Re: [HACKERS] Increased company involvement

From: Robert Treat <xzilla(at)users(dot)sourceforge(dot)net>
To: David Fetter <david(at)fetter(dot)org>
Cc: "Nicolai Petri (lists)" <lists(at)petri(dot)cc>, "Jim C(dot) Nasby" <decibel(at)decibel(dot)org>, Kris Jurka <books(at)ejurka(dot)com>, Andrew Dunstan <andrew(at)dunslane(dot)net>, Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>, PostgreSQL advocacy <pgsql-advocacy(at)postgresql(dot)org>
Subject: Re: [HACKERS] Increased company involvement
Date: 2005-05-01 01:34:13
Message-ID: 200504302134.13314.xzilla@users.sourceforge.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-advocacy pgsql-hackers

On Saturday 30 April 2005 17:40, David Fetter wrote:
> On Sat, Apr 30, 2005 at 10:37:06AM +0200, Nicolai Petri (lists) wrote:
> > I totally agree. In our preference list I would have the following tasks
> > : 1) IOT (Index Ordered Tables)
>
> Is this different from CLUSTER?
>

I think what he is looking for is a table that is maintained in CLUSTERed
order... ie. new inserts will be put in the appropriate place mimicing what
happens with indexes. This has been discussed in the past and has been shot
down before...see archives on pgsql-hackers for past discussions.

> > 4) Extending PostgreSQL's plugin support with additional hooks in the
> > backend e.g. :
> > - for adding new tablestore engines (like mysql can)
>

my$ql's multiple table handlers is about the ugliest hack I have ever seen
inside a database. I mean I used to not like it, but I've been studying it a
little bit lately and now find it absolutly dreadfull. I can't imagine that
we would ever implement a scheme like the one they have used; whatever
advantage people think they can gain from such a setup can probably be
implemented in a cleaner fasion in postgresql assuming it would be an
advantage at all (and some of those features have been dismissed in the past
as well so there's no garauntee)

> > - for adding callbacks that get's called on transaction
> > success/failure using SPI. (e.g. for housekeeping and cleanup)
>
> Would this be like an ON COMMIT TRIGGER?
>
Another idea I have seen shot down on hackers a number of times...

The point of this email being that, I think the general idea probably needs to
be that any coordinated community sponsored development needs to be centered
on picking your favorite items from the TODO list to be worked on rather than
just specifying a list of things you think should be the top priorities. Of
course there is nothing preventing people from getting new items added to the
todo list, but as a general guidline picking items from the todo gets you
halfway home right out of the gate.

--
Robert Treat
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL

In response to

Browse pgsql-advocacy by date

  From Date Subject
Next Message Robert Treat 2005-05-01 01:40:46 Re: [HACKERS] Increased company involvement
Previous Message David Fetter 2005-04-30 21:40:04 Re: [HACKERS] Increased company involvement

Browse pgsql-hackers by date

  From Date Subject
Next Message James William Pye 2005-05-01 01:37:20 pl/Python
Previous Message Mischa Sandberg 2005-04-30 23:41:26 OLAP and PG and deja-vu