[PATCH 4 of 4] revlog: separate methods for deltas and key frames compression/decompression

[Durham Goode]

On 9/26/13 3:32 PM, "Wojciech Lopata" <lopek at fb.com> wrote:

>> Do you have numbers showing what benefits an extension could get by
>>doing key frame and delta compression differently?
>
>This vary. The biggest gain of being able to apply different compressions
>for key frames and deltas I've seen for now was additional 5% manifest
>size reduction. I believe gain grows as repository grows, but I haven't
>checked what the gain is on big repositories, since with my draft, pure
>python implementation such conversion would take days.


But why can we not get the same gain from compressing both key frames and
deltas the same?  Why would we ever want to apply some compression to a
key frame, but not to a delta?