On 29 July 2011 09:46, Joerg-Cyril.Hoehle@t-systems.com wrote:
these fixes all trivial errors I mentioned recently but one, plus a few ones I discovered since then.
These need to be separate patches, of course. I'd also encourage everyone to just use git send-email for sending patches.
Henri Verbeet wrote:
On 29 July 2011 09:46, Joerg-Cyril.Hoehle@t-systems.com wrote:
these fixes all trivial errors I mentioned recently but one, plus a few ones I discovered since then.
These need to be separate patches, of course. I'd also encourage everyone to just use git send-email for sending patches.
As always a categorical "It depends!". git send-email is not for everybody.
If you're new to (Wine) development or you do a lot of janitorial work I would actually recommend to *avoid* git send-email. See it as an opportunity to look again at your patch when you manually submit it. Like a last line of defense to spare you from embarrassment. Git is powerful and it is kinda cool that you can efficiently submit the 7th version of a 20 patch series on the same day! But that doesn't means it is a great idea to actually do it; you'll only get yourself graylisted by Alexandre.
bye michael
P.S.: "You" refers to the interested reader and not to Henri.
On 29 July 2011 16:38, Michael Stefaniuc mstefani@redhat.com wrote:
If you're new to (Wine) development or you do a lot of janitorial work I would actually recommend to *avoid* git send-email. See it as an opportunity to look again at your patch when you manually submit it.
I think that's orthogonal. I certainly look patches over one last time after generating them with format-patch, before actually sending them with send-email. (And yes, it helps.)
If you happen to already have a setup that works well for you, sure, keep using that. However, if you don't, I'd argue that it's well worth the time (also for the people trying to review the patches) to just learn how git send-email works. *Especially* for new contributors it just avoids stuff like wrapping, wrong character sets, wrong mime types, forgetting to attach the actual patch, sending to the wrong list, messing up the numbering, etc. (And of course common sense still applies. I.e., try with --dry-run, try sending some patches to yourself first to see what the various options actually do, etc.)
Like a last line of defense to spare you from embarrassment. Git is powerful and it is kinda cool that you can efficiently submit the 7th version of a 20 patch series on the same day! But that doesn't means it is a great idea to actually do it; you'll only get yourself graylisted by Alexandre.
Similar to above, I don't think this has much to do with what you're using to send the patches.
-
Henri Verbeet -
Michael Stefaniuc