Get the correct roborumble.txt file here: RoboRumble/roborumble txt
Get the correct meleerumble.txt file here: RoboRumble/meleerumble txt
Get the correct teamrumble.txt file here: RoboRumble/teamrumble txt
The temporary server has a public interface on the pezius.com domain. So you must all edit your roborumble.txt files to participate in this recovery effort.
Note that not only the URL for the servlets has changed. Also the URL to reach the particpants list must be edited:
PARTICIPANTSURL=http://robowiki.net/cgi-bin/robowiki?RoboRumble/Participants(That's becuase I have no mod_rewrite privileges with the new hosting solution.)
The rankings look strange so don't look too close at them. What's important is that I accumulate battles log files on my machine and I can then mangle those slightly and turn them into uploadable battles that my client can later upload to Davids server.
What are you waiting for? Start your client now! =)
Regards,
-- PEZ
I just started it... -- Jonathan
Me too. We will need lots of clients running it to get a stable ranking... -- Axe
I can't upload the results all the time...
java.net.MalformedURLException: no protocol:
I gave the right URLs without typos. -- Jonathan
'LETS GET READY TO RRRRRRRRRRRUUUUUUUUUUUUMMMMMMMMMBBBBBBBLLLLLLLLLEEEEEEEE' -- jim
Got that error sorted out Jonathan? Are you having the problem all the time or just sometimes? -- PEZ
$ cut -d, -f3 battles_roborumble.txt | sort | uniq -c | sort -r 2177 Pulsar 1712 PEZ 1134 Axe 264 JIMDon't wait. Start your clients now please.
-- PEZ
Gah!!!! Sorry about this folks. But suddenly my temporary RR@H server is unreachable. And some other services are unreachable for me too so I can't fix it remotely. Watch this space for a message for when the bloody thing is working again. Good we have at least moved the wiki to a more stable environment! -- PEZ
Keep the clients running anyway and they will upload the results when the server is available. -- Pulsar
The temp RR@H server is alive and kicking again. -- PEZ
And. I'm watching you:
$ cut -d, -f3 battles_roborumble.txt | sort | uniq -c | sort -r 2538 Pulsar 1820 PEZ 1206 Axe 330 MUE 264 JIM 148 VIC=) -- PEZ
Don't forget to change the USER= line of the file above. --David Alves
Looking better!
$ cut -d, -f3 battles_roborumble.txt | sort | uniq -c | sort -r 2748 Pulsar 1946 PEZ 1534 Axe 390 MUE 264 JIM 223 VIC 194 David Alves 75 VIC2 60 STRIDER 16 DavidAlvesMore clients anyone? Vic, you can use the same upload name for both your machines if you like. But maybe you want to see which machine is the most efficient? =)
-- PEZ
No I just wasn't sure if I could give them the same name. Until now that name never showed up, so I never bothered checking.
Another thing: anyone who has started the client before the server went up again should restart it. My 'running' RR client could't find the Ratings file. It said: "All Ratings File not found!!!". and then it started the new iteration with "Using smart battles is false". It looked like it was running a lot of battles of old bot against old bot, and not bots that have been updated/added this last year. After restarting the client everything was fine again. I don't know if that is a freak occurence on my computer, but I suggest evryone restarts their client to be on the safe side.
--Vic
I've seen similar beaviour before. Good that you remind us! -- PEZ
My client seems to only run battles agains dft.Calliope, everything else looks normal. Is this expected behaviour? -- Strider
Dunno! My client also zooms in on selected bots some iterations. I have set ITERATE=NO in my roborumble.txt settings so the roborumble client is restarted each time. -- PEZ
Ok. I restarted the client and it ran random battles the first iteration then only battles agains tm.Yuugao. I'll set ITERATE=NO too. -- Strider
It's like that for me too. I have no clue why. ... Maybe it somehow populates the priority battles file during the first iteration and then acts on it the following? -- PEZ
$ cut -d, -f3 battles_*.txt | sort | uniq -c | sort -r 5073 Pulsar 3191 PEZ 2302 Axe 689 MUE 659 VIC 598 David Alves 495 ABC 282 JIM 262 STRIDER 261 VIC2 23 DavidAlves-- PEZ
That DavidAlves machine ... is that a ZX Spectrum? ;-) --Vic
Maybe it's a one of those [iButton Java rings]? -- PEZ
$ cut -d, -f3 battles_*.txt | sort | uniq -c | sort -r 5642 Pulsar 3516 PEZ 2423 Axe 971 VIC 819 David Alves 779 MUE 536 ABC 469 VIC2 391 STRIDER 282 JIM 23 DavidAlves-- PEZ
Join the competition: http://pezius.com/perl/ulcomp -- PEZ
What about meleerumble ??? =) Tried it and didn't worked... I think...
-- M3thos
All in due time. =) The temporary server will concentrate on getting the 1v1 rankings stable. -- PEZ
I'm having problems, basically, java is not using much cpu (15%..) in the fights.. i'm using java1.5.0 .. might that be the problem? TIA
greets
-- M3thos
No clue. But never mind it. We're happy your client is running! -- PEZ
I managed to wake up the LRP graphs too. So now if you click on one of those links in the Rankings table things should work! -- PEZ
Weird... My laptop is a 2.8GHz pentium 4, desktop is an AthlonXP? 3500+ (socket 939)... I knew my desktop was faster at running battles, but that much faster? It's 1440 uploaded vs. 113 uploaded, and I started them at the same time. --David Alves
I fixed Jim's request about presenting a default flag for bots lacking one. Now the Ranking table on the temporary server presents a pirate flag () as a default flag. How did I do it? mod_rewrite of course:
RewriteCond %{REQUEST_URI} ^/rumble/pkg-flag RewriteCond %{REQUEST_FILENAME} !-f RewriteRule .* rumble/pkg-flag/pirate.gif-- PEZ
How can the LRP applet be allowed to download the RatingDetails from a different path tree than it is itself loaded from (a tiny CGI in /perl)? mod_rewrite of course:
RewriteCond %{REQUEST_URI} ^/lrp RewriteRule ^lrp(.*) http://localhost/perl/lrp$1 [P]-- PEZ
I'm having trouble running RoboRumble. I made the changes that PEZ stated earlier on the page, and it runs and downloads all the bots fine. However, it gets to Fighting Battle 0 and then does nothing. No errors or anything, just, nothingness. -- Alcatraz
Run the robocode.bat file from yoru rumble directory to start robocode. When it finally shows up, close it out and restart RR@H. I think the robot databse gets corrupted or something when downloading more than one bot. Starting from the command line seems to clear this. -- jim
It seems to be selecting focussed competitors at random. I ran 25 battles of pi.Dark earlier, and 25 of Cigaret... both of which have plenty of battles. Any idea why and how to fix it? We'll bring all bots up to 500 battles a lot sooner if it starts choosing competitors properly. --David Alves
But did it run against competitors which it had met before? If not it is alright. -- PEZ
25 battles? My PowerBook has been (and is still) running ONLY hundreds of DogManSPE? (extreme SlowBot) battles! -- Jonathan
... and the PC stops now and then. >:-( -- Jonathan
In the last hour or so, my desktop has run 3 battles. I suspect something is wrong. I tried your suggestion, PEZ, and have gotten these three because of it. Also, my priority battle file is missing. -- Alcatraz
Yep. i noticed in my client at office, the message notifying that didnt found prioritary battles file. -- Axe
Here too, both on my PowerBook and the PC. Only the PC stops though, so my PowerBook can keep running DogManSPE?......... -- Jonathan
Still ONLY DogManSPE?! What is that prioritary file for btw? -- Jonathan
I'm getting "Parameters File not found !!!" error... I'd love to help if I could get this thing running... -- Goofy
Goofy: Did you follow the instructions?
Look at Pulsar's count!
12345 Pulsar ... 1205 JonathanMy PowerBook sped up because it chose another bot to test 100s of rounds. The PC didn't get anywhere, so I stopped it completely. -- Jonathan
PEZ: All of the functionality of the current RR@H server code is now done in my version. I ended up rewriting everything from scratch, designing for extensibility and maintainability. I suggest that we let the new bots get up to 500 battles each (probably one more day) then start working out how we're going to transfer the data. If you'd like to play with uploading data to my app I can email you instructions on how to do that. --David Alves
That sounds promising! Why don't we point our clients your way right away? I'm constantly on my toes here, worried sick something should break. Win XP Home isn't exactly the dream server environment. Can I try with my client? -- PEZ
Test some first please :) -- Pulsar
Oh, ye of little faith! But I could make some results files by tapping the battles-log on my server on just a few of the 30,000+ battles it has collected oin just a few days. (YAAAAAAAYYYY!!!) Then my client can try upload those. And if that works we can start some 1,000 clients all over the world. =) -- PEZ
It would be great if you could write a program to upload the battle data you've collected to my app. I'll email you details about how to upload right now. --David Alves
Well, I have written a program that eats the battles log and spits out results files for the RR@H client to upload. Won't that work? Do you ever read your gmail btw? I've been one-way-communicating with you there for some days now. =) -- PEZ
Err, sorry about that. I've been very busy writing a program. ;-) I've already spotted a few oddities. The last run dates are off and the battles from the database are not shown in the details view, only the ones that have been uploaded this session. These are all easy things to fix and I'll do them tomorrow. It would be great if you could try and fill the app with battles... I'm curious what the capacity is. I need to verify that performance scales well to really large numbers of battles and need to be sure we aren't going to run out of memory. I'm estimating that there will be issues at around the 1 million battle mark. Try it out and let me know. Don't upload them *too* fast though, don't want to melt the poor thing. :-)
Well, here's what I've done: Downloaded codesize.jar, installed a second instance of robocode, (renamed the folder roborumble), placed codesize.jar in /roborumble, downloaded beta 9.12 (bottom on the list at http://www.geocities.com/albert_pv/RoboRumbeAtHome.html ), unpacked it, tried running roborumble.bat, made a roborumble.txt file and copy/pasted the above text. Still getting that error. -- Goofy
Goofy, try to paste here the exact output message... "Parameters File not found !!!" is a little generic. Is it from vm or RR? Can u run regular Robocode? Have u checked the .bat file paths? -- Axe
"Parameters File not found!!!" is the message that the RR@H client gives when it can't find the settings file you specified. For example, my RR@H command line looks like this:
java -Dsun.io.useCanonCaches=false -Xmx256M -cp .;../robocode.jar;../codesize.jar; roborumble.RoboRumbleAtHome? ./roborumble/roborumble.txt
If if can't find the .txt file specified as a command line parameter, you'll get that message. My guess is you extracted to the wrong directory or moved files after extracting. Double check that roborumble.bat is in the main robocode directory and that roborumble.txt is at /robots/roborumble/roborumble.txt --David Alves
Results seems not the same as before the BigCrash?... For example:
abc.tron3.Shadow 3.25 2044.73 492
Seems that even with almost 500 battles we have differences of 20 points or so... The same with SS, its -20 pts from old score (same version)... Maybe 500 bats isnt enought to get it stable? Thoughts? -- Axe
May be 500 it's not enough, but note that the rating of a bot deppends not only on its battles, but on the ratings of all the bots it fought, so as long they don't have an stable rating, a bot will not have it neither. -- Albert
It may also have to do with the fact that there is no "standard" bot... the ratings are free to "drift" up and down. Since expected score is based off of rating difference, all the bots could be 50 points higher or lower and the predicted scores would still be just as valid. That's why a few days ago I proposed that we set an "anchor" to prevent rating drift. For example, if we set tobe.Fusion to be 1600, whenever Fusion's rating drops below 1600, all bots have their ratings raised by that much. Whenever it rises above 1600, all bots have their ratings lowered by that much. If you read Paul Evans' original document describing the rating system, he proposes doing this. Look under the heading "Standardising the Ratings" on http://www.aqre94.dsl.pipex.com/robocode/rating.shtml --David Alves
I always thought that the ranking was something like the averaged sum of all ELO ratings... Can someone explain to me how the rating is calculated? -- Axe
Just found the page RoboRumble/RankingSystem. Ill try to understand it... -- Axe
I would think that an anchor bot would be good, but only after the ratings are stabelized. This may be hard to do before then. How would you know what bot to choose? Select the wrong one an everything could go kablooey. Lets wait and see once everyone gets to 500. Maybe we need to go a bit beyond 500. We all know that pre-loaded data pushes a bot's performance up some. The whole rumble just started with pre-loaded data on almost every other bot in the rumble. I personally think the only way to correct this would be to get everyone to wipe their data directories and start again. But this is not fair as some bots were built with pre-loaded data. -- jim
Wow, just finished reading RoboRumble/RankingSystem... I didnt fully understand the reasons of the ranking based in expected avgs instead of average % only, but I believe that Albert & co. had good reasons for that. Well, lets wait more battles to see if it gets stable... Btw: that 3 * (expected % - % of score)... Where (if) does ELO calculation enter in that formula? Is that "expected %"? -- Axe
Yep. Like so:
public static double botOneExpectedShare?(double botOneRating?, double botTwoRating?){ return 1.0 / (1+Math.pow(20,(botTwoRating? - botOneRating?)/800)); }
@Jim: We can tell which one to pick by looking at the archived rankings from before the big crash. I suggest either nic.Nicator or tobe.Fusion, both are very old (so they have a lot of battles), have relatively low specialization indices, and are around 1600. As a bonus, choosing one of them would also be an honor for someone who contributed to the robocode community, tobe ran the MiniBotChallenge and Alisdair Owens (author of Nicator) wrote the SnippetBot tutorial and tutorials on AntiGravityMovement and CircularTargeting. Reading those tutorials is how a lot of ancient pre-wiki robocoders (like me) got started, and the MiniBotChallenge is what kept me interested. :-) --David Alves
I think we should wait until everybody gets 800+ battles, it looks to me like the rankings will stabilise very near to what they were before. The fact that many bots started with pre-saved data should probably change things a bit, PM bots will go down and data saving GF bots will go up. That would have been solved if everyone had wiped their robotcache before reseting the rankings, but after 250k battles it is a bit too late. Shadow is #1 in the PL now, so I can live with it... ;) -- ABC
Yes, we should have thought about clearing the .robotcache before this recovery effort... About rankings and anchor bots and such. I think many people who started worrying about this when we were at 300-400 rounds should at least now have learnt what it means whenever ABC and I play our tape with "wait for the bots to stop getting priority battles before drawing conclusions" recorded on it. In any case an anchor bot wouldn't have helped if those concerns would have had groundings. It would only help (maybe) with that drift David is talking about. A drift which I can live with at least. I only really care for relative performance anyway. Or maybe without the drift we would get better feedback with if a particular update of a bot is better or not? -- PEZ
The main thing drift correction is good for is comparing rankings from different times. Drift correction ensures that 1800 is always 1800. Without it, you can't be sure how well an old version of SandboxDT (for example) would rate today. --David Alves
But you could maybe guess how it would rank. =) Anyway, when we've got the rankings recovered and your RR@H server up and ready to serve we could add that anchor bot. Seems like a good idea. And then we can throw in both DT 2.11 and DT 1.61 a while to get a reference for our collective improvement efforts. Man; Did I, for a long while, feel DT 1.61 was almost unbeatable! -- PEZ
It seems that SS 2.42 dropped about 10 pts in this new ranking... Shadow also. Maybe iŽll try later a 2.42.reloaded, just to see if goes back to ~ 2042, but now ill try my new version. Btw: ABC, why dont u try a re-release of Shadow 3.25 just to see if the enemies data make diff? -- Axe
Any bot carrying data on Shadow is making a huge mistake. -- PEZ
Maybe. But a re-release could be a good test... -- Axe
Will there also be anchor-bots for mini, micro and nano ? -- GrubbmGait
Can some one look at the compare servelet/jsp and see why it is not working? Now that we are all releasing bots again it would be nice to be able to compare our new performance to the old. Isn't there a mod_rewrite that could fix this? =^> -- jim
OK. I'm on it now Jim. Though I've been staring at the code for two hours and can't see why it should fail. While digesting it I've put the PremierLeague in operation again. Look at Shadow's score! 294 pairings and zero, zilch, nada, no, not a single loss. So cool!
Anyway, you might experience some error messages from the servlets while I stop and start Tomcat here to test if I can get the Compare servlet working again.
-- PEZ
Found it! Some bugs are more subtle than others to track down. In this case an inner class was used where a regular class works just as well (at least from what I can tell). The strange thing was that the servlet exited and returned without complains when this inner class was instantiated. But when I moved it out of the enclosing class the servlet started to work again. Even stranger that this worked before but not on this install. Might be a bug in the servlet.jar library which I think is a bit dated on this machine. (I'm using a Tomcat distro which came with an old NetBeans I tested quite a while ago.) Anyway. Now things work. Hurray! -- PEZ
I have another theory for that diff. in Shadow & SS scoring after the BigCrash?...
It must be the third time that i mention this here: The JRE 1.4.2 SecurityException Bug...
Clients running jre 1.4.2 or higher, without the add of the option "-Dsun.io.useCanonCaches=false" in the .bat file, can leave disabled any bot using save/load of data... And the RR@Home clients for downloading arent coming with that issue. this is how their roborumble.bat should look like:
cd robots :run java -Dsun.io.useCanonCaches=false -Xmx256M -cp .;../robocode.jar;../codesize.jar; roborumble.RoboRumbleAtHome ./roborumble/roborumble.txt goto runAfter the BigCrash?, lots of clients (10 or more, and some new) where running the bots battles, and bots that access files (like SS & Shadow) can be disabled against enemies like HawkOnFire & other HOT bots, what can have a very bad effect in the ranking...
Poetic justice? =) -- PEZ
More to INjustice, dont u think? -- Axe
There's no such thing as poetic injustice. =) -- PEZ
If the rules now are "If u access data u can get disabled...", we must put this clear. Otherelse, it is simply unfair. (unfair to be polite:). Rules are rules. Dura lex, sed lex. -- Axe
I agree. I agree. Just teasing ya! -- PEZ
It will take a day or two more to sort the ISP problems out. Meanwhile you can let your RR@H clients rest. I'll let you know when my server has contact with the net again. -- PEZ
OK. My home is attached to the net again. I've got a new ip addresss for pezius.com now. Your clients will start functioning once the DNS changes I just made propagates to your part of the net. It will happen in the time frame from now (Nov 16 2004 - 20:07 CET) and two days forward. -- PEZ
The power supply to my home is very unstable at the moment. The RR@H server lives on a machine that is contantly crashing... The rankings are broken now, but they will repair themselves once all bots have run a new pairing I think. -- PEZ
PEZ, is the server unavailable again, i can't open the rankings page now, i don't know if it was my networks' problem. -- iiley
Yeah. This page should be renamed. =) -- PEZ
And now it seems Internet found its way to my home again. -- PEZ
#leave it blank or put GENERAL if you want to execute battles for all bots, or use MINI, MICRO or NANO if #you want to run only battles for this type of bot.-- GrubbmGait
RUNONLY=MINI
. -- PEZ
That was too obvious :-p --GrubbmGait
Here you have mine:
#RoboRumble@home properties #properties for executing the modules. Use YES or NOT DOWNLOAD=YES EXECUTE=YES UPLOAD=YES ITERATE=NO #It tells the program if this competitions is for teams. It is necessary since teams.jars are different from bot.jars MELEE=YES TEAMS=NOT #leave it blank or put GENERAL if you want to execute battles for all bots, or use MINI, MICRO or NANO if #you want to run only battles for this type of bot. #Use SERVER if you want an smart client that uses server info to control process of running battles (recommended) #Melee only suports a simplified SERVER mode RUNONLY=SERVER #properties for the battle engine USER=GrubbmGait INPUT=./roborumble/temp/battlesmelee.txt PRIORITYBATTLESFILE=./roborumble/temp/prioritymelee.txt ROUNDS=35 FIELDL=1000 FIELDH=1000 MELEEBOTS=10 OUTPUT=./roborumble/files/resultsmelee.txt #properties to retrieve bots from internet BOTSREP=./ TEMP=./roborumble/temp/ #properties to retrieve participants list from internet PARTICIPANTSURL=http://robowiki.net/cgi-bin/robowiki?RoboRumble/MeleeParticipants PARTICIPANTSFILE=./roborumble/files/participmelee.txt STARTAG=pre #properties to prepare the battles list NUMBATTLES=10 #properties to upload the results RESULTSURL=http://rumble.fervir.com/rumble/UploadedResults #related competitions (competitions where participants are a subset of the participants in the main competition) #Comment the following lines if you don't have codesize.jar installed or you are experimenting problems with it. CODESIZEFILE=./roborumble/files/codesizemelee.txt MINIBOTS=minimeleerumble MICROBOTS=micromeleerumble NANOBOTS=nanomeleerumble #URLs to download the rating files (and file names) #BATTLESNUMFILE is the file used to store the number of battles fought by the bots (returned by the server when results are uploaded) RATINGS.URL=http://rumble.fervir.com/rumble/RatingsFile RATINGS.GENERAL=./roborumble/temp/ratings_m_roborumble.txt RATINGS.MINIBOTS=./roborumble/temp/ratings_m_minirumble.txt RATINGS.MICROBOTS=./roborumble/temp/ratings_m_microrumble.txt RATINGS.NANOBOTS=./roborumble/temp/ratings_m_nanorumble.txt BATTLESNUMFILE=./roborumble/temp/meleebattlesnumber.txt #URL to tell the server to delete old bots UPDATEBOTSURL=http://rumble.fervir.com/rumble/RemoveOldParticipant #BATTLESPERBOT = 999
Please do set the user to your own name. This file works fine, but I suspect there is a minor bug in it. For each meleebattle each bot gets 18 battles fought instead of 9. Maybe an other meleebattlerunner could check it to be sure. Welcome (back) at the frightening world of melee! -- GrubbmGait
Cheers GrubbmGait now i can get cracking with development, I sorted the problem with versions, that'll teach me to try and uplaod when i've been drinking heavily... :) Brainfade
@Brainfade: Did you alter the 'User=xxx' line? Currently someone is running melee-battles using my name (actually I don't mind, just getting higher on the RR Uploaders Competition) and my client in not running at the moment. -- GrubbmGait
Ummmmmmmmmmmmmmmmm, yeah i forgot to change that bit. Oh well, not to worry - consider the upload boost a thank you... --Brainfade
It looks like someone is trying to run the Roborumble client for the first time. Please make sure that you read this page and make the necessary changes to your roborumble.txt file. If you run it with the default setup it will remove all participants because it can't get the participants file (wrong url). -- ABC
RoboRumble is now hosted by me (Pulsar), no changes to the client are needed. More info later. --Pulsar
Yes, changes to the client are needed in order to upload results. Just noticed it. All references to "pezius.com" in the settings files should be changed to "rumble.pulsaronline.com:8080". I've edited the example files. -- PEZ
Is there any new way to check the number of uploads per user? I could use something like that while I've got this remote box churning away on Kawigi's fix for SMP boxes, easier to check such a page every few hours than to check on the box. -- Kuuran
Regarding the RoboRumble@Home configuration: All references to "pezius.com" or "rumble.pulsaronline.com:8080" should be changed to: "rumble.fervir.com" ALSO remove the following "servlet/" part. For example: "RATINGS.URL=http://rumble.pulsaronline.com:8080/rumble/servlet/RatingsFile" should be changed into "RATINGS.URL=http://rumble.fervir.com/rumble/RatingsFile" -- Pulsar
Edited the meleerumble.txt file example that is located two pages up. --GrubbmGait
Downloading ...bvh.micro.Freya 0.3 Unable to download bvh.micro.Freya 0.3 from site. Could not download bot bvh.micro.Freya_0.3.jar Downloading ...bvh.micro.Svadilfari 0.2 Unable to download bvh.micro.Svadilfari 0.2 from site. Could not download bot bvh.micro.Svadilfari_0.2.jar Downloading ...bvh.mini.Freya 0.541 Unable to download bvh.mini.Freya 0.541 from site. Could not download bot bvh.mini.Freya_0.541.jar Downloading ...bvh.mini.Wodan 0.50 Unable to download bvh.mini.Wodan 0.50 from site. Could not download bot bvh.mini.Wodan_0.50.jar Downloading ...bvh.tyr.Tyr 1.74 Unable to download bvh.tyr.Tyr 1.74 from site. Could not download bot bvh.tyr.Tyr_1.74.jar Downloading ...cjk.Merkava 0.1.1 Unable to download cjk.Merkava 0.1.1 from site. Could not download bot cjk.Merkava_0.1.1.jar Downloading ...cjk.Merkava 0.2.0 Unable to download cjk.Merkava 0.2.0 from site. Could not download bot cjk.Merkava_0.2.0.jar Downloading ...cjm.Che 1.2 Unable to download cjm.Che 1.2 from site. Could not download bot cjm.Che_1.2.jarThere a way to fix this? Seemed like it downloaded the first 10 fine then stopped. --David Alves
Edit: never mind, the front page of robocoderepository.com shows
java.lang.OutOfMemoryError: Java heap spaceUhhhhhh... wasn't me O= )
So, I seem to have totally killed the repository by downloading those couple bots (my client only made it through the letter A and a few Bs before it died)... can someone zip up the rest and send them to me? Phoenix has been at 10 battles for hours and I'd really like to see how it does --David Alves
I compressed my whole robot directory (36 MB), you can get it at http://www.sanji2k.goracer.de/roboking/vids/robots.rar --Krabb
Geez, couldn't someone just create a new repository? I see the ccurrent one's in agony since... Well, since I started robocoding.
I think even I could start something up if it had enough support from the community. --lRem
Have a look at the RobocodeRepository page --Pulsar
Noooo!!! I don't want to see RoboCode fading away, just as I find out about it! -- A worried new user
What makes you think it's fading away? As you can deduce from the Changes page, there are plenty of us still hooked on this addictive little hobby. =) -- Voidious
Hooked! Who's hooked? I can quit at any time. Really! --Chase-san
HAHAHAHAHA! Ha! Ha ha ha! Hehehehehehe. ha! hooo. mmm. mmmmmm. he. hmmm. -- Simonton
hehe. Blimey it's still here. I had a fiddle with robocode a few years ago, and I'm giving it another go. Keen to get a RR@Home server going, but I'm hitting this error;
C:\robocode\robots>java -Dsun.io.useCanonCaches=false -Xmx256M -cp .;../robocode.jar;../codesize.jar; roborumble.RoboRumbleAtHome ./roborumble/roborumble.txt Iteration number 0 Downloading participants list ... Downloading missing bots ... Downloading ...ag.Gir 0.99 Unable to download ag.Gir 0.99 from site. Could not download bot ag.Gir_0.99.jar Downloading ...AIR.iRobot 1.0 Unable to download AIR.iRobot 1.0 from site. Could not download bot AIR.iRobot_1.0.jar Downloading ...altglass.Cruiser 2007-01-29 Exception in thread "main" java.lang.NoClassDefFoundError: robocode/util/Utils at roborumble.netengine.BotsDownload.downloadBot(BotsDownload.java:217) at roborumble.netengine.BotsDownload.downloadMissingBots(BotsDownload.java:157) at roborumble.RoboRumbleAtHome.main(RoboRumbleAtHome.java:49)I guess I may need to change my classpath, or I have the wrong version of robocode? - I would try a later version, but I'm 1.3.1 which is recommended for RR? cheers : YellowAlienBaby
Welcome back ;) As for version, I use Robocode 1.1.3 for RoboRumble. I can't remember if it's still on SourceForge?, but I do remember I have it on my web space: [robocode-setup-1.1.3.jar]. The newer versions have tons of great development options, but Fnl is still working on integrating the RoboRumble client into the main Robocode app. Also, it seems to me you should have backslashes in your -cp argument if you're on Windows, maybe that's the problem. I also usually put the whole classpath in quotes, but that might not matter. -- Voidious
I sorted the above problem by placing robocode.jar in the robocode dir. That got me as far as,
Downloading ...ag.Gir 0.99 Exception in thread "main" java.lang.NoSuchMethodError: robocode.util.Utils.copy(Ljava/io/File;Ljava/io/File;)Zthen swapping my classpath slashes around (well noticed ;P) gets me as far as;
Downloading ...ag.Gir 0.99How long should it stay like that? Ive also installed 1.1.3, and I get the same behaviour.
Ahh.. just seemingly random timeouts. It's working it's way through a list trying to download them, succeeding on a few. Sweet :)
The RobocodeRepository goes down a lot, that's the reason for all the timeouts. When it's down for long times, I actually edit BotsDownload?.java and recompile so that it skips those bots, then I re-enable it later. Somewhere on the wiki is info on how to do that. Also somewhere on the wiki, someone has posted a .zip file with all the bots from sometime recently - you'd still need those updated since then, but it would help sav eyou some time. When I get home, maybe I'll have time to look around and give a little better info on where those pieces of info are =), or maybe someone else will chime in before then. Also, if you set your name in the Preferences, your updates will show up as you on the Changes page. -- Voidious
S'all working nicely now. I think I need to twiddle with BotsDownload?.java . Cheers :) - hey, does this thing count failed conciousness checks? I hope not :/
The link to the zip file is here: RoboRumble/StartingWithRoboRumble -- Kinsen
And the RoboRumble .class files that skip the repository: [roborumble_norepository.zip]. The discussion and original link is on the Voidious page. -- Voidious
So..I guess this is no longer a 'Temporary Server'... :) --Starrynte