Timings

[Suilven]

Hi all,

I'm new to AutoIT, so apologies if I should have found the answer elsewhere (I did look, honest!). I'm using the current AutoIT3 and Scite and trying to automate a Lotus Notes application. I'm doing this across an RDP link, if that is relevant. I think I have 2 problems -

1) Can I slow the execution of the script down? It seems to be too quick for poor old Notes to get some of the field data sometimes.

2) Sleep(5000) seems to be ignored sometimes and messageboxes do not always wait the 10 seconds that I ask. It is inconsistent with the same script behaving differently on different runs.

I have tried to create a small script that shows this behaviour but can never get anything that is consistent.

Any pointers or ideas (or RTFM comments) will be deeply appreciated.

TIA Dave

[kaotkbliss]

at the top of your script you can put

Opt("SendKeyDelay", 45)

45 can be changed to any number you wish, it is the delay in milliseconds between automated key presses. Also instead of using sleeps, maybe controlclick, controlsend, winwait?

[Suilven]

at the top of your script you can put

Opt("SendKeyDelay", 45)

45 can be changed to any number you wish, it is the delay in milliseconds between automated key presses. Also instead of using sleeps, maybe controlclick, controlsend, winwait?

kaotkbliss,

Many thanks - I totally missed the SendKeyDelay option, so I'll try that straightaway. re Sleep(), at one point, I have to wait until Notes sends some data back to the screen. Unfortunately with Notes, controls and text are mostly invisible to AutoIT and the window caption does not change, so I want to wait 60 seconds before restarting the script. Without Sleep(), what's a good way of putting that order of delay into the system?

TIA Dave

[kaotkbliss]

if winexists or winwaitactive? depending on how notes works

[Suilven]

if winexists or winwaitactive? depending on how notes works

That was my first try, but AutoIT sees no change in the window between the state when I press a Notes button to ask the server for an id and when the id appears on the screen. The discovery tool reports exactly the same parameters for both states. I've looked at controls, visible text and invisible text. My only hope is to wait 60 seconds. Sleep(60000) should be the answer but seems buggy. Do others see it as buggy/inconsistent? Or is this me misinterpreting the symptoms?

Notes, BTW, is to be avoided like WW3. Goodness knows how it works, when it works. My job is to hide the whole system from users so they can have their lives back )

Thanks for you help. Dave

[kaotkbliss]

so when Notes is done doing whatever it is doing, there is no change in the window? If there is even a text change that wingettext or a button that becomes dissabled or enabled then maybe controlcommand?

[Suilven]

I can see no change with the Find Tool. Is that using WinGetText? I might try that directly and see, but almost all windows text is invisible to teh Find Tool. Notes has its own GUI and is a bit like a browser in the way it reorganises the window if resized. But, as I write, I'm wondering if the text is on a child surface. I've got WinSpy somewhere here - I'll poke around with that.

But what about Sleep()? I can't imagine this is broken when it would be a straight win32 call. Does anyone else have problems with timings or am I barking up the wrong tree?

Dave

[kaotkbliss]

Sleep does seem to be funny. I created a program to keep track of how much time was spent on a project and in 1, 8 hour work day it recorded 10 hours of work

[jchd]

@Suilven

Welcome to the forum.

I just tried the test below and it does give very consistent results for me, either local or running thru a TS session (XP x86 SP3 everywhere).

Local $t
For $i = 1 To 24
    ConsoleWrite("start" & @LF)
    $t = TimerInit()
    Sleep(5000)
    ConsoleWrite("stop - start = " & TimerDiff($t) & @LF)
Next

I tried various sleep values and, beside a small variation and overhead of other calls, they work as expected for me.

@kaotkbliss

How do you count time using Sleep()?

@both

It would be interesting to post a sample creipt where Sleep() clearly fails, in order to track down the issue.

[kaotkbliss]

@jchd

I don't but I assume that both sleep and timer functions use the same timekeeping method (a count to 10 ms in one is the same as a count to 10 ms in the other)

[jchd]

That's true but Sleep introduces a delay in the flow of your program and I'm just curious about how you can use it for timing run time. I believe TimerInit / TimerDiff is better suited for such measurement.

[Suilven]

@Suilven

Welcome to the forum.

I just tried the test below and it does give very consistent results for me, either local or running thru a TS session (XP x86 SP3 everywhere).

Local $t
For $i = 1 To 24
    ConsoleWrite("start" & @LF)
    $t = TimerInit()
    Sleep(5000)
    ConsoleWrite("stop - start = " & TimerDiff($t) & @LF)
Next

I tried various sleep values and, beside a small variation and overhead of other calls, they work as expected for me.

@kaotkbliss

How do you count time using Sleep()?

@both

It would be interesting to post a sample creipt where Sleep() clearly fails, in order to track down the issue.

Well, it is very reassuring NOT to hear anyone else having trouble with Sleep(). I have tried to create a small example that fails consistently but I cannot create one. When it fails, it fails totally, so a 10 seconds sleep() is just ignored or a messagebox with a 5 second wait in it just blinks before disappearing. The same script will work sometimes and at other times fail. The work I did on the previous AutoIT version did not show this problem, so I'm going to reinstall, just in case. Would have done this sooner but I have to cajole someone at the client end to do the work...

Thanks for everyone's help on this - makes a big difference when you know you're not on your own!

Dave

[jchd]

Not being left alone Sleep($ing) in the dark in the whole purpose of a forum.

More to the point: can you have it fail in a local session (no TS)? FYI, my test were only on a LAN, no Wan/VPN/you-name-it involved.

Which OS do you run and could it be a virtualization side-effect? (I'm not GRU nor KGB, just trying to understand!)

[IchBistTod]

Proposed solution:

Very low cpu usage ~9MS off using sleep(not relative to amount of time).

func _Sleep($ms)
$timer = timerinit()
while timerdiff($timer) < $ms
sleep(1)
wend
endfunc

Max CPU usage, exact time, down to the nanosecond(not relative to amount of time).

func _Sleep($ms)
$timer = timerinit()
$ms = (($ms-0.1)+0.07)
while timerdiff($timer) < $ms
;nothing
wend
endfunc

tell me if you have any luck with it.

Green= My Sleep

Red=AU3 Sleep

Test(1000ms):

1000.00107962209

1001.00960277494

1000.00256635174

1000.00359289244

1000.0010265407

1000.00667251453

1000.00256635174

999.984088619205

1000.00110616279

1000.00872559592

1000.0010265407

999.297846163638

1000.0010265407

999.998973459304

1000.00153981104

999.993840755821

1000 <<<<<<<<<<<<<<<< PERFECT!

1000.01745119184

1000.00359289244

1000.00051327035

1000 <<<<<<<<<<<<<<<< PERFECT!

999.909664418714

1000.00153981104

1000.01539811045

1000.00205308139

999.993840755821

1000 <<<<<<<<<<<<<<<< PERFECT!

1000.01283175871

1000.00153981104

1000.00564597383

Test(30000ms):

30000.001539811

30011.3386552625

30000.0010265407

30003.2941690949

30000.0020530814

30008.3391033475

as you see my sleeps are never more than ~.003 off where as AU3 sleeps can be up to ~1.0 off on a simple 1 second sleep.

AU3 sleep shows much variation so when used for larger amounts of time,it can be very inaccurate.

AU3 sleep is offset relevant to the amount of time the sleep is called for, neither express this attribute.

Edited March 23, 2010 by IchBistTod

[Suilven]

Proposed solution:

Very low cpu usage ~9MS off using sleep(not relative to amount of time).

func _Sleep($ms)
$timer = timerinit()
while timerdiff($timer) < $ms
sleep(1)
wend
endfunc

Max CPU usage, exact time, down to the nanosecond(not relative to amount of time).

func _Sleep($ms)
$timer = timerinit()
$ms = (($ms-0.1)+0.07)
while timerdiff($timer) < $ms
;nothing
wend
endfunc

tell me if you have any luck with it.

Green= My Sleep

Red=AU3 Sleep

Test(1000ms):

1000.00107962209

1001.00960277494

1000.00256635174

1000.00359289244

1000.0010265407

1000.00667251453

1000.00256635174

999.984088619205

1000.00110616279

1000.00872559592

1000.0010265407

999.297846163638

1000.0010265407

999.998973459304

1000.00153981104

999.993840755821

1000 <<<<<<<<<<<<<<<< PERFECT!

1000.01745119184

1000.00359289244

1000.00051327035

1000 <<<<<<<<<<<<<<<< PERFECT!

999.909664418714

1000.00153981104

1000.01539811045

1000.00205308139

999.993840755821

1000 <<<<<<<<<<<<<<<< PERFECT!

1000.01283175871

1000.00153981104

1000.00564597383

Test(30000ms):

30000.001539811

30011.3386552625

30000.0010265407

30003.2941690949

30000.0020530814

30008.3391033475

as you see my sleeps are never more than ~.003 off where as AU3 sleeps can be up to ~1.0 off on a simple 1 second sleep.

AU3 sleep shows much variation so when used for larger amounts of time,it can be very inaccurate.

AU3 sleep is offset relevant to the amount of time the sleep is called for, neither express this attribute.

Thanks for all the ideas. I'm working on something else today while AU gets reinstalled. But I'll be back to let you know how it pans out.

Thanks again, Dave