Re: Perf patches, libmpg123 patches, and a new release of LAME?

Am Tue, 05 May 2020 16:17:48 +0200
schrieb Alexander Leidinger via Lame-dev <[email protected]>: 

> There is also a patch for a faster CRC routine:
>     https://hydrogenaud.io/index.php?topic=115900.0
> Anyone up to have a look at this?

Btw. I'd happily accept a patch to libmpg123 for optionally checking
the CRC values Lame produces … so far, it only knows how to skip
them and rather deals with invalid data inside the decoder. Fuzzing the
decoder would be a bit more involved otherwise, when you have to
crack/fake the CRC first;-)

But to the actual discussion, I want to have the analyzer decision
clarified:

> 2. libmpg123:
> Also in terms of libmpg123 and the API for the analyzer... as nobody  
> seems to have stood up into looking into the analyzer part, it doesn't  
> seem to be that interesting anymore. So I can imagine two reactions to  
> this, either remove the analyzer part (given that it uses gtk1 maybe  
> the preferred solution?), or to commit the switch to libmpg123  

So, if you integrate the libmpg123 patch, you'll keep the analyzer
around, as it still builds given an Gtk1 install and could be ported to
a newer toolkit. If not, you'll remove it instead?

> (assuming there will be a release of libmpg123 with this API close  
> enough to the decission which way to go).

The respective API is scheduled to appear in mpg123 1.26rc1 this week,
in hope that _someone_ makes use of it. After all these years, now
would be a last chance to suggest changes. I might update the lame
patch to simplify libmpg123 usage a bit with mpg123_open_fixed(), which
avoids clients having to deal with the possibility of concatenations of
differing MPEG streams.


Alrighty then,

Thomas


_______________________________________________
Lame-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lame-dev