[FFmpeg-devel] [PATCH] avio API cleanup II

[Michael Niedermayer]

On Sun, Feb 20, 2011 at 01:10:33AM +0200, Anssi Hannula wrote:
> On 19.02.2011 22:41, Michael Niedermayer wrote:
> > On Sat, Feb 19, 2011 at 07:46:46PM +0100, Anton Khirnov wrote:
> >> On Sat, Feb 19, 2011 at 04:01:22PM +0100, Michael Niedermayer wrote:
> >>> On Sat, Feb 19, 2011 at 12:42:44PM +0100, Anton Khirnov wrote:
> >>>> Hi,
> >>>> due to BBB being lazy, here's the updated batch of renames.
> >>>>
> >>>> The basic idea is that functions working on AVIOContext (formerly
> >>>> ByteIOContext) get an avio_ prefix, while functions working on
> >>>> URLContext (to be renamed to AVURLContext) get an av_url_ prefix.
> >>>>
> >>>> Comments welcome.
> >>>
> >>> You are making hundreads of internal functions part of the public API
> >>> The correct prefix is ff not av for quite a few i think
> >>> That said if you make things part of the public API have you thought about
> >>> the structs being ready for this? Can they be extended without breaking ABI?
> >>
> >> They aren't internal, being in an installed header makes them de facto
> >> public. As the attached log (huge thanks to Flameeyes for making it) shows,
> >> many/most of those are actually used by other projects.
> >>
> >>>
> >>> old team: technical reviews
> >>> new team: political reviews because they plain dont know they code for which
> >>>           they review, the people knowing have been kicked out
> >>>
> >>
> >> If you have some specific functions that should be kept private in mind,
> >> please say which ones. But keep your insults to yourself.
> > 
> > If you feel insulted by a comparission of past and present reviews and code
> > quality then maybe you should look into improving the present and not ask
> > the messenger to shutup.
> > 
> > about flameeyes list, id love to see the actual cases where some things are used
> > ill start with xbmc:
> >    virtual void put_byte(ByteIOContext *s, int b) { ::put_byte(s, b); }
> >    virtual void put_buffer(ByteIOContext *s, const unsigned char *buf, int size) { ::put_buffer(s, buf, size); }
> >    virtual void put_be32(ByteIOContext *s, unsigned int val) { ::put_be32(s, val); }
> >    virtual void put_be16(ByteIOContext *s, unsigned int val) { ::put_be16(s, val); }
> > ...
> > 
> > they are just some odd wrapers that are unused
> 
> Those four are used in
> xbmc/cores/dvdplayer/DVDCodecs/Video/DVDVideoCodecVDA.cpp.

This code is copied out of libavformat/avc.c and converted to C++
I dont think we should base our decission on what to make public on that.


[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

There seems to be only one solution to NIH syndrom, ... a shooting squad
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20110220/d39d363e/attachment.pgp>