*That would be the Thrice-Married Virgin, although I’ve recently seen versions in which the poor unfortunate was married 12 times.  The last husband on the list is always a computer or software salesman, who keeps telling her how good it is going to be.  I first heard  the joke from Flip Filipowski. I decided it must not be too terribly sexist after hearing Sandy Kurtzig tell it to a group stock analysts.

Am I missing anything major?

Edit: I emailed the company on May 8, asking what Expressor had in the way of customers.  There has been no response.

So there are a whole lot of other worthy subjects I doubt I’ll be getting to any time soon.  In some cases, of course, other people are doing great jobs of writing about same. Here are pointers to a few links that I am glad to recommend:

• I wrote recently that I’ve discovered a number of different in-memory OLAP engines. Cindi Howson far outdid that, writing at length for Intelligent Enterprise on in-memory analytics, in an article that seems to itself be a teaser for a longer, free white paper on the subject.
• CouchDB posted an eye-catching, risque slide presentation promoting CouchDB and, more generally, key-value stores, at least for internet applications.  And yes, they’ve integrated MapReduce.
• Merv Adrian posted favorably about Birst, with special reference to its OEM efforts.  As previously noted, I was highly unimpressed with Birst’s end-user BI story at the time of its September roll-out, and Jerome Pineau’s recent examination did nothing to reassure me.  But perhaps OEM is a different matter.
• Merv also offers an interesting post about data integration upstart Expressor, and a highly favorable one about “visualization” vendor Tableau.
• Ann All interviewed Nigel Pendse, who grumped that BI features are overrated, and what end users really want is great query performance. I’m not so sure about the features side of that, but I’m hugely in agreement about the performance. That’s a big part of why the analytic DBMS industry is so vibrant. It’s also why in-memory OLAP is suddenly so hot.

• Ab Initio sells high-end data integration software.
• Ab Initio commonly costs $1/2 million or so.
• Ab Initio’s core claims include:
  • “It just works”
  • Ab Initio has great performance, even on big tasks.
  • Unlike many competitors, Ab Initio has an integrated product line written from scratch. (Hence the “Ab Initio” name.)
• Like most data integration software – Talend is an exception – Ab Initio includes an execution engine.
• Everybody agrees that Ab Initio’s software has great performance, although Talend claims to come close and Expressor claims to be faster yet. But rivals assert that besides having high license fees, Ab Initio’s software is also very consumptive of hardware resources. Certainly I’d suggest checking that aspect carefully if you ever get into an Ab Initio POC.  Perhaps that’s what Ab Initio means by saying its software uses any and all hardware resources. )
• Price isn’t the only regard in which Ab Initio is hard to do business with. Another is secretive business practices. For example, Ab Initio – confident in the quality of its software – pushes prospects toward POCs (Proofs-Of-Concept). But it wraps so many NDA requirements around these that some prospects walk away.
• Not surprisingly, Ab Initio has added lots of features over the years, especially in response to prospect or customer requests. Examples I was given include:
  • IBM OS/390 support (including COBOL copybooks, etc.)
  • SOAP/XML support. Associated with that is a story that boils down to “With great encapsulation, one can change a complex system of data integration processes incrementally without going crazy.”
  • A compressed file system that can directly store 100s of TBs of user data, with very fast query performance. Apparently, this is not at the extremes of inflexibility, as it is realistic to have up to 5-6 keys on a table (at least). Associated with that is a story that boils down to “Hey, if you’re only getting at something via web services, you’re limited in how you can query it anyway. Worst case – your needs expand and you decide to put the data back in a true DBMS after all.”
  • Similarly, Ab Initio claims that its software is easier to use than rival, cobbled-together products. While Ab Initio may not dispute the existence of products that can get data integration tasks done more simply, it argues that these products do a lot less than Ab Initio does.

• Expressor was founded in 2003 and funded in 2007. Two rounds of funding raised $16 million.
• Expressor’s first product release was in May, 2008; before that Expressor built custom integration tools for a couple of customers.
• Michael believes Expressor will have achieved 5 actual sales by the end of this quarter, as well being in 25 “highly active” sales cycles.
• Whatever Expressor’s long-term vision, right now it’s selling mainly on the basis of performance and affordability.
• In particular, Expressor believes it is superior to Ab Initio in both performance and ease of use.
• Expressor says that parallelism (a key aspect of data integration performance, it unsurprisingly seems) took a long time to develop. Obviously, they feel they got it right.
• Expressor is written in C, so as to do hard-core memory management for best performance.
• Expressor founder John Russell seems to have cut his teeth as Info USA,which he left in the 1990s. Other stops on his journey include Trilogy (briefly) and then Knightsbridge, before he branched out on his own.

Expressor’s real goals, I gather, have little to do with the performance + price positioning. Rather, John Russell had a vision of the ideal data integration tool, with a nice logical flow from step to step, suitable integrated metadata management, easy role-based UIs, and so on. But based on what I saw during an October visit, most of that is a ways away from fruition.

What I know already is that our numbers will between 7 and 8 min to load one TB of data and will set another world record for the tpc-h benchmark.

The whole blog post has a delightful air of skepticism, e.g.:

Sometimes the mention of a join and lookup are documented but why? If the files are load ready what is there to join or lookup?

… If the files are load ready and the bulk load interface is used, what exactly is done with the DI product?

My guess… nothing.

…  But what I can’t figure out is what is so complex about this test in the first place?