[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gnash-dev] revno.h bashing

From: Rob Savoye
Subject: [Gnash-dev] revno.h bashing
Date: Thu, 09 Dec 2010 15:38:21 -0700
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv: Gecko/20101027 Fedora/3.1.6-1.fc13 Lightning/1.0b2 Thunderbird/3.1.6

Email from gnu,org seems a bit behind today... So this may overlap with
other messages...

Ok, I've fixed how revno.h gets generated yet again... I'm not 100% sure
what you were trying to do, but the way it is now revno.h only gets
rebuilt after a git commit or git pull, since those are the only two
actions that change the revision data. I believe this was the original
intent, to not rebuild revno.h unless it had to be. To try to avoid
argument, here is what I changed and why.

revno.h goes in the *build* tree, so it works with NFS mounted
filesystems that may be read-only.

revno.h only gets regenerated on a git action that changes the repository.

In configure, I use AC_SUBST(GIT) to put the value of git in the
Makefile, rather than an automake conditional, as this allows for the
default to using the date.

If .git or git itself isn't found, it defaults to the date. This way
"make deb" works from a source tarball + debian files, and not just from
a git tree. Default values are used in this case for the branch and
comm_id, so nothing else breaks.

Any changes to how the revno,h target works *must* be discussed after
this so we're all on the same wavelength so we can stop stomping on each
others toes. Discussing it on irc when I'm asleep doesn't count, I'm 8
hours behind you, remember ?

        - rob -

reply via email to

[Prev in Thread] Current Thread [Next in Thread]