Yesterday I was listening to the IT Conversations Podcast and heard a presentation given by Michael Stonebraker from the Money:Tech conference. Stonebraker (I keep wanting to call him Ballbuster) talks about how the elephant database engines (SQL Server, Oracle, DB2, etc) are never going to make it in financial trading engines because their approach to databasing (like freebasing only with data) is antiquated. It hit me when I was listening to the presentation that I don’t really know about the different types of databases. I’ve always wondered why SAS can be so damn fast at certain database functions and Oracle or SQL Server so slow at the same thing. I understand a little better now. I’ve been trying to get SQL Server to act like some other kind of DB, when in reality I just need a different type of DB… like a column oriented DB, maybe.

Since I don’t need the type of super high performance DB that Ballbuster talks about, I am wondering if I can efficiently use some type of open source Column Oriented DB. But I have a lot of investment in learning SQL which I would like to not lose. Looks like I need to research my options.

Below are some sources for database technology information beyond the big three elephants:

• John Sichi’s blog. He is the co-founder of LucidDB
• Julian Hyde’s Blog - Founder of the Mondrian open-source OLAP engine, and chief architect of the SQLstream streaming SQL engine
• The Database Column - A multi-author blog on database technology and innovation.

More to come on this topic…

I love ya, SAS!

Tibco acquired Insightful last year.  Many folks have reported that S-Plus (the closed source implementation of the S language) was dang near financially killed by the success of R (the open source implementation of the S language).  I may be slow on the update, but today I learned that a company named World Programming is making a product, WPS, which is an alternative implementation of the SAS base language. WPS costs a small fraction of what SAS charges. In addition, a smart chap named Phil Rack has created a program for WPS called Bridge to R which, as the name implies, bridges WPS to R. In an interview with Phil Rack on the blog Decision Stats, Phil poins out that while WPS is only an alternative for base SAS and not the statistics tools like SAS/STAT, ETS, OR, etc by having a bridge to R a SAS programmer can get the data handling abilities of SAS and the statistical tools of R through WBS and R without having to pay our friends in North Carolina any cash. Very nice!

If there was only an open source implementation of the SAS language and base abilities… oh I can’t even continue the thought. It’s just too good to be true.

Kudos to Ajay at the blog Decision Stats for opening my eyes to WPS.  Keep up the good work, Ajay. But I agree with Phil, if you have 6300 contacts in LinkedIn, you, my good man, are a link whore. Not that I have a problem with that.

PS:

Let me clarify that I am not anti-SAS. SAS is a tool, and it does a good job at what it does. I learned to program SAS in grad school and used it for years, first in academia and later as a consultant.  However, I honestly believe that their licensing model is totally outdated and stifles innovation. I also know that I was VERY excited when my previous employer chose SAS as their BI provider. After over $500,000 in fees and 18 months we still had no BI system. That really pissed me off. After using R for a couple of years I now realize how good statistical programming can be and I don’t miss SAS one bit. By effing up their BI build at my previous employeer, SAS forcefully pushed me into R. *sing* And that’s the way we all became the Brady Bunch.. doo dee doot doot dooooo!