Thread: Starting the 8.2 release process ...
All,
Well, it's that time again. In the next couple of days, we'll be starting to
draft the 8.2 release. Here's what needs to be done to draft the release:
THEME: Bruce has suggested "usability" since this is a bit of a
"consolidation" release, without a lot of big-name features. I'm not sure
about that but I don't have a counter-suggestion. Ideas?
QUOTES: We need at least two quotes, preferably including one about the theme
and one which highlights something cool in 8.2. I can probably get one from
Sun or a partner. Suggestions?
FEATURES: AFAICT, there are only four big-name features worth mentioning in
the release:
on-disk bitmap indexes
d-trace / GMF
stored procedure debugging
standby databases (new PITR)
Anyone know anything else? Performance (sort, joins, etc) improvements would
be worth mentioning but *only* if someone can attach a figure to them.
Based on the suggestions on this list, I'll throw a draft onto pgFoundry CVS
and we can take it from there.
--
Josh Berkus
PostgreSQL @ Sun
San Francisco
>
> FEATURES: AFAICT, there are only four big-name features worth mentioning in
> the release:
> on-disk bitmap indexes
> d-trace / GMF
> stored procedure debugging
> standby databases (new PITR)
> Anyone know anything else? Performance (sort, joins, etc) improvements would
> be worth mentioning but *only* if someone can attach a figure to them.
updateable view? if the bernd's patch get applied of course...
--
regards,
Jaime Casanova
"Programming today is a race between software engineers striving to
build bigger and better idiot-proof programs and the universe trying
to produce bigger and better idiots.
So far, the universe is winning."
Richard Cook
On Wed, 2006-08-02 at 21:59 -0700, Josh Berkus wrote: > Well, it's that time again. In the next couple of days, we'll be starting to > draft the 8.2 release. It might be worth waiting till the set of features that will actually be in 8.2 is well-defined -- that is, until the patch queue has been cleared (circa the end of August). For example, updateable views and bitmap indexes may or may not make it into the release, last I checked. > FEATURES: AFAICT, there are only four big-name features worth mentioning in > the release: > on-disk bitmap indexes > d-trace / GMF > stored procedure debugging > standby databases (new PITR) > Anyone know anything else? Performance (sort, joins, etc) improvements would > be worth mentioning but *only* if someone can attach a figure to them. There's been quite a lot of useful performance work: off the top of my head, Tom's work on SMP scalability, for example. I'd expect Tom's outer join query optimization work will help a lot of people. The improvements to sorting performance are also likely to be significant. I don't have figures, although we can surely produce some before the release if need be. Online index builds will be very useful for some people. -Neil
Josh Berkus wrote:
> All,
>
> Well, it's that time again. In the next couple of days, we'll be starting to
> draft the 8.2 release. Here's what needs to be done to draft the release:
>
> THEME: Bruce has suggested "usability" since this is a bit of a
How about "polish" or "maturity"?
Joshua D. Drake
--
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
On Thursday 03 August 2006 01:57, Joshua D. Drake wrote:
> Josh Berkus wrote:
> > All,
> >
> > Well, it's that time again. In the next couple of days, we'll be
> > starting to draft the 8.2 release. Here's what needs to be done to
> > draft the release:
> >
> > THEME: Bruce has suggested "usability" since this is a bit of a
>
> How about "polish" or "maturity"?
>
One theme is that this release should make it easier to migrate to postgresql,
by providing syntax similar to that in other databases. (default updatable
views, multiple single inserts, multiple single update syntax, if exists,
etc...) of course that list is dependent on what actually gets in :-)
--
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL
Josh Berkus wrote: > FEATURES: AFAICT, there are only four big-name features worth > mentioning in the release: > on-disk bitmap indexes > d-trace / GMF > stored procedure debugging > standby databases (new PITR) > Anyone know anything else? Performance (sort, joins, etc) > improvements would be worth mentioning but *only* if someone can > attach a figure to them. Most of these features are still in doubt. You won't get a useful feature list before beta. -- Peter Eisentraut http://developer.postgresql.org/~petere/
josh@agliodbs.com (Josh Berkus) writes: > FEATURES: AFAICT, there are only four big-name features worth mentioning in > the release: > on-disk bitmap indexes > d-trace / GMF > stored procedure debugging > standby databases (new PITR) > Anyone know anything else? Performance (sort, joins, etc) improvements would > be worth mentioning but *only* if someone can attach a figure to them. Did the "WITH clause" (recursive queries) make it in? That might be worth mentioning. It probably isn't worth mentioning "multiple tuples in one INSERT"... -- let name="cbbrowne" and tld="cbbrowne.com" in name ^ "@" ^ tld;; http://www.ntlug.org/~cbbrowne/spreadsheets.html Rules of the Evil Overlord #80. "If my weakest troops fail to eliminate a hero, I will send out my best troops instead of wasting time with progressively stronger ones as he gets closer and closer to my fortress." <http://www.eviloverlord.com/>
On 8/3/06, Chris Browne <cbbrowne@acm.org> wrote: > Did the "WITH clause" (recursive queries) make it in? Nope, I didn't have time to get that one into a good enough state for submission. -- Jonah H. Harris, Software Architect | phone: 732.331.1300 EnterpriseDB Corporation | fax: 732.331.1301 33 Wood Ave S, 2nd Floor | jharris@enterprisedb.com Iselin, New Jersey 08830 | http://www.enterprisedb.com/
On Thursday 03 August 2006 12:01, Chris Browne wrote: > Did the "WITH clause" (recursive queries) make it in? That might be > worth mentioning. It probably isn't worth mentioning "multiple tuples > in one INSERT"... I've been thinking about that: it would be good for the MySQL community to know that we would now be supporting multiple tuples in one INSERT.
Robert Bernier wrote:
> On Thursday 03 August 2006 12:01, Chris Browne wrote:
>> Did the "WITH clause" (recursive queries) make it in? That might be
>> worth mentioning. It probably isn't worth mentioning "multiple tuples
>> in one INSERT"...
>
> I've been thinking about that: it would be good for the MySQL community to
> know that we would now be supporting multiple tuples in one INSERT.
>
Yes it would.
Joshua D. Drake
> ---------------------------(end of broadcast)---------------------------
> TIP 2: Don't 'kill -9' the postmaster
>
--
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Joshua D. Drake wrote: > Robert Bernier wrote: > >> On Thursday 03 August 2006 12:01, Chris Browne wrote: >> >>> Did the "WITH clause" (recursive queries) make it in? That might be >>> worth mentioning. It probably isn't worth mentioning "multiple tuples >>> in one INSERT"... >> >> I've been thinking about that: it would be good for the MySQL >> community to know that we would now be supporting multiple tuples in >> one INSERT. > > Yes it would. Actually I think people will find some interesting uses for the related "VALUES clause in the FROM clause" capability too. Joe
Chris Browne wrote: > It probably isn't worth mentioning "multiple tuples > in one INSERT"... On the other hand if it's worded "Initial support for SQL standard feature id F641 (row value constructors)", that is one area where database comparisons ding postgresql today [1]. [1] http://troels.arvin.dk/db/rdbms/#insert-multiple
Neil, > It might be worth waiting till the set of features that will actually be > in 8.2 is well-defined -- that is, until the patch queue has been > cleared (circa the end of August). For example, updateable views and > bitmap indexes may or may not make it into the release, last I checked. We can swap out individual features as things are finalized. The feature part of the release actually requires less work than some of the other parts. Historically, it's taken about a month to compose the release and a month or more to translate it. So we can't wait until the beginning of August to start in case we have a short beta. > There's been quite a lot of useful performance work: off the top of my > head, Tom's work on SMP scalability, for example. I'd expect Tom's outer > join query optimization work will help a lot of people. The improvements > to sorting performance are also likely to be significant. I don't have > figures, although we can surely produce some before the release if need > be. Yes -- volunteers? Anyone? I have some figures from Sun but I'm not sure I can publish them. > Online index builds will be very useful for some people. Oh, right. --Josh
+1 on behalf of O'Reilly and other internet companies dumping from MySQL with intent to load into Postgres. The most popular mysqldump format emits the "INSERT INTO foo VALUES (...),(...),(...)..." - Luke On 8/3/06 10:01 AM, "Robert Bernier" <robert.bernier5@sympatico.ca> wrote: > On Thursday 03 August 2006 12:01, Chris Browne wrote: >> Did the "WITH clause" (recursive queries) make it in? That might be >> worth mentioning. It probably isn't worth mentioning "multiple tuples >> in one INSERT"... > > I've been thinking about that: it would be good for the MySQL community to > know that we would now be supporting multiple tuples in one INSERT. > > ---------------------------(end of broadcast)--------------------------- > TIP 2: Don't 'kill -9' the postmaster >