Brain Research:

In the early 90's in 7th grade I was a bolo addict on the old Mac LCII. We had an awesome Appletalk lab at school - the latest technology! So, probably like all the rest of ya, I'm rediscovering an old game for the sake of nostalgia and some fun.

I do have a little coding experience in my past, and I'd like to give brain construction some effort. I'd like an interesting computer project and after stumbling across this site, it reminded me how much fun I had playing against allied brains back in the days of macbolo. (yes I know in winbolo they can't yet ally themselves, I can do it manually if I want)

So, I tried to download the sample workspace someone mentioned in a previous post, but the link is broken. Does anyone have a good VS workspace already set up with a compiling brn I can start from? Also, does anyone have the source code for the macbolo indy 2.02, 3.1, ladma 6, or any of the other decent brains so I'm not reinventing the wheel here? I'm sure there are some pretty slick algorithims for pathfinding etc. that would take me a long time to put together myself.

I looked back through the old posts here and can see there are probably some hurdles to success as compared to the old macbolo brains, but I've got spare time and a lot of interest. My goal is to build upon some of the old brains and get something put together that could at least capture all the bases on Everard island, let alone put up a good fight or take a pillbox.

The only thing I know right now is how much I don't know about this, so I look forward to getting some input from the few of you who've experimented in the past...

Thanks!

You can find a number of brains and borgs for download at http://www.winbolo.us/downloads.asp as well as the source for some of them.

I wish you luck with this.

Building on LRL's post, specifically, "btemp.zip" is the Visual Studio workspace. It includes a DX7 input wrapper.
Scroll down to find "btemp.zip" and download it from it's mirror.

The only Mac code I could find was the authors original documentation that included some sample C, as well as the standard autopilot that comes included in WinBolo's sample code folder.
As for being Windows\Linux specific, Ryan2 offers it's source but doesn't include anything too different from Stuart's autopilot.

I've built a couple of brain's that I havn't released, it will definately be a challenge making something that doesn't suck. Keeping track of pill's is a interesting prospect in the bot interface.

Min

Build a bot that cuts set-up and bitching time to somewhere around 30 seconds.

Nova wrote:

Build a bot that cuts set-up and bitching time to somewhere around 30 seconds.

Impossible.

Min

OMG FALLACY!!!!!!!!!!!!!!!!!

Nova wrote:

OMG FALLACY!!!!!!!!!!!!!!!!!

OMG!!! Looks like Nova needs to do some brain research... and invest in a new one. Get the new model.. the one with the candyapple-red paint and the 17" chrome wheels. Make sure you get the super-sporty package, the big 6.2L V8. That'll sure turn those gears in your head... and will probably turn other people's heads too...

Once you got the gears going to adequate velocity I'd be constantly spinning. I'm afraid it's a lost cause, mate. Thanks for the consideration anyway.

You know what's sad? That is 119% true.

Yeah, they tried making the gears lighter but by the time they got it to non-spin mass the gears ended up snapping. It was a really neat party trick to sneeze stripped gears but there's only so much of one thing a man can take.

Min wrote:

I've built a couple of brain's that I havn't released, it will definately be a challenge making something that doesn't suck. Keeping track of pill's is a interesting prospect in the bot interface.

Min

Keeping track of them in what way? I'm assuming there are functions that return number of built, carried, etc. pills, right? Just as a player could scan his pill window and see each pills status?

Just got started messing with the the template in VS...

Operon wrote:

Keeping track of them in what way? I'm assuming there are functions that return number of built, carried, etc. pills, right? Just as a player could scan his pill window and see each pills status?

there isn't any functions to return pill information... the bot interface is given less information about pill status than a player is. Which is why its interesting, its also why I havn't bothered completeing a bot, why bother when it doesn't even get the basic information that a human gets.

Min

Min wrote:

there isn't any functions to return pill information... the bot interface is given less information about pill status than a player is. Which is why its interesting, its also why I havn't bothered completeing a bot, why bother when it doesn't even get the basic information that a human gets.

Min

Just out of curiousity, was this information even available to the Mac Bolo brains? He must have cut a chunk out of the brain writing info file Stuart Cheshire provided if it was... Unfortunately, I don't have access to a mac now and I can't seem to find the Mac bolo version of a brain writing guide to check...

Not time to mess around with stuff now anyway, will check back later...

http://bishop.mc.duke.edu/bolo/links/

Some broken links there but I think there's still a Mac brain or two handy.

the key is in how the object list is mantained, there is really no way to know what sort of information the object list gave macbolo'ers, without loading up a mac and coding something to output it.

Min

Min wrote:

the key is in how the object list is mantained, there is really no way to know what sort of information the object list gave macbolo'ers, without loading up a mac and coding something to output it.

Min

Just for grins, I sent an email off to "Jolo" via a form on his bio page... Maybe he knows how to contact Paul Joswig or other contributors who may still have copies of old source from the more advanced brains.

I only took a cursory glance at the Standard Autopilot code originally written by Stuart Cheshire, but not enough to determine what data it had to go off of... For that matter, was that modified by the creator of Winbolo only to show the code he used in the Winbolo version rather than all the latest Macbolo code?

Ok. Now we all know that everybody is constantly saying that the brain interface gets less information than a player does, but let's look on the bright side.

The brain interface supplies this information that is NOT given to a player:

-LGM coordinates (Including parachute coordinates)
-Weither or not LGM is touching a wall
-a bunch of crap i forgot about

So there we have it. NOW, not only can the brain builders bitch about the brain interface BUT THE PLAYERS can also bitch about the human interface.

The more to complain about, the merrier.

:lol: :D :) :wink: :P

DANTHEMAN! wrote:

...The more to complain about, the merrier.

:lol: :D :) :wink: :P

Well, apparantly, this JM(Elvis?) guy did a pretty good job of passing all the same objects to a Winbolo brain that Macbolo sent off. At least, from what I see in the last macbolo release documentation, (found a working Mac LCIII emulator) the BrainInfo struct is almost identical to the sample provided by JM.

From what I've read, some of the later Macbolo brains incorporated moderately succesful strategy engines. Going off the assumption these were accomplished with data similar to what Winbolo provides, I think there's still significant room for development...

Too bad none of the old source is available, but I suppose half the fun will be developing strategy and tactic engines myself...If I can even get a tank to move around on the map!

Operon wrote:

Well, apparantly, this JM(Elvis?) guy did a pretty good job of passing all the same objects to a Winbolo brain that Macbolo sent off. At least, from what I see in the last macbolo release documentation, (found a working Mac LCIII emulator) the BrainInfo struct is almost identical to the sample provided by JM.

seems you misunderstood what I said, the object list isn't given the nessicary information. Thats not the braininfo structure, its the object list itself, there is -no- way to know exactly how the information was presented in the object list itself in macbolo, without building a brain on a mac, and having it output it. In winbolo, this information isn't presented in a way thats useful imo.

Min

DANTHEMAN! wrote:

-LGM coordinates (Including parachute coordinates)
-Weither or not LGM is touching a wall
-a bunch of crap i forgot about

so your saying if you scroll around, you can't see were a lgm is on your screen? .... same for weither a lgm is touching a wall ...... you can't see it touching a wall? ... perhaps your eyesight should be checked.

Min

Min wrote:

seems you misunderstood what I said, the object list isn't given the nessicary information. Thats not the braininfo structure, its the object list itself, there is -no- way to know exactly how the information was presented in the object list itself in macbolo, without building a brain on a mac, and having it output it. In winbolo, this information isn't presented in a way thats useful imo.

Min

Probably I did... So tell me, this information is there, even if it's a lot of extra work to process, correct? As I said, I only took a cursory look at the files... Also, many of these terms are new to me as this project serves to motivate me to learn c, which I only have a rudimentary knowledge of at present. Not exactly starting with "hello world", but I I haven't done anything fancier than sorting multi-dimensional arrays...

I also have never done anything with windows before, so even some of the menu handlers are foreign to me. All my coding experience is from APCS high-school classes using pascal - the last year before the APCS program switched to c :( - on a Xenix time-share environment in a computer lab that was already ancient.

So yeah, I don't expect to be actually coding a working brain for months, but this project is my motivator to keep me interested and provide material for practice. In other words, you might have to spell things out for me, at least until you're completely disgusted with my ignorance. :)

i think u and min will get along very well.

Min wrote:

DANTHEMAN! wrote:

-LGM coordinates (Including parachute coordinates)
-Weither or not LGM is touching a wall
-a bunch of crap i forgot about

so your saying if you scroll around, you can't see were a lgm is on your screen? .... same for weither a lgm is touching a wall ...... you can't see it touching a wall? ... perhaps your eyesight should be checked.

Min

So, I'm saying that if I scroll around, SOMETIMES you can't see your lgm on the screen... same goes for an lgm that is touching a wall... you can't see it touching a wall. Perhaps your brain should be checked.

You know, sometimes your lgm can be on a different area of the map.

Operon wrote:

Min wrote:

seems you misunderstood what I said, the object list isn't given the nessicary information. Thats not the braininfo structure, its the object list itself, there is -no- way to know exactly how the information was presented in the object list itself in macbolo, without building a brain on a mac, and having it output it. In winbolo, this information isn't presented in a way thats useful imo.

Min

Probably I did... So tell me, this information is there, even if it's a lot of extra work to process, correct? As I said, I only took a cursory look at the files... Also, many of these terms are new to me as this project serves to motivate me to learn c, which I only have a rudimentary knowledge of at present. Not exactly starting with "hello world", but I I haven't done anything fancier than sorting multi-dimensional arrays...

I also have never done anything with windows before, so even some of the menu handlers are foreign to me. All my coding experience is from APCS high-school classes using pascal - the last year before the APCS program switched to c :( - on a Xenix time-share environment in a computer lab that was already ancient.

So yeah, I don't expect to be actually coding a working brain for months, but this project is my motivator to keep me interested and provide material for practice. In other words, you might have to spell things out for me, at least until you're completely disgusted with my ignorance. :)

I've expiremented with building a brain in the past, not seriously, but merely as a recreational activity and like you, to learn some C. I understand Min's complaints about the brain interface, but I don't think they will be a real obstacle for somebody who is merely attempting to have some fun. However, Min feels the need to bitch, and I have no issue with that. I don't believe that winbolo brains will ever be as good as human players unless some serious, paid, proffesional full-time developers are put on the project, which is highly unlikely. Therefore, I see no reason for a perfect brain interface to be implemented.