William A. Rowe, Jr. wrote: > On that note, it sounds like all of the bucket, ssl, proxy and include > changes > work together, and everything is nearly healthy again? If so, I'll > simply push > .34 to head tommorow, incorporating these changes [none appear to be > trouble], and let Greg Ames start playing it against daedalus. > > Any observations before I do so?
Just make sure you have the latest CGI changes from today (mod_cgi.c and util_script.c); the version of mod_cgi in the original CGI tag had a bad bug (returning a 200 when a script failed). > Are remaining segfaults in includes the > exception [bad script] or of a more general nature? I *think* the remaining include segfaults are rare, based on Paul R's description of his test case that's failing. --Brian