http://bugzilla.spamassassin.org/show_bug.cgi?id=3599
------- Additional Comments From [EMAIL PROTECTED] 2004-07-28 13:11 ------- +1 on 2192. '2183 adds a test which makes a compilation fail if there are any warnings in spamc -- so for the normal build it will just warn until you run "make test" which will then fail at that test. The rationale behind this is that most people simply ignore compiler warnings these days because they are so common in other projects.' BTW, I'm -1 on this idea. It's *great* for us developers to get warnings, but end-users shouldn't have to care -- occasionally some OSes will be so far out there that they will produce a warning for something very minor, and a build failure for that will result in a lot of aggravation for people on those OSes, and us in terms of bugs reported as a result. All *we* have to do is watch out for warnings. And if we want to use a "fail on warning" switch, we can do that ourselves by editing the Makefile or passing in Makefile.PL switches. (Don't forget: the PRSPAMD bug had nothing to do with a bug that would have been caught by a warning. Instead it was caused by the incorrect workaround to *avoid* a warning ;) I am also -1 on adding 2183 and 2184. It's far, *far* too late in the 3.0.0 cycle to go changing that stuff heavily. 2183 in particular looks like a portability nightmare, which I don't want added to our test suite days before a major release.... ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
