[deanerh]

BTW,

The files are not stored in the library as being unavailable. If you exit xbmc and then start it again, and don't do a library update, the files can be played through library mode.

I get the same "unavailable" message on my ATV2 with files I've connected to using WebDAV . Besides starting and restarting XBMC, I also found that if I do a long press instead of a short one when picking the file to play, it pulls up the detailed information page on the file (tv shows in my case) and if I choose the "Play" option from that screen, it works fine. Don't know if the issue I'm seeing is the same as with FTP sources and if this will work as well, but I thought I'd throw it out there.

[kno1]

Now I'm at a complete loss. I can't seem to be able to reproduce the problem anymore. Not that I'm really complaining since it was annoying of course, but I don't understand how it can have disappeared, not after having been able to reproduce it so consistently.

Could it be that the movie scraper was updated and that that solved the issue? I'm now at version 3.0.6, anyone any idea when that was released?

Remco

[kno1]

And now it's back again. But I can't reproduce it consistently anymore. I had it happen after adding a source that didn't contain any actual media files (but I did set the content to movies). But this was just the once and on windows. Though I've also encountered the problem again on the ipad, but even there I'm unable to consistently reproduce it now. I've been trying to create a xbmc config in which I can reproduce it consistently, so I could post an archive of the userdata folder, but no luck so far...

Remco

[kno1]

[jmarshall]

If you can reproduce it on windows, I've started a build for win32 with some extra debugging. It'll be done in a couple hours - will post the link here in the morning. If you could test this extensively that would be great.

[kno1]

@jmarshall: No problem

[jmarshall]

It'll be here:

http://mirrors.xbmc.org/test-builds/win32/

In about another hour or so - it'll be called something like:

XBMCSetup-20120126-65d29ed-excessive_debugging_to_hunt_ftp_exists.exe

Cheers,
Jonathan

[kno1]

Do I activate debug logging or is it already logging enough without that?

[kno1]

I've been testing with the debug build, but unfortunately I'm unable to reproduce it with the debug build. I reverted back to beta2 and I was able to reproduce it again. Then just to make sure I tried the debug build again, but no luck reproducing it in there.

Is this the debug build the same as beta2, with just the extra logging? Or does it already contain new commits? If it already has new commits, then I guess one of these will have fixed the bug (or reverted the code that broke it)

Remco

[jmarshall]

Yes - that is possible. In particular there were 2 commits related to this that may have changed something. I'll revert both and rebuild that branch so that we can be completely sure.

[jmarshall]

Here you go - see if you can repro with this (hopefully you will be able to!)

http://mirrors.xbmc.org/test-builds/win3...exists.exe

[kno1]

Thanks,

I've been able to reproduce the issue with that build:
http://pastebin.com/zAQNtfVF

Like I said before, xbmc keeps an ftp connection open, and once that connection times out (or if I kick it from the FTP server) xbmc doesn't find the files anymore.

Here's the last copuple of lines of FTP fileserver as well, showing you the connections timing out, you see just the end of xbmc updating its library. Then I'm waiting after 2 minutes the connections time out, the last connect seems to be xbmc connecting again, but not logging in and then disconnecting again.

Code:

(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> 250 CWD successful. "/Zzzzz" is current directory.
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> CWD 8
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> 250 CWD successful. "/Zzzzz/8" is current directory.
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> PASV
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> 227 Entering Passive Mode (82,215,51,135,6,158)
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> LIST
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> 150 Connection accepted
(000297)27-1-2012 8:23:08 - xbmctest (188.203.200.66)> 226 Transfer OK
(000295)27-1-2012 8:24:54 - xbmctest (188.203.200.66)> 421 Connection timed out.
(000295)27-1-2012 8:24:54 - xbmctest (188.203.200.66)> disconnected.
(000297)27-1-2012 8:25:09 - xbmctest (188.203.200.66)> 421 Connection timed out.
(000297)27-1-2012 8:25:09 - xbmctest (188.203.200.66)> disconnected.
(000316)27-1-2012 8:25:13 - (not logged in) (188.203.200.66)> Connected, sending welcome message...
(000316)27-1-2012 8:25:13 - (not logged in) (188.203.200.66)> 220 Welcome to kno1
(000316)27-1-2012 8:25:13 - (not logged in) (188.203.200.66)> disconnected.

If you need anything else, let me know.

Remco

[jmarshall]

Great. I think we can call the problem solved in that case

[kno1]

Great. I think we can call the problem solved in that case

Just curious, but you mean that you now now what caused the issue and one of the new commits already fixed this? Or you just fixed it in a new commit?

Of course, the most important thing I'm taking away from this is that it'll be fixed in the next beta/release.

Thanks for your effort John

[jmarshall]

Yes, one of the commits already in master has fixed it (there was a timeout related commit there).