Comparison of Oracle vs. Sybase?

I understand that it is a FAQ to ask for a comparison between Oracle and Sybase. So, I'll give some extra information to describe our situation at the Documentation project at the Arts faculty, University of Oslo. This project will develop a common database system for the archive and collection departments of the University.

Each of the institutions in the project should have its own local database, but all the databases should be combined into one (distributed) database for the entire faculty. We will have about 1.5Gb of text, and probably a lot of images later on. The data will be rather static, but may get a fair amount of queries.

We would like to be able to combine relational databases with free text searching. PAT is a free text search system we have looked into. We use SGML to code the data, so it would be nice if the database was SGML aware... (Whatever that should imply.)

We will run the server(s) on Unix-hosts. We have DECstations at the moment. We want to make clients for PC, Mac, and Unix/X11 which will access the data, so it is important that there is good development tools available. (We have heard rumors of a forthcoming Smalltalk-based tool from Sybase: Calypso. Does anybody know anything about that?)

Questions:
  Is it obvious that Sybase or Oracle it the right choise for us?

  How easy is it to make distributed databases in the two systems?

  How much better have Oracle become in the upgrade v6 -> v7?

  How easy is it to integrate the databases with free text search   systems?

  Which one is the more open (as in: can be replaced later, if a   better product becomes available).

  Who has the best development tools?

Thanks in advance,

Dag Asheim
Documentation project
University of Oslo
Norway Received on Tue Feb 23 1993 - 13:34:32 CET