Feedback on svn-1348

[fizze]

well, this is a little OT…

My playlists, that worked with 1200something dont work with 1348.
I set the case sensitive = 1, and its recognized by the web interface, so I guess its right.

Just triggered a full rescan from the interface, maybe that helps.

Other than that, any clues ?

[grommet]

Set debug logging to 9 and see what happens when it walks the Playlist? I can’t repro this. But I don’t have a lot of playlists on my test environment.

[rpedde]

@fizze wrote:

well, this is a little OT…

My playlists, that worked with 1200something dont work with 1348.
I set the case sensitive = 1, and its recognized by the web interface, so I guess its right.

Just triggered a full rescan from the interface, maybe that helps.

Other than that, any clues ?

what kind of playlists? m3u, itunes, or smart?

[grommet]

I think I repro’d part of the broken playlist problem using an old build and upgrading to the latest nightly… keeping the old database. I guess I’ve been blowing away databases too much. 8)

[rpedde]

@grommet wrote:

I think I repro’d part of the broken playlist problem using an old build and upgrading to the latest nightly… keeping the old database. I guess I’ve been blowing away databases too much. 8)

Yeah, me too. Found it, though. Full scans through the web interface weren’t forcing scans on playlists, and I should have forced a full scan anyway when I added the case stuff. In any event, it’s fixed in svn now.

Stopping the server, whacking the db, and starting the server will also fix it, as a short-term fix.

— Ron

[fizze]

No worries, since playlists werent working (right) for me for too long, I didnt get used to it. Yet. 😉

[rpedde]

@fizze wrote:

No worries, since playlists werent working (right) for me for too long, I didnt get used to it. Yet. 😉

I think maybe I need to make some playlists that I use. Since I always use smart playlists for everything, I don’t have any m3u playlists I depend on.

I really do need to build some kind of regression suite though. It’s just that there is so much other stuff to do, that kind of falls to the back burner.

— Ron

[grommet]

Ron, I haven’t had time to beat on it, yet… but so far, with only 2 attempts, something is slower in svn-1359:

2006-08-20 17:46:33 (f3f11f6d): Scanned 20814 songs in 505 seconds

2006-08-20 17:46:33 (f3f11f6d): Serving 20814 songs.  Startup complete in 505 seconds

Now, not that dreadful 40 minutes we had before… but slower than the svn-1344 record benchmarks I had at the beginning of this thread.

Oh yeah, and Sample Rate and Size is still missing from WMA. 8)

[rpedde]

@grommet wrote:

Ron, I haven’t had time to beat on it, yet… but so far, with only 2 attempts, something is slower in svn-1359:

2006-08-20 17:46:33 (f3f11f6d): Scanned 20814 songs in 505 seconds

2006-08-20 17:46:33 (f3f11f6d): Serving 20814 songs.  Startup complete in 505 seconds

Now, not that dreadful 40 minutes we had before… but slower than the svn-1344 record benchmarks I had at the beginning of this thread.

That’s peculiar. I know what changed there, but not why it should change the scan speed. I thought I had saw that, but then wasn’t sure.

Oh yeah, and Sample Rate and Size is still missing from WMA. 8)

Argh!!!

[Author]

Posts