Open Source Ingres for Linux

[Christopher Browne]

Francois Ouellette wrote: 
> I have supported Ingres database servers running 24x7 with 700 interactive 
> users. I haven't seen too many Postgres or MySQL installations of that size! 

I help support several PostgreSQL instances of comparable size.

Hugh Redelmeier wrote: 
> | From: Francois Ouellette <fouellet-cpI+UMyWUv9BDgjK7y7TUQ at public.gmane.org> 
>  
> Sorry, what you say strikes me as hype, not information.  That makes 
> me nit pick.  OK, I admit it, I nit pick anyway. 
>  
> | >From what I know Stonebreaker was the guy who first put in practice the 
> | theory developed by Codd and others and made it a commercial success. 

Actually, that's Michael Stonebraker.
 
> I seem to remember System R predating Ingres.  Not a commercial 
> success, but I think that it evolved into DB/2, which is successful 
> (not sure -- I don't pay much attention to database systems). 
>  
> | Postgres is a spinoff of Ingres and a research project, Ingres was the 
> | commercial version of the same concept. 
>  
> My understanding at the time was that Postgress was the successor as a 
> research project.  So it ought to have shiny new ideas (no longer new 
> now). 

The history is a bit different than that.

-> Stonebraker's group at UCB developed University Ingres

   http://s2k-ftp.cs.berkeley.edu:8000/ingres/

-> Stonebraker founded Ingres to commercialize this.  One of the
notable developments was that commercial Ingres added in an SQL
processor in addition to QUEL.

One of the interesting things there is that [so I hear] the query
optimizer for Ingres was pretty good particularly when they typically
involve a transformation from SQL to QUEL.

-> Stonebraker continued research work, creating Postgres

"Object relational," blah, blah, blah...

-> Postgres was commercialized as Illustra

-> Andrew Yu and Jolly Chen (grad students at UCB) added an SQL
processor to Postgres, creating Postgres-95

-> Subsequent work on Postgres-95 took place outside the university
setting such that it is no longer at all fair to consider PostgreSQL
to be "university-related."

-> Informix bought out Illustra, and integrated it into Informix
Universal Data Server

-> IBM bought out Informix, so that "Postgres" is now called DB2 :-).

> | More like an early "open source" 
> | before its time! 
>  
> There is no "before its time" for open source.  There has been open 
> source as long as there has been source. 
>  
> | Ingres has matured as a commercial enterprise-class software product while 
> | Postgres remained a university project. 
> |  
> | If you want the real thing, then go for Ingres! 
>  
> There has been a lot of work on Postgress making it a practical tool. 
> Outside UCB. 

It hasn't been a UCB project since about 1995 or 1996...
  
> I was hoping for a serious comparison, not just sloganeering. 

Indeed.

The way I see it, the "open sourcing" of Ingres looks like a
last-ditch attempt to get "free developers" to do maintenance work on
a product that CA is finding isn't viable to continue to support.

It looks plenty like the SAP-DB situation...

- Software AG had a database called Adabas-D, of similar maturity to Ingres.

In some ways, it's comparable to Oracle, but they weren't getting
particularly impressive sales.

- They sold it to SAP AG, who were interested in having "their own
database" as an alternative to Oracle, which is useful as a "stick"
any time license negotiations with Oracle go badly.

- SAP AG discovered that it was a rat's-nest of painful-to-maintain
code, and wasn't particularly interested in paying for maintainers. 
They released it under the GPL with libraries under the LGPL,
evidently hoping maintainers would appear out of the woodwork.

With code written in a mixture of Mainframe style with German
"mnemonics" alongside build tools that were arcane, maintainers didn't
appear.

- MySQL AB came to the table, proposing that they'd sell a
"recommercialized" version of SAP-DB, now called MaxDB, and that
they'd do some form of porting to bring it into being somewhat
compatible with their product line.

SAP-DB is "open source," but there is certainly no public community
around it.  Just like with MySQL(tm), OpenOffice.org, Mozilla, and
other such products, they are _really_ commercial products with some
veneer of "try it for free."

I can't see that Ingres is that much different.  It may be that its
source code is less unreadable than SAP-DB, but it would seem
surprising for it to become of any great interest to the community at
large particularly when there are already vibrant communities of
developers surrounding other systems (notably PostgreSQL).
-- 
http://www3.sympatico.ca/cbbrowne/linux.html
"The true  measure of a  man is how he treats  someone who can  do him
absolutely no good." -- Samuel Johnson, lexicographer (1709-1784)
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml