[FFmpeg-devel] [PATCH] Enable swscale by default

[Michael Niedermayer]

On Mon, Sep 15, 2008 at 04:03:12AM +0100, M?ns Rullg?rd wrote:
> Michael Niedermayer <michaelni at gmx.at> writes:
> 
> > On Sun, Sep 14, 2008 at 09:44:26PM +0100, Robert Swain wrote:
> >> 2008/9/14 Guillaume POIRIER <poirierg at gmail.com>:
> >> > On Sun, Sep 14, 2008 at 3:56 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
> >> >>
> >> >> The second part that exists is the begin of a failed yuv2rgb rewrite, here
> >> >> the student choose to use wiki instead of git to keep track of the code,
> >> >> the code contained assumtations like a^b being pow(a,b).
> >> >
> >> > How is it possible that this guy has been accepted since qualification
> >> > tasks are supposed to at least ensure that the student know C and is
> >> > able to program?
> >> > I'm puzzled.
> >> 
> >> I hope this is a miscommunication and the wiki notation Keiji used is
> >> not intended to be C but rather shorthand development of ideas.
> >> 
> >> If patches were submitted that contained no pow() and only ^ in place
> >> of pow() and this was questioned, maybe Keiji would have admitted that
> >> he either did not know (unlikely, hence everyone's outrage) or used
> >> the wiki for quicker notes where ^ might more traditionally, outside
> >> the programming world, be considered as a power raising operator.
> >> 
> >> Regardless, aside from being sure that people are reasonably competent
> >> with C through the qualification tasks and application process, I
> >> don't think this discussion has any real merit but to poke fun at
> >> something which is most likely a misunderstanding.
> >> 
> >
> >> Why jump on someone when nurturing them could lead to improvements in
> >> the project which is ultimately what we all want?
> >
> > true, it was not my intent to "jump on someone" at all, i just had to reply
> > to mans as he belived that there are some patches that could be merged and
> > i think there are not.
> 
> The question remains, what new code needs to be written to enable an
> lgpl swscale?  Is there anything other than the table generator?

Iam not aware of any other GPL code except the intermingled SIMD under
#ifdef HAVE_MMX and similar, and maybe some special converters in seperate
files that we do not need.
That said i did not backtraced where every line came from.



> 
> BTW, does the table generator need to be particularly optimised?  I
> would expect the answer to be no, but I've been wrong enough times
> about swscale, so I'd better ask.

no, the generator does not need to be particularly optimised, its
just run when building a context which generally should be once per
program execution for most applications (assuming no video size changes).

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Everything should be made as simple as possible, but not simpler.
-- Albert Einstein
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20080915/34abd829/attachment.pgp>