Compiles and seams to work fine on amd64 here.
as you can see pycairo depends of >=cairo-0.9.0 and this version is hard masked. So we need to wait until >=cairo-0.9.0 is unmasked to mark pycairo for Testing. Thanks for the Report
Yes, I can clearly see that. However, other ebuilds contain ~amd64 even though they are hard masked. A package being hard masked should not prevent a keyword from being added to indicate that it will work on a particular platform when it is unmasked, imo.
I will keyword it when all deps are unmasked.
I agree with Morgan here. For testing purposes, people tend to edit packages.mask or packages.unmask (not withstanding this is not something everyone should do) and try out new versions of software. A good recent example is Gnome 2.11.92. A lot of people in the Gentoo community already tried this pre-release and they SHOULD too, because this is how bugs get filtered out of the beta's. This topic: http://forums.gentoo.org/viewtopic-t-364794.html has had almost 10.000 views and over a hundred posts. People are interested in trying out new software, and pycairo is such a thing. All deps seem to have the ~amd64 flag, so why make it harder to try out this software and iron out the bugs by refusing to add ~amd64 to the ebuild? If something is hard-masked and flagged unstable, it should scare away the people-who-don't-know. The people who do know show test software. That's what they're for ;-) This is just my
I agree with Morgan here. For testing purposes, people tend to edit packages.mask or packages.unmask (not withstanding this is not something everyone should do) and try out new versions of software. A good recent example is Gnome 2.11.92. A lot of people in the Gentoo community already tried this pre-release and they SHOULD too, because this is how bugs get filtered out of the beta's. This topic: http://forums.gentoo.org/viewtopic-t-364794.html has had almost 10.000 views and over a hundred posts. People are interested in trying out new software, and pycairo is such a thing. All deps seem to have the ~amd64 flag, so why make it harder to try out this software and iron out the bugs by refusing to add ~amd64 to the ebuild? If something is hard-masked and flagged unstable, it should scare away the people-who-don't-know. The people who do know show test software. That's what they're for ;-) This is just my 2/100, of course.