On Nov 1, 2011, at 19:31, Jeremy Lavergne wrote: > Since we're starting to be able to unify our ports, do we want to revisit how > we're calculating the quantity of ports available? > > I'm curious if we'd like to move away from saying "py25-distribute" and > "py27-distribute" are different ports. > > If the db schema supports it, we could probably count the distinct portfiles > listed rather than the number of entries.
It sounds like you're saying we should count all of a port's subports as a single port. While I agree that might be helpful for the python and perl portgroups, those portgroups are only the beginning of the possible uses of subports. For example, there are currently separate ports for graphviz, graphviz-gui and gvedit, but they all come from the same distfile so I feel they're ideal candidates to consolidate into a single portfile using subports; that'll let me consolidate the livecheck information in one port instead of having three ports tell me they need to be updated to the same version, and will mean I only have to update the version and checksums in one place. This will be even more helpful in the php5 port; currently when I update php5 I have to update the php5 port and 33 other ports, one for each bundled extension; consolidating these will mean I only have to livecheck and update one port instead of 34. But as far as the user should be concerned they are all separate ports. _______________________________________________ macports-dev mailing list [email protected] http://lists.macosforge.org/mailman/listinfo.cgi/macports-dev
