MPD 0.24.2 update

[nosferatu_cat]

I did the original 9.30 update on 4 Pis, 3 OTA and 1 fresh install.  On all of them the MPD database was broken.  On the fresh install the creation stopped with the error I reported in my earlier post on this thread.  On the 3 OTA updates what had been complete databases ranged from no visible files to a small subset.  I tried the MPD update to the new version on the 3 Pis I had updated to 9.30 OTA.  On all 3 I did the install, rebooted, and regenerated the database.  It succeeded with one.  The other two failed completely not recognizing the NAS share.  The share on those two was created successfully but the database was not built.  I'm running a regenerate on both currently.  I'll know how it went in 5 or 6 hours.   On the remaining Pi i did a vanilla MPD install on the regular Bookworm OS.  The apt install version is .231.  It created the database without a problem.  Both the MPD log file and MALP show a complete selection of files.  

I'm pretty sure the problem is with MPD.  The error I get that causes the database build to time out or lock up was actually thrown by FFMPEG as shown in my post earlier in the thread.  The one thing that puzzles me is that one of the three MPD upgrades to the new version of .24x worked.  It's great to have something like this to occupy my time in retirement.  Part of the fun is that it's not my code and my job isn't at risk.  Like everyone else who had done coding I know that debugging is a significant part of software development.  The puzzle of discovering what part of the implementation wasn't quite right is fun in a way.

[Caesar]

All is good here.

[the_bertrum]

Added some new tracks, played a few things, then did a full regen with no issues on the 0.24.2 on a Pi3A+

[TheOldPresbyope]

I didn't mention updating MPD on several 9.3.0 systems because it was, as my doctor would say, unremarkable.

I also see some ffmpeg errors messages thrown as MPD scans my library. I haven't traced them down to specific files yet but my MPD scans run to completion.

This is true even with the problematic album containing the zero-length track added to the library. Didn't fail with MPD 0.24 and still doesn't fail with MPD 0.24.2. (Nice listening to Glen Gould again, by the way.)

Still looking for a definitive test case.

Regards,
Kent

[droner3r]

Hi!
I’m glad to report that my issues have been fixed after this update.
Now everything’s fine.

Thank you Tim , thanks to the forumers for their help.

[Tim Curtis]

Glad to hear its fixed :-)

[yannig]

everything's fine after this update on my Raspberry 3B (allo Boss dac)

[nosferatu_cat]

I did the original 9.30 update on 4 Pis, 3 OTA and 1 fresh install.  On all of them the MPD database was broken.  On the fresh install the creation stopped with the error I reported in my earlier post on this thread.  On the 3 OTA updates what had been complete databases ranged from no visible files to a small subset.  I tried the MPD update to the new version on the 3 Pis I had updated to 9.30 OTA.  On all 3 I did the install, rebooted, and regenerated the database.  It succeeded with one.  The other two failed completely not recognizing the NAS share.  The share on those two was created successfully but the database was not built.  I'm running a regenerate on both currently.  I'll know how it went in 5 or 6 hours.   On the remaining Pi i did a vanilla MPD install on the regular Bookworm OS.  The apt install version is .231.  It created the database without a problem.  Both the MPD log file and MALP show a complete selection of files.  

I'm pretty sure the problem is with MPD.  The error I get that causes the database build to time out or lock up was actually thrown by FFMPEG as shown in my post earlier in the thread.  The one thing that puzzles me is that one of the three MPD upgrades to the new version of .24x worked.  It's great to have something like this to occupy my time in retirement.  Part of the fun is that it's not my code and my job isn't at risk.  Like everyone else who had done coding I know that debugging is a significant part of software development.  The puzzle of discovering what part of the implementation wasn't quite right is fun in a way.

It took a while to regenerate the DB on the misbehaving Pis but everything is working now after upgrading MPD.  I got the fun of watching the MPD log as records were added to the DB.

[Tim Curtis]

I've updated the post-release patch section of the 930 release announcement.

[Stephanowicz]

FYI

ver 0.24.2 (2025/03/26)
* input
- io_uring: fix stuck seek after I/O error
* decoder
- flac: enable chained Ogg decoding
* player
- fix stalled playback after queue modification
- fix deadlock with "single" mode
* Linux: fix build failure with ancient kernel headers

https://raw.githubusercontent.com/MusicP....24.2/NEWS