jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Nuescheler <david.nuesche...@gmail.com>
Subject Re: RepositoryImpl's UUID constants
Date Sun, 06 Feb 2005 22:01:41 GMT
hi tim,

> I noticed this the other day (or more):
> +    private static final String SYSTEM_ROOT_NODE_UUID =
> "deadbeef-cafe-babe-cafe-babecafebabe";
> +    private static final String VERSION_STORAGE_NODE_UUID =
> "deadbeef-face-babe-cafe-babecafebabe";
> +
> I think real UUID's should be used. Preferably version 1 (time based.) On
> one hand these would never conflict with a version 1 or version 4 uuid since
> the version and variant fields appear to be invalid, but on the other hand
> if a uuid implementation were going to try to validate those fields it would
> cause an issue. How big of a deal would it be to change them at this point?
> Maybe it's not a big a deal? Then again, if it's not too big a deal to
> change...
the idea was that one could immediately recognize those "special
uuids" when debugging etc. while i am not sure if it makes a big 
difference whether those uuid's are in fact "real uuid's" or not, 
since i wasn't able to come up with a usecase where someone 
would validate the uuid's (but i think you are right, we should
be more careful)

> Two candidate replacements if needed:
> 9fbcb710-7734-11d9-914f-0010afec213c
> af5af9c0-7734-11d9-914f-0010afec213c
i have no issue using valid uuid's but maybe we
can use something distinctive, that could be 
recognized by a human? something like:
00000000-0000-4000-0000-000000012007
do you think that makes sense? i am not
attached to any uuid in particular.

regards,
david

Mime
View raw message