va!n wrote:Have you tried following test with the example source?
; Another way to solve this... Just only copy and paste the
; complete "Part2" section between EndProcedure and before
; starting the "Part1" section! Can someone confirm and explain
; this?
Then the compiling will be very fast! A bit strange... Tested on a brand new installed WinXP System (2 days old)
I hadn't tested it but i have now and switching part 2 before part 1 does indeed make the compilation almost instant!
We've got different hardware too - i've got an Intel 3 ghz on an Intel motherboard using on-board graphics (after my Radeon 9800 broke) with Win XP SP 2 so not much in common there except the OS.
Here i compile in less than .5 secs the example posted by Mr.Mat, with or without debugger.
The problem is not what the source is, neither its size. And is not related to debugger on or off. It happens time to time, i don't know what's is the cause, but i can say that these behaviours comes since POLINK messages appeared, i.e. since linking was changed (about 3.8x version if i am not wrong).
@Psychophanta:
You have to test the full source i posted... i can reprocduce it here everytime!!! If you change the *4 to something like *43 it could be that the compiler need 2 seconds longer for compiling...
btw, how is it possible to speed up the compiling speed just by exchange part1 and part2...? :roll: (seems my old posting / thread about this has been deleted!?)
Sorry va!n. I've not problem here with your example. Less than 2 secs the first time i compile it, and less than 1 sec the next times.
I guess we are talking about different things :roll:
@va!n: The slowdown is definitely occurring with polink. I'm linking the .obj file generated by pbcompiler/fasm (which is created very fast) but it is the linking that is taking about 10 seconds.
On my desktop (Athlon XP 3000+) it takes 3 sec and on my notebook (Athlon M 1500+) it takes much less than 1 sec. I'm not running any programs on my desktop so there must be another reason for this. If POLINK is the cause, why does it take longer on certain machines?