making emerge faster

[Wolfgang Rohdewald]

On Freitag 14 Januar 2011, Patrick von Reth wrote:
> hm true, but we call Python sub instances multiple times and I
> have no idea how to cache it for the parent processes.

don"t. Just cache it within one process. When I do
emerge --fetch kdelibs
I get hundreds of 

emerge debug: GCC Target Processor:i686-w64-mingw32

scrolling by. That is what I want to eliminate.
And it seems to work here.

meanwhile I got rid of almost all Python error messages and
warnings but I did not commit yet anything because I now need
to test emerge a lot. I would not have wanted to do that for
every single commit. When I feel confident I will have to
rearrange my patches because a few fix typos I did before,
such that no single patch should break anything. Still I
think I should push all patches at the same time. Right
now I have 71 single commits in my local git. They do
fix some bugs but most of them in rarely used code paths.

actually I told pylint to just not warn about many things
to make this manageable. They can be re-enabled step by
step later.

there is still one problem with things like configure()
overriding each other but having very different
signatures. I have ideas about making that easier to
understand, but later.

(emerge kdelibs just finished without faults)

-- 
Wolfgang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.kde.org/pipermail/kde-windows/attachments/20110114/92aac62f/attachment-0001.htm