opaque ids vs. natural keys

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

...
>> I can see the desire for having natural keys that we show the
>> user, but history has shown (over and over) that having opaque
>> unique ids provides a better identity story.
>> 
> 
> Having natural keys that the user sees and opaque, surrogate keys
> for DB identity are not mutually exclusive :-)
> 
> 

I view it a bit as the revno vs revision_id that bzr went with
 revno: context sensitive 'natural' ids that is reasonably easy for a
        human to grasp
 revision_id: actual unique identifier for the object

I especially like having UUIDs for machines when you start talking
about cross-environment relations. Since at that point you can start
reasoning about any object in any environment without having to also
pass around all the context for the identifier all the time.

However, this all seems very much state-breaking changes. Which we've
been explicitly cautioned to be careful with.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (Cygwin)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlGlm1gACgkQJdeBCYSNAANOEACgmUgNaoolZtyYuQvYoY/HC19z
YF4AoNbF83yBuoUth4JXWhxDQ2OLEKBh
=W8JZ
-----END PGP SIGNATURE-----