On 2016/01/22 13:49, Nigel Taylor wrote: > On 01/22/16 12:32, Stuart Henderson wrote: > > I'm running amavisd/spamassassin with the updated IO::Socket::IP > > with no problems. > > > > Here are all the hits from a search of .pm and .pl files from ports > > source unpacked at the end of August - (require|use).*IO::Socket::IP. > > We could add deps to these pretty easily. > > > True but it hadn't been done from what I could see. > > Now just need to check all the scripts in base, that might have started > using IO::Socket::IP - yes I know probably none.
Is anybody going to go that far (with all of the modules..) when we update Perl after unlock? I think it's unlikely. We'll do builds and some people will test things that they particularly care about but I can't see anything more than that happening as a deliberate effort. (I don't recall anybody doing that when we moved from the previous ports version of IO::Socket::IP either, which was a bigger change (moving *back* from 0.34 to 0.29).. There are some areas in cpan where I'd be very wary of doing this without more extensive testing (especially when a set of modules are fairly tightly bound together) but this doesn't seem to me like one of those situations? > *.pl/ *.pm doesn't mean there aren't just plain filenames like corelist, > get_iplayer I'll run a full search then. (I don't suppose any Go users fancy figuring out what's up with codesearch on OpenBSD do they? ;-)