On Mon, Mar 18, 2013 at 12:06 PM, Douglas Regehr <[email protected]>wrote:
> > > > On Mon, Mar 18, 2013 at 3:03 PM, J. Liles <[email protected]> wrote: > >> >> >> On Mon, Mar 18, 2013 at 11:58 AM, Douglas Regehr <[email protected] >> > wrote: >> >>> Here's another session that I can't load. The console output ends with: >>> >>> ../nonlib/Loggable.C:331 update_id(): Attempt to create object with an >>> ID (0x295D) that already exists. The existing object is of type >>> "Audio_Region", the new one is "Audio_Region". Corrupt journal? >>> >>> I did a pull this morning, so I think I'm running the latest. Looking >>> at the history file, I can clearly see that 0x295D is created three times, >>> each time with identical parameters. I had a crash just before this, but >>> unfortunately I didn't get a trace, so I can't help you figure out how the >>> journal got corrupted like this...If it happens again I'll try to get a >>> trace... >>> >>> I was able to work around the problem by deleting the duplicate creates >>> in the history. There were also duplicates for 0x2961, which was created >>> twice. The second create was followed by a destroy, so I just deleted the >>> create-destroy pair...if that makes sense... >>> >>> Question: When does non-daw make snapshots? Is there some way to force >>> it to take a snapshot? >>> >> >> So, just so we're clear... Did you create this session *after* I pushed >> the fixes? Or could the duplicate entries predate those changes? >> >> As to the snapshot--it is created whenever the project is closed normally >> (i.e. upon quit/close). You can also force the history to be erased and >> replaced with a snapshot by using the 'Project/Compact' option. >> >> >> >> The session is quite old. Definitely not created after you pushed the > fixes. So if I compact this session then I can be reasonably certain I > won't run into this problem again? > Reasonably. I would keep a backup of the history just in case though.
