Re: PostgreSQL flamage on Slashdot

-----Original Message-----
From: Merlin Moncure [mailto:merlin.moncure@rcsonline.com]
Sent: Tuesday, March 18, 2003 5:34 PM
To: Jason Hihn
Cc: PostgreSQL Advocacy
Subject: RE: [pgsql-advocacy] PostgreSQL flamage on Slashdot

...

MySQL works just fine for your typical web app and is optimized for that
purpose. For example, they added a query caching feature which IMHO is
useless except in a web environment. Their catering to the web crowd
has helped establish them in the public eye.

Postgres is the database of choice for business apps. This is a tough
market to crack because of the conservatism of business planners and
their well funded ability to buy top dollar software. Databases are
like boxing: if two players go the distance, the judges give it to the
champ.

I think in the long run, the 'business first' strategy will prevail.
Database developers who can solve business problems usually make a lot
more money than web developers. Since the dot.com fallout, this
disparity has increased. This means a despite a smaller user base, the
postgres community will be better funded and enjoy more of the top
talent. Also, postgres has enjoyed and will continue to enjoy better
support from the academic community because of its closer adherence to
the relational model, maybe the closest of any major dbms.

You make an astute observation that I think should become a strategy of the
advocacy team. That is to portray MySQL as a "hobby" database, but Postgres
as a "production" database. I think this is a very easy stance to take,
since I've always thought that. Anyone caught arguing that MySQL is better
will show how much they don't know about Postgres or real databases.

Also with this stance, the two DB's will be able to cooperate and help
publicize OSS successes without having to cut each other down. Then though,
we must concede that MySQL is "good enough" for at least "armature" (hobby)
usage; and that is another statement that I agree with.

As for the companion apps, Postgres is much less reliant on external
tools than other dbs because of the extremely powerful client
interfaces. I would like to see better dts services and better xml
support (xml/edi) for adoption by businesses. A win32 port is
absolutely crucial to start zapping sql server, which has virtually no
direct competition for 2nd tier database deployments, and also to the
embedded market which now relies on mysql, access, or FoxPro.

Both win32 and replication are big "requirements". I can delay the win32
requirement for a while, and I can suffer through huge pg_dumps for the time
being. If that wasn't the case, I'd be else where!

Also, integration with Delphi/Kylix could use some press. This is a
fantastic rad environment for win32/linux.

I've heard of it, but never used it. I took a look long enough to know that
I'd need to take a longer look later ;-)

-J

Guys,

You make an astute observation that I think should become a strategy of the
advocacy team. That is to portray MySQL as a "hobby" database, but Postgres
as a "production" database. I think this is a very easy stance to take,
since I've always thought that. Anyone caught arguing that MySQL is better
will show how much they don't know about Postgres or real databases.

And why are we "competing" against MySQL, exactly?

Look, the people who use MySQL aren't going to use PostgreSQL. They're
looking for a fast, simple database with no DBA requirements, which IS NOT
POSTGRESQL. Such databases have their niche just as we have ours.

Our "competitors" are MS SQL, SQLAnywhere, Oracle, and DB2. Business-class
databases. The tech press likes to focus on MySQL vs. PostgreSQL because
they haven't caught up to the idea that an OSS database could compete with
commmercial offerings. When *you* focus on MySQL vs. PostgreSQL, YOU ARE
BUYING IN TO THEIR IGNORANCE, and helping the press compartmentalize Postgres
as an alternative to MySQL.

Now, enough, already.

--
Josh Berkus
Aglio Database Solutions
San Francisco

-----Original Message-----
From: Josh Berkus [mailto:josh@agliodbs.com]
Sent: Wednesday, March 19, 2003 12:20 PM
To: Jason Hihn; Merlin Moncure
Cc: PostgreSQL Advocacy
Subject: Re: [pgsql-advocacy] PostgreSQL flamage on Slashdot

Guys,

You make an astute observation that I think should become a

strategy of the

advocacy team. That is to portray MySQL as a "hobby" database,

but Postgres

as a "production" database. I think this is a very easy stance to take,
since I've always thought that. Anyone caught arguing that

MySQL is better

will show how much they don't know about Postgres or real databases.

And why are we "competing" against MySQL, exactly?

Good question.

Look, the people who use MySQL aren't going to use PostgreSQL. They're
looking for a fast, simple database with no DBA requirements,
which IS NOT
POSTGRESQL. Such databases have their niche just as we have ours.

I don't believe that is the case. I think many people are being mislead
through it's frequent mention in the press.

Our "competitors" are MS SQL, SQLAnywhere, Oracle, and DB2.
Business-class
databases. The tech press likes to focus on MySQL vs. PostgreSQL because
they haven't caught up to the idea that an OSS database could
compete with
commercial offerings. When *you* focus on MySQL vs.
PostgreSQL, YOU ARE
BUYING IN TO THEIR IGNORANCE, and helping the press
compartmentalize Postgres
as an alternative to MySQL.

The two share the same development style (open source). Unfortunately, those
who don't already know, read these articles and get the wrong impressions. I
know Postgres is more appropriately compared to Oracle, but they don't. If
they see all the MySQL press and then see how bad it is as a DB, they will
probably assume that Postgres falls into the same category.

And that is the key. That's why it's not enough already - Because newbies
often rely on publicity to make decisions for them. That's why we have the
advocacy group.

And sir, when I "buy into their ignorance" it is only to grow the dichotomy
of the two. Maybe I am wrong, maybe we shouldn't have a MySQL separation
effort. Maybe we should just let ourselves be lumped in with them. *Shudder*
I think not. Therefore we must engage in these "battles" against MySQL. I
think we'd be worse off not fighting them.

We could go on battling away against MS SQL, Oracle, DB2, but people will
still ask "How does it compare to MySQL?" That will be the case for everyone
who works with DBs as a hobby and not professionally.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

And why are we "competing" against MySQL, exactly?

Look, the people who use MySQL aren't going to use PostgreSQL. They're
looking for a fast, simple database with no DBA requirements, which IS NOT
POSTGRESQL. Such databases have their niche just as we have ours.

Sorry, but I must disagree with this. MySQL is as much our competitor as
Oracle is. Simply look at the number of open-source apps that use
MySQL instead of PostgreSQL to see why this is so. I also see plenty
of crossover as far as being a DBA is concerned: real DBAs are using
both products, everyday, on a wide variety of technical projects. People
who use MySQL *are* going to use PostgreSQL - the conversion happens
all the time. We should be encouraging the use of Postgres in all niches.

Remember, having a successful product is about much more than being
the best from a scientific/technical standpoint. Decisions are made
on a number of factors, such as supported platforms, market saturation,
availablity of skilled workers, interface tools, use in other
existing products, commercial support, etc. Every decision on which
database system to use affects the Postgres project, directly or indirectly.

I'm also tired of the attitude that MySQL is somehow "beneath" us,
and a "hobby" database. Sure, PostgreSQL is more feature-rich, but
that growth is asymptotic, and MySQL is moving forward and adding
features all the time. In almost every other aspect except the actual
database engine, MySQL is the better project. MySQL are PostgreSQL
are going to grow together quicker than they grow apart. We should
treat them with respect as a friendly rival.

- --
Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200303191259

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE+eLNavJuQZxSWSsgRAu6+AJ9UFXGvOAJUUWj9BP0EfgQbOac17wCgwyOJ
IWcmrQ4EZGT5YdM1a86zpQ8=
=XDI2
-----END PGP SIGNATURE-----

Jason, Greg,

The two share the same development style (open source). Unfortunately, those
who don't already know, read these articles and get the wrong impressions. I
know Postgres is more appropriately compared to Oracle, but they don't. If
they see all the MySQL press and then see how bad it is as a DB, they will
probably assume that Postgres falls into the same category.

<snip>

And sir, when I "buy into their ignorance" it is only to grow the dichotomy
of the two. Maybe I am wrong, maybe we shouldn't have a MySQL separation
effort. Maybe we should just let ourselves be lumped in with them. *Shudder*
I think not. Therefore we must engage in these "battles" against MySQL. I
think we'd be worse off not fighting them.

But you see, that's exactly my point. We will *not* differentiate ourselves
from MySQL by slamming them at every opportunity. We will only build
PostgreSQL's user base by championing Postgres' features and
standards-compliance.

To give you an analogy: Three years ago, Willie Brown and Client Reilly were
the two lead candidates for Mayor of San Francisco. They spent the first six
months of their campaign slinging mud at each other ... to the point that
many San Franciscans said that they would stay home from the polls rather
than vote for either of them. Then, on the last day of the filing deadline,
Tom Ammiano announced that he was running ... and grabbed 49% of the runnoff
vote.

Or another analogy: Microsoft recently hired a PR think tank to analyze their
efforts to stamp out Linux through their various FUD campaigns against Linux,
OSS, and the GPL. The think tank found out that MS's attempt to slam Linux
had, in fact, made people more aware of OSS, and more interested in it, than
they would have been on their own.

I think you can see the analogy, yes? We should promote Postgresql by
promoting Postgresql ... NOT by slamming MySQL.

--
-Josh Berkus
Aglio Database Solutions
San Francisco

On 19 Mar 2003 at 9:20, Josh Berkus wrote:

Guys,

You make an astute observation that I think should become a strategy of the
advocacy team. That is to portray MySQL as a "hobby" database, but Postgres
as a "production" database. I think this is a very easy stance to take,
since I've always thought that. Anyone caught arguing that MySQL is better
will show how much they don't know about Postgres or real databases.

And why are we "competing" against MySQL, exactly?

Look, the people who use MySQL aren't going to use PostgreSQL. They're
looking for a fast, simple database with no DBA requirements, which IS NOT
POSTGRESQL. Such databases have their niche just as we have ours.

I have a growing feeling that the basis of perception that postgresql is slow
stems from the arcane and age old defaults postgresql ships with. I know that
they will be bumped in the next release but that does not help much to heal the
mind share loss over the years..

Besides postgresql requires shared memory which is a system resource. I don't
know much about mysql but my guess would be they can get as much resource
runtime without intervention of root or kernel. That definitely plays to their
advantage when people run out of box installations..

Our "competitors" are MS SQL, SQLAnywhere, Oracle, and DB2. Business-class
databases. The tech press likes to focus on MySQL vs. PostgreSQL because
they haven't caught up to the idea that an OSS database could compete with
commmercial offerings. When *you* focus on MySQL vs. PostgreSQL, YOU ARE
BUYING IN TO THEIR IGNORANCE, and helping the press compartmentalize Postgres
as an alternative to MySQL.

Agreed. Druming our features is the way to go. Not slamming other *OR*
defending ourselves.

Bye
Shridhar

--
byob, v: Believing Your Own Bull

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jason, Greg,

[snip quote from Jason]

But you see, that's exactly my point. We will *not* differentiate
ourselves from MySQL by slamming them at every opportunity.

First, you should not address both of us but only respond to Jason's
points. I in no way advocate slamming MySQL, but we should not ignore
them either. Your initial argument was that we should not be competing
against MySQL. The comparison between the two will continue to be
made by many people for a long time to come. We cannot simply decide they
are not worthy of our time and stand on the sidelines, waiting for the
PostgreSQL vs. Oracle match.

- --
Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200303201017

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE+edzzvJuQZxSWSsgRAot0AKCupYTLjXj7jnODTVlSn0Tq0I4ILACgnbjr
rJMA+UY37fcI8Y8Uzt7zhN0=
=j9oR
-----END PGP SIGNATURE-----

On Wed, 2003-03-19 at 22:42, Shridhar Daithankar wrote:

On 19 Mar 2003 at 9:20, Josh Berkus wrote:

Our "competitors" are MS SQL, SQLAnywhere, Oracle, and DB2. Business-class
databases. The tech press likes to focus on MySQL vs. PostgreSQL because
they haven't caught up to the idea that an OSS database could compete with
commmercial offerings. When *you* focus on MySQL vs. PostgreSQL, YOU ARE
BUYING IN TO THEIR IGNORANCE, and helping the press compartmentalize Postgres
as an alternative to MySQL.

Agreed. Druming our features is the way to go. Not slamming other *OR*
defending ourselves.

The truth is that the reason we have many of these postgresql vs. mysql
is because advocates of mysql often use old information and bad database
theory in their arguments. For example, if you look at other open source
databases like firebird or sapdb, you never hear arguments that
transactions, triggers, foreign keys, views, subselects etc... are
unnecessary. IMHO "competing" with mysql is pointless, because I think
we already "beat" them on our own merits; but we do need to defend
ourselves when people make false claims about postgresql, and we need
should advocate sound database fundamentals as well, whether they come
from mysql, oracle, m$ or whoever.

Robert Treat

-----Original Message-----
From: pgsql-advocacy-owner@postgresql.org
[mailto:pgsql-advocacy-owner@postgresql.org]On Behalf Of
greg@turnstep.com
Sent: Thursday, March 20, 2003 10:24 AM
To: pgsql-advocacy@postgresql.org
Subject: Re: [pgsql-advocacy] PostgreSQL flamage on Slashdot

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jason, Greg,

[snip quote from Jason]

But you see, that's exactly my point. We will *not* differentiate
ourselves from MySQL by slamming them at every opportunity.

First, you should not address both of us but only respond to Jason's
points. I in no way advocate slamming MySQL, but we should not ignore
them either. Your initial argument was that we should not be competing
against MySQL. The comparison between the two will continue to be
made by many people for a long time to come. We cannot simply decide they
are not worthy of our time and stand on the sidelines, waiting for the
PostgreSQL vs. Oracle match.

Whoa, I never advocated 'slamming' either. In my 1st message I said I was
glad that we don't have to that. The word "hobby" is not nessasailry
slamming. There are lots of good "hobby" things, like Linux, that are great
products and leaders in their field. (Note: Linux is making the crossover
to "professional" now, please, no flames) (and in time, MySQL may make that
crossover too) I use 'hobby' vs. 'professional' to indicate robustness. I
don't know about you, but I'd not use a bank that used MySQL. If they used
Postgres, ok.

Most of the "battling" or "slamming" is done by the vocal uninformed few
that use incorrect arguments. "MySQL has transactions"+"MySQL is faster than
Postgres"!="MySQL is faster than Postgres when transactions are used", but
you'll find most people are arguing just that.

Furthermore, these same MySQL people don't bother tuning Postgres. We all
know how the config file is conservative. I think they are going to address
that in the next release: there have been looong threads about that already
in the PG-General list. In all my usage of the two, they seem to be
comparable in speed, YMMV. But I'd rather lose 10% maybe even 20% and have
all those great features.

How many people tune their disks with hdparm? How many people tune their NFS
block size? You can expect an even fewer amount to tune a much more complex
beast like Postgres.

-J