On Aug 31, 1:30 pm, "tom.andersen" <
tom.ander...@gmail.com> wrote:
> Hi,
>
> I checked in code that allows people to NOT make backup files, it just
> checks for a file called "No Backups Please.txt" in Library/
> Application Support/OpenMeta/backups
This leaves people open for losing data from the Adobe metadata
stripping problem. If they are not writing backup files then a Restore
Metadata will not help them. And doesn't this negate any usefulness
that OMFix would have?
>
> I checked in support for doing a backup of all open meta data to a
> single file. (I think that the file will be about 200MB for 100k
> entries). I also made a restore from this file format. (omsback)
This may be a philosophical difference but I am more comfortable
managing many files than "putting all my eggs in one basket", so to
speak. I fear the potential of losing all the backed up Tags at once
than losing a few here and there. Also, is this single file a
sparsebundle for easier Time Machine backups or something? I guess I
don't understand why the change.
>
> I changed the way restore is done.
> Restore is only done for keys of the form org.openmeta:* .
> Dates are compared on a field by field basis to merge in changes.
> This helps in for instance restoring openmeta backups over top of say
> a Time Machine restored file, which may have old tags.
>
> --Tom
Jim