Hmmm I sorta did think about that when we discussed that the last
time. Then, it seemed to me that it would be a bad idea for memory
consumption and such. However, this would only have to be changed for
the method signatures of the GenericQueryManager class, not for the
identifier property types.
What do others think of this?
On 05 Aug 2006, at 09:25, Steven Grimm wrote:
A while back there was discussion about how RIFE could support both
"int" and "long" for IDs of database objects.
A thought just occurred to me: if supporting both is difficult, why
support "int" at all? If a smaller value is needed in the database
table (for space reasons, etc.), it can be specified using a
maxValue() constraint on the ID field, or maybe some other new
constraint.
If IDs had been "long" from the get-go, I seriously doubt you'd
have seen a single request from someone who wanted *fewer* possible
primary key values. I don't really see any significant downside to
using "long" for all identifiers.
Well, there's one downside to *switching* to "long": obviously it
would be a non-backward-compatible change. However, I suspect it'd
be a change that would not take people much time to cope with. And
it would certainly lead to less confusion than any solution I've
seen discussed for supporting both key types at the same time.
-Steve
_______________________________________________
Rife-users mailing list
[email protected]
http://lists.uwyn.com/mailman/listinfo/rife-users
--
Geert Bevin
Uwyn "Use what you need" - http://uwyn.com
RIFE Java application framework - http://rifers.org
Music and words - http://gbevin.com
_______________________________________________
Rife-users mailing list
[email protected]
http://lists.uwyn.com/mailman/listinfo/rife-users
