Directions in Advocacy

Hi all,

I've been thinking about advocacy for a fair while now. Here are some
ideas. What follows is going to be critical :-(.

1) Advocacy Strategy

Currently, there doesn't seem to be any strategy about PostgreSQL
advocacy. I think a strategy (ie, TODO list) will allow those interested
in Postgres advocacy to see where we are and where we want to be. It will
encourage a bit of creativity, track progress, determine whether or not
something is working or not.

2) Case studies

There aren't a hell of a lot of case studies on the advocacy Web
site. There is no information about how to add your *own* case study.

I think the case studies need to be augmented by smaller testimonials
which the advocacy group has *permission* to use/print/what ever. This
will provide a resource for people putting together papers on PostgreSQL
for specific applications: education, private sector, government, etc. It
will help consultants choose testimonials specific to a tender that
they're putting together. The ability to show a client that someone else
is using this for the same application is very handy.

How do we get these testimonials? Not a week goes by without someone
writing to pgsql-hackers about how much their shop loves Postgres. Someone
needs to follow this person up, go through the process of getting a
testimonial that their employer signs off on. This can take some
convincing, but we must stress that 1) they're getting a lot out of
Postgres, why not return some 2) Their testimonial will lead to increased
usage which leads to an even better product in the future.

3) Advocacy Web site

Style

I think the advocacy Web site is preachy. I don't like things like "Join
the PostgreSQL revolution", "Better support than the proprietary vendors",
"Legendary reliability and stability" and:

"Unlike many proprietary databases, it is extremely common for companies
to report that PostgreSQL has never, ever crashed for them in several
years of high activity operation. Not even once. It just works."

This is a huge call. Where's the evidence? Certainly not in the case
studies ;-). This kind of thing doesn't play well when the other
professional requirements are missing (see below).

Advantages of Postgres

i) Immunity to over-deployment

This is a *huge* point. It isn't covered properly.

Over-deployment is what some proprietary database vendors regard as
their #1 licence compliance problem. With PostgreSQL, no-one can sue
you for breaking licensing agreements, as there is no associated
licensing cost for the software.

It rarely happens that companies like Microsoft, Oracle, IBM and Sybase
*actually* sue people to recover unpaid license. Their software is
designed to only run for the number of users/CPUs you have purchased. The
end.

I think that this would be better covered as total cost of owner ship. No
licensing, cheaper consultants, runs on cheaper hardware and cheap
operating systems, has the kinds of features you only find in
expensive databases, well documented (for free), etc.

ii) Legendary reliability and stability

A huge call with nothing to back it up.

iii) Extensible

This, incorrectly, looks at source code level extensions. What about the
depth of extensibility which can be added at the SQL level by unskilled
programmers? (ie, they don't know C). User defined functions, triggers,
rules. If they know C, they can go even further. How about mentioning all
the supported procedural languages?

iv) Designed for high volume environment

MVCC doesn't just allow this. Its designed to allow concurrency users to
be isolated. Its about data integrity. People using Access and looking for
a better database *really want to know about this*.

v) ANSI compliance

Since when? I don't know of any database which is ANSI compliant.

Web site URL

advocacy.postgresql.org? Why not just integrate it into the existing Web
sites? Pointing someone to this URL suggests to them that they shouldn't
believe most of what they read ;-).

4) Professionalism

The real problem with the advocacy site is that it is unprofessional. It
is not clearly/correctly targeted. It uses the kinds of buzzwords you
would read in the latest copy of CIO magazine, but it doesn't come through
with the goods. It also isolates, by doing this, 95% of the people
evaluating Postgres: actual developers, not management. Here's what's
missing.

i) URL to press releases

ii) URL listing events/conferences which will involve Postgres. This will
require the people speaking at conferences to send in
dates/times/URLs. This can only serve to increase attendance numbers and
impress potential users with Postgres's popularity :-)

iii) List of awards Postgres has won

iv) !!! Media kit !!! If you want to get written about, you need to have a
media kit. Most IT journalists know little to nothing about IT. They will
not write about Postgres if it is too hard. A media kit has the following:

a) Overview of the project, recent history, etc
b) fact sheet: what Postgres is, latest version number, companies
involved, how it is funded, major developers, estimate of number of users
c) URL to testimonials
d) URL to awards
e) URL to press release for most recent release
f) Dumbed down list of features introduced in most recent release
g) URL to *printable logos* (300 dpi, CMYK'd, list of pantone colours)
h) Trademark information/list of names (PostgreSQL? PostgreSQL,
inc? PostgreSQL Global Development Team? .. which one?)
i) Contact details

v) Developer's section

Most open source software comes in through the back door. Developers are
going to find and evaluate Postgres. We need to tell them, in their
language, what Postgres has and what it can do. When then need to give
them the information to answer management questions. The current site
addresses only middle management types.

vi) URLs to recent papers on Postgres

5) Future directions

Here's a list of things going on now/ideas for the future to kick off the
advocacy strategy.

i) Big Nerd Ranch

Bruce is training people at a multi-day event called 'Big Nerd
Range'. This is *dedicated* to PostgreSQL. Though it is limited to 20
places it is a serious training event and will do a lot to advocate
PostgreSQL -- provided, of course, there is the right follow up: a Press
release issued in conjunction with the Big Nerd Ranch people, information
on the PostgreSQL Web site, information for the media.

See: http://www.bignerdranch.com/Classes/Postgresql.html

ii) PostgreSQL mini-conference

Christopher Kings-Lynne and I will be organising a one day PostgreSQL
conference to take place on the 20th of January. This will be attached to
the Linux.Conf.Au conference, where I will be giving a more technical
tutorial on Postgres.

This will force Christopher and I to create resources (presentations,
logistical information) which will allow other people to replication the
conference in other countries.

http://www.linux.conf.au

iii) Enterprise Solution Center

A company called Wild Open Source (http://www.wildopensource.com) is
working with Linux International to create a kind of opensource pavilion
called 'Enterprise Solution Center'
(http://www.wildopensource.com/ESC/index.html). The goal of this project
is to deploy open source software in all roles inside a company: desktop,
back office, productivity, accounting as well as the obvious ones, like
Web site, file server, mail server, etc. We *have* to get involved in
this. Someone will need to contact Wild Open Source to see what we need to
do to get involved. The Centre will first come into action at LinuxWorld
New York, 21-24th January. It would be best if the person(s) who contact
Wild Open Source can actually attend this conference. It will, however,
need to be a group effort to determine what software will run on top of
Postgres, actually setting it up, etc. Regardless, the hard work will
definitely be worth it, since the pavilion will be redeployed around the
world, as far as I can tell.

iv) Database taxonomy project

The DBMS Reference Guide is an idea put together by Zak Graent
(zak(at)php(dot)net). Zak is an employee of MySQL AB. We get together three or
four times a year and talk databases. Zak is a firm believer in the idea
that competition in open source can only serve to increase the market for
it, there by benefiting all competitors. So, even though he works for
MySQL AB, a lot of what he does serves us all, because it gets people
talking about open source databases.

The project is designed to provide guidelines to developers about major
databases. Amongst other things, this will be an invaluable resource for
people wanting to migrate databases (hopefully to Postgres). Better yet,
Zak tells me he has had plenty of interest from developers at proprietary
companies. If you're interested, email him.

v) Conferences/talks/tutorials

I like to talk. I talk at conferences, at universities, at user group
meetings, at pubs. If you like to talk, consider submitting proposals to
one of the many open source orientated conferences; or at the university
you attended; or at a local open source/UNIX focussed user group.

For those interested in conferences, here is a brief list of those I have
spoken at/know of:

EgovOS (March, Washington) - http://www.egovos.org

O'Reilly Open Source Convention (July, Oregon) -
http://conferences.oreilly.com

FOSDEM (February, Brussels) - http://www.fosdem.org

Linux.Conf.Au (January, Perth) - http://linux.conf.au

UK Unix Users Group Developers Conference (June/July, Edinburgh) -
http://www.ukuug.org/events/linux2003/

LinuxTag (July, Karlsruhe, Germany) - http://www.linuxtag.org

International PHP Conference (April/November, Amsterdam/Frankfurt) -
http://www.php-conference.de)

SAGE-AU (August, Hobart, Australia) - http://www.sage-au.org.au

It would probably be useful to provide guide lines to those who are
thinking about this. That is, how to submit a proposal, how to prepare the
slides, how long a talk should go for, etc.

vi) Write tutorials for IT magazines

As I said before, most IT journos know little to nothing about IT. If they
knew more, they'd stop being a journalist and start earning real
money. Its a simple fact. This means that when IT mags need more indepth
stuff, they have to farm it out to freelancers. If you know how to write
and have some experience, write to the editors of some magazines published
in your country offering to write about PostgreSQL. Particularly given the
release 7.3, now would be a great time.

A word of advice: suggest possible articles: 'What's new in PostgreSQL
7.3', 'Installing and running PostgreSQL 7.3', 'Open Source Database
Systems', etc.

vii) Write papers/Documentation/etc

If you're using PostgreSQL for some specific purpose, if you're really
interested in, say, "PostgreSQL in Education", write a paper about it. No
matter how long or short, the written word carries weight. The more we
increase the knowledge base surrounding PostgreSQL, the bigger it will
become.

---------------

That's it for now.

Thanks,

Gavin