We are doing great things with PostgreSQL

I took some time last night to go to DejaNews (www.dejanews.com), and do
a search for PostgreSQL, ordered by date.  I looked at Usenet posting
from the past two weeks that mention PostgreSQL.

I did this about a year ago, and was disappointed in what I saw.  There
were not many postings that mentioned us.   When people asked for
comparisons between us and others, like MySQL, people usually
recommended MySQL because it was faster.

One year later, huge difference.

First, there were hundreds of postings.  I read several hundred, then
stopped because I got tired of reading.  Much more discussion about us.

Second, I saw no negative reports about PostgreSQL.  I'm sure they
exist, but I didn't see any.  (I did see a few Alpha compile problem
reports, though.)

Third, in cases where PostgreSQL was compared to other databases, we
were always fairly considered.  Almost universally, people would say
PostgreSQL is a reliable database, with most SQL functionality, like
subqueries, transactions(these two were always mentioned), views, free,
major features added in frequent releases, user-extensible functions,
many interfaces, and good mailing list support.  A few even took our
mantra, saying "They are the most advanced open-source database.
Period".

People recommended MySQL only when speed was most important.  Many
people reported that MySQL's lack of transactions, subqueries, and other
SQL standard features were a real liability, and if you choose MySQL,
you had better be sure you don't need these features, because they may
never be added (particularly transactions).  They stated that the MySQL
documentation was clear about these limitations.  I personally have seen
official MySQL documenation that recommends PostgreSQL if transactions
are required.

In fact, people universally mentioned PostgreSQL as having better SQL92
support than MySQL.  In the past, that was the reverse.

Even in comparisons against Interbase, we were recommended.  In fact, I
didn't see many postings that clearly stated another database over
PostgreSQL.  If was usually, "PostgreSQL, but if you need speed, MySQL".

An interesting thing is that most of the people recommending us were not
names I recognized as regular maillist people.  They were just regular
users.  Some of them were names I knew, and I was glad to see that.  It
is a big help in telling people about PostgreSQL.

There is clear momentum.  This is the fruit of the several years of
development.  For some time, I was wondering why people on Usenet were
not taking more interest in our work.  The word is clearly out, and
people like the releases were are generating.

Sometimes, when things are quiet, I wonder if people are still using us.
Now I know, things are quiet because PostgreSQL works, and people are
telling others about us.

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

In February, we start beta on 6.5, which will have multi-version
concurrency control, thanks to Vadim.  For those of you who did not like
our table-level locking, you will be really surprised.  Not only do we
have row-level locking, we have something even better.  In row-level
locking, a reader can not read a row that is locked by a writer.  In the
6.5 locking system, READERS ARE NEVER BLOCKED, and writers block only if
they try to modify a _row_ that another writer has modified and not
committed.

This is a major feature, better than many commercial databases.  There
is also no running out of locks like commercial systems, because the
transaction ids take care of that.  A true NUMERIC type with
user-specified precision is also planned for 6.5, thanks to Jan.

--
  Bruce Momjian                        |  http://www.op.net/~candle
  maillist@candle.pha.pa.us            |  (610) 853-3000
  +  If your life is a hard drive,     |  830 Blythe Avenue
  +  Christ can be your backup.        |  Drexel Hill, Pennsylvania 19026