Which is faster? (PBDLL or PBCC)?

[Scott Turchin]

If you run these two code segments, you will see that PB/DLL is doing the test in 3/4 the time that PB/CC does...

I would have thought it was the other way around....

PBCC: 8.67 seconds
PBDLL: 5.78

PBDLL
'
#Compile Exe
#Include "win32api.inc"
Function PbMain () As Long
Local lTimer As Double
SetPriorityClass GetCurrentProcess, %HIGH_PRIORITY_CLASS  'To speed things up a bit
lTimer = Timer
! mov eax, 1000000000
deceax:
! dec eax
! jne deceax
Msgbox "Elapsed time: " + Str$(Timer - lTimer) + " seconds"
End Function

'PBCC
'
#Compile Exe
#Include "win32api.inc"
Function PbMain () As Long
Local lTimer As Double
SetPriorityClass GetCurrentProcess, %HIGH_PRIORITY_CLASS  'To speed things up a bit
lTimer = Timer
! mov eax, 1000000000
deceax:
! dec eax
! jne deceax
StdOut "Elapsed time: " + Str$(Timer - lTimer) + " seconds"
End Function

------------------
Scott
mailto:[email protected][email protected]</A>

[Kev Peel]

Scott,
I think could it be a case of console loading times?



------------------
Kev G Peel
KGP Software, Bridgwater, UK.
www.go.to/kgpsoftware

[Peter Lameijn]

Scott,

If you add a print or stdout command to the pb/cc code (BEFORE the SetPriorityClass statement) then both are equal.
It seems that the console handle is opened at the first console command.
So the SetPriorityClass didn't work in the PB/CC example.


------------------
Peter.
mailto[email protected][email protected]</A>

[Steve Hutchesson]

Scott,

From all of my own test code and from the technical data from PowerBASIC
that I have seen over time, its the same machine under the hood so the
performance should be much the same. I think Peter has a very good point
and all I would add to that is to put the test code in each example into
a function or sub and get the initialisation code complete for both before
running the test.

Regards,

[email protected]

------------------