I completely agree that tests should not be disabled. On the other hand, very few of the fink developers have any experience with any of the perl testing stuff. This makes it hard to even know how to report a failure.
My saga yesterday evening started when -- in the interests of trying out somebody's recent change to the code -- I attempted to install the fink package manager from CVS HEAD. Problem #1 was that the most recent commit had a bug which kept stuff from compiling. So I backtracked one step in CVS, and then ran into problems with the tests in failure.t . The testing suite has now been fully integrated into fink, so that (for example) anyone trying to install from CVS HEAD will be unable to do so unless "make test" succeeds. (And by the way, why are we always finding that there is "1 unexpected success" during testing? Shouldn't an "unexpected success" count as a failure?) That's all fine and good, but creates problems for working on the code if there is a bug in the testing suite that I don't understand in the least. I checked on IRC, and most folks did not experience the bug but a few of us did. I find the failure output from "make test" to be rather cryptic, and I don't even know what would constitute a useful bug report. My personal workaround was to comment out the "make test" line from fink.info.in, allowing the new fink to be installed to that I could continue to work. As Ben Reed pointed out to me on IRC, the failure was happening in the new Command.pm module which isn't even being used in fink yet. -- Dave ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ Fink-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fink-devel
