| 1 |
On 2008/06/10, Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> wrote: |
| 2 |
|
| 3 |
> Currently we don't touch the ebuild's content *at all* for metadata |
| 4 |
> operations, except where there's no or stale metadata cache (which is |
| 5 |
> rare). We can get away with this currently because 0 and 1 have |
| 6 |
> identical cache layouts and PMS has some (necessary) weasel wording; |
| 7 |
> future EAPIs likely won't, so we're back to the chicken / egg problem. |
| 8 |
> |
| 9 |
> So... We either have the EAPI from the extension (which we already |
| 10 |
> have, since we have to read the dir to know what versions are |
| 11 |
> available), in which case we know how to read the metadata cache file. |
| 12 |
> Or we have to open up a file that would otherwise not be opened, just |
| 13 |
> to parse one line so we know how to read the cache file. |
| 14 |
|
| 15 |
Or you apply to future EAPI's cache formats one of the solutions that |
| 16 |
have been proposed for the ebuild side of the very same "chicken / egg |
| 17 |
problem": for instance, you could use $EAPI as cache filename extension. |
| 18 |
|
| 19 |
And that's it, you can get EAPI from the cache, hence no more extra |
| 20 |
reading of ebuild files, and no more perfs issues. |
| 21 |
|
| 22 |
It sounds so simple, am I missing something obvious? |
| 23 |
|
| 24 |
-- |
| 25 |
TGL. |
| 26 |
-- |
| 27 |
gentoo-dev@l.g.o mailing list |
Archives