Sumoman

another weird problem i'm having: for some reason the line Send("!1") doesn't always work. But sometimes it does....

Wait... i thought the primary colors were Red, blue, and yellow.... don't blue and yellow make green? But the colors you listed make sense for my results (the text color i'm having trouble with is greenish). Still... the hex values for my text are all over the place, I want to be able to distinguish between the White text, the green text, and the black-grey-blue background. Any ideas of how to get this green text?

my other problem with this shade variation is that the hexa string for some of the colors is really complex (i.e. 175B19) and i'm having trouble averaging all of the numbers.... here's some raw data i'm getting from just 1 color of text 2D9A2D 175B19 271100 21FF7C 21FF57 I'm gonna work on averaging these and see what i can come up with >.< edit: If I did all the calculations right (which i prob didn't >.<) then the colors would look like this: 46 155 46 24 92 26 40 18 0 34 256 125 34 256 88 That's a pretty big range o.0

edit - stupid question >.< here's a better one how does pixel search determine whether to search in Deci or hexa? Also: my search came with FFFFF8 and FFFFFF, that = 8 variation right? so if I go with 16 would that work? edit... AGAIN >.< is it the "0x" FFFFFF that's the flag for hexa? edit.... just in case i haven't edited it enough how do i set @error back to 0 if pixelsearch sets it to 1

Can i use deci to determin my shade variation? Because the shade variation i was seeing with decimal was changing even the 2nd digit, so i don't know if 255 would be enough... edit: i'll try hexa first though, how do i convert from hexa to deci though? is 0-9 then a-f?

I'm having a very hard time trying to use autoIT's pixelsearch and pixelgetcolor to read text in a graphics window. (I know, I know. I have been previously told that autoIT is not optical recognition software, but unless autoIT can work with an OCR [which i'll have to research in the meantime] then i'm going to have to stick with pixelsearch). My problem is that the text always appears to be white, but the spywindow gives me a different deci-color everytime the text is refreshed. So I was wondering how I make pixelsearch for 1 pixel, and also how to work with the shade variation variable (i.e. how large of a variation is large enough to cover all of these different whites)

Yes, i'm trying to read text from a graphics window. >.< sounds like this is gonna be hard First of all, my problem is that i can't get exact pixels with the spy window. I mean there's 800 increments for just 1 direction on the screen Any tips on getting exact numbers.... and any tips on autoIT tricks for reading graphic's text? OH here's a brainstorm.... how about I use autoit to move the mouse to the exact pixel i want >.> tricky tricky... but i might just work

I understand that autoIT comes packaged with a nice pixelcolor/location getting program, but I cannot get very detailed results from this program. Also what program do you use to take screen shots? I use fraps right now, is there a better one out there? My problem is that despite my best efforts i cannot seem to get my autoIT script to read text in a program and respond accordingly. I am also using 800x600 resolustion thank you!

errrr nm that was a problem with my script itself, @error does reset itself.

hmmmm @error=0 is invalid, and if i don't reset it to 0 it always hits true in a loop... any ideas on how to reset it?

>.< oh well, thanks for the help anyway

what i meant was, why is it: IF NOT @error instead of : IF NOT $error

What's the significance of an @var compared to a $var?

I was wondering if you can use the pixelsearch command in an IF THEN statement. something like: IF Pixelsearch(12,40,100,200)=1677555 THEN If not, then would: DO PixelSearch(1,1,100,100,1556765) IF NOT @error THEN dfaefdasfesdageaefd EndIF @error=0 UNTIL 1=2 reset @error everytime it goes through the loop? would that work?