Subject | RE: [IB-Architect] Re: IB 6.1 Whish: boolean field |
---|---|
Author | Jim Starkey |
Post date | 2002-02-15T15:59:55Z |
At 10:49 PM 2/14/02 -0500, Leyne, Sean wrote:
requiring only a change to the ODS minor version.
The large issue is whether a boolean datatype is worth anything.
I don't think it is. The only difference between a boolean and
a smallish int is that boolean is constrainted by values 0 and
1 (presumably). That's not much. What folks are often looking
for is simpler SQL syntax for a conditional like:
select * from somewhere where active
where "active" is boolean field. Syntactically, however, this
doesn't work in the SQL grammar without violating some fairly
critical language design rules.
Jim Starkey
>Julien,An additional datatype is generally an upwards compatible change,
>
>A Feature Request for a Boolean datatype has been recorded for the
>Firebird project (quite some time ago).
>
>One of the main constraints to the new datatype, is that its
>implementation would require a new ODS (On Disk Structure) in addition
>to the changes to the engine logic.
>
requiring only a change to the ODS minor version.
The large issue is whether a boolean datatype is worth anything.
I don't think it is. The only difference between a boolean and
a smallish int is that boolean is constrainted by values 0 and
1 (presumably). That's not much. What folks are often looking
for is simpler SQL syntax for a conditional like:
select * from somewhere where active
where "active" is boolean field. Syntactically, however, this
doesn't work in the SQL grammar without violating some fairly
critical language design rules.
Jim Starkey