Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!

Bug 55679

Summary: Request for /etc/portage/package.features
Product: Portage Development Reporter: Christian Bock <christian.bock>
Component: Enhancement/Feature RequestsAssignee: Portage team <dev-portage>
Status: RESOLVED DUPLICATE    
Severity: enhancement CC: jonas.baehr, radek, sascha-gentoo-bugzilla, solar
Priority: High    
Version: unspecified   
Hardware: All   
OS: Linux   
Whiteboard:
Package list:
Runtime testing required: ---

Description Christian Bock 2004-06-30 07:09:53 UTC
There a really good reason for this. Lets think about ccache.. its useless if you get a few gig of cache, finding will get even slower than compiling w/a searching. Lets think of cvs-ebuilds, 'app-misc/krusader-cvs ccache' in a file called package.features makes sense. I don't need ccache for xorg, which I am building exactly once, so almost everyone will deactivate it globally... Please consider adding this one. Regards, Christian Bock

Reproducible: Always
Steps to Reproduce:
1.
2.
3.
Comment 1 Jonas Bähr 2004-06-30 09:28:14 UTC
I already proposed that (see http://bugs.gentoo.org/show_bug.cgi?id=51023 )
It was rejected with the reason that features should be global and I should use /etc/portage/bashrc (but I've no idea how this solves my problem).
In my eyes, ccache is something that does not make sence to apply global but it's defenitly a feature. As Christian said, ccache should speed up the build-process. In order to do so, the cache has to be small; ergo it it only usefull if applied only to a few packages which are often rebuild (like cvs-ebuilds or other software with a frequent release-schedule).

I think that packages.features is nessesary.
Comment 2 SpanKY gentoo-dev 2004-08-02 19:28:57 UTC
i dont think we should have a package.features, i think we should have a package.env

then we wouldnt have to have a specific file for every single env variable that could possibly affect an emerge

*** This bug has been marked as a duplicate of 51023 ***