Wednesday, 2007-02-28

*** tpb has joined #tp00:00
*** ChanServ sets mode: +o tpb00:00
*** mithro has joined #tp00:32
mithrohowdy people00:42
mithronash: btw, tpserver-cpp has the following bug00:43
nashheyo00:43
nashmithro: Nothing hey?00:43
mithro[18:53] <mithro> well the following bug exists, Ship A (single frigate) is 1 Turn away from Planet it's trying to colonise, Enemy Ship B is 1 Turn away from defending the Planet00:44
mithro[18:53] <mithro> What will happen will depend on the object ids,00:44
mithro[18:53] <mithro>  if Ship A is a lower id, it will arrive and colonise the planet, Enemy Ship B arrives and has combat with an Enemy Planet00:44
mithro[18:53] <mithro>  if Enemy Ship B is a lower id, it will arrive and see an empty planet, then it will have combat with Ship A00:44
mithronash: hopefully Lee will fix very soon (just after 0.4.1)00:44
mithrois there a place we can get your AI to see if we can reproduce your bug?00:44
nashIf you like00:45
nashYou'll need to install evas, ecore, eet and edje to run it first however00:45
nashTell me when you've got them, and I'll make sure I've got a public git archive set up, or mail it to your or something ;-)00:45
nashThey are all at enlightenment.org (or get-e.org)00:45
mithronash: do you need the development versions of the versions in debian/ubuntu okay?00:45
nashSorry?00:46
tpbhmm... Sorry is riding home from work, nash00:46
* nash kicsk tpb00:46
nashThere aren't any in debian officially at the moment :-(00:46
nashHowever there are some unofficial packages floating around - they should be fine if < 6 months old00:47
nashOlder will probably work - although some colours may be farked00:47
mithroIE whats the minimum version of your evas, ecore, eet and edje you need00:47
mithro?00:47
nash0.99.026 I believe...00:48
mithrodoes anyone use Enlightenment anymore?00:49
mithroi use to use E back on redhat 5.200:50
* nash uses E00:55
nashEnough people use E00:55
nashE17 will be released this year00:55
* nash will beat carsten otherwise00:55
mithronash: now here is a question, can you compile your AI without the GUI?00:56
mithrowhat advantages of using E over a normal window manager00:57
nashNot currently00:57
nashmithro: Fast00:57
nashConfigurable00:57
nashsmall memory usage00:57
nashPretty00:57
mithroout of intrest, what does a window manager do apart from move around windows?01:00
nashThese days... lots01:00
nashIt includes all your docs and other crud as well don't forget01:00
mithroIE?01:01
nashMultiple desktops01:01
nashLaunch apps01:01
nashclocks01:01
nashThats what mine is doing now01:01
* nash points mithro to carstens talk at LCA01:01
mithroclocks!? why does a window manager need a clock?01:01
mithroguess i'll have to watch it then01:02
nashmithro: What is on your desktop now?01:03
nashDo you have a clock as part of gnome or kde or something?01:03
mithrognome01:03
nashSo all the crud (except applications) that come with gnome e replaces01:03
mithroas part of gnome on the panel01:03
nashE has it's own panel to replace gnome-panel01:04
mithroso why does a window manager do panels? why not have a seperate application?01:04
nashAnother time.  But they need to talk lots anyway, and they are part of the E project.01:05
nashAs is gnome-panel part of the gnome project01:05
mithronash: I think it's probably philosophy difference - I prefer the do small amount of things well01:08
mithronash: on a side note, I only need the E libs right?01:08
nashAs does E.  Anyway01:08
nashThats correct01:09
mithronot need to get the applications?01:09
nashyou don't need to run E01:09
nashNo.01:09
mithrookay, getting E now, might be a bit lagged01:09
nashedje, ecore, evas, eet should be enough01:10
nashembyryo as well sorry01:10
nashedje needs that01:10
mithrowish cvs had a progress dialog01:13
nashheh01:17
nashI wish e would finally stop using cvs... but thats another story01:17
nashmy company will have git mirrors up soon publically accessable01:18
mithronash: can/will you be able to compile your AI without the GUI?01:21
mithrohave to move now01:22
mithrobblr01:22
nashI can make it so it runs without gui (it did for a while, but I remvoed it)01:22
nashokay01:22
nashNot sure if there is anything to be gained to compile without evas support - I still rely on ecore for events and communication01:26
*** mithro has quit IRC01:49
*** mithro has joined #tp02:07
mithronash: ping?02:10
nashpong02:44
nashmithro:02:44
nashpong02:44
mithrohttp://www.thousandparsec.net/tp/comp.php#criteria02:47
tpb<http://ln-s.net/JP:> (at www.thousandparsec.net)02:47
nashmithro: So?02:49
tpbRumor has it mithro: So is there any update on the homeworld issue - is there an official way of doing it, nash02:49
* mithro points to the -> Usability, how easy it is to setup and get the client running on a standard Ubuntu box. 02:50
nashI can assure you my code is readable, extendable and simple.02:50
mithronash: can you log a bug report for tpb "He is really annoying"?02:50
nashWho runs it?02:51
mithroi guess 3 out of 4 aint bad :)02:51
nashAnd the other is "download some software"02:51
nashPfft02:51
nashA few CVS downloads should not be hard02:51
mithromight not be hard, but harder then an apt-get02:53
mithrobut at the moment your competition isn't even up to the other parts I'm guessing02:53
mithroI think i need to pimp harder02:54
nashSo idd you install the libs yet?02:55
nashmithro: Any luck installing E17 libs then?02:59
mithronash: still downloading02:59
mithrocurrently on the equiv to a 5k modem link03:00
nashright...03:00
* nash notes apt-get would also be painful on this link03:00
nashmithro: What version of ubuntu do you use?03:01
* nash points mithro to: http://www1.get-e.org/Main/News/_articles/365.html03:02
tpb<http://ln-s.net/JPB> (at www1.get-e.org)03:02
mithroedgy eft03:03
nashSee the link then03:03
nashWho runs tpb?  Can someone just get op and ban him?03:04
mithroi run tpb03:06
nashThen kick the damn thing03:06
mithrohe does logging and stuff03:06
mithroi just need to turn of the ?03:06
mithrobit I find the link and the other stuff useful03:07
nashThought you may,03:07
nashMakes installation on a standard ubuntu box pretty simple ;-)03:07
mithros/of/off03:07
mithronash: looks that way :)03:07
mithronash: the sf bug stuff doesn't appear to be working either - need to figure that out03:08
mithro~sf bug 1885603:08
tpbI don't have a clue, mithro.03:08
nashheh03:08
mithro~help sf03:08
tpbmithro: Error: There is no command "sf".03:08
nash~help03:08
tpbnash: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.03:08
nash~help plugins03:09
tpbnash: Error: There is no command "plugins".03:09
nash~help plugin03:09
tpbnash: (plugin <command>) -- Returns the plugin(s) that <command> is in.03:09
mithro~help sourceforge03:09
tpbmithro: Error: There is no command "sourceforge".03:09
nash~help plugin sf03:09
tpbnash: Error: There is no command "plugin sf".03:09
mithroahh, well so much to do so little time03:09
nash~help lsit03:09
tpbnash: Error: There is no command "lsit".03:09
nash~help ls03:09
tpbnash: Error: There is no command "ls".03:09
nash~help list03:09
tpbnash: (list [--private] [<plugin>]) -- Lists the commands available in the given plugin. If no plugin is given, lists the public plugins available. If --private is given, lists the private plugins.03:09
nash~list03:09
tpbnash: Admin, Alias, Anonymous, AutoMode, Babelfish, BadWords, Channel, ChannelLogger, ChannelRelay, ChannelStats, Config, Ctcp, Currency, CyborgName, Darcs, Dcc, Dict, Dunno, Ebay, Factoids, Filter, Format, Freenode, Freshmeat, Gameknot, Games, Geekquote, Grasshoppaz, Herald, IgnoreFormatting, Infobot, Insult, Internet, Karma, Lart, Later, Limiter, Linux, Lookup, Markov, Math, Memo, Minstrel, Misc, MoobotFactoids, (1 more message)03:09
nashTheres a fwe there03:09
nashtpb:03:10
nash~list03:10
tpbnash: Admin, Alias, Anonymous, AutoMode, Babelfish, BadWords, Channel, ChannelLogger, ChannelRelay, ChannelStats, Config, Ctcp, Currency, CyborgName, Darcs, Dcc, Dict, Dunno, Ebay, Factoids, Filter, Format, Freenode, Freshmeat, Gameknot, Games, Geekquote, Grasshoppaz, Herald, IgnoreFormatting, Infobot, Insult, Internet, Karma, Lart, Later, Limiter, Linux, Lookup, Markov, Math, Memo, Minstrel, Misc, MoobotFactoids, (1 more message)03:10
nash~list s*03:10
tpbnash: Error: 's*' is not a valid plugin.03:10
nash~list s03:10
tpbnash: Error: 's' is not a valid plugin.03:10
nash~more03:10
tpbnash: Network, News, NickCapture, Nickometer, Note, OSU, Owner, Plugin, Poll, Postman, Praise, Protector, Python, Quote, QuoteGrabs, RSS, RawLogger, Relay, Reply, Rhyme, RootWarner, Scheduler, Seen, Services, ShrinkUrl, Sourceforge, Status, Stock, String, Success, Supybot, Tail, Time, Todo, Topic, TracBot, URL, Unix, UrbanDict, User, Utilities, Vim, Weather, Web, Webserver, WordStats, and Zipinfo03:10
nash~list Sourceforge03:10
tpbnash: bug, bugs, fight, patch, rfe, rfes, sfstats, total, and tracker03:10
mithro~total03:10
tpbmithro: (total {bugs,todo} [<project>]) -- Returns the total count of open bugs or todo. <project> is only necessary if a default project is not set.03:10
mithro~total bugs03:10
tpbmithro: Error: Could not find bug statistics for thousandparsec.03:10
nash~Sourceforge bug 1885603:11
tpbnash: Error: Invalid Tracker page snarfed: http://sourceforge.net/support/tracker.php?aid=1885603:11
mithro~help bug03:11
tpbmithro: (bug <id>) -- Returns a description of the bug with id <id>. Really, this is just a wrapper for the tracker command; it won't even complain if the <id> you give isn't a bug.03:11
nash~Sourceforge bug 166884903:11
tpbnash: Bug #1668849: Ships sometimes fly into space with colonisation orders <http://sourceforge.net/support/tracker.php?aid=1668849>03:11
nash~Rhyme03:13
tpbI don't know, nash.03:13
nash~help Rhyme03:13
tpbnash: Error: There is no command "rhyme".03:13
nash~list Rhyme03:13
tpbnash: rhymes03:13
nash~rhymes pants03:13
tpbnash: 30 rhymes for pants: ants, aunts, bantz, chants, frantz, gantz, glantz, grants, grantz, granz, hantz, implants, jantz, kantz, krantz, lantz, mantz, mccants, nantes, nantz, plants, plantz, rantz, schantz, schrantz, shantz, transplants, tschantz, yantz, and zaentz03:13
nash~rhymes orange03:13
tpbnash: No rhymes found.03:13
nash~rhymes mithro03:14
tpbnash: No rhymes found.03:14
nash~rhymes nash03:14
tpbnash: 48 rhymes for nash: abash, asch, asche, ash, ashe, basch, bash, brasch, brash, cache, cash, clash, crash, dasch, dash, flasch, flash, frasch, furash, gash, gnash, guasch, hasch, hash, kasch, kash, lasch, lash, lukash, masch, mash, noncash, palash, pash, precrash, rasch, rasche, rash, rehash, sash, slash, smash, splash, stash, tasch, tash, thrash, and trash03:14
mithro~rhymes nash03:14
tpbmithro: 48 rhymes for nash: abash, asch, asche, ash, ashe, basch, bash, brasch, brash, cache, cash, clash, crash, dasch, dash, flasch, flash, frasch, furash, gash, gnash, guasch, hasch, hash, kasch, kash, lasch, lash, lukash, masch, mash, noncash, palash, pash, precrash, rasch, rasche, rash, rehash, sash, slash, smash, splash, stash, tasch, tash, thrash, and trash03:14
mithro~rhymes lee03:14
tpbmithro: 401 rhymes for lee: ab, abductee, abee, absentee, abt, ac, achee, acree, addressee, adee, adoptee, adoree, agee, agree, ailee, aimee, akey, albee, allee, amc, amputee, andree, apc, appointee, aretsky, artsy, askey, atp, attendee, awb, b, baldree, banshee, barbaree, barbee, bbc, be, bea, beaudry, bee, benami, bibbee, bibee, bibi, blea, bouchey, bouie, bouley, bourgeoisie, branitzky, brea, bree, bresee, brie, (6 more messages)03:14
nashCache does not rhyme with nash.03:14
mithroif you pronouce cache like cash03:15
mithrolee the absentee amputee abductee03:16
mithroand other rhymes to make your ears bleed03:16
mithroit apepars that e17 deb stuff doesn't exist anymore03:17
nashHmm03:19
nashgive me a moment03:19
mithrobrb03:19
jothama lot of people pronounce it cash,03:24
jothamboo, the dictionary even says it's rhymed with cash03:24
jothami pronounced it differently :\03:25
jothamguess i better americanise myself03:25
nashDoesn't make it right03:27
mithroc-aye-sh03:29
* mithro slowly fixes electrical check error03:30
mithros03:30
jothami can't do the a with a /\ over it in this terminal03:30
jothambut that's the sound the dictionary wants me to use (like pat)03:30
jothami usually say c-aye-sh03:31
jothami think there are a few things worth adopting from american english though (much to my mothers horror)03:36
jothamone example is oriented instead of orientated03:36
jothamand i think color instead of colour is more asthetically pleasing, but i can't be drawn in by the american use of z instead of s03:36
* mithro wishes the autorouter wasn't so dumb :/03:48
jothamwish my nvidia drivers didn't keep crashing03:58
mithrobblr04:00
mithrogoing to get food04:00
nashhave fun04:01
*** mithro has quit IRC04:20
*** nash has quit IRC05:27
*** tpb has joined #tp06:00
*** ChanServ sets mode: +o tpb06:00
*** mithro has joined #tp06:08
mithrohowdy people06:10
mithro~seen nash06:10
tpbmithro: nash was last seen in #tp 2 hours, 8 minutes, and 39 seconds ago: <nash> have fun06:10
jasminenash was seen 2 hours, 8 minutes and 39 seconds ago.06:10
jothamre mithro06:11
mithrohowdy jotham06:12
mithrohow goes the battle viewer?06:12
jothamgood will have a patch tonight hopefully, if the cold doesn't get me before i get it done (its making me tireder than normal)06:13
jothamand hopefully a mockup of the result screen for you to give your OK to before i build it06:14
jothamhave got oceampgui's event manager etc in now06:14
jothamheh tireder <- my mother would kick my ass for using such english06:14
jothami don't know what to do about the fact that different battles will require different sized screens, just prescan the battle events and determin their maximum movement rect or something i guess06:15
mithrojotham: well none of the battles actually have any movement stuff in them06:16
jothami mean for their layout06:16
jothamthere is a movement comamnd when they go into position06:16
jothambefore that they don't have a coordinate as far as the viewer is concerned06:16
jothamscrolling would be better, but that would involve a minimap etc, which should be done in another phase of the project06:16
mithrojotham: yes, but none of the generated battle data has any move commands06:16
jothamoh i see, yeah just checked the xml, i haven't got them to run yet, just looked at them last night and saw they wouldn't run until the parser and the viewer were modified06:17
jothamok so i should lay out the ships using some asthetically pleasing algo?06:18
jothamor is there some kind of specific method that you had in mind06:18
jothambrb dinner06:24
mithrojotham: asthetically pleasing is good06:27
*** llnz has joined #tp06:29
llnzdinner, bbs06:41
mithrohey llnz06:54
*** llnz has quit IRC07:15
*** llnz has joined #tp07:18
*** nash has joined #tp07:23
* nash waves07:24
llnzhi07:36
llnzback07:36
*** llnz2 has joined #tp07:44
*** llnz has quit IRC07:45
*** llnz2 is now known as llnz07:45
*** llnz2 has joined #tp07:52
*** llnz has quit IRC07:53
*** llnz2 is now known as llnz07:53
llnzgrrr... unstable adsl07:54
nashheyo07:59
llnzany luck yet?08:00
nashQuick question - if the client sends extra terminating nil characters in a message for a string (and includes them in the len) what happens?08:00
nashllnz: With what?08:00
llnzto answer you first question: they disappear from the string in the C++ server, tpserver-py would preserve them though08:01
llnzthe second is the colonising fleet flying away from target bug08:01
llnzneither mithro or i have been able to replicate it yet08:02
nashHow have you been trying?08:03
llnzusing tpclient-pywx08:03
nashAll I need to do is set a fleet of a single frigate to colonise a world (>1 turn away) with a move order.08:03
nashSo move(coords of planet), colonise(planet)08:03
nashGet yourself a version of evas, edje and ecore and I'll give you client code...08:04
llnzdoes it work on any planet?08:05
nashSometimes08:05
nashSeems to be longer trips that cause the problem08:05
nashAlso when there are multiple ships - never seen it with 108:05
llnzand it occurs the turn after the move order finishes?08:09
nashyes08:10
nashI get the move complete message08:10
mithronash: i have evas and stuff installed08:17
mithrodo you have somewhere I can get the code?08:18
nashSource is not yet open, jsut for the record... It will be, but not until I'm happy with it08:19
nashDo you want a git archive or a tar.gz?08:20
mithronash: either is fine git archive in tar.gz? :)08:21
jothamproblem with inheritance is it's a bitch to find things in other peoples libraries sometimes08:22
mithrojotham: abc.__file__ ?08:22
jothamnaah just looking where something is defined, i found it but while looking found this one class has 5 levels of ancestory08:22
jothameach level containing quite a few functions08:23
jothamgrep gave me simply what i was after, but after piecing togeather how the two files were associated i was surprised how convoluted it was :p08:23
nashmithro: Got it?08:32
mithronash: not quite yet08:33
mithrobloody rate shaping :/08:34
nashmithro: I'll pack it next time08:35
mithrollnz: so do we only have nash's bug too fix before 0.4.1?08:36
llnzyeah08:37
mithronash: getting at about 4k/s08:38
JLPmorning all08:39
nashmithro: I can repack to reduce the size - but you'll need to restart08:40
nashhow much do you have?08:40
mithroopps just finished then08:42
mithrohey JLP08:42
llnzhi JLP08:42
nashThats okay - you have my entire history ;-)08:42
mithrohow do you find git?08:43
nashBrillient08:43
JLPyay I finally got the openSUSE build account, now let's se how it works08:43
mithrohow do I use the client?08:43
nashI was using arch until a month or so ago... moved to git, and it works perfectly08:44
mithrogive it a url?08:44
nashNot yet.08:44
nashopen tpe_gui.c08:44
nashJump to line 157, and uncomment one of the comm_connect lines08:45
nashOr it should be set up to use my machine by default08:45
nashClick on the green square to connect08:45
mithrohrm isn't doing anything08:46
nashHave a light grey window?08:46
mithrono window or anything08:46
nashSo you run it, and you have nothing?08:47
mithroyes08:47
mithroi'm using the following line ->tpe_comm_connect(tpe->comm, "localhost", 6923, "nash", "password");08:47
* nash bets he knows08:47
mithrort_sigaction(SIGTERM, {0xb7ef0990, [], SA_RESTART|SA_SIGINFO}, NULL, 8) = 008:47
mithrogettimeofday({1172652588, 888681}, NULL) = 008:47
mithroselect(1, [], [], [], NULL08:47
mithroit's stop in a select08:47
nashls /usr/local/lib08:48
nashIs there a lib ecore_evas there?08:48
mithrotim@ultraslim:~/nash/eclient$ ls /usr/lib/libecore_evas.so -l08:49
mithrolrwxrwxrwx 1 root root 22 2007-02-28 19:07 /usr/lib/libecore_evas.so -> libecore_evas.so.1.0.008:49
nashHmm..08:49
nashcg-update master08:49
nashThen make it again08:49
mithrotim@ultraslim:~/nash/eclient$ cg-update master08:50
mithroUpdating from a local branch.08:50
mithroApplying changes...08:50
mithroBranch already fully merged.08:50
nashSorry - cg-update origin08:50
nashActually cg-update should jsut work for you08:50
mithroecore_event_handler_add(82, 0x804e770, 0x8057c08, 0xb7eb1703, 118529)        = 0x805813008:50
mithroecore_main_loop_begin(0x8056888, 48, 0xbfa8eac8, 0x804f019, 0xb7bc18b608:50
nashls /usr/local/lib/evas/modules/engines/08:50
mithrotim@ultraslim:~/nash/eclient$ ./tpe08:51
mithroCOuld not create ecore_evas_xll.08:51
mithroCheck you built evas and ecore with x11 support08:51
mithro:)08:51
nashThat was what I just added08:51
nashDo that ls - see if there is a software_x11 directory08:51
mithrono engines are installed08:52
mithroinstalling now08:52
mithroseems to be running08:55
nashGot a light grey window?08:56
mithrooh no, it's alive!?08:56
jothamdo ships always have one kind of weapon, or is there the potentual for a technology tree08:56
mithrookay managed to get the bug08:56
nashNot hard ;-)08:56
mithrojotham: ships will always have 1 type of weapon08:57
mithro2007-02-28 19:28:39 < Debug > Object(13)->updatePosition(): Moving object to [-249954075, -1505066966, 0] (which is the destination? No)08:57
mithro2007-02-28 19:28:39 < Debug > Object(13)->updatePosition(): Velocity is now [0, -200000000, 0]08:57
nashmithro: What? Only one weapon?  How the hell are we supposed to play stars?08:57
mithrothats the bug right?08:57
mithronash: step by step :)08:57
jothamplay stars?08:57
nashmithro: If you click on a object, it dumps some info to stdout08:57
mithro2007-02-28 19:28:39 < Debug > Object(13)->Move->doOrder(): Moving 2694967296 at 200000000 speed (will take about 13 turns)08:58
mithro2007-02-28 19:28:39 < Debug > Move->doOrder(13): Velocity is [0, -200000000, 0] (will arrive at [-249954075, -1505066966, 0])08:58
mithrojotham: Stars! is the game which nash wants to clone with Thousand Parsec08:58
jothamahh08:58
mithrojotham: for now, assume ships only have 1 weapon08:59
nashOne of the best 4X games around.  A little technical, but fairly well blaanced (CAs excluded) and quite a few viable strategies for advanced players08:59
jothamcool08:59
jothamonly game i really play a lot of is openttd08:59
nashmithro: Speed wise - what to you think?08:59
mithronash: seems okay09:00
mithronash: no real noticable "speed"09:00
* nash dares mithro to compare to the released python client... for connecting09:00
nashI test with 1000 odd systems09:00
mithronash: it's not like you see anything on the screen when it updates :)09:01
nashYeah.. but I don't need a progress bar ;-)09:02
mithrobut I can't go "wow that progress bar is so fast :)"09:06
nashIt's a bit strange over a slow link - the stars popup in groups...09:07
mithrothe fonts are a bit funny09:10
nash?09:10
nashWrite a new theme ;-)09:11
nashIt's vera09:11
mithroI think i know whats causing this09:11
nashAlso there is a small shadow?09:11
nashclient issue?09:11
mithronash: the move issue09:13
* nash wonders if mithro is going to keep us in suspense for ever...09:13
mithronash: yes!?09:13
* llnz returns from being afk09:13
jothamdo either of you use intern() (built in) in python?09:15
jothameither/any09:15
nashjotham: Every non-trivial python program I've ever written has contained at least 5 intern calls09:15
* nash wonders what it does09:15
jothamoh...i was about to ask if i was right in understanding it as i did09:16
* nash shuts up and lets someone answer the question seriously09:16
jothami believe it adds the string you intern to a table and returns a pointer, so if you use it for dictionary keys its faster than rehashing/comparing all the time09:16
nashAhh...09:16
* nash would use that all the time09:16
jothamwas just looking for confirmation09:16
* nash uses a similar system in c09:17
llnzjotham: from what i know of python, that sounds about right09:17
llnzjava has a similar system09:17
llnz(and a similar name for it)09:17
jothami encountered it for the first time a few days ago09:17
llnzit also reduces memory usage when a string is repeated a lot09:17
nashIn C, it also changes strcmp to ==, which is nice09:18
nashuse it three times you'll generally save09:19
jothamthat's cool nash09:19
* llnz waits for mithro to start dropping clues so he can help09:20
mithrollnz: i might be going off in the wrong direction09:20
* nash is wondering if mithro has deduced the problem is "there is a bug"09:20
mithrohrm09:22
mithro2007-02-28 19:54:36 < Debug > IGObject::setNumOrders09:22
mithro2007-02-28 19:54:36 < Debug > Object(67)->Colonise->doOrder(): Target id is 6709:22
mithro2007-02-28 19:54:36 < Debug > Object(67)->Colonise->doOrder(): Current position is [1298040000, -1646002296, 0]09:22
mithro2007-02-28 19:54:36 < Debug > Object(67)->Colonise->doOrder(): Target position is [1298040000, 2648965000, 0]09:22
llnzwtf?09:23
* llnz thinks of something to check09:23
mithrothose print outs are wrong09:23
llnzoh?09:24
* nash notes that indeed seems to be the case09:25
mithrollnz: the Object(67) is wrong09:25
nashIt;s a pretty big warp of one coord - exactly what we I am seeing09:25
nashIt fits perfectly09:26
mithrobut the coordinates problem is exactly the effect09:26
mithrojust get you the right details09:26
mithro2007-02-28 19:59:24 < Debug > Object(73)->Colonise->doOrder(): Target id is 909:27
mithro2007-02-28 19:59:24 < Debug > Object(73)->Colonise->doOrder(): Current position is [-249954075, 294933034, 0]09:27
mithro2007-02-28 19:59:24 < Debug > Object(73)->Colonise->doOrder(): Target position is [-249954075, -4000034262, 0]09:27
mithro2007-02-28 19:59:24 < Debug > Object(73)->Move->doOrder(): Moving 4294967296 at 200000000 speed (will take about 21 turns)09:27
mithro2007-02-28 19:59:24 < Debug > Move->doOrder(73): Velocity is [0, -200000000, 0] (will arrive at [-249954075, 94933034, 0])09:27
mithrollnz: want me to commit the logging?09:28
mithroi have to head to dinner now09:28
llnzsure09:28
nashbe back later09:30
nashheading for dinner09:30
mithrollnz: when you fix up the order processing, 90% of the colonise doOrder goes away09:30
CIA-20[email protected] * tpserver-cpp/modules/games/minisec/ (8 files): Extra logging to find Nash's bug.09:30
llnzwhy>09:30
mithroanyway off to eat steak, will explain when I get back09:30
llnzwhy does is "go away"?09:30
llnzok09:30
* llnz notes bug in IGObject::updatePosition()09:32
jothambath, bbl09:35
*** tpb has joined #tp09:54
*** ChanServ sets mode: +o tpb09:54
mithro2007-02-28 20:26:25 < Debug > Object(60)->Colonise->doOrder(): Target id is 3409:54
mithro2007-02-28 20:26:25 < Debug > Object(60)->Colonise->doOrder(): Current position is [-1623927296, 474965000, 0]09:54
mithro2007-02-28 20:26:25 < Debug > Object(60)->Colonise->doOrder(): Target position is [2671040000, 474965000, 0]09:54
mithro2007-02-28 20:26:25 < Debug > Object(60)->Move->doOrder(): Moving 4294967296 at 200000000 speed (will take about 21 turns)09:54
mithro2007-02-28 20:26:25 < Debug > Move->doOrder(60): Velocity is [200000000, 0, 0] (will arrive at [-1423927296, 474965000, 0])09:54
llnzselect posx,posy,posz from object where objectid=60;09:54
mithro2007-02-28 20:26:25 < Debug > Object(60)->updatePosition(): Moving object to [-1423927296, 474965000, 0] (which is the destination? No)09:55
mithro2007-02-28 20:26:25 < Debug > Object(60)->updatePosition(): Velocity is now [200000000, 0, 0]09:55
mithroopps don't have mysql running09:56
* llnz thinks of a way to try locally and quickly09:58
mithro2007-02-28 20:29:53 < Debug > Object(84)->Colonise->doOrder(): Target id is 6709:58
mithro2007-02-28 20:29:53 < Debug > Object(84)->Colonise->doOrder(): Current position is [-1289880000, -719072296, 0]09:58
mithro2007-02-28 20:29:53 < Debug > Object(84)->Colonise->doOrder(): Target position is [-1289880000, 3575895000, 0]09:58
mithroObject ID 84 Type 4 Name: 'A Fleet'09:59
mithro        Size 209:59
mithro        { x = -1289880000, y = -719072296, z = 0 }09:59
mithro        { dx = 0, dy = 0, dz = 0 }09:59
mithro        0 children:09:59
mithro        5 orders: 3 1 0 4 509:59
mithro        1 current orders09:59
mithro        OrderSlot 0 Type[ 3]: Colonise09:59
mithro        Turns: 22  Resources: 009:59
mithroColonising object 67:09:59
mithroObject ID 67 Type 3 Name: 'Desiderata 3'09:59
mithro        Size 209:59
mithro        { x = -1289880000, y = -719072296, z = 0 }09:59
mithro        { dx = 0, dy = 0, dz = 0 }09:59
mithro        0 children:09:59
mithro        0 orders:09:59
mithro        0 current orders09:59
mithroEnd col09:59
mithroso it's something strange with the target position09:59
*** tpb has joined #tp10:06
*** ChanServ sets mode: +o tpb10:06
mithrollnz: dunno, and it's not always the same component either10:07
llnzit's always a negative -> positive change though....10:08
mithrosomewhere it's getting into a UInt32 i think10:09
llnzmust be before line 91 of move.cpp10:14
mithrohey JLP10:16
mithroit's not move.cpp10:16
mithroit's on   IGObject* target = Game::getGame()->getObjectManager()->getObject(object->getObjectId());10:16
mithros/on/in10:16
llnzcan you be any more specific?10:19
llnzObjectManager::getObject() looks fine10:19
mithrowell, it's wrong when it comes out of that10:20
llnzis it right before it comes out of that?10:22
mithrodunno10:22
mithrollnz: I'm guessing it's effecting all positions, just most of the time the values is less then Int32?10:30
llnzmaybe10:31
llnzi could be that this bug ony affects x86, not my x68-6410:31
mithropossibly on x86-64 are all ints 64 bit?10:32
llnzyes10:32
llnzbut all the position code uses long long, and shouldn't be a problem....10:33
llnzdoesn't appear to affect me here10:34
* llnz which could be why i hadn't noticed it10:34
llnzthere is one unsigned long long in Vector3d, if i made it signed, the universe could be no more than 3,037,000,500 units in diameter10:38
llnz3e13 Km (roughtly, at current scale)10:39
mithrohrm...10:39
mithroI don't think that has anything to do with it?10:41
llnzneither, but it's the only unsigned int near the code at all10:41
mithroi think it's probably an int?10:41
llnztry replacing all the long long in Vector3d (except the unsigned long long) with int64_t10:42
* llnz notes the the unsigned long long makes the max diameter 4,294,967,29610:43
llnznot really that much bigger10:43
mithrocould it be in persistance somewhere?10:44
llnzvector3d.h will need #include <stdin.h> at the top10:45
llnzshouldn't be10:45
llnzdata only comes out of persistence after a server shutdown10:45
mithro2007-02-28 21:19:28 < Debug > Object(61)->Colonise->doOrder(): Current position is [-1130887296, 1325897296, 0]10:47
mithro2007-02-28 21:19:28 < Debug > Object(61)->Colonise->doOrder(): Target position is [3164080000, -2969070000, 0]10:47
* nash has returned10:48
*** llnz2 has joined #tp10:50
*** llnz has quit IRC10:50
nashSo I see you guys are closing in...10:50
*** llnz2 is now known as llnz10:50
llnz <llnz> data only comes out of persistence after a server shutdown10:50
llnz[23:48] <llnz> posx BIGINT NOT NULL, "10:50
llnz[23:48] <llnz>                     "posy BIGINT NOT NULL, posz BIGINT NOT NULL10:50
mithrollnz: not using tpmysql10:51
llnzthere the object never touchs persitence10:51
nashI am not using persistence on the server10:51
nash(or the client)10:53
mithronash: we could use another set of eyes10:56
llnz> <llnz> try replacing all the long long in Vector3d (except the unsigned long long) with int64_t10:56
llnzmaybe that will help?10:57
mithrollnz: nope10:57
nashmithro: Looking now10:57
llnzhumm...10:57
mithrois there a way to get C++ to "trap" when you overflow or similar?10:57
mithrollnz: making all int64 and doing clean rebuild11:00
llnznot that i know of without getting into assembler11:00
llnzcool11:00
mithroobjectmanager.h:    std::set<uint32_t> getObjectsByPos(const Vector3d & pos, uint64_t r);11:02
mithroobjectmanager.h:    std::set<uint32_t> getContainerByPos(const Vector3d & pos);11:02
mithroobjectmanager.h:    std::set<uint32_t> getAllIds();11:02
mithro?11:02
mithronope, was reading that wrong11:02
mithrollnz: made no difference11:04
llnzhummm....11:05
*** jotham has joined #tp11:07
nashJust to confirm - it's just a single thread right?11:15
llnzyes11:15
llnzthe only thing i can think of is that the guile GC is stuffing around where it shouldn't11:15
llnzbut it shouldn't be active11:16
llnzsomething has zero'd the top 32 bits of the value11:16
nashIs guile being called at this point?11:16
llnznope11:16
llnzonly in design validating11:16
* nash notes that there is an active GC when he tries to valgrind the program11:17
llnzhumm...11:17
llnzpthread is compiled in....11:17
llnzand i think guile requires is.....11:17
mithro2007-02-28 21:48:44 < Debug > Object(76)->updatePosition(): Moving object to [1714007296, 380932296, 0] (which is the destination? Yes)11:17
mithro2007-02-28 21:48:44 < Debug > Object(76)->updatePosition(): Velocity is now [0, 0, 0]11:17
mithro2007-02-28 21:48:44 < Debug > Container object 111:17
mithro2007-02-28 21:48:44 < Debug > Object(76)->updatePosition(): Position is now [1714007296, 380932296, 0]11:17
nashRunning this under valgrind, the GC segfaults11:20
llnzhehe, odd11:21
llnznash: what distro are you on?11:21
nashdebian11:21
mithro2007-02-28 21:50:14 < Debug > Object(35)->updatePosition(): Moving object to [-2580960000, -3914035000, 0] (which is the destination? Yes)11:21
mithro11:21
mithro2007-02-28 21:50:14 < Debug > Object(76)->updatePosition(): Moving object to [441215092, -891859908, 0] (which is the destination? No)11:21
mithro2007-02-28 21:50:14 < Debug > Object(76)->updatePosition(): Velocity is now [-141421356, -141421356, 0]11:21
mithro2007-02-28 21:50:14 < Debug > Container object 111:21
mithro2007-02-28 21:50:14 < Debug > Object(76)->updatePosition(): Position is now [441215092, -891859908, 0]11:21
mithro11:21
mithro2007-02-28 21:50:24 < Debug > Object(76)->Colonise->doOrder(): Target id is 35 (35)11:21
mithro2007-02-28 21:50:24 < Debug > Object(76)->Colonise->doOrder(): Current position is [441215092, -891859908, 0]11:21
mithro2007-02-28 21:50:24 < Debug > Object(76)->Colonise->doOrder(): Target position is [-2580960000, -3914035000, 0]11:21
mithronash: can we lock the client to only create 1 ship?11:23
nashYeah11:24
nashgive me a minute or two11:24
nashOnly one ever/11:24
nashmithro: Pull that11:25
mithrowtf...11:27
llnzwtf....11:27
*** Demitar has quit IRC11:27
nash?11:28
nashShut up tpb11:28
llnzhumm...11:29
*** Demitar has joined #tp11:30
nashmithro: Yes, no - does that help?11:30
mithronash: rm......11:31
nashNow I'm confused... I've pushed up a patch that forces my AI to only make one ship (ever).  if you cg-update you should get it11:32
nashOtherwise, I have NFI what is going on11:32
mithros/rm/hrm/11:33
mithrostruct vector {11:35
mithrouint64_t x,y,z;11:35
mithro};11:35
mithroumm, nash - that should be int64_t ?11:36
nashProbably11:36
* nash hides11:37
* nash is still confused though - how did that cause that bug? 11:38
* nash thinks he sees how :-/11:42
nashshit11:42
* llnz ponders11:48
mithrollnz: i think your off the hook11:48
nashI belive he is11:48
* nash offers his sincerest apologies11:48
mithronash: i think you owe Lee a cold one ;)11:48
llnzoh, i see, hehe11:48
nashmithro: Or 611:48
* llnz can't find eclient on the category list for bug tracker.....11:49
llnzoh well, i'll just close it11:49
nashPlease do11:49
nashPlace it in the 'nash is an idiot' category11:49
nashDamn it... Can someone plase document the "colonise will also move" thing11:50
nashplease11:50
mithronash: i have a feeling it will go away soon11:54
* llnz starts preparing the release11:54
JLPllnz: consider making the archives bz2 instead of gz11:58
llnzJLP: any hints as to how with autoconf/automake?11:58
JLPllnz: have no idea, i thought the sources are compressed by hand12:02
* llnz just does make dist12:02
nashtypical - I just added code to make orders more generic - and now it doesn't work..12:04
nash2007-02-28 23:13:11 < Debug > isEnoughRemaining, checking for 24, have 412:13
nashThat would cause the the order to fail right?12:13
llnzyes12:13
nashHmm... :-/12:13
nashCan you or mithro try colonizing a world using the py client.12:14
* nash is worried al this debugging has broken somehting12:14
* nash is trying12:16
nashnow12:16
nashTurns too long12:17
llnzon the console:12:18
llnz> turn end12:18
llnzto immediately end the turn12:18
mithrollnz: cool12:19
nashNope - bug at my end12:19
llnz31 line ChangeLog12:20
mithrollnz: so when you rework the order stuff, you can remove the combat initiation - the move stuff, as you will know that combat and move has already occured12:21
llnzahh, cool12:21
llnzso at that time, should colonise not make the fleet move?12:22
mithrollnz: yes12:22
mithroyou just check current position == object position12:22
mithroyou don't want objects to move after the move phase12:23
CIA-20Lee Begg <[email protected]> * tpserver-cpp/tpserver/ (object.cpp object.cpp object.cpp):12:25
CIA-20Safer use of Objects and correct doneWithObject calls:12:25
CIA-20 Just incase the doneWithObject calls do something, like delete the object to make12:25
CIA-20 space in the cache.12:25
mithroand it makes combat simple - you just find all objects at the same location and check if they should do combat12:26
llnzyeah12:27
llnzLee's Dev game12:27
llnzRunning on tpserver-cpp (Version: 0.4.1)12:27
llnz:-)12:28
mithroso colonise doesn't need to worry about anything but colonisation :)12:30
mithroi like your little ships nash :)12:30
nash:-)12:30
mithrohave you tried a vs yet?12:31
nashAnyway - I'm off to bed.12:31
nashOnce again, sorry for the trouble :-(12:31
mithronash: we found plenty of other bugs12:32
nash:-)12:32
nashYeah - but this one was the longest12:32
*** nash has quit IRC12:33
CIA-20Lee Begg <[email protected]> * tpserver-cpp/ (ChangeLog NEWS configure.ac configure.ac): Prep for 0.4.1 release12:36
CIA-20Lee Begg <[email protected]> * tpserver-cpp/: (log message trimmed)12:37
CIA-20Safer use of Objects and correct doneWithObject calls:12:37
CIA-20 Just incase the doneWithObject calls do something, like delete the object to make12:37
CIA-20 space in the cache.12:37
CIA-20]12:37
CIA-20[Extra logging to find Nash's bug.12:37
CIA-20[email protected]**20070228093155]12:37
mithrollnz: darcs can produce xml output12:39
llnzthat is darcs-cia, not what i wrote12:40
mithrollnz: i was talking about the mail stuff12:40
mithroi've been looking in to how tailor works12:41
mithrollnz: if you match on the hash - rather then the comment you wouldn't get the horrible burst when the comment is the same12:42
llnzmithro: i know12:43
llnzi will look at making it better some day12:43
mithroi'll look into modifying the script12:43
llnzyou go for it12:43
mithromight do some more work on getting the svn stuff working first12:44
llnztpserver-cpp 0.4.1 on sf12:47
mithrocool12:48
llnzthe tar.gz is 781kB12:49
mithrois that big?12:56
llnz0.4.0 was 442 KB12:57
CIA-20Lee Begg <[email protected]> * web/ (4 files in 2 dirs): Releasing tpserver-cpp 0.4.112:58
mithrolibltdl?13:00
llnzmostly, there is over 1,000 new lines of C++ too13:00
tpbNew news from http://www.thousandparsec.net/tp/rss.php: TPServer-cpp 0.4.1 released.13:02
llnzhummm.... my file servers data disk appears to be read-only again13:04
llnznot good13:04
mithrono13:05
mithroi should probably be in bed13:08
llnzso should i13:09
mithroso do you have a timeline for the new order stuff? I could help?13:13
llnznot yet, will think about it tomorrow13:15
* llnz wanders off13:19
llnzlater all13:19
*** llnz has quit IRC13:19
*** _JLP_ has joined #tp17:25
*** JLP has quit IRC17:41
*** bobbens has quit IRC20:18
*** bobbens has joined #tp20:18
*** _JLP_ is now known as JLP21:00
*** zipola has quit IRC21:19
*** mithro has quit IRC21:33
*** nash has joined #tp22:04
* nash waves22:04
jothamWhat is a Pirate's favorite thing about Programming?23:03
jothamArgs!23:03
tpbaloril has quit worldforge (purple.worldforge.org blue.worldforge.org)23:03
tpbjasmine has quit worldforge (purple.worldforge.org blue.worldforge.org)23:03
nashAhh... pirate jokes...23:06
nashThe hoy of 439 different jokes...all with the same punchline...23:06
jothamhah23:06
nashDid you hear about the pirate movie?  It was rated 'R'23:07
nashsorry... rate 'Arrg'23:07
nashHow is life?23:08
jothami have the flu bad and couldn't go to work today23:09
nashSounds good23:09
nash:-)23:09
jothammy normal must-work-while-sick trick of 2 painkillers and a huge coffee didn't get me togeather today23:09
jothambut otherwise everythings awesome23:09
* nash only read the last 5 words...23:09
nashLOL23:09
nashSo what are you working on in TP, if you don't mind me asking?23:10
jothama battle viewer in pygame23:10
nashAhh... cool23:10
jothamshould have done it in opengl, but that's another story23:10
* nash mutters about doing it in Evas... but thats another matter ;-)23:10
jothamyeah i read about Evas23:11
* nash is trying to work out the whole status of the project23:11
nashHow is it going?23:12
jothamgood, realised my event management approach was naive and have just reworked it all, it's almost bug free, was a major re-work23:13
jothamonce that's done the rest should be pretty trivial23:13
jothami'd say 40% of the way overall, maybe more once the new event management is working properly23:13
nashSo how long before a release?23:14
jothamas long as work doesn't eat my life (sometimes it is prone to taking me from 9am->11pm regularly) it should be usable/quasi-useful sometime late march23:15
jothami've been getting sick a lot lately which hasn't helped much23:15
nashReally?  I get lots of personal projects done when sick...23:16
jothamwhen you have colds/flus?23:16
jothammy brian turns to mush23:16
* nash can code when drunk, tired, sick... whatever23:17
nashI just need to pick my target23:17
jothamheh cool23:17
jothami make a mess unless i'm sober23:17
nashI write fairly simple code, and always design first - keep the nesting down, keep things as simple as they can be - so even when not in a good state I can write.  When I'm awake and sharp, I tend to design.23:19
jothamthat's a good approach23:20
nashWhen you maintain C libraries - it's the only way ;-)23:21
nashAt least if you don't want to spend your entire life debugging23:21
nashjotham: The battle viewer protocol - is that documented anywhere?23:28
jothamit's an xml file with a DTD23:30
jothamthere are other xml files that define various graphic attributes of the entities23:30
jothamfor instance what PNG file to use, what weapon the vehical has, etc23:30
nashIs it on the web anywhere or something?23:31
JLPnash: there is a BATTLEFORMAT file in the darcs repository for battleviewer23:32
jothamdarcs get http://darcs.thousandparsec.net/repos/battleviewer23:32
tpb<http://ln-s.net/JRW> (at darcs.thousandparsec.net)23:33
nashokay23:33

Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!