This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - Trovaner
Pages: 1 ... 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 ... 64
781
« on: April 10, 2012, 11:56:37 AM »
I think its possible but it wouldn't be easy. Also, since I've never done anything like it, I can't say for certain that it would even work. First off, it would need to be separated into four components (two motors, one wheel, and one housing unit). It would go from housing unit to motor to motor to wheel. The housing unit would pretty much look like your current rim but the collision mesh would have a slightly bigger hole than the display mesh. There would also be an AP along the inside of one of the sides. The motors would be identical (just to simplify things a little). Attaching the motor to the shell would put the body of the motor along one of the adjacent sides (meaning the Generic_M end would be floating in space at 90 degrees from the actual motor). Attaching the second motor to the first would have it go to another adjacent side (0 or 180 degrees from the shell's Generic_M AP). The axle of each motor would be right in front of the motor's body. Unfortunately, it wouldn't work right unless you locked the axles in place (like the DSL tread wheels are). The sphere would have a traction zone like any other wheel. I'm pretty sure that you would only need to use a slightly larger sphere as the traction zone for it to work realistically. Unfortunately, with the axles locked on the motors, this wouldn't actually rotate. If the visual rotation is necessary, you could turn the housing unit into something like Serge's free moving axle but the other stuff would still be necessary. IIRC, there are like five different types of hinges listed in the EXE but we don't know the syntax for all of them. If we did then this might have been easier to implement.
782
« on: April 09, 2012, 02:44:01 PM »
Be careful what you do here because these are still being enforced.
783
« on: April 05, 2012, 05:34:46 PM »
The thing is, the SnapperII wasn't accepted because of its ability to stack. It was accepted because, though it weighed less, it required more chassis weight and didn't have enough power to do much more than hold a wedge. It has since been abused to the point of being unbalanced. The same is starting to be said for the axle mount but I doubt it will go so far. The extender starts out unbalanced and dominates all other options. With time it can only get worse. The SnapperII is not without its drawbacks which inevitably make you consider other options. I personally prefer to use a Snapper 2 but other options include the axle mount, servos, pistons, or even baseplate anchors. Each has its own advantages and disadvantages but they come out pretty even IMO. More to point, there are no plausible alternatives to the 140cm extender. If people want to continue discussing this, please do it here because we are drifting off topic.
784
« on: April 05, 2012, 02:25:48 PM »
But the SnapperII was never used for its original intention. It was supposed to be a burst motor, but people instead used it to create super-compact Snapper-loaded designs. The 140cm extender might be broken because of its weight, but you pay the price in terms of fragility. It's such a big target, and one decent hit will easily take it off.
Actually the SnapperII wasn't being used for stacking purposes until way after it was accepted. I have a feeling that if it had been, it wouldn't have been accepted. One "decent" hit is also much harder to do on a well-built bot than you are giving it credit for. Humans and AI have a hard time aiming for something that thin and it would never be used for widening except when making trappers (which would be screwed either way if you managed to get on its side). As far as random chance goes, most builders could find a use for the weight loss to reduce the chance of receiving significant damage. It also would usually take more than one hit to the extender for it to break off (since it is impossible to break something off with only one weapon hitting it (it requires at least two weapons or two hits from a single weapon)). Also to take it one step further, here is some data that I took for HWs in multiple packs: Pack: AverageDamge | MaxDamage | Damage>=100
BBEANS AI: 59.3 | 669.4 | 17.6%
ReStocked V0.1b: 45.8 | 478.5 | 9.9%
Pys AI: 52.5 | 597.3 | 13.2%
Starcore V4a: 53.6 | 680.0 | 14.7%
Comprehensive: 54.1 | 680.0 | 14.7%
AI vs AI with no bot participating more than twice.
Excludes the trinity bot (Starcore V4a would have gotten 53.9 | 1554.4 | 14.8% and the Comprehensive would have gotten 54.2 | 1554.4 | 14.7%).
Doesn't include anything less than 10.
The test was conducted by telling the game to save all damages to a text file. By having the game do it, I got 100% accurate results and stacked weapons would be grouped together whenever they hit at once.
So just to be clear, only about 3 of every 20 attacks ever exceed 100. This means that most 100HP with 15 fracture components take 3 hits to break off. Decent no, lucky yes. I agree with ACAMS that the 5kg 140cm extender is far too unbalanced. Every time that I have used it, I always ended up using two of them (which is 18kg less than if I had used 20cm round extenders instead). If I didn't have the weight to add protection, e.g. two blades (20kg), then I could easily use two 20cm extenders with razors instead (18kg). Also, doing it this way gave me the exact same HP as using purely 20cm extenders (if I had the additional 2kg than I would have had 600HP more) but with the added benefit of delivering damage with the razors. I should also note that though I'm referencing 20cm extenders as the alternative, the rule of 7 gets in the way of actually being this efficient. Changing gears, it is also important to understand how you use them. They are long enough to put snowplows on them with minimal effort making them excellent for trappers (again saving 18kg that could be used elsewhere). Since they are often longer than our chassis, they can take advantage of APs that would otherwise be too far back and, if done right, the extender is covered up by the chassis. If you used a SnapperII than each resulting AP costs 4.5kg (admittedly this is due to the SnapperII being unbalanced when using it for stacking purposes) and is for the most part protected by the chassis (so comparatively speaking, it is .5 kg less than using a baseplate anchor (chassis space and the eFFe glitch aside)). If you wanted to use it for height purposes or wedges then you could still protect the extender as previously mentioned or use less protection to keep it light (hammer). If you think about it, all bot types would benefit from using a 140 extender but that is because they are far less balanced than their alternatives. Any bot not using them would be at a disadvantage to a similar bot that is. If they were balanced, their impact on bot efficiency would be much more subtle.
785
« on: April 02, 2012, 12:18:30 PM »
Wait a minuete.
I just read the rule.
It says No More than 2 people per challenge.
That's a pointless rule. Pointless rules are made to be shattered, and as long as you provide enough reason for doing so you are in the right.
To an extent, I would have to say that I disagree. Rules that are pointless should be removed, not ignored. I also wouldn't consider this rule pointless because it keeps the naming convention and scoring system simple. The naming convention allows the global scoring to be easily (re)calculated. The unified scoring system also helps with this and has the added benefit of keeping the challenges comparable (in terms of both builder skill and the weight of the challenge on the scoreboard). Every rule has a reason and if we allow people to ignore some, but not all of the rules, it is both hypocritical and confusing. If the rule was made to be broken then it wouldn't have been listed. The reason why it was kept after the rule change was because we had already gone through numerous scorekeepers since the first set of rules. Therefore, the point of enforcing a rule like this is to make their job easier. I agree that given enough reason some exceptions may be made but just stating that you want to do it your way is not enough. "The only way to..." or "to be fair..." are examples of how you would construct a solid reason (as long as you are being factual). Exceptions are not meant to be overused though so, for the most part, people shouldn't expect the rules to be any different than they are stated. Questioning something is fine because it oftentimes leads to gradual improvement. Ignoring something only leads to the deterioration of the whole. For this reason, I would encourage people to follow the rules as they are stated but request/suggest alternatives for future rule changes. Since this is Sage's section and I have no real power in it, you don't have to listen to me but I would expect these courtesies in any of my sections. Seeing as it doesn't impact me in any way, I would consider my opinion to be pretty, but not completely, unbiased.
786
« on: March 30, 2012, 08:03:18 AM »
Nope, its completely possible. The only reason why it has never been done before is because there is no easy way of changing the starting point without swapping out the whole arena. It also hasn't been used in any arenas because realistically, the bots should never start upside-down unless you specifically designed it to perform best under those conditions. In which case, you would realistically flip your own bot because it isn't normal.
787
« on: March 29, 2012, 08:50:43 PM »
Mad made a special garage with a second level for flipping but I can't find the link anymore and my copy seems to have disappeared. IIRC, Click didn't want it for DSL so it was being converted into the SC4's garage. This could mean that there has already been work done to the DSL's garage but I can't say for certain.
The DSL "flipper" obstacle is hard to use and doesn't work the way most people think. As the wheels of your bot make contact with the ramp's surface, their associated direction is reversed. So to flip your bot, you are suppose to hold the forward key to gain both speed and momentum so that the sudden swap in direction causes your bot to flip itself. The problem is that most bots aren't fast enough or top heavy enough to make this an effective solution. I personally would have liked a normal ramp that is that tall so that you can fall off the front.
Since this has been an issue for both me and everyone else, I added an option to start upside down in my garage. I'm only about 40% done with it though and I've been working on it off and on for nearly 2 years. Though that sounds like a lot of time, I've had many other projects keeping me busy.
Gropaga is right that you can't add bots (or even an arena flipper) to the test garage but that doesn't mean that their aren't solutions. As far as the test garage goes, anyone could mess with gravity ("plus.gravity(X,Y,Z)") or even apply a force that only affects your bot ("plus.force(BotID,X,Y,Z)"). Mods could include a button that does any of those, a ramp, a ledge, or an upside down starting point.
As far as other ways of testing your bot, you could fall off the walkway in the obstacle arena, fall off the bridge in the bridge arena, step on one of the force thingies in the stunt arena, etc. You are not limited to getting flipped by your opponent. If you were, I guess, you could turn off the countdown clock ("plus.practice(Boolean)") whenever you faced an AI that didn't move but don't forget to turn things back on when your done.
788
« on: March 28, 2012, 02:44:45 PM »
i thought there was a rule of 2 contestants per match
Its actually been a rule since this board was started. Its best if it is limited to 2 because of the naming convention and the complications when dealing with strange rules (I'm looking at you Naryar). The first set of rules weren't enforced very well so we haven't been cracking down on it unless you posted a challenge after the rules were updated.
789
« on: March 27, 2012, 01:49:59 PM »
I feel that there would be much more bias (because it is obvious who posted what), less creativity (HS and popups are over made and showcased), and less challenging (part of the fun is trying to follow the criteria given). Not to mention, I wouldn't be able to participate without making a showcase :P
790
« on: March 25, 2012, 11:02:18 PM »
IMO, 720p would look better than 1080p. Without custom status bars, they get to be too small to read at 1080p. Other than that, it looks pretty good and I'm looking forward to seeing how this tourament will progress.
791
« on: March 25, 2012, 04:41:46 PM »
This threads location should be reviewed.. it is after all in the modification ideas section... just sayin'
If you desire to mod your own thread without interruption from me Click, just let me know.
I generally don't mess with moderator or administrator threads unless I see obvious misconduct (sustained off-topic-ness, language, etc.). Thats why it hasn't been moved. Personally, this thread should be devoted to actual previews of DSL3 (and moved) and another thread should be for people's suggestions. As a community mod, it is important to get feedback and ideas regarding DSL. also are wheels going to be more durable
Click just got done hinting that he was going to update one of the least acknowledged tabs. Considering that weapons, extenders, power, and mechanics tabs have all be reformulated numerous times, its a safe bet that wheels are getting the much needed attention (though its been a long time since I played DSL, I'm pretty sure the extras tab got about as much attention as the extenders). If I may I'd like to request a re-texturing on connectors so that each connection point is "marked" with a circle or something identifiable. This would be most useful on the red & black one (multi-extender), but all the circular ones have the same problem.
I don't think its as cosmetically pleasing to add indicator skins when we could just add more APs.
792
« on: March 15, 2012, 12:13:04 PM »
Except I PMed you back 20 hours ago saying... Unfortunately, I don't think I have enough time to host for at least a couple weeks. Sorry, maybe next time.
BTW, the rules state that if you are going to have a host, it needs to be determined and confirmed before the actual thread is started.
793
« on: March 13, 2012, 12:22:54 PM »
Pon2n uses a hovercraft but IIRC it is using more than one.
The unique behavior of flamethrowers, magnets, cannons, and hovercrafts are hard coded to behave the way they do. We currently don't know how to edit them without modifying the EXE. If you want something more customizable than the hovercraft but less useful look at how Firebeetle made the Rocket Jet.
794
« on: March 12, 2012, 11:42:57 AM »
TTYTT, it doesn't look like you changed that much. From what I can tell, you changed the default spin_range (which was already adjustable via the bindings) and made Ram the main tactic (which has already been done inside OmniTrueRam.py). Though by building upon Spinner.py, you have a specialized version of RobotInRange code that isn't found inside of any of the Omni variants (IIRC).
Issues:
1. Remove the first LostComponent method. I am aware that it was in Spinner.py but it is overridden anyways so its just wasting space.
2. Fix your __init__ method so that everything is indented by 8 spaces.
3. Fix your LostComponent method so that it properly switches tactics when all the weapons fall off (replace "Engage" with "Ram").
4. Due to an issue with how RA2's AI uses switches, you'll want to change the self.Input("Spin", 0, 1) to self.Input("Spin", 0, 100). This makes it compatible with Analog controls while still allowing it to work with buttons.
For anyone interested in using this, ram is not preinstalled in stock or DSL so you'll need to download Clickbeetle's upgraded tactics.py to use it (it may be found in BBEANS AI). You may also want to wait until he fixes the above issues.
795
« on: March 11, 2012, 01:16:24 PM »
A python file is basically a text file that has the .py extension (as far as you need to concern yourself). Python files are read by an interpreter which in RA2's case I believe is version 2.3. If your game crashes during startup, you usually have a syntax error. If this is the case, IDLE or some unix based terminal can tell you whats wrong with the "grammer."
Notepad++ is in no way required to program especially in python but its my favorite text editor (scite is also pretty good).
796
« on: March 09, 2012, 08:39:52 PM »
Read the procedure regarding challenges. A host should be determined before starting a challenge thread. The rules changed since your last challenge so note the changes. The staff is trying to get this section running smoothly and easier to maintain so that's why there is now an example procedure.
797
« on: March 08, 2012, 02:58:12 PM »
Well, I didn't have as much time over the weekend as I thought but I did manage to get something put together before the deadline. My original intention was just to upload a bare-bones video though as you'll soon discover, I didn't. On a side note, I would be interested in hearing what people think of the Action Cam.
Note that it goes up to 720p.
798
« on: March 01, 2012, 01:19:14 AM »
I've received Scrap's entry (like 9 hours ago) so I should have recorded and ready to upload by the end of the weekend.
799
« on: February 28, 2012, 12:03:20 AM »
I've received and successfully tested Pyromaniac605's bot and AI so now we are waiting on Scrap Daddy. Everything was in order by the 22nd so he has until the 29th to get something in.
800
« on: February 27, 2012, 11:03:01 AM »
The splash is gonna be on the front page though, so it's important that it look nice.
I completely agree. Though it is important to make a decent bot, it is also important to make it presentable. If people really think this is a problem, I can make a program that generates the splash automatically (I made one a while back but I'm certain I could do better and make it work for all resolutions). I'm a little bummed out that I couldn't participate this month but thats fine because we have a few good ones. I ended up voting for 7 but 3 was a close runner up.
Pages: 1 ... 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 ... 64
|
Show Posts