Hi Thiago, On Mon, Jun 13, 2011 at 12:46 PM, Thiago Macieira <thi...@kde.org> wrote: > If we removed the commit authors and commit messages and removed the extra > file > paths that were never released to the public, it would be 95% ok. But who > would do the other 5% of the work? Does anyone want to sign an NDA and work > for free reviewing 131k commit? > > $ git rev-list qt-start..public-4.5-start public-4.6-start | wc -l > 131215 >
I understand that we have a problem trying to filter out all these commits. What if we restrict ourself to just opening src/*, rest of it is not very interesting anyway. If it's a lot, maybe just corelib/gui/sql/svg/xml or similar. Just handpick modules and increase them as we go. Note that I don't agree with removing commit messages; we should just change them to remove any sensitive info. Without commit messages, the history is not very useful. For the sake of argument (so feel free to ignore this :-) ), if it's around 50k commits and we get say 50 trolls to sign up for this project, it's about 1000 commits to review each. If they review 10 commits a day, we can open the repo, in say, 5 months ;) That's not too bad, is it. Of course, more the trolls who can volunteer, the lesser the time and less work for all. For curiosity sake, how did you guys handle scanning of all the files when the git repo was opened? Or even the tests/*? Girish _______________________________________________ Opengov mailing list Opengov@qt-labs.org http://lists.qt-labs.org/listinfo/opengov
