eds.activemath.org

The question of storage of OMDocs keeps coming… why the hell do you, in ActivMath, use your own storage solution for OMDoc fragments and not one of the classical SQL or XML databases? Below is a short answer, helped by Eliotte Rusty Harold.

For SQL databases, the support would just be a way to store structured documents and find them back… the answer is simply… that for this purpose, Lucene is used, it is a pure java clean-open-source, super-scalable solution… that didn’t exist in SQL (until Derby) and still potentially beats about any SQL database in scalability.

For XML databases, Eiotte Rusty Harold posted a the state of XML databases and the answer is “it’s not really ready for prime-time” or “good commercial ones are there”. Among the conclusions: no non-viral-open-source that scales.

It is nice to see someone caring to be realistic for such reviews… I’ve seen so many reports and claims that it’s a shame to have your own XML solutions!! In ActiveMath, our needs are fairly simple: a few children extraction (a handful), pre-processing (we have to do anyways, this includes links in both ways), and a few attribute extractions, and all this, please, 300 queries per second.

Although I look at it from time to time, I’ve never met an XML storage solution that can do it.

paul

PS: Eliotte, it’ so nice to see this quote “doesn’t reveal exact pricing until the salespeople play golf with your C-level execs so I have to guess at the price here”… it’s describing so well the very many computer-business approach nowadays! ;-) (and it includes Adobe for some of their products).