I don't think every difference in behavior is a bug. I'm also not not sure which you classify as a bug, that a log is created or that it isn't?
Maybe it's if a force quit happens with an Open Record, maybe it's the OS version, maybe .log files are hidden in 10.10. It could be a whole bunch of things. I would say force quitting is not closing correctly.
FYI you also get a recover.log file when uploading files to FileMaker Server v13 using FileMaker Pro/Advanced v13 via the File menu>Sharing>Upload to FileMaker Server.
I believe Server runs a consistency check whenever it opens a file. A consistency check creates a Recover log, but I wouldn't consider it a "Recover" in the scary sense of the word.
That is a good question. FileMaker Client should run a Consistency Check if a file was not closed properly ( but perhaps it somehow was )
Which FileMaker Version ?
I just tried a FileMaker 11 force quit on OS 10.8.5, and got no Recover Log when the file was re-opened
When you get a log, what does it say ? is it a Consistency Check ?
Also, if the file is hosted elsewhere, a force quit may not matter
> I have done it several times on 10.10 and no log so far. So are they being closed correctly with a force quit or is this an ugly bug?
I'm not sure exactly how 10.10 does this...but Windows 7 and Windows 8 attempt to close the file gracefully first. Then does the classic "clothes-line" takedown if that doesn't work. I usually only get a recover log in the 2nd scenario.
I guess the force quit is sometimes closing the file 'cleanly', but when I did the same file set in 10.9 it had files that were not closed properly. I remember from years of devcons that the one of the issues always it that the number of 'clean' closes <> opens, then the client runs a consistency check. I will rerun some more force quit tests soon with a known open record and see how they behave.
Sometimes things happen but we have 'clean' masters, that are not clones of backups, but are files that have not been under load on the server. I want to know if I have a problem offline, do I need to redo to the master. We had significant file corruption from earlier developers a number of years ago, and have no desire to ever return to those days, and the work of redoing files.
You really don't need "masters", just keep regular sets of backups, so you can revert if required
Please note that finding a good backup can be tricky, if the corruption went unnoticed for any length of time
I have collected info here:
> Sometimes things happen but we have 'clean' masters, that are not clones of backups, but are files that have not been under load on the server. I want to know if I have a problem offline, do I need to redo to the master