Computer Associates Sells Ingres DB Tech

Christopher B. Brown writes to tell us Network World is reporting that Computer Associates is selling their Ingres database technology to a private equity firm called Garnett & Helfrich Capital. From the article: "CA released Ingres last year as an open source project, reviving interest in the dormant software. Still, databases have never been a core part of CA's portfolio. CA CEO John Swainson cast the Ingres sale as part of CA's larger effort to streamline the vast collection of applications it amassed through a decade of heavy acquisitions in the 1990s. Ingres came to CA through its 1994 buyout of ASK/Ingres"

Comments

Ah, the tale of many an application...

[penguin_asylum]

First, it's considered interesting. It gradually reduces in attention given to it, until... They release the source code! Revel ye cupids, for the code hath been releas-ed! Next day, no one's heard of it again...

Re:Ah, the tale of many an application...

[Doc Ruby]

Could it be that the app was only interesting because of ignorance, both of the quality of the code and of the business value of the app? First the biz value is revealed to be less, which is what causes the release of the source code. Interest declines in the biz value, but possibly rises on the value of the code itself. But then inspection of the code shows its quality to be low, and nothing stops the decline of interest. The secrecy that propped up the app value before is gone, and its true, low value is now clear. It makes more sense when the cause and effect are considered in the proper order.

I think it's pretty safe to say...

[Sheetrock]

That these acquisitions of modular lightweight technologies appear to be part of a larger dynamic approach to enterprise-class offerings. I can't imagine it stops with an RDBMS and network clients -- I wonder if there are any underappreciated server hardware platforms that they've got their eyes on.

Why bother?

[HermanAB]

I don't get it. Since PostgreSQl is the successor to Ingres and is properly funded by DARPA - why would anyone bother with the older version? It feels like Linus making a big whoopdedoo about a release of kernel version 1.0 under a BSD license...

Re:Why bother?

[Anonymous Coward]

PostgreSQL is not the successor to Ingres, it is a branch of the source code from the early eighties. The commercial Ingres product was maintained and improved by a large team of full time engineers for 20+ years since the branch. It is arguable the superior branch.

Re:Why bother?

[buckhead_buddy]

HermanAB wrote:

Since PostgreSQl is the successor to Ingres and is properly funded by DARPA - why would anyone bother with the older version?

Please note that this company is not specifically a technology company; their focus seems to be on investing. My guess is that they're going after the Intellectual Property and need the ownership of the proprietary code to set up a basis for sales, licensing, or litigation.

Perhaps they have other interests. Though they don't appear to be, they might be the Investor equivalent of Sanford and Son and see the Ingres code as having future value even if it isn't developed further. I don't see the appeal, but someone surely saw something in it. Their strategy may not be clear, but they certainly didn't buy it out of nostalgia.

Re:sad attempt...

[Anonymous Coward]

You have no tools available in the Open Source market that compare with Ingres. If you think otherwise you are deluding yourself.

Ingres is an RDBMS used by large organisations and small, and has a very good technical support system behind it with actual people you can talk to - some even local.

Try that with MySQL, Postgres, etc - the only way you can get equivalent support for them is if you live in the same city as the developers. Computer Associates have offices with support people around the world.

Re:PostgreSQL Lineage

[Wugger]

The lineage of PostgreSQL and Ingres is pretty clear. They are descendents of separate research projects of Prof. Michael Stonebreaker at UC Berkeley. Ingres descends from the an earlier project, which was a proving ground for pure relational technology.

PostgreSQL (note the play on words, "post" gres comes after "in" gres) descends from the follow-up project which extended relational concepts into an early "object-relational" system. Stonebreaker lays out his goals for the Postgres project in this 1986 paper.

So, Ingres is based on an older design that PostgreSQL. It has also spent 20 years in the corporate world being changed, upgraded, and improved, so evaluating it based on its lineage is like evaluating Oracle 10g based on your knowledge of Oracle 1.0. Interesting historical note: one of Oracle's first substantial competitors (and an early market leader) was a company called "Relational Technologies" that sold a cutting edge relational database named... Ingres.

How many posters have actually used Ingres ?

[RobWalker]

I can tell you, I've used both Ingres and many other database (including MS-SQL, MySQL, DB2, etc. etc.).

Despite it being considered "old" it, Ingres is a fast and stable database - certain comparable and up to the performance levels and features of many current databases, MS-SQL and MySQL included.

And several cite how much better MySQL is an open source database - to which I'd say:

- read the Ingres and MySQL licenses and tell me which is more open source, and less restrictive? To my eye (and IANAL) the Ingres one poses fewer constraints on use of Ingres as an open source product within commercial products

- ask one of the many big Sun sites who still run very large, very stable applications on Ingres whether they'd like to swap for MySQL?

I have no axe to grind here, but Ingres is a decent database and a proper open source contribution. Just because it has CA's name associated with it, doesn't make it bad