+1: rjung
rpluem says: Now t/security/CVE-2005-3352.t fails. Not sure if this is a real
regression or if just the test is wrong, but this should be investigated.
+ rjung: The test sends a Referer '">http://fish/'.
+ The original code returns '<a href="http://IP/">http://fish/">'
+ The patched code returns '<a href="http://IP/%22%3ehttp://fish/">'
+ This seems to be even better IMHO. 2.4 also returns the percent encoded
+ variant, so the test should fail there as well.
PATCHES/ISSUES THAT ARE STALLED