[theora] Indexing Ogg files for faster seeking

[Chris Pearce]

On 11/15/2009 4:39 AM, Jason Self wrote:
>>> Okay: I managed to find a file that results in:
>>>
>>> WARNING: Recorded more keyframes than there's room for in the index. 
>>> We must have miscalculated index size. Not writing index.
>>>
>>> ...but the input file is almost 3.4GB. I'm sure you don't want a 
>>> file that large, so I'm using dd to create a smaller version. A 
>>> small 10MB version didn't report this problem so it looks like I 
>>> need to experiment to find a size that reports this problem, while 
>>> also not being the entire file. More soon (hopefully, depending on 
>>> how long this process takes.)
>
> On Nov 11, 2009, at 7:26 PM, Chris Pearce wrote:
>
>> Hi Jason,
>>
>> Thanks for trying to reproduce this. I have spotted a potential cause 
>> for this problem, and I've attached a patch... Would you be able to 
>> apply the patch to your local tree and see if the error still happens?
>
> Sorry for the delay; work has kept me very busy. I've applied your 
> patch and it does seem to have solved it. I'll continue testing with 
> other files and let you know if anything else pops up.
Great! Thanks finding that and testing my patch Jason.

J: Would you be able to take the attached patch in ffmpeg2theora?

Thanks,
Chris P.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: index-record-sample.patch
Url: http://lists.xiph.org/pipermail/theora/attachments/20091115/3e30f85b/attachment.txt