Having occasion to work with some Siebel software, I read up a little about its history.
Siebel, of course, is a CRM system. That's Customer Relationship Management: software for managing sales, customers, opportunities, contacts. (Also known as Constituent Relationship Management when speaking of the fundraising or not-for-profit sphere. Someone might like to make that connection for Wikipedia, which is uncharacteristically sparse in this area).
Siebel was founded in 1993 by Tom Siebel, an ex-Oracle executive who was unhappy that Oracle's Larry Ellison wouldn't invest in his idea. To cut a long story short, it was very successful in a relatively short space of time, and Tom Siebel made his billions.
But apparently it suffered from two faults, compared to the other CRM software that had emerged in its wake. First, it was particularly CRM-focused, and didn't integrate terribly well with other software - quite a no-no in this era. Second, it was rather too flexible. You could do so much with it, that implementations ran for years, and suffered chronic scope creep. This is the bane of an implementer's or project manager's life: people wanting to add more functionality, do more with it, blowing out the project's costs and timespan. And you could do this with Siebel, because it was so powerful, so flexible. You could mould it to fit your business processes - which also blew out implementation costs. Whereas its ultimately more successful competitors required companies to re-align their business processes to fit the software - perversely making it easier to manage. (SAP, anyone?)
So, ironies, Siebel was eventually sold to... Oracle. Yet another good technology falling victim to a turning marketplace.
Postscript 1: Of course, I have to qualify those positive epithets. Bottom line, integration (of a company's disparate software systems) is a seriously important issue - one of the biggest of the decade - and integration difficulty should rightly have doomed it eventually. Still, in theory at least, there would have been nothing wrong with a good buyout that could have made use of its strengths while addressing its weaknesses.
Postscript 2 (updated 4-Aug-06): The metastory is itself amusing. I got the above story from Answers.com. Now they suck a lot of their content from Wikipedia, including this one. But since Oracle took over Siebel, they seem to have sanitised the Siebel entry on Wikipedia. This is not always the case: a person or organisation can insist on telling their story their way, but of course the official story doesn't always win out. See for example the discussion on Jaron Lanier's bio, where Lanier pushed one angle, and others pushed another until there was... eventually (sigh) consensus.
Having said that, the reason for the Oracle version prevailing would be available on the history and discussion tabs behind each article, since the development of Wikipedia's information is always transparent. On the other hand, it's very interesting what occasionally results from this process... for example, the alternative ("out-of-date") version of Siebel's story being available at answers.com.