Re: pgAdmin III commit: Fix the removing of an inherited table

On Tue, 2011-09-20 at 03:17 -0500, Dave Page wrote:
> On Mon, Sep 19, 2011 at 2:48 PM, Guillaume Lelarge
> <guillaume(at)lelarge(dot)info> wrote:
> > On Mon, 2011-09-19 at 14:28 -0500, Dave Page wrote:
> >> On Mon, Sep 19, 2011 at 2:18 PM, Guillaume Lelarge
> >> <guillaume(at)lelarge(dot)info> wrote:
> >> > On Mon, 2011-09-19 at 14:10 -0500, Dave Page wrote:
> >> >> That doesn't seem right - the user shouldn't see quoting, except in SQL
> >> >> queries.
> >> >>
> >> >
> >> > We show the user the schema and the table names. It would be weird to
> >> > display foo.bar.something if the schema name is foo.bar and the table
> >> > name is something. We could use two columns, of course, but I don't
> >> > really see the point.
> >> >
> >> > Anyway, it was already displayed that way. I just fixed the issue. So,
> >> > if you want to fix the display, be my guest :)
> >>
> >> My point is that you haven't actually fixed the original bug; you've
> >> made it worse by further propagating the original issue.
> >>
> >
> > So, what should we do? display two columns? how will that work on the
> > combobox?
> >
> > I agree to work on this once we've found a good way to deal with it, but
> > we don't have one right now. And actually, the current patch fixes the
> > OP's issue, and that's good enough for me.
>
> Just show them unquoted as we do elsewhere. I think the
> stupid.schema.name.tablename issue is a corner case that can be safely
> ignored (it wouldn't be wrong per se, it just requires a little
> thinking on the part of the user, which frankly serves them right :-)
> ).
>
> For an example, look at dlgForeignKey, which shows the unquoted names
> in the Reference field.
>

Could work that way, sure. I don't have time right now, but I guess
we'll have to do it. Thanks for the pointer.

--
Guillaume
http://blog.guillaume.lelarge.info
http://www.dalibo.com