A couple of thoughts here:
Why use the hash or other computed value. The real issue is that you
want a uniquely identifiable value that represents the file. This key
value should be one that would not be generated by any other audio file
in the system. So why not just use a simple sequential number.

The reasoning behind this is that only if you start using something
like EAC and secure rip do you start to get a value that could be
considered universal any transcription errors from the CD would cause a
different hash to be generated. If you rip to a lossy format then the
hash would be different to that for a lossless rip. You couldn't store
the original WAV hash because you couldn't recreate it.

As there are currently no tools that we seem to be aware of that would
calculate a hash for audio portions of say mp3 then such a tool would
need to be written. Essentially how is this different from any other
tagger. Instead of writing general purpose tags it performs a single
function and writes the hash into a tag on the file. You can then run
this hash tagger independant of SC just as we do now for normal tags.

In a similar way if you just used a sequential ID in a tag the tagger
would just need to assign a unique ID. It would need access to the
database to obtain the unique ID but that doesn't sound too difficult.

Using a uniqueID could also be widened a little in that if you have an
MP3 rip and a FLAC rip of the same track you could conceivably assign
them the same ID thus the system would be aware that they are the same
track. This would make it more easy to have multiple copies of the same
track for different purposes and even allow SC to pick the most
appropriate for different circumstances.

All of this doesn't mean that SC has to be a music manager it is just a
collection of the meta data. How the physical files are arranged is
still within the remit of the person or an independant music manager.

SC doesn't even need to manage the database, whilst it could continue
to read tags it could also hand database entry over to a different
module also accessed through the web server in which the full
capabilities of a database can be used to tag and othewise manage the
metadata. 

It even seems like it would be possible to implement such a scheme with
little if any change to SC code. The only thing that you would want to
stop it doing is any scans, either explicit or implicit.


-- 
Zaragon
------------------------------------------------------------------------
Zaragon's Profile: http://forums.slimdevices.com/member.php?userid=14577
View this thread: http://forums.slimdevices.com/showthread.php?t=43851

_______________________________________________
ripping mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/ripping

Reply via email to