Thank you for taking the time to do this.
On Sun, Jun 12, 2011 at 11:52, Richard Hartmann
wrote:
> * no use strict in validate_metadata?
How embarassing! Thank you for pointing that out. I have no idea how
that happened. I *always* start writing my perl scripts with 'use
strict; use warnings;' ... excuses excuses
> * validate_metadata uses -e, not -f to check of .metafile exists
Good point, fixed.
> * .metafile seems to reside in $GIT_WORK_TREE/.git/hooks, not in
> $GIT_WORK_TREE?
I'm not sure where you're seeing this. build_metafile blindly opens a
file relative to the directory 'git --rev-parse --show-toplevel'
reports. post-commit calls 'build_metafile .metafile' which will
create .metafile in the git repo toplevel directory.
> * post-commit sets $OLDPWD, but does not use it. Why?
It's a relic of the original project I scavanged from. This has been fixed.
> * post-merge seems to be buggy and a no-op?
post-merge is, or should be, a symlink to post-commit.
> * post-commit will always clobber all local data. This seems to be unsafe.
I'm not sure what you mean. .metafile will be overwritten with the
current information but since its being committed to the repository
the changes will always be accessible. Nothing else is affected.
--
Harley J Pig
___
vcs-home mailing list
vcs-home@lists.madduck.net
http://lists.madduck.net/listinfo/vcs-home
