Thread: Database comparison ideas

I think that the people on this list tend to make a mistake.

They try to pick apart the competition by focusing on their weak points.

From a growth standpoint, I think it is a much better idea to focus on
their strong points.  Look at the things each competitor can do best.
Try to think of ways to get the same functionality from PostgreSQL.  If
it is impossible [or currently infeasible] to meet the functionality,
then close the gap.

Suppose (for instance) that MySQL were faster at some particular
operation by a factor of 5.  If the difference cannot be eliminated or
overcome, can the gap be narrowed so that it is a factor of 2?  If DB/2
has some special security feature, can the same feature be added to
PostgreSQL?  If there is an administrative tool for Oracle that provides
essential functionality, can the same tool be created for PostgreSQL?
By careful examination of the *strong* points of the competition, you
can form a strategy to close the gap.  By focusing on what they do
poorly, how will progress be made?

The weak points are always going to be there, for any database system.
But the way to expand the functionality of PostgreSQL best would be to
focus on the *strong* points of the competition and try to achieve the
same level.  For weak points, it is better to focus on the weak points
of PostgreSQL than that of the competition.  Admit they exist, and form
a plan to eliminate them.  I would like to see the day when PostgreSQL
is on every desktop in the world, as a superior replacement for Foxpro,
MS Access, etc.  I would also like to see the day when Postgresql is on
every server in the world as a superior replacement for DB/2, Oracle,
etc.  I think the best way to meet those goals is to be realistic and
aim in the right place for strategic decisions.

IMO-YMMV.

Dann,

> From a growth standpoint, I think it is a much better idea to focus on
> their strong points.  Look at the things each competitor can do best.
> Try to think of ways to get the same functionality from PostgreSQL.  If
> it is impossible [or currently infeasible] to meet the functionality,
> then close the gap.

You are, of course, correct.     We will have to prioritize which "gaps" mean
the most to us.   For example, if I was to make a "top six list":

-- Lack of comprehensive GUI admin tools
-- Lack of replication and point in time recovery
-- PL/pgSQL does not 100% replace PL/SQL or T-SQL Stored Procedures
-- Miscellaneous speed/optimization issues
-- Need good GUI installer, including installer for Postgres+PHP+Apache
-- Win32 Port

But what order would we want to tackle these in?  For that matter, don't
forget about Postgres goals to acheive features that nobody else has:

-- 98% SQL-99 Compliance, including Schema, Domain, etc.
-- 100% support of all data types and operators
-- etc.

All of this is a moot point, though.  Programmers work on what they want to
work on ... so even if, say, a GUI installer is really important to *me*, it
ain't gonna get done unless I do it myself.


--
-Josh Berkus

On Wed, 2002-06-26 at 20:54, Josh Berkus wrote:
> 
> Dann,
> 
> > From a growth standpoint, I think it is a much better idea to focus on
> > their strong points.  Look at the things each competitor can do best.
> > Try to think of ways to get the same functionality from PostgreSQL.  If
> > it is impossible [or currently infeasible] to meet the functionality,
> > then close the gap.
> 
> You are, of course, correct.     We will have to prioritize which "gaps" mean 
> the most to us.   For example, if I was to make a "top six list":
> 
> -- Lack of comprehensive GUI admin tools
> -- Lack of replication and point in time recovery
> -- PL/pgSQL does not 100% replace PL/SQL or T-SQL Stored Procedures
> -- Miscellaneous speed/optimization issues
> -- Need good GUI installer, including installer for Postgres+PHP+Apache
> -- Win32 Port
I know I (not knowing Oracle PL/SQL) have a hard time find enough docs
on PL/pgSQL, even with buying Bruce's and the German PG Developers
books. 

I'm personally having a hard time learning all the in's and out's of the
trigger/rule stuff.  I know I can use more of them, but have a hard
time. 


-- 
Larry Rosenman                     http://www.lerctr.org/~ler
Phone: +1 972-414-9812                 E-Mail: ler@lerctr.org
US Mail: 1905 Steamboat Springs Drive, Garland, TX 75044-6749