|> bugs and lack of some features that would come in handy. But it is a
|> product that you can pick up and get started in very quickly.
|>
Given the size and development cycles of these DBMSes, it would be
crazy to assume software errors would not exist and some features would
not be available. But I concur with you, Informix is a much easier
product to use, from the gitgo.
Informix is a UNIX DBMS as far as their main engine (translation: highest
performance) is concerned, however you can get the Informix line of products
for the DOS operating system with there low end engine (Informix SE).
Its been my experience (I've administered both Oracle and Informix) that
the Informix is easier to administrate than Oracle and there is very little to do to tune the Informix On-line engine to increase performance. Of course, if
you use the Informix SE (standard engine) there is virtually no administration
necessary.
Informix spent alot of time up front optimizing their engine to run on
UNIX boxes, and is playing catch up ball with some of the end user products
that present X-based user interfaces, where Oracle and Ingres went to the
market with these products before Informix (although Informix just released
a preprocessor for their 4GL product that generates GUI code. I guess
it comes down to want kind of application and data environment your working
in. Most of my work has been in large scientific databases that are more
of a decision support environment, embedded SQL in 3GL languages, and
modestly complex data models. Informix's support of binary and text streams
(BLOBs) has proved to be solid in our development of large databases and
the explicit support (in 5.0) of the primary/foreign key relationships
has allowed us to provide some very dynamic user interface environments
that shelter engineers from having to know about underlying data models
and/or data access (SQL).
Of course, all the hype about Informix goes out the tube if your not
on one their supported platforms.
These are just a few, disjointed comments from an old, biased