Re: Not going beyond symbolic links

Hi,

On Mon, 4 Aug 2008, Linus Torvalds wrote:

On Mon, 4 Aug 2008, Junio C Hamano wrote:

quoted

quoted

The rewritten 'has_symlink_leading_path()' should do ok, but it 
migth still be a huge performance downside to check all the paths 
for things like "git add -u".

Not yet.

I think this is a necessary "correctness" thing to do regardless of 
the performance impact, and adding the logic to stop at submodule 
boundary (aka gitlinks) should come before optimization.

Well, "performance" is a feature too, and it's not correct to say that 
"X should be fixed before optimization". If "X" slows things down, the 
question should be whether it really needs fixing..

Yes, we find symlinks when we do _new_ files, but is it really so bad to 
assume that existing directories that we have already added to the index 
are stable? It can easily be seen as a feature too that you can force git 
to ignore the symlink and see it as a real directory.

Actually, whatever you want, it needs fixing.

I vividly remember being quite pissed by Git replacing a symbolic link in 
my working directory with a directory, and instead of updating the files 
which were technically outside of the repository, Git populated that newly 
created directory.

However, please note that Junio's patch affects git-add, AFAIR, not 
git-update-index.

Ciao,
Dscho