Merlyn pointed out, again, that the version numbering for modules in SPOPS were retrograding with version increases. The problem was that I used a crappy CVS Revision formatter which treated '2.7' as '2.70' rather than '2.07'. Thanks to the CPAN testers, I also found that one of my tests failed on 5.8 because I was (one step removed) relying on hash key order. So 0.67 and 0.68 went out in short order, just after releasing 0.66 earlier this week. Doh!
You still gotta roll a major revision at least once each 100 minor revisions though. {grin}$VERSION = do { my @r = (q$Revision: 2.21 $ =~/\d+/g); sprintf "%d."."%02d" x $#r, @r }; # must be all one line, for MakeMaker
Re:For those playing at home
Dom2 on 2002-08-29T14:12:01
Bleargh. I like to keep the VERSION numbers to myself, and specify the module version number as a whole entity in the Makefile.PL.
-Dom
Re:For those playing at home
merlyn on 2002-08-29T17:56:19
If you don't mind a lot of manual work where automation can help, go right ahead.Me, I've got better things to do with my time.
Re:For those playing at home
Dom2 on 2002-08-29T22:41:41
It's not so much that, it's the idea that CVS/RCS revision id's aren't necessarily the best things to use as the version for your module (perhaps distribution would be a better term than module) as a whole anyway. I'd rather bump the version number when I want to, not when I correct a typo in the POD.
I've got a full on woody for automation everywhere else.
:-) -Dom
