File origin metadata

[Joe Shaw]

Hi,

Sorry, I am joining this discussion a little late.

Saving contextual information about a file is something that I think is
really important and something which has been on the burner of Dashboard
and Beagle for some time.  Today applications lose buckets of contextual
metadata that could be stored.  I think that using extended attributes
for much of this makes a lot of sense.  Two common use cases that come
to mind are:

        * Who sent me this attachment via email, and what was the
        subject of the thread?
        
        * From what web site did I download this file, and is it
        associated with a person?
        
There are many, many others.
        
On Sat, 2006-07-01 at 09:21 +0200, Claes at work wrote:
> On 6/30/06, Pat Suwalski <pat at suwalski.net> wrote:
> >
> > Would there be interest in starting a more complete document that has
> > common metadata attributes by mime type?
> 
> Can you explain more what you mean with this? Would you like to
> document what metadata is part of files of different mime types, for
> example the metadata that is part of an ID3 tag for mp3 files?

We need to be careful about our scope here.  It doesn't make sense to
create a single store in which all metadata goes.  A lot of data sources
and file formats have their own support for various metadata, and we'll
never be able to override (nor should we try).  In many cases, there
might not be a single file backing a resource on which to place an
xattr.

It *does*, however, make a lot of sense to have desktop indexers monitor
and index this metadata, so that applications and users have a
centralized place to search that metadata.

There is no "one size fits all" solution to such a large and nebulous
idea as metadata.

Joe