Stan's slow problem

[Stan Busk]

>> Yes, but not today, I need to be fresh to do that so lets wait
>> tomorrow or better this week-end so I have no interferences and I can
>> take the time to do it. Anyway I found out that all the places where
>> I get very big slow downs are places where I perform several queries
>> in a loop. I don't know if it has something to do. My newsreader
>> however is slow as hell but I am using threads, more queries you
>> perform slower it gets.
>
> May be RAM is eaten?
>     Try watch in Activity monitor

I can't really see that effect so far but will test.

I found out that all the slow queries have something in common, the  
use of COUNT() or SUM() like:

SELECT Count(*) FROM tblArticle WHERE
Article_MessageID LIKE 'aReference'
AND Article_Status NOT LIKE '%D%'
AND Article_Status NOT LIKE '%S%'

SELECT Count(*) FROM tblArticle WHERE
Article_References LIKE '%aReference"%'
AND Article_MessageID NOT LIKE 'inMessageID'
AND Article_Status NOT LIKE '%D%'
AND Article_Status NOT LIKE '%S%'

including simple ones like:

SELECT Count(*) FROM tblArticle WHERE
Article_Newsgroups LIKE '%inGroup%'

I personally think Valentina 2.3 has a serious speed problem with  
COUNT() and SUM() functions. As soon as you revert to Valentina v2.2,  
it works at normal speed again. So my question is, what happened  
between v2.2 and v2.3 that could affect COUNT() and SUM() ? Right now  
it is faster for me to count and sum manually with a for...next.

I have the problem in 3 different projects with about 30 queries  
using COUNT() and SUM() functions. Could you compare v2.2 and v2.3  
speed for those functions?

Stan