[FFmpeg-devel] [PATCH] Implement the function cmdutils.c:parse_int_or_die
Wed Feb 20 13:43:44 CET 2008
On Wed, Feb 20, 2008 at 03:05:31AM -0500, D. Hugh Redelmeier wrote:
> I would recommend that the project standardize what is expected of
> FPUs, at least in these dimensions. That would allow builders to know
> what pieces of 754 conformance can be turned off (for speed) and it
> would allow coders to know just what the can use.
Or in other words, lets waste time because the CS professor said it.
I do prefer some down to earth reasoning ...
Down to earth reasoning would be:
What rules/limits exactly are you talking about?
What exactly do we gain if the compiler is told that we are within these
limits/rules (If it cant be benchmarked its not relevant and thus has no
gain aka its a waste of time)
Which FPUs exactly have problems with what exact parts of the IEEE spec.
And as a good rule of thumb, the zero gain, shitload of work assumtation
applies (always a safe bet when it comes to such stuff).
Also the normal (outside the academic dreamworld) way to handle float is
not to assume floating point operations to be exact and keep in mind that
they are represented by a finite number or mantisse and exponent bits.
That worked quite well until the CS people found out that one can abstract
that to the point of not having any relation to mathematics nor reality
anymore and write 100 page papers about it.
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
It is dangerous to be right in matters on which the established authorities
are wrong. -- Voltaire
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: Digital signature
More information about the ffmpeg-devel