As I understand it Panorama is ram-based, and consequently unsuitable
for large databases. Being memory resident naturally gives it greater
speed, but limits it to small scale applications.
I note that Dr Hipp quite clearly makes the point that Sqlite does not
even map the Sqlite file into virtual memory, thus avoiding addressing
limitations and permitting the database to be limited in size only by
the addressing limit of the file. A wise design condition.
It would be interesting to compare the performance of Panorama with
Sqlite using ":memory".
JS
Richard wrote:
Well, finally import 9,337,681 records into sqlite3 test2.db
and ran the test.
import sqtest4.txt : 2 min 28 seconds
select A, '32.0833' From T ; 9 min 20 seconds
--
I use another database for MacOS X
also works on Windows, Call Panorama
Did another test, comparing database...
import sqtest4.txt : 36 seconds
select from Field A contains 32.0833 / 55 seconds found 4322 records
out of 9,337,681
Still have found no sql program yet, that can beat Panorama in speed.
Regards-
Richard Nagle
CMS
