Brian Aker wrote:
Hi!
On Aug 12, 2008, at 9:47 AM, Daniel Gary wrote:
Dear god yes, although I would still hope for a default size, but if
I put in INT(1) I'm pretty sure I would know what values I'd be
limited to.
INT(1) would be a 1 byte INT.
Is that too gear head'ish for the average DBA though?
Said it before, I'm not a representative audience, statistically
speaking, but I'd do int(1) over tinyint any day given the choice, but
then again, I'm not a "normal" DBA or person for that matter.
But the question is (kind of my development divining rod), does it take
more hassle in the long run to keep it, or to change it?
So really, if both were available... Eh, yeah, I retract my previous
"dear god" statement.
hmm, how about data type aliases as a module? keeps it out of core,
but could add some sort of "backwards compatibility" in a sense.
I am not sure how to do this, without keeping it in the parser.
Might be something to think about, pluggable datatypes, I'm sure those
"super giant ultra mega bigint" folks would love an unrealistically
large INT, or Spatial types, etc, that I don't really see being in core,
but its not my call.
But then again might be out of scope for the average web application,
though on the cluster side? I'm sure we'd see more Geometry and
impossibly large number crunching.
I am not sure how to do this either, I was always horrible at writing
parsers.
Cheers,
-Brian
-Daniel
_______________________________________________
Mailing list: https://launchpad.net/~drizzle-discuss
Post to : [email protected]
Unsubscribe : https://launchpad.net/~drizzle-discuss
More help : https://help.launchpad.net/ListHelp
