On Jun 1, 2012, at 5:41 PM, George Bosilca wrote:
> So basically to achieve the same goal the change moved from a fictitious
> email address to our real one. I agree with you that both the old and the new
> way allows one to know exactly who was the committer. But as you highlighted
> in your email, the commit messages are intended for a well informed audience,
> people who know who did the commit based on the snv_id.
Pop quiz: who's vvenkatesan?
(no cheating to look them up)
> Btw, the commit emails __are__ archived on the web, they appear on ohloh,
> cia.vc and so on. I would be careful on what information we disclose on them.
I thought they only archived the *SVN commit messages* -- which are quite
different than the emails that we sent out. The SVN commit messages only
contain the SVN username, not any of the extra stuff that we put in the emails.
E.g.:
http://cia.vc/stats/project/OMPI
Can you show an example of someone web archiving our commit emails?
Let's also not forget that hg and git include email addresses in the commit
messages. So it's a pretty common practice. The world already knows your
email address, George. :-)
Don't get me wrong -- if you hate this, I will take it out. But I do want to
clearly enunciate what this *is* and *is not* before making a decision to take
it out (that may have been formed based on incorrect information/assumptions).
--
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to:
http://www.cisco.com/web/about/doing_business/legal/cri/
