[FFmpeg-devel] git blame authorship

[Vladimir Pantelic]

Kostya wrote:
> On Wed, Feb 09, 2011 at 11:30:37AM +0100, Vladimir Pantelic wrote:
>>  Fran?ois Revol wrote:
>>  >
>>  >Le 9 f?vr. 2011 ? 11:24, Felipe Contreras a ?crit :
>>  >
>>  >>  On Sun, Feb 6, 2011 at 6:21 AM, Michael Niedermayer<michaelni at gmx.at>   wrote:
>>  >>>  Hi
>>  >>>
>>  >>>  A quick redone count of lines of authorship of my local tree (interrested
>>  >>>  parties can redo that with various other trees which should show lower numbers
>>  >>>  for me)
>>  >>>
>>  >>>  generated with stuff like:
>>  >>>  for i in *.asm ; do git blame -w -C -M $i ; done>ba
>>  >>>  and
>>  >>>  cat ba* | sed 's/[^(]*(\([^2]*\)2.*/\1/;s/  */ /g'  | sort | uniq -c | sort -n
>>  >>>
>>  >>>  That attempts to follow whitespace changes, copied and moved files
>>  >>
>>  >>  How convenient to leave .S files out. Here's the numbers with those:
>>  >
>>  >In any case, this is not authorship since the imported svn only knows about committers.
>>
>>  AFAIK, Janne went great length to attribute the author when doing the svn to git
>>  conversion, even I appear in git blame and I don't have/had commit rights...
>
> Nobody cares about really old commits. For example look at log messages for my
> first decoders (TSCC, Ultimotion, Indeo 2, JPEG-LS).

I'm not saying that I care, I just explained what was done. How to interpret all
that git blame data and what conclusions to draw from it and whether it "helps"
I do not dare judge upon...