Dave Page wrote:
>
>
> On 6/11/05 9:11 am, "Chris Velevitch" <chris.velevitch@gmail.com> wrote:
>
>
>>On 11/6/05, Sim Zacks <sim@compulab.co.il> wrote:
>>
>>>However, if your main database activity is changing the data type of a
>>>column in ways that are disallowed by PostGreSQL, then you are
>>>correct. PGAdmin is pointless.
>>
>>My main database activity is refactoring and version control of a
>>database. This activity should be supported by a tool like pgAdmin.
>
>
> It is, or at least it was. It looks like you just found the first bug in
> 1.4.0 :-(
Well, this is not a really bug but an incompletely supported pgsql8.x
feature; 1.4 works the same way as 1.2.
We're allowing only binary compatible type changes (from pre 8.0 ages
where we did type changes by direct catalog access), while with 8.x
castable changes are allowed. We must have missed that in the 7.5 cycle.
Changed in svn for 1.4.1 and HEAD now, thanks for reporting.
Regards,
Andreas
