List Readers may have followed the thread mentioned above where I was
very critical of drawing of graphics objects in RB2006 R3.
Can I apologise for my unkind words about RB and R3. I WAS WRONG!!
By fixing the problem described below the performance of R3 is
comparable to 5.5.5. Clearly the thread was hogging the CPU to the
detriment of RB's redrawing of windows/object/updates
The problem was having a thread running (a simple print spooler)
Code as follows
do
if backprntstatus(backpoutindex) = 2 then //if full
k=touchSell.print40line(backprntline(backpoutindex),10) //
then call print
backprntstatus(backpoutindex) = 0 // set empty
//increment
backpoutindex=backpoutindex+1
if backpoutindex>100 then
backpoutindex=1
end if
if backprntstatus(backpoutindex) = 3 then
exitl=true
end if
loop until exitl
Under 5.5.5 on Win this was not a problem. But as I am now catching
up by reading archives etc handling of threads is very different in
RB2006 than 5.5.5. and also on WIN
It appears that (unlike as I thought) the above thread does not
appear to yield at the end of the loop to the main app. (I am still
learning - but this is the effect anyway). Hence when RB attempts to
do complex work - it is very slow in execution. This only effects WIN
platform, Mac is fine
Anyway I now have good performance out of R3 on WIN and I plan to
move my commercial work to it.
If anyone has a more erudite explanation of the problem I would be
pleased to hear it
Tim James
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
