[Open Source] Lucky's Spacebot

[vapor311]

I need a bigger inbox on this site damn it lol! its over full and i get cannot send messages to often

[Cystacae]

Ill fix that with Word. Just wanted to make sure It covered our bums. I even sounded all legal and stuff in it!

GJ there mate, cheers!

[phagotron]

Welcome!

No worries, I was planning on keeping the new set separate for testing. I'd like to standardize the naming scheme for the timer images to something straightforward as well. Once the images are ready (along with the alt set you're getting) I figure we can update and test each mission. I envision something along the lines of:

If $result=WaitForMissionTimerInit($mission_name  & " 7:55 " , "pictures/55a.png", "pictures/55b.png", "pictures/55c.png", 7000) = 1 Then
   $mission_secs = 480  ; (7:55)
Else
   Send("{ESCAPE 1}")
   space_ende()
   Return 0
EndIf

This would cause us to rewrite each mission & to rename each pic, also it is hardly possible to deal with false positives then... I had something in mind that should make things a bit easier, just uploaded it to the SVN (yes, back from trip ) see misc.au3, Jabiim and Polith mission. The new sets would go into Set-A and Set-B in pictures folder. The set selection in user interface follows, will be a dropdown one to support multiple image sets. Anyway, testing can instantly start now.

[TMichael]

This would cause us to rewrite each mission & to rename each pic, also it is hardly possible to deal with false positives then... I had something in mind that should make things a bit easier, just uploaded it to the SVN (yes, back from trip ) see misc.au3, Jabiim and Polith mission. The new sets would go into Set-A and Set-B in pictures folder. The set selection in user interface follows, will be a dropdown one to support multiple image sets. Anyway, testing can instantly start now.

Right now the picture folder is a chaotic jumble because of the naming conventions used and repeated files. For example, Pakuuni, Nez, and Kalee each have their own images of "10". I'm guessing they were added as needed while the missions were being completed.

I think it's time to consolidate and simplify. Having pictures with the times in the descriptor would be much easier, which means implementation would be a breeze. I was planning to do this as part of the mission code cleanup anyway.

For example, this code from the IS mission didn't make sense when I first read it, referencing other missions:

$result=WaitForMissionTimerInit($mission_name  & " 8:00 " ,"pictures/space_hydian00.png",7000)
$mission_secs = 480 ; (8:00)

if $result = 0 then
   $result=WaitForMissionTimerInit($mission_name & " 7:57 " ,"pictures/space_polith.png",3000)
   $mission_secs = 477 ; (7:57) 
Endif	   

if $result = 0 then
   $result=WaitForMissionTimerInit($mission_name & " 7:55 " ,"pictures/space_hydian55.png",3000)
   $mission_secs = 475 ; (7:55) 
Endif

If $result=0 Then
   Send("{ESCAPE 1}")
   space_ende()
   return 0
EndIf

Whereas one call with clear image names would be easier to maintain and expand upon in the future as the game releases more space missions. This should also make it unnecessary to check at three different times just to guarantee an image hit:

If $result=WaitForMissionTimerInit($mission_name  & " 7:55 " , "pictures/55a.png", "pictures/55b.png", "pictures/55c.png", 7000) = 1 Then
   $mission_secs = 480  ; (7:55)
Else
   Send("{ESCAPE 1}")
   space_ende()
   Return 0
EndIf

IMO implementing one check to catch all possibilities, as opposed to having the user choose a specific picture set, which still might or might not work 100% of the time, would make more sense.

But ultimately, you guys can decide. I'm just making my case here. I'm good with Photoshop and really don't mind the work, but I'm happy to go along with whatever system is put in place.

[Schadwe]

Aetan Mission isn't working for v5.34. So far The biggest culprit is the timing. The mission file needs to be changed so that the over all mission is 322 sec instead of the 300 in the file:

$mission_secs = 322

Looks like someone added a second $mission_secs line that overrode the initial value.

I'm still tracking down an issue with Func Aeten 4. It seems that the mission does not fire the blasters (only the missiles) in that function.

[phagotron]

Right now the picture folder is a chaotic jumble because of the naming conventions used and repeated files. For example, Pakuuni, Nez, and Kalee each have their own images of "10". I'm guessing they were added as needed while the missions were being completed.

I think it's time to consolidate and simplify. Having pictures with the times in the descriptor would be much easier, which means implementation would be a breeze. I was planning to do this as part of the mission code cleanup anyway.

You are dang right, its the result of some ancient copy&paste orgy. Since some missions had high fail rates with only one image, I added as necessary. To have at least a bit order in it, I named them with the timer index. So far the story. Since I havent tried I cannot say if it works or not, but unless they already occur with mission specific pics, I would say its no big guess to see false positives with common images.

IMO implementing one check to catch all possibilities, as opposed to having the user choose a specific picture set, which still might or might not work 100% of the time, would make more sense.

But ultimately, you guys can decide. I'm just making my case here. :) I'm good with Photoshop and really don't mind the work, but I'm happy to go along with whatever system is put in place.

You are a member of the team, so we all should decide together. Unless someone puts scrum master hat on (and fun is over:). Else discussion wont make much sense, and its your time so you decide. Who else could? And - we are working on the same function. If that would go without discussion Windows would be Open Source a day later.

My only issue is that your code fragment pointed out a solution but isnt backward compatible. I have such a backwarded thing ready, but no more testing, sun is rising. Will post when its on the svn.

[Cystacae]

You are dang right, its the result of some ancient copy&paste orgy. Since some missions had high fail rates with only one image, I added as necessary. To have at least a bit order in it, I named them with the timer index. So far the story. Since I havent tried I cannot say if it works or not, but unless they already occur with mission specific pics, I would say its no big guess to see false positives with common images.



You are a member of the team, so we all should decide together. Unless someone puts scrum master hat on (and fun is over. Else discussion wont make much sense, and its your time so you decide. Who else could? And - we are working on the same function. If that would go without discussion Windows would be Open Source a day later.

My only issue is that your code fragment pointed out a solution but isnt backward compatible. I have such a backwarded thing ready, but no more testing, sun is rising. Will post when its on the svn.

Well shit, if we are working on stuff together, how about we all work together on 6.0 I scrapped most of my idea because I am just coming up with dead ends. I was going to use some pre-existing code actually bits and pieces. However, it seems I am running into issues with trying to piece together a new work but in essence it is a total backwards compatible thing. You guys want to pow wow a new decision? I am just too annoyed and frustrated with trying to do these links and bits and pieces of code. Maybe it is because I spent so much time away form it my ideas got flushed out of my head. Shit, my girlfriend is finding out my memory is about the worst anyone has for short term. She also says I may have ADHD just because I bounce around a lot and do 500 million things at once. I have some ideas but they are even fragments atm. Also a lot of the stuff on my pc I some how lost... I think it happened when I had corruption on the OS and had to reinstall because I couldn't save it.

[TMichael]

You are a member of the team, so we all should decide together. Unless someone puts scrum master hat on (and fun is over. Else discussion wont make much sense, and its your time so you decide. Who else could? And - we are working on the same function. If that would go without discussion Windows would be Open Source a day later.

My only issue is that your code fragment pointed out a solution but isnt backward compatible. I have such a backwarded thing ready, but no more testing, sun is rising. Will post when its on the svn.

Yeah, the image name changes might make more sense in a new, clean version 6.0 if we want to keep backwards compatibility.

My own experience with image detection issues: When tinkering with the Archenar image (which is one mission that wouldn't start for me), tolerance changes didn't seem to help at all. I tried editing the image and replacing more of the background with black, but still no luck. The mission script just wouldn't get past the starting image check. So, I took a screenshot, cropped it, edited it, and saved it into the Pictures folder. The new image looked much brighter than the original, and upon use, it immediately fixed my problem. Hopefully this means we're on the right track.

I'll work on some of the image clean up tonight, leaving the originals in the main Pictures folder and the putting the copies into A or B. Perhaps we can put the sharper, brighter images into B (the images in there now look quite sharp/bright) and the toned down copies into A. In Photoshop, I should be able to replicate the difference in color saturation/vibrancy/brightness between the image types currently in use.

Anyway, I'll test some of the mission starts after doing some image populating, and I'll keep an eye out for your update.

[TMichael]

Well shit, if we are working on stuff together, how about we all work together on 6.0 I scrapped most of my idea because I am just coming up with dead ends. I was going to use some pre-existing code actually bits and pieces. However, it seems I am running into issues with trying to piece together a new work but in essence it is a total backwards compatible thing. You guys want to pow wow a new decision? I am just too annoyed and frustrated with trying to do these links and bits and pieces of code. Maybe it is because I spent so much time away form it my ideas got flushed out of my head. Shit, my girlfriend is finding out my memory is about the worst anyone has for short term. She also says I may have ADHD just because I bounce around a lot and do 500 million things at once. I have some ideas but they are even fragments atm. Also a lot of the stuff on my pc I some how lost... I think it happened when I had corruption on the OS and had to reinstall because I couldn't save it.

Don't feel too bad - my wife likes to refer to my memory as "alzheimer-y".

Yay, new version! I love tinkering, improving, and optimizing, so I vote aye. Off the top of my head, some potential ideas for the 6.0 task list:

• More concise, pretty, user-friendly UI
  
  • Win/Loss/Bonus tally per mission
  • Simplified user experience - it should just work
  • The user shouldn't have to do much more than choose missions and crew tasks in most cases
• Every mission updated with an 80 - 90%+ success rate
• Code clean-up/optimizations
• More stuff I can't think of

[Cystacae]

Don't feel too bad - my wife likes to refer to my memory as "alzheimer-y".

Yay, new version! I love tinkering, improving, and optimizing, so I vote aye. Off the top of my head, some potential ideas for the 6.0 task list:

• More concise, pretty, user-friendly UI
  
  • Win/Loss/Bonus tally per mission
  • Simplified user experience - it should just work
  • The user shouldn't have to do much more than choose missions and crew tasks in most cases
• Every mission updated with an 80 - 90%+ success rate
• Code clean-up/optimizations
• More stuff I can't think of

This is intended to be a code rewrite completely... we will keep Lucky as the respect he deserves for giving us the foundation to go on. Just instead of calling the Lucky's Spacebot we are going to go with Lucky Spacebot or something like that. Since this bot is technically a complete rewrite to make sure we compensate for problems and make sure the code remains clean. I do want to add a module to the bot so when it is compiled with it, it will unlock additional features for donators that will give them some form of skinning ability and more of a personalized touch. Donations do not require a minimum amount but this module will be closed to the general public. However, since it only does skinning and personalization it doesn't cripple the functions of the bot for non-donators. I just want a way to show our appreciation for their financial support. I know I can never fully protect this code but I want it never to be released in source and if someone donates we can provide them with a link to download it once they donate and that version will be compiled with that module. It is the only way I can think of to reward them. I am thinking of you donators. We do appreciate it that our work is getting recognized and that you appreciate our hard work enough to donate your hard earned money. Also I will provide a link to donators for 6.0 once it goes to alpha phase and you can close test all the new goodies. We'll talk more... I know Phago doesn't want to skype it he has ventrilo or something like that. TM we can connect up with him and brainstorm all over again with you involved for 6.0. I will also be assisting you guys since you are helping me with 6.0 to aiding in some of the stuff for 5.3 line and what is left to go for it. I am sure some of the code from 5.3 will still have some usefulness in 6.0 anyways. Biggest thing I want to tackle is getting the bot to work for all classes.

As far as the UI for 6.0 I have been working on it. I am trying to compact all things down but leave the bot with as many goodies as possible. I am using a lot of pull down combo boxes, main menu systems, context boxes for dailies instead of checking you just click on the mission and add it to the other context box via and arrow. I am also keeping shit hidden and not a tabbed setup until the user calls for that specified thing. Biggest thing is a huge rewrite to the targeting system. We need to to eliminate the urge the bot has to want to commit suicide, no suicidal tendencies anymore. We want a less emo and more friendly bot.

Also check out the modded bot I got from another user... it is off 5.33 but some of the code maybe useful. I also want to split the code so that the menuing system and the call for function start() is in the first file and the rest of the bot is broken up to core functions, targetting functions, crew skill functions, etc. So that way the code isn't one jumbled mess. I think I started doing that with 6.0 and I have the basic idea I want in my PC atm. I just broke up a 5.34 cause that is what I was working out of at the time. I was using it like a roadmap instead of building off of it.

[TMichael]

Sounds great. For 5.35 I'll continue to help with the image additions and mission code clean-up. For 6.0, ventrillo works for me as well.

[Cystacae]

Sounds great. For 5.35 I'll continue to help with the image additions and mission code clean-up. For 6.0, ventrillo works for me as well.

Phago we should get to the meeting time and your ventrilo. So give an idea so we can all pow wow, Vapor you are invited too along with your helper if they want.

[phagotron]

Well shit, if we are working on stuff together, how about we all work together on 6.0 I scrapped most of my idea because I am just coming up with dead ends. I was going to use some pre-existing code actually bits and pieces. However, it seems I am running into issues with trying to piece together a new work but in essence it is a total backwards compatible thing. You guys want to pow wow a new decision? I am just too annoyed and frustrated with trying to do these links and bits and pieces of code. Maybe it is because I spent so much time away form it my ideas got flushed out of my head. Shit, my girlfriend is finding out my memory is about the worst anyone has for short term. She also says I may have ADHD just because I bounce around a lot and do 500 million things at once. I have some ideas but they are even fragments atm. Also a lot of the stuff on my pc I some how lost... I think it happened when I had corruption on the OS and had to reinstall because I couldn't save it.

First they love it, then they joke about it, then they blame you for it until they find out how to use it. If at stage 4 you usually are married or at least engaged. Be prepared:)

I can only speak for myself. I would feel bad if I'd leave 5.35 as it is and dive into something new. I know that you badly want 6.0 to pace up but what would it help if we start over with 7.0 in same way one day leaving behind some half baked code behind. Maybe I'm a bit too much on rails to split my attention, maybe things go a bit too fast for me, so unless 5.35 isnt done I wont be much help for 6.0. Also maybe I feel a bit pushed atm. Maybe its the "just make this feature work, no matter how" thing where I only know too well what it leads to.

[phagotron]

Btw - any reason why 5.35 shouldnt go GPLv3? Since 6.0 will be a complete rewrite there should be no conflict if it has a different license and its about time to have a working one for 5.35.

[TMichael]

Btw - any reason why 5.35 shouldnt go GPLv3? Since 6.0 will be a complete rewrite there should be no conflict if it has a different license and its about time to have a working one for 5.35.

That makes sense to me.

I can only speak for myself. I would feel bad if I'd leave 5.35 as it is and dive into something new. I know that you badly want 6.0 to pace up but what would it help if we start over with 7.0 in same way one day leaving behind some half baked code behind.

No worries, I feel the same way about 5.35. I'm right there with you until it's ready to go out into the big bad world on its own.