Ok
we can do that with my package. We should tweak it a bit. Now today  
is full of lectures (as yesterday was) for me and after
I'm dead ;) But I will check what we can do. If someone wants to help  
please say it.

Stef

On 22 sept. 06, at 04:49, Andreas Raab wrote:

> stephane ducasse wrote:
>> Andreas I should be able to produce the same list for 3.7, 3.8 and  
>> 3.9 (I'm just flooded and dead tired...imagine sleeping at 9 :))
>> Is it enough to get the initial and numbers of change?
>> Do you count the author of the last version only?
>
> Yes, that's probably fine. The stats I sent also included the  
> intermediate versions but that was a byproduct of the method of  
> counting them (via analyzing the change sets that went into the  
> system).
>
> Also, don't forget that the goal is not to establish "who wrote how  
> much of what" but rather to have mapping from initials to  
> individuals. I only sent these stats as a starting point and to  
> emphasize that certain initials are more important to figure out  
> than others. For example, if you don't know who goes by "len" it is  
> important since there is a big contribution (900+ methods) in 2.7.  
> Meaning some significant feature was added which was written by  
> this person (happens to be the Speech synthesis by Luciano  
> Notarfrancesco) and looking at 2.7 we could probably figure out  
> what it was and what to do about it even if that identity couldn't  
> be established. Being able to draw these conclusion is worthwhile  
> because it also allows us to deal more effectively with small-scale  
> contributors (there are about 100 people with less than 10  
> contributions in my stats). If you can't figure out the identity of  
> "m3r" for example (having a single contribution in Squeak 2.3) it's  
> probably not quite as important as figuring out who "len" is.
>
> Cheers,
>   - Andreas
>