On Tuesday, July 20, 2010 15:45:22 François Revol wrote: > Le 20 juil. 2010 à 21:26, Mike Frysinger a écrit : > >>> instead of "fighting the system" and ultimately doing it wrong, use the > >>> tools already provided since they are known to do it right. > >> > >> No, it does it how some people decided was right, but well. > >> I'll just have to git reset, fix, recommit and try to get git to send > >> mails instead of just using the native app for this and having it give > >> me a diff. > > > > when working with git trees, following the standard git behavior keeps > > everyone on the same page and flowing smoother. when people start trying > > to do their own thing, it slows down everyone else. this isnt a matter > > of taste, it's simply the way it works in reality. > > Well, it might work for regular contributors who know the project enough, > but for people just trying to get something fixed it's just painful, > because they never know if their first try will be accepted, and if they > just post asking how they should do they usually get answers like "just > send a patch". Then someone just says "oh the commit log is too short" > (but then if you make stats you see 80% of the project's commit logs are > <= 2 lines (happened to me on QEMU)) And as I said, trying to do it right > in git is painful. the feedback i gave here i would have given for any project using git and e- mail as their main form of development -mike
Attachment:
signature.asc
Description: This is a digitally signed message part.
