| 1 |
On Thu, Oct 16, 2008 at 10:01:40PM +0100, Steve Long wrote: |
| 2 |
> Ciaran McCreesh wrote: |
| 3 |
> > On Wed, 15 Oct 2008 20:28:43 +0100 |
| 4 |
> > Steve Long <slong@××××××××××××××××××.uk> wrote: |
| 5 |
> >> Fernando J. Pereda wrote: |
| 6 |
> >> > A big gain in the context of ebuilds and source packages. Well done. |
| 7 |
> >> > |
| 8 |
> >> Yes, almost as important as not sourcing any ebuilds, so let's all |
| 9 |
> >> stick an EAPI extension on the end of the filename. |
| 10 |
> > |
| 11 |
> > If you really think that EAPI as an extension has anything to do with |
| 12 |
> > performance |
| 13 |
> |
| 14 |
> You mentioned performance a few times in that lovely thread when it got shot |
| 15 |
> down, I believe in the context of metadata generation: |
| 16 |
> |
| 17 |
> "Performance hit" (when discussing an alternative) [1] |
| 18 |
> "The GLEP is not about performance, but any solution that forces the |
| 19 |
> introduction of a slowdown of more than, say, 20%, isn't viable." [2] |
| 20 |
> "It's several more directory reads. This is a measurable performance |
| 21 |
> hit" [3] |
| 22 |
> |
| 23 |
> Are you now saying performance doesn't matter? |
| 24 |
|
| 25 |
No, performance is just a side effect it has never been part of the |
| 26 |
motivation of the GLEP. You are the only fighting over stupid pico |
| 27 |
optimizations. |
| 28 |
|
| 29 |
- ferdy |
Archives