[FFmpeg-devel] [PATCH] Add a MAXDIST parameter to tiny_psnr (for 1-off fate tests)

Måns Rullgård mans
Mon Jul 12 09:45:17 CEST 2010

Rafa?l Carr? <rafael.carre at gmail.com> writes:

> On Fri, 09 Jul 2010 02:51:05 +0100
> M?ns Rullg?rd <mans at mansr.com> wrote:
>> Rafa?l Carr? <rafael.carre at gmail.com> writes:
>> > On Thu, 08 Jul 2010 21:06:55 +0100
>> > M?ns Rullg?rd <mans at mansr.com> wrote:
>> >
>> >> > What about, just... abs()?
>> >> 
>> >> I'll bet a beer there's at least one system where it's missing.
>> >
>> > I accept the bet (although Vitor said it was "present in all
>> > supported archs" so perhaps it's too late, up to you)
>> To win the bet you must show that no system exists where this function
>> is missing.  This could prove rather difficult given the forking rate
>> of BSD (and their penchant for randomly leaving out standard
>> functions).
> All the BSD I checked have had abs() since the start and it was not
> removed.
> However I found that Haiku lacked an abs() prototype, until October 2002
> http://svn.berlios.de/viewvc/haiku?view=revision&revision=1649
> And some versions of HP-UX failed to declare abs() under certain
> conditions (can't find the link again)
> - including both stdlib.h and math.h in some specific order
> - the file being compiled being C++ (C was not affected)


> - Plan9 is ok (though i couldn't find a stdlib.h, plan9 doesn't aim to
> standards compliance)

If it's not in stdlib.h it may as well not be there at all.  Guessing
headers isn't a game I particularly enjoy.

> So is the bet won/lost by one of us? If you hesitate we could vote

Or we could just have some beer anyway.

M?ns Rullg?rd
mans at mansr.com

More information about the ffmpeg-devel mailing list