On 04/10/12 09:43, Symeon Breen wrote: > Ok so perhaps there is this and that. My point is a database like SQL server > is architected in a very different manner to U2
But that's why U2 is superior :-) - it is the actual central > sqlserver.exe process that does the majority of the work, it accesses the > tables and presents them to the api, its caching often used data, Which U2 actually does far better - (a) it leaves it to the OS, and (b) with a "disk hit rate" approaching 100% it doesn't need the cache that often (that's badly put, but I know what I mean :-) it obeys > the schema, transaction, referential, indexing rules, it map reduces > collations and totals, it looks after security, pools, licencing a whole > plethora of things. Thereis a lot more out of the box inside SQL server than > there is from U2 (again not saying it is better, just different) That's what gets me - U2 is often assumed to be inferior because it doesn't have the things SQL databases have, when the reality is it doesn't have them because it doesn't need them! The central > U2 processes look after licencing, locks, shared memory subroutines and a > few other things. It is the udt or uvsh client process in which your > databasic runs that has built into it the DB access code this is the key > difference I feel, and hence why I feel clever unibasic code has a right to > do clever things with sockets and xml etc. > But it would be nice if more of these things (even if written in BASIC, which is a *good* thing) actually came "out of the box" with U2. Cheers, Wol _______________________________________________ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
