[Home]RoboRumble/ReportedProblems

Robo Home | RoboRumble | Changes | Preferences | AllPages

Archive for March 4th, 2007: RoboRumble/ReportedProblems03042007

I should like you to try out Robocode 1.2.6 Beta, where one of the main goals is to replace the 1.1.3 version for RR@Home. The 1.2.6 has solved the issues listed here. If you find more issue when trying to run 1.2.6 Beta with RR@Home, then please notify me. --Fnl t i Darn FireFox, even with its improved crash recovery, I just lost my whole post about this... Anyway...

I copied robocode.jar and robocode.properties from a 1.2.6 Beta install into a working RR@home install, deleted and rebuilt the robot.database and .robotcache, and launched Robocode once from that directory to make sure it was working OK. Then I tried running RoboRumble and got this error:

Executing battles ...
Fighting battle 0 ... repositorio.NanoStep 1.0,simonton.nano.WeekendObsession_S 1.6.2
Exception in thread "Application Thread" java.lang.NoSuchMethodError: robocode.util.Utils.setLogListener(Lrobocode/control/RobocodeListener;)V
        at roborumble.battlesengine.RobocodeEngineAtHome.runBattle(RobocodeEngineAtHome.java:94)
        at roborumble.battlesengine.BattlesRunner.runBattles(BattlesRunner.java:86)
        at roborumble.RoboRumbleAtHome.main(RoboRumbleAtHome.java:94)
Here are lines 92 - 95 of RoboRumbleAtHome?.java:
System.out.println("Executing battles ...");
BattlesRunner engine = new BattlesRunner(parameters);
engine.runBattles();
engine = null;
I think those steps are all I should need to do to upgrade a RoboRumble client, and it's all that I've done previously, but maybe there's some other steps I should need to do that I'm unaware of.

-- Voidious

By the way, my first thought is to try deleting the .class files for RoboRumble and recompiling them. If I try that, I get these errors:

% javac -cp ".:../robocode.jar:../codesize.jar" roborumble/RoboRumbleAtHome.java

./roborumble/battlesengine/AtHomeListener.java:7: package apv does not exist
import apv.TeamCompetition;
           ^
./roborumble/battlesengine/RobocodeEngineAtHome.java:9: cannot find symbol
symbol  : class FileSpecificationVector
location: package robocode.repository
import robocode.repository.FileSpecificationVector;
                           ^
./roborumble/battlesengine/RobocodeEngineAtHome.java:85: cannot find symbol
symbol  : class FileSpecificationVector
location: class roborumble.battlesengine.RobocodeEngineAtHome
                FileSpecificationVector v = robotRepository.getRobotSpecificationsVector(false, false, false, false, true, false);
                ^
./roborumble/battlesengine/RobocodeEngineAtHome.java:85: cannot find symbol
symbol  : method getRobotSpecificationsVector(boolean,boolean,boolean,boolean,boolean,boolean)
location: class robocode.repository.Repository
                FileSpecificationVector v = robotRepository.getRobotSpecificationsVector(false, false, false, false, true, false);
                                                           ^
./roborumble/battlesengine/RobocodeEngineAtHome.java:94: cannot find symbol
symbol  : method setLogListener(robocode.control.RobocodeListener)
location: class robocode.util.Utils
                Utils.setLogListener(listener);
                     ^
./roborumble/battlesengine/RobocodeEngineAtHome.java:97: cannot find symbol
symbol  : method startNewBattle(robocode.battle.BattleProperties,boolean)
location: class robocode.manager.BattleManager
                manager.getBattleManager().startNewBattle(battleProperties, false); 
                                        ^
Note: ./roborumble/battlesengine/RobocodeEngineAtHome.java uses or overrides a deprecated API.
Note: Recompile with -Xlint:deprecation for details.
Note: Some input files use unchecked or unsafe operations.
Note: Recompile with -Xlint:unchecked for details.
6 errors
That first one is an unused import that can just be commented out, but I think the rest are real things that have changed inside Robocode.

-- Voidious

Some additional info: robots are downloaded and put into the 'roborumble\temp' and 'robots' directory. Although the bot-entry is made in de .robotcache directory, the bot itself is not extracted there. For the rest the same remarks as Voidious. -- GrubbmGait

And some more info: RoboLeague does not work either with 1.2.6beta. It does run the first battle successfully but when collecting the results the error below is shown. -- GrubbmGait
Next grouping: Yngwie 1.0 vs. GresSuffurd 0.2.4 (9 remaining).

 Exception in thread "Thread-5" java.lang.NoSuchMethodError?: robocode.control.RobotResults?.getRamDamageBonus?()I
	at roboleague.GroupingEngine?.run(GroupingEngine?.java:161)
	at roboleague.DivisionEngine?.run(DivisionEngine?.java:53)
	at roboleague.gui.RoboLeagueGui?.bg_runSeasons(RoboLeagueGui?.java:675)
	at roboleague.gui.RoboLeagueGui?.access$700(RoboLeagueGui?.java:21)
	at roboleague.gui.RoboLeagueGui?$5.run(RoboLeagueGui?.java:807)

Hmm.. the big issue for RoboRumble@Home is that is accesses internal Robocode classes, instead of accessing the public control API only, i.e. the robocode.control classes (where the RobocodeEngine? is the main entry). I also believe it is OK to access robocode.util.Utils methods, although this have never been the intension. Infact, the security manager of the current Robocode versions disallow robot to access internal Robocode classes except for robocode.util.Utils (as many robots uses e.g. normalAbsoluteAngle?). It will not be possible to continuesly evolving Robocode if the internal classes cannot change because other programs assumes that it is OK to access Robocode internals. The internals of Robocode will always change in order to enhance it, but also to gain better performance etc. That's the reason for having a API for controlling Robocode, that stays the same. If the robocode.control classes for controlling Robocode are not rich enough when it comes to controlling Robocode, then these should come as feature requests, and I will add it as fast as I can, if they make sense. The API for the control classes will not change, meaning that public classes, methods signatures, etc. will stay the same, except that additional classes and methods of course be added in the future! Existing classes, methods etc. will never be removed or changed. This way program like RoboRumble@Home can count on using the control classes without worrying about which version of Robocode it controls. --Fnl

The big question is how to proceed from here, so that RR@Home will be able to run the current but also comming versions of Robocode. One suggestion from my side is that I try to update RR@Home where the issues arise, and add missing functionality to the control classes of Robocode. --Fnl

That makes perfect sense. RoboRumble@Home is in bad need of refactoring anyway. It's more like a prototype to prove a concept. Now that the concept has been proven beyond any doubts anyone could have had, it is time we take it from the prototype phase. We... of course I seriously doubt I will have much time with it. But maybe Voidious and I can have a look at it together and try assess how much work it is. -- PEZ

Sure, I'm certainly willing to contribute some time and effort to the cause. I know David has also expressed some interest to me in doing a rewrite of RR@Home, but I don't know what kind of time he has for it right now. -- Voidious

Please notice that I intend to remove the robocode.util.Constants class (again) and that robocode.util.Utils should not use copy() and similar stuff. It should "only" be the robocode.control classes that RR@Home make use of. If some additional functionality is really needed for robocode.control.RobocodeEngine?, then send me a mail and I will add it the control API. --Fnl

If you have any problems with RR@Home regarding Robocode, I will of course also do everything that I can to help you out. It has my top priority that Robocode and RR@Home works together soon and in future versions (of both applications). --Fnl

I am trying to build a new version of RoboRumble@Home based on the available 'roborumbleathome_beta9_server_b12.ZIP'. I will fix all the places where there are issues with the Robocode client. My question is: Is anyone else doing updates on RoboRumble@Home, and if so, who? If it has already been updated, I suggest that the sources are merged and to make a new release of RoboRumble@Home. --Fnl

I know I mentioned updating the RR@Home code as part of this conversation, but I don't have anything done that would warrant merging it. I don't think anybody has updated the RR@Home code since Albert. -- Voidious

I have some code that lets you run divisions from the command line. It can also set variables in System.properties - I've used it to try different combinations of tuner variables & view results. I don't know if it's something anyone else would use, or if it should be a part of an actual release. I think links of from my wiki page. -- Simonton

All I have are quick and dirty hacks, through and through. Best you do it. Once you fimish i'll run it indefinitely given my new dsl and the aility to idle programs. --Chase-san

I know some people (Martin f.e.) have solved an issue concerning uploading results of meleebattles. In the current version of RR@Home the results are uploaded twice (nr1 against all, nr2 against all (including nr1), nr3 against all (including nr1 and nr2) etc. etc.), while every result should be uploaded only once. It is only a minor issue, but since you will change the source, it is easy to take along. Please let me know when you need some testing of intermediate versions, time permitted I am happy to try them out. -- GrubbmGait

I have now corrected a lot of issues, and can compile all RR@Home sources for Robocode 1.2.6 Beta. It is running on Java 5. However, I currently get some exceptions that I need to fix. But it seems like I will soon be able to run it without too many errors. When this is possible I will upload the files so you guys are able to try it out! :-) --Fnl

Oh, hey, I also have code to let users run focused competitors offline by modifying the RUNONLY variable in the properties file. That seems like it might be usefull for someone else. If you'd like the source, let me know how to get it to you. This is for RR@Home, my previous comment was for RoboLeague. Sorry if that was confusing. -- Simonton

Simonton: I should like all your modifications, that is if I continue developing on RR@Home. ;-)  --Fnl

Well.. I finally got a new (alpha) version up and running that runs fine with the Robocode 1.2.5A client. It requires Java 5 like the Robocode client. You can download the new version from here: [RoboRumble.zip] --Fnl

Looks like only one round is fought instead of the number in the roborumble.txt file. BTW, I use v1.2.5. -- GrubbmGait

 G:\Applicaties\RoboHome125?>roborumble
 G:\Applicaties\RoboHome125?>cd robots
 G:\Applicaties\RoboHome125?\robots>java -Dsun.io.useCanonCaches=false -Xmx512M -cp  .;../robocode.jar;../codesize.jar; roborumble.RoboRumbleAtHome? ./roborumble/roborumble.txt
 Iteration number 0
 Downloading participants list ...
 Downloading missing bots ...
 Downloading ...simonton.micro.WeeklongObsession 1.6.3
 Unable to download simonton.micro.WeeklongObsession 1.6.3 from site.
 Could not download bot simonton.micro.[WeeklongObsession 1]?.6.3.jar
 Downloading ...simonton.mini.WeeksOnEnd 1.10.2
 Unable to download simonton.mini.WeeksOnEnd 1.10.2 from site.
 Could not download bot simonton.mini.[WeeksOnEnd 1]?.10.2.jar
 Ignoring simonton.micro.[WeeklongObsession 1]?.6.3.jar: .\simonton.micro.[WeeklongObsession 1]?.6.3.jar (Het systeem  kan het opgegeven bestand niet vinden)
 Ignoring simonton.micro.[WeeklongObsession 1]?.6.3.jar: .\simonton.micro.[WeeklongObsession 1]?.6.3.jar (Het systeem kan het opgegeven bestand niet vinden)
 Ignoring simonton.mini.[WeeksOnEnd 1]?.10.2.jar: .\simonton.mini.[WeeksOnEnd 1]?.10.2.jar (Het systeem kan het opgegeven bestand niet vinden)
 Ignoring simonton.mini.[WeeksOnEnd 1]?.10.2.jar: .\simonton.mini.[WeeksOnEnd 1]?.10.2.jar (Het systeem kan het opgegeven bestand niet vinden)
 Downloading rating files ...
 Removing old participants from server ...
 Preparing battles list ... Using smart battles is true
 Prioritary battles file not found ...
 Executing battles ...
 Fighting battle 0 ... dft.Cyanide 1.90,kano.gamma.KanoGamma 1.8
 RESULT = dft.Cyanide wins 125 to 41
 Fighting battle 1 ... timmit.nano.TimDog? 0.33,kawigi.mini.Coriantumr 1.1
 RESULT = kawigi.mini.Coriantumr wins 134 to 32

 G:\Applicaties\RoboHome125?\robots>cd ..

Aww, sorry for spamming y'all's outputs, guys. Sounds like I'll soon be able to let you at those bots, though! Thanks & keep it up, Fnl. -- Simonton

I also noticed that the roborumble.txt, meleerumble.txt and teamrumble.txt are not equal to what I use regarding several URL's. You can find correct working versions here: RoboRumble/roborumble txt -- RoboRumble/meleerumble txt -- RoboRumble/teamrumble txt
Tip: If you set 'UPLOAD=NOT', the results are not uploaded but kept locally in a file, so you can easily check them without contaminating the rankings. -- GrubbmGait

GrubbmGait: Thanks for the hint. I will definitely use the real onces. Looking into all the pages here on RoboWiki for RR@Home is so "confusing", so I have a hard time finding the right info and files. --Fnl

Simonton: I will have a look with the issue you mentioned. I think it might be related to what GrubbmGait told. --Fnl

Yeah, even the RR@Home info is rather scattered. Nice moment for a re-write when the 'new' RR@Home is ready. You can forget about Simonton's bots, they are not downloadable on purpose as they use (too) much memory. Btw, I mailed you Fnl a bit more info, I hope it is usefull. -- GrubbmGait

GrubbmGait: Thank you. Your info is very helpful. I am working on fixing as much as possible. I am also trying to get in contact with Albert. That is, if it is okay that I update it or perhaps even put it on SourceForge? as an independent project beside Robocode. --Fnl

TEll me when you finish, i'm itching to finally support and not cause to much dead lag on my resources. --Chase-san

I have now made an [1.3 Alpha] (unofficial) version that I should like you to test. This is Robocode with RoboRumble@Home built-in! In the coming Beta, RoboRumble should be easier to startup and configurate, so this version is a very basic version. --Fnl

Okay, its downloaded and I am attempting to run it now, i'll tell you what I find out. --Chase-san

Only a few minutes into use I noticed it "cannot" download missing robots, this may be because the program itself or because the Repository is down, but if its the latter it should skip the download after a certain length of time. --Chase-san

Yes, it seems like the Repository is down, which is good for making a timeout-mechanishm for ignoring the Robocode Repository if it is not responding. ;-) --Fnl

I have replaced the [1.3 Alpha] with a new improved version, which should not hang when attempting to download robots. This was a hard one as I had to rewrite the part used for the file transfer to be multi-threaded and do proper cleanup and error handling. I sure hope this works much better! At least it did while I tested it against Robocode Repository, which "luckily" did not response to my download requests. --Fnl

I'll give it a shot when I get home (this compy doesn't have java on it) --Chase-san

Thanks to Chase-san a lot of issues have been spotted and have now been fixed. I have compiled the last [Robocode 1.3 Alpha] version, and the next version will be a Beta. It turned out that Robocode was not able to play team battles thru the robocode.control.RobocodeEngine? class, which might be the reason that Albert was accessing internal classes (which is a hack! ;-) in Robocode. This issue is now fixed, and also the "number of played rounds" issue has been fixed in the new Alpha. You guys should give the new version a try, and tell me if there are still "blocking" issues left. Later on, my plan is to improve the GUI of Robocode to make it easier to use RoboRumble@Home for newbies. --Fnl


At the moment it is advised not to use Robocode 1.3_Beta for running RR@Home. The problem is that a 100% radarlock with 1.1.2/1.1.3 no longer is a 100% radarlock with 1.3_Beta. Some bots without radarlock-lost recovery (like stefw.Tigger) therefor perform very badly. Other bots may be influenced as they regularly miss information (GresSuffurd gets around 6 to 7 radarsweeps per round instead of zero).

As a result, bots that use the 'Rules'-class will not be updated in the ranking, as the previous stable version 1.1.3 does not support this class. -- GrubbmGait

Okay, okay, (gets 1.1.3), the new ones aren't working, so. Also btw, a small bug.

Downloading ...bayen.nut.Squirrel 1.621
java.util.zip.ZipException: error in opening zip file
Downloaded file is wrong or corrupted:bayen.nut.Squirrel_1.621.jar
Could not download bot bayen.nut.Squirrel_1.621.jar

--Chase-san

I seem to have more success with 1.2.4, it seems stable enough, bots like SandBoxDT? and Kabuto don't malfuntion, and the newer bots seem to work like they always did. Now that is just for regular battles, I'm not sure how well it plays with the rumble. --Ne

Yes, I use later versions for development, but 1.1.3 is the latest version that works with the RoboRumble client. Versions 1.1.4 and 1.1.5 will work, but they are buggy and seem to give very tainted ratings. (Dookious gets an extra 10-15 rating points if I process its rumble battles with 1.1.5.) Soon the RoboRumble client will be integrated into Robocode (with 1.3, thanks to Fnl's hard work), but until then the Rules class is a problem for rumble bots. -- Voidious

Have you guys tried out the latest version 1.3 Beta 2? This version fixes some of the issues, among other things the radar lock problem. --Fnl

I have checked and the 'known problems' (radar, SandboxDT) are solved. As soon as the SPI problem for RR@Home is solved I'll be happy to do some more testing regarding RR@Home. -- GrubbmGait

Okay. The SPI problem (ImageUtil? is accessed, even if it shouldn't) has been solved, but not released yet. --Fnl

Is it in the CVS? --Chase-san

Yes, but now it is released, i.e. Robocode 1.3 has just been released. ;-) --Fnl

I released Robocode 1.3.1 yesterday. Still, I found a bug with SittingDuck, that gave problems when the java.io.FileReader? is used. SittingDuck is denied access to this class when running without a GUI. I have now fixed this for the coming 1.3.2, and I think this will potentially fix problems with existing robots in RoboRumble which writes data to it's data file. Just so you all know. --Fnl

 Fighting battle 1 ... rc.yoda.Yoda 1.0.5a,ar.[QuantumChromodynamics 1]?.1 1.5
 Exception in thread "Application Thread" java.lang.ArrayIndexOutOfBoundsException?: 1 >= 1
        at java.util.Vector.elementAt(Unknown Source)
        at robocode.peer.ContestantPeerVector?.elementAt(ContestantPeerVector?.java:61)
        at robocode.battle.BattleResultsTableModel?.getValueAt?(BattleResultsTableModel?.java:84)
        at roborumble.battlesengine.BattlesRunner?.runBattles(BattlesRunner?.java:95)
        at roborumble.RoboRumbleAtHome?.main(RoboRumbleAtHome?.java:94)
I think it doesn't like the "_1.1 1.5" in QuantumChromodynamics's name. It does this every time it tries to run a battle with that bot. -- Simonton

Actually, the bot name is listed as QuantumChromodynamics 1.1_1.5, which came about when I had to create a new jar compiled under Java 1.5 because roborumble didn't like my 1.6 compiler. I guess I'll rebuild the jar under the name QuantumChromodynamics 1.2 and post it again. Why does roborumble hate this bot so much? -- AaronR

If you don't pack it via robocode, there could be many reasons. ;) I try to keep my versions rather straight forward. --Chase-san

Any progress on rebuilding that bot? It's actually slowing down my client a fair amount, because it takes it a minute to figure out that it doesn't like that name every time it tries to run a battle with it. -- Simonton

Yes, there is definitely a problem with underscores with priority battles in RoboRumble. Unfortunately, it requires a redesign of how robots are represented in RoboRumble. Currently they are modelled like strings, where dots are replaced by underscore from time to time and vice versa. This is definitely not good. E.g. "1.5_2.0" could become 1_5_2_0 and when it is reversed it becomes 1.5.2.0. See the problem? I will look into this issue soon, I promise! :-) --Fnl

Feel free to put it at the bottom of the priority list - no underscores in bot names and verisons is something we can all live with, I think. =) Thanks as always, though, Fnl, you rock. -- Voidious

Dump it, and put it on the list for Robocode 2 (it hasn't been supported yet, why start on the nearly last version?). ;) --Chase-san

Okay. It's dumped! ;-) --Fnl

I have had the bot rebuilt for days now, but, as Murphy's law dictates, every time I try to get to the Robocode Repository it isn't working. I'll try to upload it one more time. --AaronR

Yes, it is quite anoying when the Robocode Repository is down (often). --Fnl


hennes.SimpleBot? 1.0 is getting a lot of 0 scores. Perhaps it uses the Rules class, or one of the new setAllColors? methods? Or some real bug? Mr. Hennes, please make sure your bot runs with RoboCode version 1.0.7 (available from the sourceforge page, where you downloaded robocode). This is the earliest version used in the rumble, and does not support all the features in the newer versions. Until then, your bot will perform very badly :). --Simonton

Um...For some reason my client can't find 'participmelee.txt' and throws an UnknownHostException? even though i am connected.

>java -Dsun.io.useC
anonCaches=false -Xmx256M -cp .;../robocode.jar;../codesize.jar; roborumble.Robo
RumbleAtHome ./roborumble/meleerumble.txt
Iteration number 0
Downloading participants list ...
Unable to retrieve participants list
java.net.UnknownHostException: robowiki.net
Downloading missing bots ...
Participants file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\participmelee.txt (The system
cannot find the file specified)
Battles input file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\participmelee.txt (The system
cannot find the file specified)
Downloading rating files ...
Unable to ratings for meleerumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for minimeleerumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for micromeleerumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for nanomeleerumble
java.net.UnknownHostException: rumble.fervir.com
Preparing melee battles list ...
Participants file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\participmelee.txt (The system
cannot find the file specified)
Uloading results ...
Can't open result file for upload
Updating number of battles fought ...
Can't open # battles file ... Aborting # battles update
--Starrynte

I get almost the exact same output running 1v1:

java -Xmx256M -Dsun.io.useCanonCaches=false -cp libs/robocode.jar;libs/codesize.jar;libs/roborumble.jar roborumble.RoboRumbleAtHome ./roborumble/roborumble.txt
Iteration number 0
Could not load properties file: ./roborumble/files/codesize1v1.txt
Downloading participants list ...
Unable to retrieve participants list
java.net.UnknownHostException: robowiki.net
Downloading missing bots ...
Participants file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\particip1v1.txt (The system cannot find the file specified)
Battles input file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\particip1v1.txt (The system cannot find the file specified)
Downloading rating files ...
Unable to ratings for roborumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for minirumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for microrumble
java.net.UnknownHostException: rumble.fervir.com
Unable to ratings for nanorumble
java.net.UnknownHostException: rumble.fervir.com
Could not load properties file: ./roborumble/files/codesize1v1.txt
Could not load properties file: ./roborumble/temp/ratings_roborumble.txt
Could not load properties file: ./roborumble/temp/ratings_minirumble.txt
Could not load properties file: ./roborumble/temp/ratings_microrumble.txt
Could not load properties file: ./roborumble/temp/ratings_nanorumble.txt
Preparing battles list ... Using smart battles is false
Participants file not found ... Aborting
java.io.FileNotFoundException: .\roborumble\files\particip1v1.txt (The system cannot find the file specified)

I use Robocode 1.3.4, and I have not modified my roborumble.txt file at all other than changing the user name and making NUMBATTLES = 30.

 -- AaronR

Under the PL (micro) I see that Decado is missing a pairing...Grinnik also. Any idea what's happening? Is there a bot in the rumble that isn't being run? -- Skilgannon

Both of them have not faced timmit.micro.TimXJ? 0.22, but don't ask me why :P --Krabb

Ok, I'm using version 1.1.3, and roborumble works for me now :D --Starrynte

Is the problem in your previous comment for 1.4? I have no problem with 1.4 once I put the roborumble.txt file in place, which it was missing. I know that WeeksOnEnd will die on 1.1.3 because of the lack of the Rules class - Simonton has been waiting a while for a stable RR client that included the Rules class to make it into circulation... -- Voidious

Are you still getting the UnknownHostException? with the version 1.4? If so, under what OS'es are you running? --Fnl

Exception in thread "main" java.lang.NoSuchMethodError: robocode.util.Utils.copy
(Ljava/io/File;Ljava/io/File;)Z
        at roborumble.netengine.BotsDownload.downloadBot(BotsDownload.java:217)
        at roborumble.netengine.BotsDownload.downloadMissingBots(BotsDownload.ja
va:157)
        at roborumble.RoboRumbleAtHome.main(RoboRumbleAtHome.java:49)
Dunno why... --Starrynte
Iteration number 0
Downloading participants list ...
Downloading missing bots ...
Downloading ...amk.Punbot.Punbot 0.01
Unable to download amk.Punbot.Punbot 0.01 from site.
Could not download bot amk.Punbot.Punbot_0.01.jar
Downloading ...apv.MicroAspid 1.8
Unable to download apv.MicroAspid 1.8 from site.
Could not download bot apv.MicroAspid_1.8.jar
--Starrynte

My brand new 1.4 installation generates this output when executing Roborumble:

C:\Users\Aaron\Applications\Robocode\1.4>roborumble

C:\Users\Aaron\Applications\Robocode\1.4>java -Xmx256M -Dsun.io.useCanonCaches=f
alse -cp libs/robocode.jar;libs/codesize.jar;libs/roborumble.jar roborumble.Robo
RumbleAtHome ./roborumble/roborumble.txt
Iteration number 0
Downloading participants list ...
Downloading missing bots ...
Trying to download bayen.nut.Squirrel 1.621
Could not connect to http://www.freewebs.com/bayen/files/bayen.nut.Squirrel_1.62
1.jar
Could not download bayen.nut.Squirrel_1.621.jar
Trying to download bvh.fry.Freya 0.81
Could not find bvh.fry.Freya 0.81 from http://home.kpnplanet.nl/~B.van.Hest@kpnp
lanet.nl/jars/bvh.fry.Freya_0.81.jar
Could not download bvh.fry.Freya_0.81.jar

... "can't download"s snipped ...

Ignoring winamp32.micro.MicroMacro_1.0.jar: .\robots\winamp32.micro.MicroMacro_1
.0.jar (The system cannot find the file specified)
Ignoring winamp32.micro.MicroMacro_1.0.jar: .\robots\winamp32.micro.MicroMacro_1
.0.jar (The system cannot find the file specified)
Downloading rating files ...
Removing old participants from server ...
Preparing battles list ... Using smart battles is true
Prioritary battles file not found ...
Executing battles ...
Fighting battle 0 ... ntc.Evader 1.2,drm.Magazine 0.39
.robotcache.abc.Shadow_3.66d.jar_.abc.a: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.a
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\a.class (The system cannot find the path s
pecified)
.robotcache.abc.Shadow_3.66d.jar_.abc.b: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.b
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\b.class (The system cannot find the path s
pecified)
.robotcache.abc.Shadow_3.66d.jar_.abc.c: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.c
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\c.class (The system cannot find the path s
pecified)
.robotcache.abc.Shadow_3.66d.jar_.abc.d: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.d
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\d.class (The system cannot find the path s
pecified)
.robotcache.abc.Shadow_3.66d.jar_.abc.e: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.e
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\e.class (The system cannot find the path s
pecified)
.robotcache.abc.Shadow_3.66d.jar_.abc.f: Got an error with this class: java.lang
.ClassNotFoundException: Could not find: .robotcache.abc.Shadow_3.66d.jar_.abc.f
: java.io.FileNotFoundException: C:\Users\Aaron\Applications\Robocode\1.4\robots
\robotcache\abc\Shadow_3\66d\jar_\abc\f.class (The system cannot find the path s
pecified)

... etc ...

This output continues for every class in every robot JAR in my robots directory. My guess from looking at the output and the structure of the robots/robotcache directory is that it is related to the translation of dots into slashes. Anyone seen this before?

Robocode: 1.4; Java: 1.6; OS: Windows Vista; roborumble.txt: Copied from a 1.3.5 jar (since 1.4 doesn't come with the file).

-- AaronR


Could someone explain why this is happening? It looks to me like Waylander should have won....or else the scores are backwards.

Fighting battle 14 ... kc.micro.Thorn 1.142,jk.Waylander 0.2.3 RESULT = kc.micro.Thorn 1.142 wins 2583 to 2617

-- Skilgannon

An even more funny result...I don't know if this is normal...but this is a definate bug. (not the score, but who it says won)

Fighting battle 11 ... jk.Waylander 0.1.4,abc.Shadow 3.66d
RESULT = abc.Shadow 3.66d wins 2463 to 2468
-- Skilgannon

The Roborumble (1.4.3) just hangs on me. I am using the same folder to run console battles, so it shouldn't have anything to do with building the robot database. I get as far as the first matchup listing and my CPU load drops to nothing. I have a dual CPU machine, but I set the affinity for [cmd] and [java] to one CPU (as I did with Roborumble@Home) but no dice. /shrug -- Martin

1.4.4 fixes it. Check on the Robocode/News page. -- Skilgannon

What the heck?

D:\Robocode_1.4.5>java -Xmx256M -Dsun.io.useCanonCaches=false -cp libs/robocode.
jar;libs/codesize.jar;libs/roborumble.jar roborumble.RoboRumbleAtHome ./roborumb
le/roborumble.txt
Iteration number 0
Downloading participants list ...
Downloading missing bots ...
Trying to download bayen.nut.Squirrel 1.621
Could not connect to http://www.freewebs.com/bayen/files/bayen.nut.Squirrel_1.62
1.jar
Could not download bayen.nut.Squirrel_1.621.jar
Trying to download cjk.Merkava 0.3.0
Could not find cjk.Merkava 0.3.0 from http://www.krill-labs.com/robocode/robots/
cjk.Merkava_0.3.0.jar
Could not download cjk.Merkava_0.3.0.jar

... "can't download"s snipped ...

Ignoring winamp32.micro.MicroMacro_1.0.jar: .\robots\winamp32.micro.MicroMacro_1
.0.jar (The system cannot find the file specified)
Ignoring winamp32.micro.MicroMacro_1.0.jar: .\robots\winamp32.micro.MicroMacro_1
.0.jar (The system cannot find the file specified)
Downloading rating files ...
Removing old participants from server ...
Preparing battles list ... Using smart battles is true
Prioritary battles file not found ...
Executing battles ...
Fighting battle 0 ... projectx.ProjectNano 1.0,shu.nitro.LENIN .T34
Exception in thread "Application Thread" java.lang.ArrayIndexOutOfBoundsExceptio
n: 0
        at roborumble.battlesengine.BattlesRunner.runBattles(Unknown Source)
        at roborumble.RoboRumbleAtHome.main(Unknown Source)

Robocode 1.4.5, Java 1.6, Windows XP. Error usually occurs within 5 battles and does not appear to be bot-dependant. -- AaronR


Rankings have gone, it will repair itself over time when battles are fought. If someone just started his client (10-15 minutes before this timestamp), please check your settings and environment. -- GrubbmGait
Just got this while building Fighting battle 3 ... mld.Moebius 2.3,sL300.Mozart life RESULT = mld.Moebius 2.3 wins 2533 to 2644

Nice to see I can win when I lose in points! --Miked0801


So is the Rules class problem officially resolved with current version of Robocode (1.4.7)? I wanna put my robot in the rumble and start using the @Home client, but I can't see whether this problem has been addressed. -- RobertWalker

There's nothing stopping someone from running a Robocode 1.07 client or anything like that, but I think everyone running right this minute is using a more recent version. Until we revamp the RoboRumble server a bit to require a newer version, there will always be some risk associated - unless you're doing something in a CodeSize restricted division, I would just use your own copies of the methods in the Rules class, but it's up to you. -- Voidious


Something needs to be done about Frederick. It's commonly getting 0 scores against the better bots, which somehow means that when they are release they get stuck running battle after battle against Frederick until it finally gets lucky to get a couple points AND that score gets submitted to the server. DrussGT just ran 49 battles in a row against Frederick on my client, in 7 of which it scored. This is a waste of RR@H processing power. -- Simonton

Well, it might error out, but really I think it just isn't very good at all, some bots don't even fire at all, such as Serenity, so its not really an error so much as the lower, older or bots of less then outstanding quality just can't hit the big guns. It does something, but generally the bots that can't even hit the better ones run many many more battles, because they get no score at all(specially against the big and bad surfers). So what do you prepose, we remove all the bottom, non-firing and head on firing robots? This is probably also why the sample bots are not in the general rumble because they would 9 times out of 10 (100? 1000?), not get any score at all and would load down the clients. --Chase-san

I think the ideal solution, long-term, is a system that gracefully handles 0 scores - whether that's just making it register as 0.1, or a different rating calculation altogether. Not sure what's a good solution in the meantime, though... Simonton, is that the only one you've noticed getting 0 so many times? I know I see 0 scores go by sometimes (and even that specific bot), but I thought usually it was just once or twice before they got some points. -- Voidious

Well, in the next version of DrussGT I'll add some shutout prevention....basically just go to a dangerous spot on a low-power bullet in the last round if I haven't been hit yet. It stands to reason that it might actually up my score some, because then when I DO get hit, it is with a small bullet and not with one that could be, for instance, power 3, due to a bad starting position, or whatever -- Skilgannon

Somebody's client is removing bots from the rumble, with the result that other clients have to run an iteration with that bot as a focused competitor...If you want to exclude bots, could you upgrade to 1.4.9, where the exclude filter has been fixed so that it doesn't remove the excluded bots from the rumble? Thanks.. -- Skilgannon

I do not know why, but my bots are not being downloaded through the Roborumble client. I am using a URL (rather than RobocodeRepository id). If you paste the URL into a browser, it download the file. I can only guess at the cause of the problem. -- Martin

It should be normal but I updated jab.micro.Sanguijuela with the 0.4 version (repository and participants list) three days ago and the 0.3 version appears and disappears in the rumble for some reason. The robots folder of people running the rumble? any kind of bug? I don't know.

Well, it's my guess that someone is probably running a RoboRumble client with an out of date participants list. It's possible they misconfigured the participants list url sometime between when you released 0.3 and when you released 0.4. In any case I don't think there's much that can be done about it except hope that whoever is running the bad client gets it fixed. -- Rednaxela

I think that it should be some kind of bug. At the stelo.RamTrackSurfer rating details the enemies list stops at letter "b" and the same thing happens at "s" with gh.mini.Gruwel. Also apv.LauLectrik throws a ArrayIndexOutOfBoundsException? when looking its details. --Jab?

Just to let you know an incidence. jab.avk.ManuelGallegus? 0.5 has a battle against ad.last.Bottom 1.0 (the unbeatable :P ) and ManuelGallegus? get 0.0 % Score. I have run many battles in Robocode between them trying to understand what happened but I'm not getting any Exception. --Jab?

Checking my local copy of the battles file shows it as being from "darkstorm", I believe when "darkstorm" was running the 1.6.1 Beta client (see RoboRumble/RankingChat). I think the results from darkstorm are fine now that he's back on the 1.6.0 client, and you would probably be best off re-releasing ManuelGallegus? 0.5 if you're concerned that bad results from that client before are affecting your score.

Thanks for the info. About re-releasing I have no problem in waiting for the next version of the bot. I'm trying hard to put ManuelGallegus? in the Top-100 :P Yeah! After two days, ManuelGallegus? 0.6 reach the Top-100 :D --Jab?


On 11 July there was a hickup of the server, and after that several bots started from zero battles again, f.e. usa.nano.Nemo. Maybe the disk of the server is full ? I am stopping my client till further notice. -- GrubbmGait

I've noticed some weirdness as well and it seems like many of us have. At one point today, I noticed the server stopped responding to HTTP requests altogether (they were timing out) for some reason. Immediately when it started responding again, the results page lost all the bots (see comments by Tim and Voidious RoboRumble/ParticipantsChat here). In addition there was apparently some corruption on Dookious' battle list (see comment by Skilgannon in Dookious, and here's another example of a [clearly corrupted results file] listing over 20 thousand battles yet missing battles against ones like Druss and Moron and many others). Then as you notice some bots seemed to have started at 0 battles again. I think that sums up the issues we've seen with the RR server in the last 12 or 24 or so hours. I'm not sure what the cause of this issue is, perhaps a disk full issue, but I don't think a disk full problem would explain why it wasn't responding to HTTP requests for a short time. Either way there certainly seems to be corruption of results files for some reason or another. Anyone been in contact with Pulsar recently or have any other ideas? -- Rednaxela

Today's problem: I've tryed to look at the currentsRankings? page but i've got this error:

(See [here], because the wiki isn't fond off html)
Asdasd

Yeah, I just noticed that myself and was about to post about it. Also, since the wiki doesn't like html I replaced your paste of it with an offsite archive of the error. The RR server certainly seems to be acting up lately... -- Rednaxela

The same happened the last time the server got out of disk space. This could be another big crash. So many ideas to test and no rumble. :( -- ABC

Yes. I just sumbitted a new version to test too. So how do we go about trying to sort this problem out? Does anyone have access to the server? -- Tim Foden

Pulsar runs the server, and I don't know of anyone but him having access to it. I'll drop him an e-mail and see if he could take a look. Maybe I'll CC David Alves and see how that RR server rewrite is coming along, too... :-P -- Voidious

I've actually been working on it the past two weeks. =) --David Alves

Good to hear =) Any idea (even vague) on how long before we could have a barebones server to upload to/get rankings from? -- Skilgannon

Hey guys, in case you missed it, Pulsar got the server back up. He said some weird characters showed up in some data files and the RR scripts were choking on them, but he wasn't sure of the cause. Looks like the rankings are at least showing up again. Could someone check and post here if they can upload results successfully? -- Voidious

Uploading is working fine, except for a few bots. Must be because of the corrupted details files. -- ABC

It would be nice if this problem was fixed, my clients are choking trying to upload those bots' results. It is making the rumble quite slow... -- ABC

This can be fixed by manually editing the involved files on the server, or by deleting the involved files altogether. I don't see another solution. (or you could (temporarily) move those bots to the bottom of the participants page, like I did with cx.mini.Blackswans) -- GrubbmGait

One way to alleviate the problem (although not get rid of it completely) is to remove the results1v1.txt file between each iteration, so all the failed uploads from all the previous iterations don't get qued to upload this time. -- Skilgannon

What is the latest "Rumble save" version of Robocode? I'm using 1.54 in both my clients but I'm getting 0 results from bots that probably use features from later versions. -- ABC

I'm also using 1.5.4... I think Rednaxela has been using 1.6.0 though. 1.6.1beta1 was buggy, IIRC. 1.6.1beta2 fixed it, I think. The new painting stuff is kinda silly IMHO, purely from performance reasons, because it can always be called (even when not painting). But I guess it makes implementing some of the more interactive painting stuff simpler. -- Skilgannon


Erm.. it seems like a large number of entries in the team rumble have broken downloads:
CharlieN.OmegaTeam.OmegaTeam 1.0, jabteam.Enjambre 0.0.1, Krabb.sliNk.GarmTeam0.9u, Krabb.sliNk.SlartibartfassTeam 0.5, pedersen.ImWithDroidTeam 1.3, and pedersen.ImWithStupidTeam 1.3
or is this just me? Anyone able to fix the downloads for them or at least have a zip of them? -- Rednaxela

new problem, help please: java.io.IOException: Server returned HTTP response code: 500 for URL: http://rumble.fervir.com/rumble/UploadedResults Unable to upload results Asdasd

another problem: Exception in thread "Application Thread" java.lang.ArrayIndexOutOfBoundsException?: 1

	at roborumble.battlesengine.BattlesRunner?.runBattles(Unknown Source)
	at roborumble.RoboRumbleAtHome?.main(Unknown Source)
it appears random during the battle (whit bot dreadnught and shadow) and stop the current roborumble, but no newline in the partecipants list... HELP!

roborumble version: 1.6.0.1 (actualy running with firefox and azureus), OS: Linux UBUNTU, VM: sun 1.6.0(the latest), amd dual processor (cpu used in this moment: 9%), 2GB ram (used in this moment: 22%), upload band used: 40kb/s for azures (limted), roborumble takes about 10kb/s. azureus with no limit goes to 60 kb/s...(so actulally i'm "surfing te web" at 70 ks/s in upload, download's band it's only at 40 ks/s, 25 ks/s for azureus, today p2p it's slow :-( ) if you need more information i've the log file. Asdasd

Delete the robots/.robotcache/ and robots/robot.database. This should sort out that problem. The download speeds shouldn't affect robocode at all. -- Skilgannon

About the HTTP error 500, there are still some bots listed in the mini/micro/nano rumble that have corrupted details files. I'll try weeding them out later today. -- ABC

@ Skilgannon: deleted the file, now everything seems right. Still only the http error, but i think at the end of many tryes it upload the result (i hope). Asdasd

Thanks GrubbmGait for fixing your own bot links for the team rumble and uploading a zip that included all the other ones with broken links. Another problem with the team rumble seems to be cropping up however: Whenever a battle that involves ArmyOfShiz happens, the rumble client locks up and stops processing. I can't duplicate this outside of the the rumble client either. Is anyone else seeing this happen? -- Rednaxela

Hey, i have that problem with kawigi.sbf.TidalWave. But ArmyOfShiz is ok here. Haven't checked it outside the rumble jet... --Krabb

Hmm, I re-downloaded ArmyOfShiz, cleared my rumble temp files, and rebuilt the robot database, but ArmyOfShiz is still locking it up. Also ImWithStupidTeam? seems to lock it up as well. I find that interesting considering according to [this] it's missing battles against 9 opponants out of a 35 bot group, which is very large. Actually, in general, the team primere league looks quite odd. Also, despite the "battles per bot" being set to 200, a very vast marjority of the battles my team rumble client is running only involve bots with far over 200 battles. Anyone have any idea what the heck is up with this? The fequent lockups also seem to make it difficult to get many battles contributed at all... -- Rednaxela

Team battles usually consume a lot of memory, maybe that's what's causing your client to hang? You could try increasing the default 128M in the batch file. Bots with more battles than the "battles per bot" setting get chosen if they have missing pairings. I'll try running my teamrumble client later today to see if I also run into problems. -- ABC

Well, I already had an increased memory limit and that isn't the issue in this case. ArmyOfShiz locks up 2 of my rumble clients despite being identical in settings so far as I can see to my third client. ImWithStupidTeam? makes all three clients however. No problems with TidalWave like Krabb notes on any of my three clients. -- Rednaxela

One other observation, is that despite being around for quite some time, florent.XSeries.Xmen has never ever batteled ImWithStupidTeam? (update: and in fact, 9 bots have not!), and to me this indicates that there is definitely something going on odd with ImWithStupidTeam? and my clients aren't alone about this. I've tried some more but it seems to behave just fine outside the rumble despite locking the client when inside the rumble. If I figure anything else out that might help tracking the problem down I'll post here. -- Rednaxela

Lots of bad results in the last 2 days, anyone running a buggy client? I'm currently on holidays with limited net access... -- ABC

I just uploaded a new bot (djc.DynaBot? 0.1.0) which seems to be performing badly on the Rumble, though it wins consistently when I run him against the bots Rumble has him losing to. Example, against el33t.EL33tGangstarr2_2.0 my bot wins 35 rounds and outscores 5000 to 600 reliably -- djc?

I also reported a similar incidence, the problem was someone running the 1.6.1 beta client. As far as I know the solution is not easy. First we have to know who is running this server (I don't know how) to alert him or her, and then the affected bots should be re-released. I also checked my bot before reporting the problem looking for posible runtime exceptions in the bot code. Good luck! -- Jab?

I wrote my bot with Robocode 1.6.1 beta 2...I thought it was the official current release; is that incorrect? -- djc?

Thanks. I'm also noicing in Rumble that the first bot in 1v1 wins a disproportionate number of matches. Where is 1.6.0 downloadable from? -- djc?

The rankings, or better the ratings, are quite messed up, so I checked the global resultsfile. It appears that meleebattles have been fought with one-on-one participants and that the results have been uploaded to the one-on-one ranking. I don't know if that is due to the beta-version or that someone accidentaly made a mistake editing his roborumble.txt file. Please note that meleebattles have their own startup-script and files and that it is very advisable to have different 'installs' for each type of roborumble (one-on-one, melee, team). -- GrubbmGait

Heh, this kind of stuff makes me hope that future incarnations of the RoboRumble server (Roborumble.org?) will be most robust with this kind of stuff. Ideally I think it would be good if the protocol also in the future included clients reporting versions, allowed the server to blacklist versions. Anyways, that's a discussion for another page, but these kind of problems make me think it's more important than I previously thought it. -- Rednaxela

Sorry guys for the buggy beta versions of Robocode 1.6.1. A lot of stuff inside Robocode has been restructured for the future in mind, making it possible to extend Robocode with better features, and make it more stable etc. This has caused a lot of unwanted side-effects, but I think we are almost there. Don't hesitate with trying out the new Robocode 1.6.1 Beta 3, which correct the problems described on this page. :-) --Fnl

Erm... looks like there's corruption in [Coriantumr's melee results page]... and probably elsewhere too. Heh.. and lots of error 500 codes from the server when uploading any melee results also... Hopefully this could get fixed up soon... =\ -- Rednaxela

Still a bad client running? The ratings are now completely useless :(. I'll use the rumble for details comparison then, should be ok if I put an obviously wrong download link in the participants page. -- ABC

This stuff with bad clients and corrupted details files recently is indeed being highly problematic.. It's part of why lately I started to get working with Polylunar, because the team rumble seems to be the only part of the rumble that's safe from all that mess right now. Personally, I'm hoping that Roborumble.org will be done soon and it and the clients will use a new protocol that makes it easier to get rid of bad client results... but of course that doesn't seem to be going quickly. -- Rednaxela

Guys, I found a serious problem with Robocode 1.6.1 regarding the scores and rankings (Thanks Liam Noonam for reporting the issue). The bug is "small", but big enough to let some robots get higher or less scores from time to time. So DON'T use 1.6.1. Use 1.6.0 instead. I am working on a new version 1.6.1.1 that solves this bug, but also two other bugs that has no impact on RoboRumble. --Fnl

Wouldn't a simple fix be to have all future rumble clients check a page on the wiki (similar to downloding the participants list) which lists the approved versions? It wouldn't fix rogue 1.6.1 clients but could minimize this problem for future versions... -- Darkcanuck

That would be a simple fix that would work to some extent, but it wouldn't allow bad results from rouge clients to be easily cleaned. Personally I think that matters, because often we don't know a client version is bad until after it has uploaded bad results. For that reason, I'd consider a protocol update to be far more desireable. Making the clients check a wiki page wouldn't be a bad short-term measure in the meantime though I suppose. -- Rednaxela

Another check that should be made in the server (in one-on-one at least), is if the number of battles is near 35. That way you can avoid that someone is running battles in meleemode with the one-on-one settings. And I mean 'near 35' as with rambots you sometimes get a sum of 34 or 36 battles won. -- GrubbmGait

In order to avoid other kinds of goofy config errors (i.e. wrong field size), I think the client should probably send certain other configuration options like field size that the server should check. It might be good to also have the client send the participants url it's using, but at very least I think both field size and like you say, number of rounds, should be checked. -- Rednaxela

I think that you have already noticed that something weird is happening in the OneOnOne roborumble. Many bots are missing in the rankings. --Jab?

Another issue, I'm running the teamrumble version 1.6.0 and I have ITERATE=YES, RUNONLY=SERVER and BATTLESPERBOT=200. The priority battles works fine the first iteration but after that it doesn't take into account the teams with less battles than 200. May be I'm missing something. Thanks! --Jab?

Hi, I am trying to fix the ITERATE problem (and other RoboRumble issues as well). I made a 1.6.1.3 Alpha version with a fix (hopefully), and I should like to know if this version fixes the problem? You can download it from [here]. If this version fixes the problem, the ITERATE feature has been "broken" in RoboRumble since forever. --Fnl

Just tested 1.6.1.3 Alpha here with upload disabled (don't want to upload results with a little-tested version yet) and ITERATE finally seems to work nicely. Thanks! -- Rednaxela

Rednaxela: Thanks. You are welcome! =) It seems like it never worked the way it should, and the bug was hard to detect. I am very happy that we finally got rid of this bug! :-) --Fnl

java.io.IOException: Server returned HTTP response code: 500 for URL: http://rumble.fervir.com/rumble/UploadedResults
Unable to upload results roborumble,35,800x600,lRem,1226703762889,SERVER vuen.Fractal 0.55,3027,1790,17 gm.GaetanoA 2.15,2457,1202,18
java.io.IOException: Server returned HTTP response code: 500 for URL: http://rumble.fervir.com/rumble/UploadedResults
Unable to upload results roborumble,35,80 0x600,lRem,1226703810574,SERVER stefw.Tigger 0.0.23,3960,1811,30 amk.ChumbaMini 0.2,1765,1384,5
This happens most time, as I watched approximately 2 battles in 10 were OK... --lRem

The rumble.fervir.com server has been kind of broken for a while. You'd want to point your client at [this URL] (See [here] for more information). There's so new fanciness in that server too :) -- Rednaxela

Wow, missed that. Anyways, the testwiki is the live one now, yes? --lRem

As I understand, it's where new content should be going and it would probably be good if more information could be migrated there probably. Personally I still check both wikis and I think most people do, so conversation on this wiki is still probably fine I'd think. --Rednaxela


(moved from main page) Hi, When I'm running the melee roborumble it seems to be fighting the bots 1 on 1. it declares the 2 bots fighting and then the 1st and second place winner.. Can someone confirm my bot "justin.MallaisV08? (and others) are actually fighting melee (10 bots) 1000*1000 bf.... there also seems to be alot of errors while running meleeRumble ,,, is that normal?? Thanks guys -justin (jlm0924)

I just updated my version of robocode to latest (1.7.1 I believe) and entered my "name"... I didn't change the server info... if you can post where or how to copy the output from I will.. though I am late for school and will have to do first thing when I return this evening... -justin

The exact same problem happened to me (including the "2 bots fighting"). Changed to 1.6.0 and all fixed :) --Starrynte

cool ... :) well I repackaged my bot for repository and reinstalled 1.5.4 ... The getGraphics method isn't supported in 1.5.4 :P .. It took a few tries but I believe my bot is up and competing.. :)

DarkcanuckCan?, can you un-suspend my uploads now pls, and confirm the proper url to use? thx - justin

justin, I unsuspended you yesterday once the 1.7.1 rejection was working. There are three URLs: UPDATEBOTSURL, RESULTSURL and RATINGS.URL should all point to http://darkcanuck.net/rumble/... instead of fervir.com. I would suggest running your client with "UPLOAD=NOT" first to test that everything is working ok. Your bot seems to be doing well! --Darkcanuck

Thxs DarkCanuck? :) and ya I'm pleased my bot is holding up :)


Robo Home | RoboRumble | Changes | Preferences | AllPages
Edit text of this page | View other revisions
Last edited April 8, 2009 0:32 EST by CPE0011d8d68637-CM001ceacce44e.cpe.net.cable.rogers.com (diff)
Search: