So far the experience has been the opposite. To see this difference I
would assume that he used to run his own optimized instance of mySQL 

Or it could be a difference in plugins.

Or it is the testing that is wrong. To compare scan times I would never
use the times from the first run. By using the times from the second run
you make sure that both test results have the benefit of any buffering.
This way it is easy to spot changes in the code performance.

For my 10k track first run could take 6-8 mín, second run 4 min.


A couple of months ago I tested scantime on various releases:
7.3.3   20:57
7.4.2   16:03
7.5.3   16:06
7.6.0   03:49 SQLite w/high mem

On my library the high memory setting does not improve on scanning.
I have not tested 7.6.0 with mySQL.

About the improvements in artwork handling.
7.5.3   06:27
7.6.0   02:20 using SQLite for artwork, but before new code
7.6.0   00:23 with the new artwork code on x86

You should still see similar improvements in artwork handling if you go
mySQL. The artwork handling _use_ SQLite.


-- 
slate

Main: Duet (Audiocom) -> Beresford Caiman -> Carver A-500x -> B&W 704
Office: Duet -> Technics SU-V50 -> Stax SR84 Pro
Server: Zotac IONITX-A, 4 GB, 1 TB WD EADS, Win7 w. 7.6 (SQLite w. High
Mem)
Tied together by D-Link DIR-655 + DGS-1008D
------------------------------------------------------------------------
slate's Profile: http://forums.slimdevices.com/member.php?userid=30502
View this thread: http://forums.slimdevices.com/showthread.php?t=84775

_______________________________________________
beta mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/beta

Reply via email to