGW's printf(). But because we know
the tv_usec is not bigger than 999999, we can handle it as int32_t.
OTOH this is very bad because libisds is written in ISOC99 and toolchains not
supporting 16 years old standard deserve eternal damnation. Therefore I will
not adjust tests only to conform broken compilers. Actually maybe I will
revert this change completly in the future beacuse I'm not fully convinced.
Windows' PRIdMAX is not compatible with MinGW's printf(). But because we know
the tv_usec is not bigger than 999999, we can handle it as int32_t.
OTOH this is very bad because libisds is written in ISOC99 and toolchains not
supporting 16 years old standard deserve eternal damnation. Therefore I will
not adjust tests only to conform broken compilers. Actually maybe I will
revert this change completly in the future beacuse I'm not fully convinced.
| src/isds.c | diffblobblamehistory |