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 - Clickbeetle
Pages: 1 [2] 3 4 5 6 7 8 9 ... 174
21
« on: August 18, 2018, 06:14:36 PM »
(I wonder though, is there any risk of sabotage if you let strangers push buttons on your keyboard?)
in theory yes. not much we can do about it unless one party uses a controller and one uses keyboard (the host can mutually exclude these) but many people don't have controllers and it's a bit tricky to get them going even so.
What I meant was, could some dipwad hijack your PC and delete system32 or something nasty. Just seems like letting strangers on your computer is slightly risky. It's not a problem here, but if anyone ever tried to do larger-scale Parsec tournaments, it could be. I hear that AI tournament hosts have to check any custom Python scripts they get sent now for that reason. Anyway, ROUND 1 HYPE. I don't know what happened in the other fights, but there were some nail-biting moments in mine!
22
« on: August 18, 2018, 05:54:16 PM »
Have to admit, that might be an improvement over my Black Box 2. It couldn't do SnS mode but the wheels are more protected which is probably more important.
I would have made an octagon chassis with the wheels in the middle and the battery in the back, to improve turning and protect the wheels even more. But I'm sure it would be a bloody pain to do that Redbird stack again, not to mention the Nifty+control board (which I still struggle with).
23
« on: August 18, 2018, 05:04:15 PM »
In the DSL blade arena?
Your robot is too tall?
LOL, I actually thought about doing that again, since the tournament host (once again) didn't think to ban hiding under the blade. But then I thought that's the kind of tactic that's funny the first time, and borders on poor sportsmanship after that. Also, making a bot flat enough in stock would require using mini wheels, RADs/Redbirds, and small battery packs exclusively, which means it would be useless in any other capacity. damn, has it really been 5 years since drumblebee 9? time sure flies huh...
It has not... *goes back to check* Oh, darn, it HAS. This community is ripe with "Want to feel old?" lead-ins.
24
« on: August 16, 2018, 01:36:41 AM »
Bot for Reier's online tournament. He suggested building something hard to AI, and I thought, not much is harder than bots with servos. (I've done it, but it's not easy.) The idea here was that I could make a saw arm with the motor acting as a counterweight, so it can move up and down quickly. Which works. The trouble is, the weapon spins so fast that the recoil knocks the arm all the way back, so it has to swing around 180 degrees to hit the opponent again, rinse and repeat. So I need the spike in the rear, which hits the floor and prevents the arm from swinging back too far, so I can get more consistent damage. It also prevents the bot from doing wheelies and improves the wedges. Unfortunately, it also makes self-righting impractical. (It can be done, but the arm usually ends up facing backwards.) If it gets a good pin on you though, watch out. Those blades can KO bots without top armor in seconds. Have not played Ironforge in a long time and not familiar with the current metagame. Already have a couple of ideas for a possible Falling Star 3. More plows is on the list.
25
« on: August 16, 2018, 12:36:30 AM »
I wonder if running away and hoping your opponent pits themselves would be a valid strategy in this tournament? It worked well enough for bud; only backfired once.
General comment: I'm really impressed with Parsec-based online play. I had practically ZERO lag in this. Parsec online is not going to replace AI tournaments (fight scheduling is still a challenge, especially on a large scale), but this removes the biggest obstacle. Hope to see more like this in the future. (I wonder though, is there any risk of sabotage if you let strangers push buttons on your keyboard?)
26
« on: August 16, 2018, 12:08:20 AM »
shame it's gonna be stomped by the generic HS everyone is gonna enter.
In the DSL blade arena? HS can be tossed OOTA in that arena, so I'm expecting more of a popup-heavy field. Would have entered Cold Front 3 if I expected a lot of HS. ...but since when have my tournament predictions ever been accurate? Dang, now it's definitely going to be filled with HS.  didn't realize other drumblebees only had 2WD - would think 4WD is mandatory. Does it perform better?
Also, why red extenders??
Definitely performs better... not a TON better, but noticeable. Much harder to flank, and the wedges are better. Easier to get bots up the wedge and in contact with the drum. Red extenders are because I thought a red drum would look cool and different. Used tastefully (=not on tribars), I think red can look good.
27
« on: August 15, 2018, 10:51:33 PM »
Saw there was exactly one stock tournament open for signups among a flood of DSL IRL. (That's Lightning S.'s Revolution 3 in case you couldn't tell.) Had to build something for it. Here I have a perfect opportunity to arbitrarily switch to Roman numerals and call this "Drumblebee X"... but I'm taking the high road and sticking with 10. I did decide to change the traditional paint scheme though. The diagonal warning stripes were getting old. Bonus points if you know what bumble bee species the skin is based on. Drumblebee 9 is OK, but it's kind of lacking in speed, and I always wanted to get 4WD on it. But how to make weight? I've tried 4WD Drumblebees in the past and they always turn out too fragile. The answer turns out to be LESS RAZORZ. Drumblebee does not really need, and has never needed, 36 razors. The weapon motors always break before all the razors do. 24 is enough. With some caster optimization too, I fit in those extra 2 wheels AND an extra battery, which keeps it running at full power for the whole match and also helps with self-righting. The chassis is slightly more vulnerable with less caster coverage, but the extra speed makes it harder to hit, so it sort of balances out. The drum is wider for better motor protection. At first I had the traditional 2 extenders in a row with more razors per extender, but that was too fragile. Even though this has less total razors than Drumblebee 9, it's overall more resilient. Wedges are actually pretty good for once, so long as the weapon isn't spinning. Can out-wedge NWB and Hot Wheels head-on. (But not Backyard Gut-Ripper, the wedgiest bot I ever built.)
28
« on: August 15, 2018, 01:50:02 AM »
Clever idea angling the weapons like that. Even offline and AI'd, it would be good in rumbles.
On Xiphos, why not put the thwacking wedges on a static extender and keep them permanently extended? I don't see why you would need to retract them. Could have weight for better drive then.
MIR is... interesting. Kind of reminds me of King Googly even though they are nothing alike.
29
« on: August 13, 2018, 06:59:12 PM »
Fun fact: Clickbeetle has experience in online tournaments.
If you can call "getting to the semifinals with a noob bot because half the competitors dropped out due to scheduling issues, and then getting completely destroyed by a generic HS in my one and only fight, but not seeing what even really happened because lag had me jumping all over the arena and even winning at one point" experience, then I suppose I do. Really hoping Parsec improves the online multiplayer experience. I'm willing to give it a try. Winning isn't the goal here so much as testing the waters for Parsec online play, and trying to build something "cool and hard to AI"... but I'm still pleased to see no one has a big VS (the one bot type my entry has no chance against). Thrackerzod and freeziez do worry me though.
30
« on: August 10, 2018, 07:39:32 PM »
The download should come with a .bot file in "Robot Designs" and some stuff in a folder called "Components". Also some .wav files in "Sounds" which are technically optional (but cool).
You need to cut/paste ALL of these folders into your Robot Arena 2 folder and say Yes when the computer asks if you want to overwrite stuff.
If you installed it correctly and it still doesn't work, try installing it in an unmodded stock game.
31
« on: August 08, 2018, 12:53:55 AM »
Unrelated note: Kind of sad that neither Chomp nor Blacksmith were in the Desperado. I was really hoping to see at least one hammer in the top 16. But Chomp is not going to happen at 0-3 and Blacksmith is really long odds at 1-2.
especially weird considering it was a voluntary thing, you'd think the more experienced builders would want a shot at it
From what I heard, more than 8 teams entered, and production picked which ones got in. So it's possible that Chomp/Blacksmith entered but weren't picked.
32
« on: August 07, 2018, 01:25:24 AM »
From the vision it looked like a tear, so im going to say crap materials strike again
This episode was really fun though in that it let some of the lesser knowns really showcase themselves
Cool to see Kraken get a chance to put up a good fight
If im not mistaken, some robots like Gigabyte are still in with a shot with 3 matches remaining? (Would probably need to go 3-1 however) or do you forfeit your other qualifying matches for the flash tournament?
they said the entries for the mini tournament would still be able to participate in the main bracket afterwards. pretty sure the main reason for people not to participate in the desperado besides small bracket size is the quick turnaround between repair jobs irl. would be really tough.
i'm not sure if the W/L received during the desperado carry over to the main bracket however
Based on what I've seen on Reddit, the Desperado counts as one of your four guaranteed matches (so Gigabyte only gets up to TWO more fight card matches, not three, Lucky only gets one and LockJaw is done even if it didn't win), but wins and losses carry over as normal.
So best-case scenario for Gigabyte is that it goes 3-2. If both of its last fights are spectacular wins, it still has a chance.
Best case for Lucky is 3-3. Kind of iffy there.
Unrelated note: Kind of sad that neither Chomp nor Blacksmith were in the Desperado. I was really hoping to see at least one hammer in the top 16. But Chomp is not going to happen at 0-3 and Blacksmith is really long odds at 1-2.
33
« on: August 07, 2018, 01:05:32 AM »
Also, you mention my AI pack, and I don't mean to get far off-topic, but how often do you get crashes with that pack, cause I hear more crashes are coming from that. Still, good to know lots of people are still getting a fair amount of use out of that pack, even I need to really update that thing to modern standards.
I was getting random mid-battle crashes a lot, but I wasn't sure what the reason was. I *think* it's something to do with Supercell 2, but I'm not sure what. It (surprisingly) doesn't break the 7-component chain "limit". Other bots I've tried in RedAce AI have not crashed as often. Supercell 2 also crashes sometimes in standard DSL AI, though *perhaps* not as often. It's hard to test for this, not knowing what actually causes the crashes.
34
« on: August 07, 2018, 12:40:54 AM »
Bit of a late reply, but in addition to MotorID (case sensitive - make sure it is capitalized like that in Bindings), you also need the correct StartAngle (also case sensitive). There's no easy way to figure this out, but generally it will be 0, +/- math.pi/2, or +/- math.pi. Guess until you get it right.
There is no script at the moment that can handle two servos. You could perhaps wire the secondary servo to the same control and hope for the best.
35
« on: August 06, 2018, 11:48:33 PM »
I don't think it's as simple as a 1.5x buff, 2x buff, etc.
The way the burst motors are balanced, it integrates speed and power (both for firing and retracting), battery drain, and HP.
JX bursts are balanced for stronger retraction, less battery drain, and higher HP, which means less firing power. Which is probably the reason they are considered less good. I thought balancing the motors this way would make them useful for hammers, while the VDMA's are better for flippers, but it seems the VDMA's may just be better for everything.
I would suggest tweaking the balance more toward firing speed/power and adjusting battery drain or HP accordingly, rather than a flat buff, which could end up making them overpowered.
You know what, I really need to send you my balancing spreadsheet.
36
« on: August 06, 2018, 11:32:55 PM »
They're using a type of motion called Meltybrain, where you can use circular movements to move in a linear direction. As far as I know, there's no way to do this manually in RA2
In theory, you can do it manually. Try spinning in place and tapping the "Forward" control intermittently. With consistent timing, you can get something close to translational movement. In practice however, the timing is almost impossible to get down correctly. The AI is far better at that than a human could ever be, and this is more true the faster the bot spins.
37
« on: August 06, 2018, 11:27:01 PM »
I recall someone doing this for 2.1. Or maybe it was stock. Can't remember the details.
It would be messy for DSL 2.2 because most IRL builders use unofficial (at the moment) mods, while standard builders stick to stock DSL.
However, 2.3 will make some of those mods official so in theory, you could have every AI pack at once.
The bigger problem is that when individual AI packs are updated, it would be a huge pain to update them in the everything-pack.
38
« on: August 06, 2018, 11:18:56 PM »
You can join me in the Rusty Club.
I tried building a DSL-S bot the other day, and yep... scratched myself on a razor, got tetanus.
There seems to be a shortage of AI packs for 2.2, so having a new NAR AI would be a good thing.
39
« on: August 06, 2018, 01:32:53 AM »
I did a double take when I saw that on the front page. Super cool. I don't think I've seen a chassis hammer done that way before.
40
« on: August 06, 2018, 12:38:12 AM »
This is what I meant when I said I could stack extenders to make thicker weapon supports for Ephialtes. Ended up making them 5 thick to cover most of the width of the Perms. I can has IRL now? If not, how about this thing? I was never totally satisfied with original Supercell, even back when I first built it. (These days, it would never even pass IRL inspection.) It had too much power and not enough stability, and spent more time trying to self-right than not. Supercell 2 uses a 6-Mag instead of a dual Perm and a smaller-diameter tribar, both of which contribute to drastically improved stability. The big lightning bolt on top is a fancy srimech that functions similarly to Megabyte/Gigabyte's pole. It does OK against the RedAce AI, except for hammers. Any hammer bot is pretty much an insta-loss, with the chassis being such a big fat target up there. It has 3mm Ti chassis armor, and the lower frame is all aluminum components. I liked the lightning-bolt extenders so much that I went and updated Thunderclap with a lightning-bolt hammer, because I thought it would look cool. (Also, original Thunderclap would no longer pass IRL inspection. There seems to be a theme here... I wonder if ANY of my old bots would still be considered IRL today?) The rest of the bot is the same as the original, except I needed to drop to 3mm steel armor to make weight for a bulkier turret, and I changed the batteries to 2 Battlepacks. Edit: Still not used to the new image-uploading system; not sure what's going on with the thumbnails at the bottom.
Pages: 1 [2] 3 4 5 6 7 8 9 ... 174
|
Show Posts