Re: Db2, Oracle, SQL Server

Serge Rielau apparently said,on my timestamp of 21/02/2005 2:00 AM:

> Legend:
> * Most <> All

It's not most. It might be UDB and that's it. All others used similar recovery ages ago. DB2 for mainframes has had full back and forward recovery since around 1979 that I know of. Probably before. Not even mentioning others, relational or not.

> * This recovery mechanism was designed <> Recovery was invented

When it purports that "most" databases (some dating back to the 70s) use something that was only "invented" in the mid 80s...

> * Initially <> Final

It might have been. According to the article, ARIES looks like the be all end all of recovery technology. Simply not true. Most of it is just stating the obvious that was already practiced and known. So is ACID, BTW.

> One of us is interpreting this article rather liberally I'm afraid.
> Not that this is all that rare. It explains a lot of what's going on in
> this planet.

Absolutely. What the article clearly attempts to do is give credibility to DB2 by associating it with research from Almaden (correct) and then generalising the research as some form of industry "standard" that everyone followed since then (incorrect). Trying to create the image that DB2 somehow "leads in technology" (about as false as it can get).

And that's about it. Sorry, end of weekend and I'm very busy next week.

-- 
Cheers
Nuno Souto
in sunny Sydney, Australia
wizofoz2k_at_yahoo.com.au.nospam