Code review comment for lp:~jameinel/bzr/2.0-490228-gc-segfault

I would be curious to know how long you and other people involved have spent on this bug
and compare that to the time necessary to get a test infrastructure good enough to
write a test reproducing this bug...

I'm more and more convinced that each time someone says: "I don't have time to write a test"
he is just lying to himself :-(

Given that you will likely be the one writing this test in the end anyway, I'm certainly
not throwing stones as I realize that we got this code without tests to start with.
So, let's land this patch :-D

Do you believe that this bug is present upstream or that it was introduced during
the adaptation to pyrex ?