[oi-dev] gcc45 port to Hipster 2016.04 oi-userland: Why do I get these pkglint errors although all deps are installed fine?

[Мартин Бохниг]

Here the bottom of build/manifest-i386-gcc-45.depend.res :

depend fmri=developer/gcc-3 at 3.4.3-0.175 type=optional
depend fmri=pkg:/SUNWcs at 0.5.11-2015.0.2.15626 type=require
depend fmri=pkg:/developer/gnu-binutils at 2.25.1-2015.0.2.1 type=require
depend fmri=pkg:/developer/library/lint at 0.5.11-2015.0.2.15626 type=require
depend fmri=pkg:/developer/linker at 0.5.11-2015.0.2.15626 type=require
depend fmri=pkg:/library/gmp at 5.0.5-2015.0.2.0 type=require
depend fmri=pkg:/library/mpc at 1.0.1-2015.0.2.0 type=require
depend fmri=pkg:/library/mpfr at 3.1.2-2015.0.2.0 type=require
depend fmri=pkg:/system/header type=require
depend fmri=pkg:/system/library/gcc-45-runtime at 4.5.2-2016.0.1.0 type=require
depend fmri=pkg:/system/library at 0.5.11-2015.0.2.15626 type=require


Of course the gcc45 runtime is not yet in the repo nor installed via pkg (only tar'ed over).

But pkglint didn't even falsely complain about missing itself, but about other packages which *are* installed, *are* in the Hipster-2015 repo and do get used fine by other gcc compilers (4.4.4 and 4.9).


I propose I leave it like that and you can later fix it on your own devel machine.
This IPS keeps killing my really really good will.
Although at least the severe performance problems are now gone after I fixed the publisher setup (this was indeed my own mistake).
On the other side they could have added some basic logic to warn a user about illegal publisher setups (given that they prevent and boycott a million other things [IPS I mean]).


best regards,
%martin bochnig



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openindiana.org/pipermail/oi-dev/attachments/20160927/e61cf2db/attachment-0005.html>