The cost/overhead of learning kdb is higher than the benefit it reaps, in this situation, IMO. (Baptism of fire for a new guy/gal! ) Unless you are dealing with mountains of data I'd stick with the more mainstream apps.I agree with DaveAngel -I seen Xenomorph recently - and was very impressed. The apps are slick and I think I remember reading somewhere that they do something similar to kx (i.e. I think, its at least partially, in-memory). They have an app called Timescape. It's relatively agnostic to the datafeed. (they have hooks to bloomy, reuters & the usual suspects):- So you can kill two birds with one stone so-to speak.

Hi,building a system like this is not as easy as it seems at the first look. My last information is that MySQL supports transactions. At least I am using them without any problem in a backtesting db. We are using Java with plugins in Eclipse ;-)Hibernate is the other problem. On a first look it seems very easy and the examples are easy to understand, but when it is going more complex, ouy begin hating Hibernate. With transactions and RI you begin hating it. I just use it for simple metadata.I would take a standard software for your requirements. There is only one point voting for self-developed. Flexibility.ByeMartin