(2016-04-16, 16:43)pr0xZen Wrote: (2016-04-15, 21:26)BatterPudding Wrote: (2016-04-13, 22:14)pr0xZen Wrote: To me it appears so, according to the logfile. Not sure why they're so quick, but its "checking" every one of them in some manner, just 1000+ in a few seconds (in my instance). Local network ones though, there's a 5 second timeout - causing a swift job taking half an hour.
Is this really happening? It just seems so absurd.
I wonder if this explains the odd "Clean Library" effect I get? When I hit the "Clean Library" button it starts and almost instantly gives up after a few seconds and the progress bar has hardly moved. I press it a second time and now it gets right through a scan (well, the progress bar moves a lot further)
I've not been experiencing this. It does freeze alot with no graphical progession, presumably on a "string" of successive timeouts, but no involuntary aborts.
I have never bothered looking at the logs, but there is a clear abort on the first attempt for me. It barely moves the progress bar before the dialog box just disappears without error. When I try it today I find it actually freezes my GUI for 20-30 seconds after the box vanishes. (GUI freeze includes the mouse pointer and debug overlay) Second attempt it always succeeds and clearly counts up to 100% in little steps... noticeable pauses but it completes.
When I look at the literal description of "Clean Library..." it says "Remove items from your library that can't be found (either renamed, deleted, or on removable storage that is currently unplugged)". That sounds very much to me like it should be removing these duff items that were added by File Mode.
I'm off to investigate my clean library freeze... see if there is anything related in the logs...
Update: Okay... reading that log didn't give me any clue as to the freeze I am getting as the log thinks all is well even though it is getting floods of warnings from the "unsupported protocol" of plugin, rtmp, rtmpe and the other streams. It even complains about not being able to open many of those File Mode files.... but leaves it all in the database instead of cleaning it out.
I get a feeling I could use those errors and warnings from the log file to locate the majority of my dodgy links in the database.
What about this as a different test. If a file is listed in the FILES table and not referenced anywhere else in the SQL Database then surely it can be deleted? Then once those files have been removed, the same rule can be run over the PATHS table - if path not referenced elsewhere then it should be removed.