On Apr 8, 3:01 pm, dp...@pgadmin.org (Dave Page) wrote:
> On Wed, Apr 8, 2009 at 1:38 PM, valgog <val...@gmail.com> wrote:
> > Hi,
>
> > When you accidentally click on the scary tab (I managed to do it twice
> > already) the whole pgAdmin is blocked for more then 3 minutes. It is
> > not normal not to ask for a confirmation before some long lasting
> > operation (especially when the button to start this operation locates
> > directly under most used buttons on the toolbar).
>
> How big is your database? Even with 5000+ tables it only takes a
> minute on my laptop. We're aware that code needs some redesign, but I
> wasn't aware it was that bad except in a tiny minority of cases. Maybe
> we need to bring forward the proposed rewrite.
Are you testing your local database or you are connected to the remote
one? I am connected to the my database (about 3000 tables) using an
SSH tunel. I can imagine, that transferring all the table data from
the remote database to the client machine takes all the time (in case
it works fast for the local databases of a big size)
>
> > And why not to add the enable/disable switch? There is a possibility
> > to control which elements types are to be shown in the object tree.
>
> That controls what elements of the database are displayed. It doesn't
> enable or disable entire features.
>
My suggestion was also not the disable the feature completely. Making
it possible to hide it from the interface is completely sufficient.
Even a simple confirmation dialog before trying to fetch all the table
information from the database would solve the problem of people who
miss the button to often, like I do.
Actually the request about the slowness of the switching between
databases in Query window is much, much more important in our case (as
we switch from standby and live databases quite often having one file
open in the Query window). I can imagine, that this slowness is
directly related to some 'optimizations' for GQB.
regards,
-- Valentine Gogichashvili
