| 1 |
On Wed, 25 Feb 2009 09:33:44 +0100 |
| 2 |
Alexis Ballier <aballier@g.o> wrote: |
| 3 |
> That sounds like an implementation detail that you could solve by |
| 4 |
> using something else than a flat file database for metadata if |
| 5 |
> open()/read() calls are the slow part. |
| 6 |
|
| 7 |
Metadata's shipped with the tree. It's a PMS detail. |
| 8 |
|
| 9 |
If we didn't care about package manager performance, we wouldn't be |
| 10 |
shipping metadata with the tree at all... |
| 11 |
|
| 12 |
> > > I would be very surprised if that "2 times" factor happens to be |
| 13 |
> > > true, because finding a string in a file is an order of magnitude |
| 14 |
> > > simpler than sourcing said file with bash. Moreover this doesn't |
| 15 |
> > > take into account disk and os cache. |
| 16 |
> > |
| 17 |
> > No no no. *Opening* the file is the slow part, not searching. The |
| 18 |
> > file wouldn't otherwise be opened at all. |
| 19 |
> |
| 20 |
> Thus the only drawback is when you open a file, see there that you |
| 21 |
> can't handle the eapi, then close it and open an older one. |
| 22 |
|
| 23 |
Uh. No. The drawback is that you're opening around ten thousand files |
| 24 |
that would otherwise not be opened. That's a huge cost. |
| 25 |
|
| 26 |
-- |
| 27 |
Ciaran McCreesh |
Archives