| 1 |
Alan McKinnon writes: |
| 2 |
|
| 3 |
> On Wednesday 03 March 2010 13:27:45 Alex Schuster wrote: |
| 4 |
|
| 5 |
[RANT RANT RANT] |
| 6 |
|
| 7 |
> > Ah, I see the problem. It mainly scans /data/mp3/incoming, a |
| 8 |
> > directory I have NOT selected as collection folder (but most other |
| 9 |
> > directories in /data/mp3 are selected). Still, those files do not |
| 10 |
> > show up in my collection, which is fine - some time ago amarok did |
| 11 |
> > index all in /data/mp3, even if a directory was not selected. |
| 12 |
|
| 13 |
I investigated this further. Amarok seems to look for all playlists below |
| 14 |
/data/mp3, and then looks up all of their files. No idea why. |
| 15 |
|
| 16 |
> I wonder if amarok would not be better off using the strigi/nepomuk |
| 17 |
> indexing function, instead of trying to be real clever and doing it |
| 18 |
> itself. |
| 19 |
|
| 20 |
Strigi also keeps indexing parts of my /data/mp3 stuff with EVERY login. |
| 21 |
|
| 22 |
> OTOH, that might just resurrect the mother of all threads we had |
| 23 |
> recently - the one about the pros and cons of nepomuk and |
| 24 |
> semantic-desktop :-) |
| 25 |
|
| 26 |
I am pro, I like it, but again it seems those things are not yet working |
| 27 |
right. Strigi indexes stuff over and over again at every login. virtuoso-t |
| 28 |
then also runs for a while and hogs resources. dbus-daemon uses 10-15 |
| 29 |
percent of CPU time according to top. Should it do this? |
| 30 |
|
| 31 |
I enabled auto-login for KDE, so when I boot the system, at least things |
| 32 |
are already indexed when I start working with it. |
| 33 |
|
| 34 |
Now I'm going to emerge KDE-4.4.1, let's see what this will change. |
| 35 |
|
| 36 |
Wonko |
Archives