[Open Source] Lucky's Spacebot

[dsnexium]

How do I set the image tolerance to be higher than 125? For some space missions this just isn't enough for my setup.

[phagotron]

How do I set the image tolerance to be higher than 125? For some space missions this just isn't enough for my setup.

If you set it higher than 125, it would take a picture of your dog for a red talon fighter. If it seems to be necessary to set tolerance to a high value, something is wrong with your desktop or swtor settings.You better work on that, since setting tolerance too high might make a mission start but thats just it and fun is over.

[TMichael]

Nice updates, the dragright seems to work great! Most of your code is already compiled into the latest hotfix. I beautified mmove a bit since it began to look a bit weird, guess it might need some more love until hotfix 6. Also I left out loop2 as I already had some changes in there, it didnt seem to benefit from the general timeout. Your Kalee mission also has been added, hope its ok that I put you into the bots "hall of fame"

Lots of thanks for the code, I really appreciate it and look ahead to Impossible Sector!

Awesome, glad to contribute!

I'm finishing up edits on the last minefield for the IS mission. After those are complete, I can focus more on the attack areas and timings for the last part. The bot actually completed all the regular mission goals on one recent run through, but I think it just got lucky. Then it blew up with about 1:00 left.

Okay, back to work.

EDIT

I'm testing these runs with normal level 48+ ship upgrades since I'd like the mission to work for anyone who has it available.

[vapor311]

Prev 5 has some strong issues. Its over coded tbh. It now focuses to strongly on itself. It pays more attention to its own missle fire trail and ignores primary targets. Not sure where the issue came into play in the coding but something went awol.

[phagotron]

Prev 5 has some strong issues. Its over coded tbh. It now focuses to strongly on itself. It pays more attention to its own missle fire trail and ignores primary targets. Not sure where the issue came into play in the coding but something went awol.

After some rounds I can confirm your observation, but not your conclusion. Dragright seems to be the problem here,as it introduces 200ms sleep time into every search process. Together with other delays this reduces firing with missiles to 2 or 3 aims a second from 5 to 7 before. Aiming its own rockets did occur before, but now it just is more noticeable.

I'll test if lower delays will help on that, but it will be necessary anyway to work on pixel colors so they wont target missiles. I'd be glad if anyone has a good suggestion for that since I havent got an eye for this pixelwork ...

[phagotron]

Awesome, glad to contribute!

I'm finishing up edits on the last minefield for the IS mission. After those are complete, I can focus more on the attack areas and timings for the last part. The bot actually completed all the regular mission goals on one recent run through, but I think it just got lucky. Then it blew up with about 1:00 left.

Okay, back to work.

EDIT

I'm testing these runs with normal level 48+ ship upgrades since I'd like the mission to work for anyone who has it available.

FYI: I updated mmove and dragright to by default work as the old funcs did. With an optional third parameter dragright can be activated then where needed:

Func dragright($xstart, $ystart, $drdelay = 5) ; Holds the right mouse button while dragging the mouse down some pixels
   If $ystart > 710 Then  $ystart = 710
   If $ystart < 20 Then   $ystart = 20

   $ystart -= 5
   MouseMove($xstart, $ystart, 1)
   ;Sleep(50)
   MouseDown("right")
   Sleep($drdelay)
   MouseMove($xstart, $ystart + 20, 2)
   Sleep($drdelay)
   MouseUp("right")
   ;Sleep(50)
EndFunc

Func mmove($pos, $r=False, $drdelay = 0)  ; uses dragright with drdelay > 0 
Local $x,$y           
Global $SWTORPID

	If $SWTORPID = -1 then return 0 ; no client? do nothing.

	If IsArray($pos) = True Then
		$x=$pos[0]
		$y=$pos[1]

		if $pos[0] < 0 then $x=0
		if $pos[0] > 1024 then $x=1024
		if $pos[1] < 0 then $y=0
		if $pos[1] > 766 then $y=766
		MouseMove($x, $y, 1)

		if $r = False then
			If $psx_debug = 1 and $imposs == 0 Then MouseClick("left") ; ???
			if $psx_debug = 1 then MouseDown("left")
			if $psx_fire = 1 then sleep($mdelay)
		endif

		if $r = True then
			if $psx_debug = 1 then 
				MouseDown("left")
				sleep(150)
			endif
			if $psx_fire = 1 then sleep($mdelay)

			$rdiff = TimerDiff( $rsync )
			if $rsync = 0 OR $rdiff > $rdelay then
			   if $drdelay > 0 then 
					dragright($x, $y, $drdelay)  
			    else
					MouseClick("right")
			   EndIf
			   $rsync=TimerInit()
	        	EndIf
		endif
					
		If $psx_debug = 1 Then MouseUp("left")
	EndIf
EndFunc

Also I adjusted the drag area a bit, which seems to work better in some imp missions, maybe you want to set this back.

[Cystacae]

I plan to start a BH soon, but ... there is not much hope to rise, since so far I have no idea how to overcome the pixel problem with reasonable effort.

Well I have kind of a fix for this but it will involve 6.0's double target grids feature and since I just came back due to my system being back online I am a bit behind on stuff. Actually Quite a bit behind but who is counting. I will remain behind by a lot but I am hoping my check comes in soon to get a notebook to test all the stuff I need to and code on and off while I am on my IRL vacation to see my family. I just got my new car as my sister fronted me the money knowing I am getting the check soon anyways. Got me a nice Volvo s80 t6 and I friggin love it! The double grid targeting system involves the input of two ini files with the grid co-ordinates and uses letter number combination to load grids into the game and shut them off. This can be done ship by ship basis for the individual pathing or open up several at the same time to do general searches. Along with this it has color codes ini file that loads up and you can assign the color combination number to search for grid by grid. This does allow you to avoid targeting things in the grid the ship is in which would keep form the bot targeting the BH's ship. This is the ONLY way I have figured out to kill this problem. If you can figure out a less complex version lemme know. The idea is all grids are 40x40 and the second grid panel is set 20 pixel in and is also 40x40. This allows a smoother coverage and allows you to only open up grid by grid if you want in a diagonal ship movement. I would like to try to get it so that things are more automatic but I am unsure. I know this also gives a statement of why not just go auto-recorded pathing then. Well the problem with that is syncing. We could do that but the syncing would cause a lot more issues and would revert us back to that issue we basically have nipped in the bud. This is where I am at and also Phago I will get the links and the first page here updated. Just lemme know what I missed while I was shot down due to my pc.

[vapor311]

Welcome back

[Cystacae]

Welcome back

Was I fashionably late? LOL

[vapor311]

Was I fashionably late? LOL

been busy here. lol progress is progress! Glad to see you.

[Cystacae]

been busy here. lol progress is progress! Glad to see you.

Glad to have my system back... GEIL angered me... they never answer their tech line in the US.

BTW, to everyone, all links are updated on the Spacebot. We moved our SVN to Assembla, this may even change I have no idea. I just was sick of not having full control over my SVN.

Also, TMichael, if you want to be a full member of the team lemme know and I will get you setup with a folder on the SVN. Assembla is allowing me to add accounts.

[vapor311]

ive been working on some theory coding with BH. If we set focus outside of said pixel area it should stay away from the engines of the ship.

[phagotron]

ive been working on some theory coding with BH. If we set focus outside of said pixel area it should stay away from the engines of the ship.

It would need one search to estimate the ships position and then from 2 to 4 conventional searches in areas calculated from ships position per needed pixel color. Sounds a bit like 2 targets a second or less ... another option would be to run a separate prog doing nothing except estimating the ships position and covering it with a black image ... would not look that nice, but could be effective since it wont delay the bot then. Anyway, I'll go to continue on the imp missions, so someone else might address that.

[phagotron]

We moved our SVN to Assembla, this may even change I have no idea.

Assembla works great, just commited my last updates to it. Main changes are mmove updates as posted and bound the Mission control standby mode to HD mode. And wb btw :)

[Cystacae]

Assembla works great, just commited my last updates to it. Main changes are mmove updates as posted and bound the Mission control standby mode to HD mode. And wb btw

Thanks! I appreciate it... from now about two weeks till I see my grandparents as we talked about phago. I will try to get a new notebook before then that will allow me to play swtor to test code and get all the info I need as well as 6.0 code onto it. I dunno how much I will get done I anticipate we will see a 5.36, 5.37, and by the 5.38 I will have an alpha of 6.0. and maybe by 5.9 we will see beta 6.0 and then 6.0 will be the next release stable after 5.9 that is my basic road map. So hopefully I can stay on track. I am not starting college till the winter for finishing my computer science degree (focus game development). So, it should give me time once I get my vacations done. Going to see my grandparents in Texas, my best friends in South Carolina, and my brother in North Carolina. Going to be about a month worth of time out of town and then it is moving into my own place again. Yay! I gtfo of my mother's home... love her and thank her for letting me stay here after my divorce but she drives me crazy sometimes and her new husband is a moron.