Thread: Future plans

Future plans

From
Michael Meskes
Date:
Hi,

what do we have planned for the next release? In the TODO file there are
quite a lot of points but I don't like to talk about things we will
eventually do, but would like to present what we will implement for 7.3
resp. 8.0. 

IMO the most important stuff seems to be:

- replication (which is listed as urgent anyway)
- alter table drop column (for which I get asked about once a week)
- recursive views (you know, I wanted to implement this when I started my work on PostgreSQL, but never found the
time)

Michael

-- 
Michael Meskes
Michael@Fam-Meskes.De
Go SF 49ers! Go Rhein Fire!
Use Debian GNU/Linux! Use PostgreSQL!


Re: Future plans

From
Hannu Krosing
Date:
On Fri, 2002-05-17 at 16:58, Michael Meskes wrote:
> Hi,
>
> IMO the most important stuff seems to be:
>
...
> - recursive views (you know, I wanted to implement this when I started
>   my work on PostgreSQL, but never found the time)

A good start would be to make the parser recognize the full sql99 syntax
for it. Its quite big - see attached gif I generated from grammar
extracted from the standard:

Attachment

Re: Future plans

From
Michael Meskes
Date:
On Tue, May 21, 2002 at 12:35:20AM +0500, Hannu Krosing wrote:
> > - recursive views (you know, I wanted to implement this when I started
> >   my work on PostgreSQL, but never found the time)
> 
> A good start would be to make the parser recognize the full sql99 syntax
> for it. Its quite big - see attached gif I generated from grammar
> extracted from the standard:

Well, the parser seems to be the easier part. :-)

Michael

-- 
Michael Meskes
Michael@Fam-Meskes.De
Go SF 49ers! Go Rhein Fire!
Use Debian GNU/Linux! Use PostgreSQL!


Re: Future plans

From
Hannu Krosing
Date:
On Tue, 2002-05-21 at 10:18, Michael Meskes wrote:
> On Tue, May 21, 2002 at 12:35:20AM +0500, Hannu Krosing wrote:
> > > - recursive views (you know, I wanted to implement this when I started
> > >   my work on PostgreSQL, but never found the time)
> > 
> > A good start would be to make the parser recognize the full sql99 syntax
> > for it. Its quite big - see attached gif I generated from grammar
> > extracted from the standard:
> 
> Well, the parser seems to be the easier part. :-)

Sure.

My point was that we should put in the _full_ syntax in one shot and
then we could implement in smaller pieces.

------------------
Hannu