Im exploring some ideas in the land of stream databases [data which has strong sequential properties] and wonder if there are others already working in this space, or who might like to contribute/discuss/argue/hack open source code.
Ive put up some thoughts and C code, see -
- vfuncs project at google code
- quantblog.wordpress.com
enjoy,
gord.
..whereas my interest is in how to deal with the fire-hose of ~10Gb data per day, once you have 100Gb of it on disk and more streaming in all the time.
I'm aware of commercial offerings - - KDB+ [its own semi-functional terse languages - K & Q ] - StreamBase [extended SQL]
The open source ones ... - Aurora / Borealis ( MIT, Brandeis) - Medusa (MIT) - Streams (Stanford) - MonetDB (a Netherlands university) seem to embrace/extend SQL and/or XQuery and/or RDF
Im thinking there must be a better way.