MySQL

[prev] [thread] [next] [lurker] [Date index for 2003/08/27]

From: Ann Barcomb
Subject: MySQL
Date: 14:52 on 27 Aug 2003
In another mailing list, there is a bit of a flamewar going on
concerning FreeBSD vs. Linux and PostgreSQL vs. MySQL.  This has
reminded me of my dislike of MySQL.

My single biggest complaint about MySQL is that historically it hasn't
promoted itself as the right tool for some jobs, but as the right
tool for every job.  This may have changed--I haven't checked out
the marketing in recent years--but my initial impression was that
MySQL was misleading people.  Remember when they used to say that
nobody needed transactions (it's true that not everyone does...but
to say that nobody does is quite misleading)?

Therefore I dislike MySQL not because of the quality of the product
(which is certainly acceptable for a good number of uses), but
because they have marketed themselves unscrupulously.  When a
software makes misleading claims and spreads ignorance simply
to get a larger market share[1], I wonder what their ideals are.
I also don't like it when commercial software does this, but most (not
all) people have a great deal of skepticism for products that will cost
them than free software.

I don't like it when products promote ignorance--when the product is
open source, I expect more of it.  I'm disappointed in it.

The ignorance it offers also lead directly to problems for me.  People
parrot that transactions just slow things down, without even knowing
what a transaction is.  And when non-standards are pushed as standards
(although pretty much every dbm is guilty of adding non-standards), it
leads to situations like the Explorer and Netscape HTML-tags, where
you have to do at least double the work to keep everyone happy, or
people don't understand why you won't do X--it must be a standard,
because everyone else does it (never mind if is incompatible with
the standard, and seems to show not even the slightest understanding
of the intent of the standard).

MySQL may have given up its previous stance, but to me it doesn't
matter, because the damage is already done.

---
[1] Although given what they said about transactions, perhaps they really
believed it...but in that case, they should not have been writing the
software.  Discounting decades of research on the subject is at best
reinventing the wheel.


Generated at 19:00 on 27 Feb 2006 by mariachi 0.52