V4RB 2.0 beta 7

[Ruslan Zasukhin]

On 7/12/04 9:45 PM, "Sims, John" <AYU8 at CDC.GOV> wrote:

>> 
>> I'm using my 1.x code. But frankly I haven't gotten it to run 10
>> minutes without a crash, so I don't know if it will timeout when I
>> do!   :->
>> 
>> Jon
> 
> :-)
> 
> At least you've got it running.  I just know I've overlooked something
> silly.  I'm looking forward to getting into some of the new features as
> Ruslan gets them working.  I'm intrigued about the Binary Links and the
> Vsets.  Plus, I'm curious to see how much faster he has made it...if he
> makes it much faster than 1.x, it may be the first step towards building
> a time machine :->

:-)

1) beta now is very slow. Because
    a) we do not use at all cache now.
    b) we log each step and command with Flush() now


----------
2) I have told. Then main improvements was in

A) cache. Now BIG cache of size e.g. 100-200MB will work in 100-500 times
faster of Valentina 1.x cache.

B) the same is true for internal file system. For huge dbs, >500MB file
system works now in hundreds times faster than 1.x


C) also improved sorting indexes for numeric fields types.
    they improved not in speed but in life-cicle.
    Now modification operations do not require rebuild them.
    in some queries this can be one or few seconds save.

D) some algorithms improved thanks to new internal way access indexes.

E) not done yet -- I want improve format of indexes. This must speedup them
in 2-10 times on search of single/first value.



-- 
Best regards,
Ruslan Zasukhin      [ I feel the need...the need for speed ]
-------------------------------------------------------------
e-mail: ruslan at paradigmasoft.com
web: http://www.paradigmasoft.com

To subscribe to the Valentina mail list go to:
http://lists.macserve.net/mailman/listinfo/valentina
-------------------------------------------------------------