On 2012-5-11 03:29 , Bradley Giesbrecht wrote: > On May 10, 2012, at 9:49 AM, Ryan Schmidt wrote: > >> >> On May 10, 2012, at 11:46, Bradley Giesbrecht wrote: >> >>> $ >>> /opt/local/var/macports/sources/svn.macports.org/trunk/base/portmgr/jobs/port_binary_distributable.tcl >>> -v gcc46 >>> gcc46's dependency 'cloog' has license 'unknown' which is not known to be >>> distributable >>> >>> Is it intended that this script evaluates the cloog license even though >>> cloog is a non-default variant? >> >> Your ports tree is out of date. The cloog variant was removed. The recent >> versions of the gcc ports now always use cloog. > > Ok. > > My reading of the comments for mportlookup is that mportlookup uses the > PortIndex from my port tree and in my port tree cloog is a non-default. > > macports.tcl:2586: > # Returns the PortInfo for a single named port. The info comes from the > # PortIndex, and name matching is case-insensitive. Unlike mportsearch, only > # the first match is returned, but the return format is otherwise identical. > # The advantage is that mportlookup is much faster than mportsearch, due to > # the use of the quick index. > > > So it appears my question remains unanswered or I am not reading this > correctly.
The behaviour of mportlookup isn't particularly relevant given that both port rdeps and port_binary_distributable.tcl go on to open the portfile. I have no idea how things are set up on your machine, but if I had to guess I'd say the macports installation linked in /Library/Tcl and used by port_binary_distributable.tcl probably isn't the same as the one you get when you run 'port'. - Josh _______________________________________________ macports-dev mailing list [email protected] http://lists.macosforge.org/mailman/listinfo.cgi/macports-dev
