| 1 |
On Saturday 07 August 2004 20:07, Chris Bainbridge wrote: |
| 2 |
> On Saturday 07 August 2004 00:57, Jason Stubbs wrote: |
| 3 |
> > Even with all that, a full dependency tree is built with all circular |
| 4 |
> > dependencies resolved in about 6 seconds, where currently emerge takes |
| 5 |
> > about 2 seconds. I can't see a way to get USE-based SLOTs in without at |
| 6 |
> > least doubling that figure (which is incidentally the time that 2.0.50 |
| 7 |
> > took). I'm open to suggestions on how to get around it. |
| 8 |
> |
| 9 |
> Maybe I'm not understanding this fully, but couldn't you just compute the |
| 10 |
> dependency tree once and then cache it? And allow quick updates to the |
| 11 |
> cached info by only editing dependency links for a single ebuild when it |
| 12 |
> changes. You'd need to cache the dependencies for different sets of use |
| 13 |
> flags as well. |
| 14 |
|
| 15 |
The dependencies of a single package with a single set of conditions don't |
| 16 |
take long to calculate. The issue is that, real soon, packages are going to |
| 17 |
be able to specify some of those conditions for other packages. Figuring out |
| 18 |
the best combination of packages/conditions and whether it is indeed possible |
| 19 |
(ie. no conflicts) is what takes the time. |
| 20 |
|
| 21 |
Regards, |
| 22 |
Jason Stubbs |
| 23 |
|
| 24 |
-- |
| 25 |
gentoo-dev@g.o mailing list |
Archives