Randal L. Schwartz wrote:
[Move to eval/die.] No leak there.
I'm not an experienced Perl hacker, but I intuitively liked the throw/catch method better than the eval/die method.
If I read http://www.nntp.perl.org/group/perl.perl5.porters/2006/03/msg110331.html correctly, this has been fixed in recent 5.8 versions, so it would really only affect 5.6. But perhaps we don't care about 5.6. ;-) People who run into serious memory issues with 5.6 will have to upgrade, and the existing functionality that programs like git-send-email or git-svn rely on seems to work fine with the memory leaks, so using throw/catch further probably won't cause any regressions for them.
I'm honestly not too keen on sacrificing time (or code prettiness) on 5.6 compatibility, so if there are no reasons besides the memory leak to move away from throw/catch, perhaps we can just keep using it?
-- Lea -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
