Hi,
[nice to go to sleep and see a patch in the morning]
On Thu, 5 Jun 2008, Victor Bogado da Silva Lins wrote:
Please imitate what you see on this list, especially by the Git regulars.
They never include these headers.
This is a bit vague, a bit wrong, and it has the subjective "I" in it.
remote_get() _is_ assumed to return NULL when no remote with that name was
found.
And the function was fixed (3rd person, passive). At least that is how I
remember most of the rest of git.git's commit messages.
Hmm. That sounds very dangerous, using a pointer without checking that it
is NULL.
Besides, I think that the add_url_alias(ret, name) above is at
fault. Not the alias part about it, though. This used to be an add_url()
from the conception of remote.c, but then it was only used for fetching.
I think the proper solution is to add a parameter, or a function, to force
NULL if the remote did not exist.
Of course, this involves a careful review of the existing callers, to find
out which ones rely on getting a newly created remote when none was found.
In any case, thank you very much for finding that bug (the if() I pointed
out was in remote.c from the very beginning, even then never being
triggered).
Ciao,
Dscho
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html