[FFmpeg-devel] Discussion: Feature: Subtitle charenc detection

[Michael Niedermayer]

On Thu, Oct 23, 2014 at 04:18:29PM +0200, wm4 wrote:
> On Thu, 23 Oct 2014 16:12:52 +0200
> Nicolas George <george at nsup.org> wrote:
> 
> > Le duodi 2 brumaire, an CCXXIII, wm4 a écrit :
> > > I kind of disagree.
> > 
> > I kind of have gathered that.
> > 
> > >		      With your approach, all data gets (potentially)
> > > trashed on opening, and it's hard to change the encoding afterwards.
> > > You'd have to reopen the demuxer. And then you'd have to somehow cache
> > > all input data (but only when you open subs), unless you're fine with
> > > re-reading all data, etc....
> > 
> > I will ignore that irrelevant nonsense and all subsequent similar content,
> > as it only shows that you still have made no effort to understand the API I
> > propose as it should be used instead of trying to map it on your incorrect
> > assumptions.
> 
> I understand your API extremely well, and I'm telling you it's
> shortsighted.

I dont understand either side or the proposed API nearly well enough
to really argue about this but iam concerned that this fight could
lead to a deadlock again and noone implementing any API, which would
be a loss to all.

Do i understad correctly that the dispute here is about the following
scenario?
There is some input over a slow channel with interleaved subtitles
There are subtitles in the first 10min but they use almost only
characters which have a identical encoding in multiple char-encodings
thus making distingushing the encoding&charset impossible.
at 11 minutes into the movie theres a subtitle with text that allows
identifying the encoding & charset
We need to demux, decode and display the subtitles before we have
enough information to finally know the exact encoding/charset and
once known update/set it.
?

and wm4s concern is that this would not work in the API that Nicolas
envissions, while Nicolas belives theres no problem and that will
just work fine once the API is finished ?

Or maybe iam totally wrong.
But i think generally and IMHO
Whoever implements the API should decide how its implemented.
The oppinions, concerns and wishes of API users should be considered
and supported, and sometimes its easier and less work to support
something than to argue about it being useful.
API users are important

This also reminds me a bit about AVFMT_FLAG_NOFILLIN, which ive always
had the feeling is not really usefull but many people wanted it, its
quite possible i was wrong and it is usefull for some.


Thanks

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

Awnsering whenever a program halts or runs forever is
On a turing machine, in general impossible (turings halting problem).
On any real computer, always possible as a real computer has a finite number
of states N, and will either halt in less than N cycles or never halt.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20141023/649cc34b/attachment.asc>