PostgreSQL Training,
Consulting & Support
General Bits
By A. Elein Mustain

09-Aug-2004 Issue: 81

Archives | General Tidbits | Google General Bits | Docs | Castellano | PortuguÍs | Subscriptions | Notifications | | Prev

General Bits is a column loosely based on the PostgreSQL mailing list pgsql-general.
To find out more about the pgsql-general list and PostgreSQL, see

All About O'Reilly Open Source Conference
Overview 01-Aug-2004

As with all events, we wish you could have been there. So here are some Pictures to help you visualize some of the people you see on the lists and IRC all of the time.

The handouts included the PostgreSQL and Slony-1 CD (BitTorrent version available at created by Robert Bernier and funded by Affilias, PostGIS's nice brochure and SRA's windows version of Powergres on CD. Flyers sponsored by Varlena, LLC were: General PostgreSQL Flyer (pdf) and PostgreSQL and the Open Source Maturity Model (pdf). Slony stickers and other little gadgets generously by Affilias and Jan Wieck.

The PostgreSQL track talks were very successful. The ones I attended were well prepared and full of useful information. Some of the presentations are available in Tidbits and more will be added as speakers give me permission to do so. Some presentations are also available on Techdocs. The official OSCON presentations are also now available but the site is incomplete.

A high point for me, personally, was that O'Reilly did not close the women's restroom on the main floor as they did last year. And in fact, I encountered a line waiting for the restroom. This is a very good sign for more diverse participation. I and other women felt more welcome and integrated into the hubbub of the conference. And some men I spoke with were happy that more women were attending.

Editor: elein at
Comments from attendees
PostgreSQL People at the Conference 01-Aug-2004

How was our visibility at OSCON this year?

Visibility was very good, thanks to pre-planning by the pg team.
Bill Harper
Anyone see the mozilla and gnome folks (hint: they were right behind our crowd)
Reece Hart
This year we bootstrapped ourselves into the "everything but Perl" echelon of visibility, on a par with MySQL, Linux, or Python for "mindshare".
Josh Berkus
I've attended OSCON now for 4 years in a row. Each time the visibility of the Postgres community seemed to gradually grow -- until this year when it seemed to explode relative to last year. I was very encouraged and happy/honored to be part of it all.
Joe Conway
As an OSCON freshman, it seemed very good; at least as good as mysql. It was unfortunate that in at least one keynote, the speaker (the gentleman from Novel) equated open source DB==mysql. This also seemed to be the case in a couple of the PHP sessions I attended.
Bill Montgomery
Fantastic!! Lots of people around the booth. What more could you ask for (more booth space?).
David Reese
Others chimed in with "Extremely good" and "Great!". For me it was telling that Robert Lefkowitz, one of the key note speakers, reported that he kept "running into PostgreSQL people everywhere." This is particularly good to hear from someone both in the upper echelon and outside of the immediate PostgreSQL circle.

What "aha!" moments did you have?

When several people stood up during the BOF [Birds of a Feather]-- including sys admins from the City of Garden Grove and the National Gallery -- and said, "PostgreSQL is great and **you can quote us on that.**" This is a huge advance from last year; I can tell you for the 7.4 release that I emailed over a dozen known corporate PostgreSQL users and got the same "I'd like to give you a testimonial but my boss won't let me."

What this means is that, like Linux, we are no longer considered experimental technology by everyone in business, but rather as a legitimate database option which they're not ashamed to admit to.
Josh Berkus

Seeing the details of Slony-1 was an eye opener. It is a killer system that rivals (beats out, actually) commercial ones I've seen.
Joe Conway
When I realized that Slony-1 could be considered the "general solution" to single-master asynchronous database replication for PostgreSQL. I can't give enough kudos to that team, and I'm excited to get involved with them and deploy the software.
Bill Montgomery
Definitely Joe's extension workshop on the first day. Just the detailed reminder of all the things one can do right inside the database. The database can be turned into a full-on application integration environment, if one is so inclined. I thought -- hey, our GIS data file loader / dumper utilities do not have to be external programs, they could be built-in functions that return record sets!
Paul Ramsey
Greg's talk about playing nice with developers (the "documenting in the db" bit was great). Also, the discovery of so many Informix porters. (Anyone else get the impression that Informix users are getting pressure to go away?)
Andrew Sullivan
Aha, so this is Jan Wieck. He is a lot nicer in person than the impression you get from his emails. Jan is a one of a kind. In the replication panel, Jan said something like "I didn't prepare anything because if i did we would be here for a couple of days."
George Essig
Opening presentation by O'Reilly and Lefkowitz completes my belief that open source is headed into the enterprises in a big way. Last year, there were probably 80% macintosh laptop users. This year there were numerous (30-40%) PC laptops. From what I could see, I believe 50% of those laptops were running Windows. Since this is the enterprise laptop model, it's good to see.
David Reese

Funniest thing that happened.

When David Fetter and [Josh] were eating breakfast in the American Grill (upstairs from the convention) and the waitress, approaching from the back, called us "Sir and Miss" (I [Josh] have long blond hair).

Second funniest moment was having 2 engineers in MySQL shirts at our booth.
Josh Berkus

Not exactly funny, and not PostgreSQL related, but riding an elevator with Larry Wall, playing pool with ESR, standing in line for food with Ken Coar; being my first OSCON it did seem funny to me to be in such proximity to the people I've read about for years, and whose software I rely on every day, but have never even seen.
Bill Montgomery
For me the moment I walked into the exhibit hall and saw the PostgreSQL poster (World's Most Advanced Open Source Database) then turned around and saw the MySQL poster (World's Most Popular Open Source Database). At least everyone knows where they stand. :)
Paul Ramsey
Robert Treat and Greg Sabino Mullane playing hacky sack in the middle of a couple hundred people waiting for the next presentation.
George Essig

What was the best talk you attended (and why)? And the worst? (and why?)

"Best" can mean more than one thing. I'd break it into at least these three categories:
- "Coolest" -- Paul Ramsey's Spatial Smarts with PostGIS
- "Most Interesting" -- Reese Hart's Structural Biology
- "Most Enlightening" -- Jan Wieck's Slony-1 workshop
Joe Conway
The most valuable (if that means best) was Jan's Slony workshop. Chris Shiflett's PHP security sessions were very good, too. I can't single out the worst talk, but it was very frustrating to spend 45 minutes listening to someone who was clearly not prepared.
Bill Montgomery
Greg Sabino Mullane's tutorial about keeping everything close to the data in the database was thought provoking. I think it's difficult to do this in practice though.
George Essig
Best: Plone tutorial. Went over installation, configuration, implementation and enhancements. This technology 'looks' good.

Worst: Java discussion panel moderated by Tim O'Reilly. It was the worst bantering session led by the open source people and covered by Sun (Tim let it get out of control). It would have been nicer to have the panel just answer questions instead of arguing amongst themselves.
David Reese

Comments, Reviews of any Pg talks

Bruce did a great job of summarizing the roadmap for 7.5. I was particularly impressed by the pragmatism of 7.5's features set. I often glibly dismiss Windows and tell people to anchor their boats with such boxes. But he's right: PostgreSQL must be taken to those users on their turf. This kind of big picture pragmatism is essential.
Reece Hart
Be animated! Get up, walk around, gesture! If you show your audience that you are excited about the subject matter, some of them will get excited too.

Oh, and don't include distracting pictures in your slides. Learned that one the hard way. :-(

We also had somewhat more people at the State of PostgreSQL than the State of MySQL had; I think this was partly timing, but it's still not a bad sign. Both "State of" talks were somewhat sparsely attended though.
Josh Berkus

Josh also reports that Robert Bernier's moderator role in the Advanced Search talk worked particularly well.
I thought 'Build Your Own MapQuest! Adding Spatial Smarts with PostGIS' by Paul Ramsey highlighted the power of PostgreSQL's custom data types. PostGIS implements geographic data types for PostgreSQL. I don't understand how other databases could implement a similar feature.
George Essig

Comments, Reviews of any other talks

It seemed like a mini-theme of the conference was virtual machines. There were several talks on .NET, Mono, and Parrot.
George Essig

Comments on the PostgreSQL Birds of a Feather (BOF)

We had too much competition for the evening hours on Wednesday. I thought that Elein ran the BOF well and prevented it from becoming a dialog between the PG hackers with the newbies excluded.
Josh Berkus
BOF was nice, and it was particularly nice to hear all the testimonials. I thought that was a good format, and Elein moderated it well.
Paul Ramsey

Things that went well for PostgreSQL (iy(ns)ho). Things that did not go well (iy(ns)ho).

I was impressed at how many people showed up for booth duty & at the booth--lots and lots of pg tracks.
David Fetter
Best thing was interest by good-sized IT shops considering Pg. ... The *strongest* impression from the conference for me was the silent interest in Pg. Representatives from some organizations considering Pg spoke up at various sessions. Outside of sessions, I heard from two organizations who were considering Pg and they were both impressed with the technology.
Bill Harper
The exhibit hall was great. What really worked for us was that our people hung out at our table, which means that there were always enough people there to attract more people. One pivotal moment was when we had about 15 people in front of the table, and someone from the University of Wisconsin walked up to me [wondering what the fuss was about and if we had free beer or something].

This is not just a matter of outreach to the people you talk to. Michael Tiemann (Red Hat) and VPs from IBM and Novell, and Jay Lyman (NewsForge) and Dan Gilmour (SJMN) were watching our booth. Having lots of people interested looks really, really good for us; it shows to others that we have "momentum" and can't be ignored.
Josh Berkus

The Postgres community was well represented, and seemed to have as many people and as much interest as the MySQL crowd at least from my vantage point). However, I got the impression that O'Reilly is closely aligned with MySQL (actually have had that impression for some time now). I hope someone will impress on O'Reilly that they ought to at least treat the two communities as equals, e.g. give us proper billing on their onlamp site, etc.
Joe Conway

What worked well at the Booth (and what did not)

Assertive presence by at least one person at the booth at all times was good for Pg. Pg people were generally in front of the table in order to ensure contact with passers by. (The Mozilla booth was almost smothered by the Pg people!)
Bill Harper
I was glad we had stuff to give away. It allowed us to thrust something at everyone approaching, which gave us an opportunity to talk to them & find out what they're doing.

For what it's worth, this was my second OSCON, and I've been terribly happy with both of them. I feel that I learn a great deal there, mostly due to all the folks who show up and who are much smarter than I.
Andrew Sullivan

It was great to have a booth, banner, demo CD, and T-shirts. My only suggestion is to have the demo CD running on a notebook at the booth.
George Essig
The demo CD was up and running much of the time, but that I guess it should have been more constant and visible.
Paul Ramsey's GIS demo looked great. On seeing it, several people including Reece Hart realized that their work would also demo well. Next year we should also call for demo-able applications.

We must recognize that Affilias' generosity of providing the CD copies, Slony stickers and gadgets as well as their people's time contributed to the success of the booth and the conference for PostgreSQL.

And several people swear they would have worn the elephant costume if we had one.

Interesting and related Blogs

Contributors: Bill Harper bharper at, Josh Berkus josh at, Reece Hart hart.reece at, Joe Conway mail at, Bill Montgomery billm at, Paul Ramsey pramsey at, George Essig george_essig at, David Reese dmreese at, elein at

Open Source In Action
How It Works Well 01-Aug-2004

There was some discussion brought up by Michael Tiemann of Red Hat at Bruce Momjian's State of PostgreSQL talk regarding the planning methods that we use to organize our activities and progress.

Companies and corporations usually get directives handed down from marketing to engineering for the decisions on what to do for the next release; this is usually called Top Down Planning. In PostgreSQL, we use "bottom up" planning where features are driven by people's interest in having the features and learning the code. Our development is also dispersed, unfunded or funded by various companies rather than one company. So for many reasons the "bottom up" method works well for us.

In my experience at Ingres (Relational Technology, Inc.), Illustra and Informix development was more engineering driven than marketing driven. In fact these companies were known for "Great Technology, Lousy Marketing." These days in PostgreSQL, we have an additional marketing support by the grass roots Open Source Community via the pgsql-advocacy list. This enables us to have our appropriately managed, great technology while we work harder to create a market presence.

Research done by Karim Lakhani on several Open Source projects, including PostgreSQL, shows the efficiency of more openly and lightly managed projects. Quality and innovation are products of enthusiasm for good technology. Good community carries much more weight than management directives. bs innovations implemented in a timely manner. Research by Rix Groenboom presented at NLUUG: Open Source in Business 2004) has also shown fewer bugs and higher quality code in Open Source projects.

PostgreSQL's technology is widely accepted as better technology. This tells me that we are using the right planning technique for our community.

Linux World 2004
Linux World San Francisco 08-Aug-2004

We have Pictures from Linux World, too! In the next issue of General Bits we will have more comments and ideally more pictures from Linux World in San Francisco.

Contributors: Josh Berkus josh at elein at
PostgreSQL 8.0 Goes Beta!
Time to down load and test! 9-Aug-2004

PostgreSQL Version 8.0 is ready for Beta! The downloads should be available from the PostgreSQL Mirrors on Monday August, 9.

You should be downloading and preparing to test your applications on this significant release of PostgreSQL. Your participation will enable a solid and timely production release!

Yes, it is 8.0 and not 7.5. And do not forget to review the RELEASE NOTES.

Contributors: elein at

Comments and Corrections are welcome. Suggestions and contributions of items are also welcome. Send them in!
Copyright A. Elein Mustain 2003, 2004, 2005, 2006, 2007, 2008, 2009

Search General Bits & Search WWW