Sunday, 2009-03-29

*** tpb has joined #tp00:00
*** ChanServ sets mode: +o tpb00:00
landonwrt the 3d client, what would the tp04 priorities be?00:01
landonor just "all of it needs done" :)00:01
llnzall needs to be done00:02
landonok00:02
llnzmost of it is low level, not graphics00:02
landonright, earlier I was thinking of implementing tech trees on the wxwidgets client, but it was suggested that the 3d client is a higher priority00:03
* landon is finally getting started on applications :)00:06
jnengland77hi all00:07
llnzhi jnengland7700:13
jnengland77I'm interested in GSoC and your project.  I think making another ruleset for tpserver-cpp (C++ is what I've been learning in my classes) would be neat.  I'm not entirely sure what game I would make though. I used to play a lot of C&C games, but I've introduced to 4X games( Dominions 3) about a year ago or so by a friend... it's bit addicting. :P00:21
jnengland77a bit addicting.*00:23
llnzcool00:24
llnzyeah00:25
jnengland77I haven't had a lot of time to play games though so I haven't played it in a couple months. I think the last time I was playing a multi-player game, and was getting beat by the AI...00:28
*** ninja-kage has joined #tp00:35
*** ninja-kage has quit IRC00:53
mithroheyo people01:57
Greywhind_hey Mithro02:01
Greywhind_i made some updates to my application, stemming from your comments02:01
mithroGreywhind: will look at your application again in a moment02:01
Greywhind_mithro: ok, thanks! no rush, just wanted to let you know02:01
mithroezod: I'm assuming you are no longer around02:13
*** jnengland77 has quit IRC02:18
*** jnengland77 has joined #tp02:26
mithrohey, I'm back02:48
mithroGreywhind: wow, it's much easier to read with a little bit of formatting :)02:49
*** jnengland77 has quit IRC02:56
*** landon has quit IRC02:58
Greywhind_mithro: I hadn't even noticed the formatting options the first time :P03:00
Greywhind_i guess i'm too used to all web forms just being text boxes03:00
Greywhind_anyway, i have to go to sleep03:00
Greywhind_it's 3 AM03:00
*** jmtan has joined #tp03:08
mithroGreywhind: no problem03:08
mithrohey jmtan03:08
mithrojmtan: did you ever fix your mac problem?03:08
jmtani managed to compile an older version of python-ogre for the mac03:10
jmtanthe latest build (end feb i think) didn't work for me03:10
jmtanhowever my build manages to work with the client no problem03:12
jmtanwith an occasional boost.python error03:13
*** Landon has joined #tp03:41
Landonwhat happened to the TP repository I keep seeing mentioned in documents?03:53
LandonI get a 404 when I try to get the key for it03:53
* Landon is trying to get pyogre as per http://wiki.python-ogre.org/index.php/LinuxBuildV203:53
tpb<http://ln-s.net/2VC0> (at wiki.python-ogre.org)03:53
llnzLandon: let me see if i can find it03:57
llnzlooks like the key isn't there03:59
BCarlyon|HomeLandon: the TP key.asc was missing but I'm installed and running TP fine.04:00
llnzyou should be able to install without it04:00
Landonodd, apt-get update was giving me a 404 on the repo itself as well04:00
LandonW: Failed to fetch http://packages.thousandparsec.net/intrepid/dists/intrepid/repository/source/Sources.gz  404 Not Found04:00
llnzintrepid isn't in the repo, only etch, gutsy, hardy and sid04:01
Landonjust noticed04:01
* Landon goes to fis that04:01
Landongot it now04:03
mithrofis?04:03
Landonyeah04:03
Landonit's like fixing04:04
Landonbut uncovering another error04:04
Landon:)04:04
Landon;)04:04
Landonrunning into dependency hell here04:05
Landonpython-ogre depends on libopenal0a, which was replaced by libopenal1, but apparently that's not good enough for pyogre04:05
Landonanyone know the way around that?04:06
llnznope04:07
llnzthough you could grab an old version from etch or something like that maybe?04:08
Landonthat's a good idea04:09
*** BCarlyon|Home has quit IRC04:12
*** barrycar1yon has joined #tp04:17
barrycar1yonHmm ubuntu asus eee dual screen fail04:17
*** barrycar1yon is now known as BCarlyon|Home04:17
Landonheh, so is this the guaranteed way to get into gsoc? "If you can get the 3d client installed from source, you're in."04:24
BCarlyon|Homerofl04:26
LandonI think there's a problem with the python-ogre available in the TP repos04:26
LandonThe following requirements where not met04:26
Landon['python-ogre/ogre == 1.6.1']04:26
Landonfrom the package04:26
LandonPackage: python-ogre04:26
LandonVersion: 0.0.0-704:26
Landonno wait, I fail at context :)04:27
mithroLandon: the problem is that the client has moved to a new version of python-ogre04:30
mithrowhich I'm still trying to package04:30
Landonahh04:31
Landonwell, I'm going to bed now so I might as well get ogre compiling04:34
Landon:)04:34
BCarlyon|HomeLater Landon .04:36
Landonyou know what would be cool?04:42
Landonsome sort of compiling gui thing04:43
Landonwhere it prettifies all the compiling messages into a game or something04:43
* Landon is really off to bed now though04:43
mithroLandon: colorgcc is pretty good04:50
mithroezod: poke me when you get up05:06
CIA-1joel tpserver-cpp * r063eb1524196 / (5 files in 2 dirs): (log message trimmed)05:23
CIA-1Added shave support to build system05:23
CIA-1Shave is is a set of scripts to make autotools05:23
CIA-1output more friendly. This commit adds shave,05:23
CIA-1as well as the necessary bits to enable shave05:23
CIA-1by defualt when building. To disable, pass V=105:23
CIA-1to make when building05:23
CIA-1llnz tpserver-cpp * r577cccc33ae6 / (5 files in 2 dirs):05:24
CIA-1Merge commit 'shenki/pretty-build'05:24
CIA-1Added shave support, thanks shenki.05:24
CIA-1Conflicts:05:24
CIA-1Makefile.am05:24
shenki\o/05:27
llnz:-)05:29
llnzworks fine for me05:29
shenkiyeah, i just gave it a go on my machine. works here05:30
mithroshenki: yay!05:37
mithrollnz: libltdl updated because you have a newer version05:39
llnzmithro: ok05:39
mithro-libltdl_la_LDFLAGS = -no-undefined -version-info 4:4:105:39
mithro+libltdl_la_LDFLAGS = -no-undefined -version-info 4:6:105:39
*** prady has joined #tp05:46
mithrohi prady05:47
*** prady has left #tp05:47
* BCarlyon|Home is having a go at getting the phplib running05:54
mithrophplib? as in libtpproto-php ?05:57
BCarlyon|HomeYip mithro05:58
mithroit still needs a load of work05:58
BCarlyon|HomeIm not a python coder.05:59
BCarlyon|HomeWhich means part of my application will need to include work on libtpproto-php05:59
*** cd_d has joined #tp06:05
llnzhi cd_d06:05
cd_dhello06:06
mithroEpyon: ping06:06
mithrohi cd_d06:06
cd_dis the community interested in any sort of .NET project?06:09
BCarlyon|Homeeww microsoft rofl06:10
mithrocd_d: not really06:11
llnzcd_d: if it's cross platform and on our idea's list, then we could think about it06:12
llnzhaving said that, .net is not something that would not be high priority or would continue to support06:20
cd_dthanks llnz, I won't push the idea any further. .NET is not very cross platform by itself06:23
Patrick`mithro: it looks like there's been some efforts to integrate zeroconf into twisted already, I'll look into it06:30
mithroPatrick`: yes, but I have yet to see someone succeed06:31
Patrick`one thread claims that it's something the OS should be doing in the background - how much weight is there to this? It looked a bit like an excuse to me.06:32
Patrick`"not very much"06:34
mithroPatrick`: well both Linux and Apple have a daemon which you just poke the correct details into06:35
mithrobut on windows you need to run your own06:36
Patrick`so we always run it06:36
mithrosee avahi and bonjour06:36
Patrick`reading up on them now06:36
mithroPatrick`: you should probably use them where avaliable06:36
mithroPatrick`: see how the zeroconf stuff already works06:37
Patrick`currently it does that?06:37
mithroyes06:37
Patrick`I've not read all that code yet06:37
Patrick`anyway, I'm getting bogged down in later details. You think it's doable and that's fine for me :)06:37
mithroPatrick`: it has a common interface then imports the correct one06:37
mithrofalling back to a full python implimentation06:38
mithroin a background thread06:38
Patrick`I'd make what code we have, work with twisted rather than bringing another implementation probably06:38
Patrick`and it's that background thread that can run in twisted but doesn't06:38
mithroall the avahi/bonjour stuff is even driven06:39
mithroso you would have to intergrate their event loops into the twisted one06:40
Patrick`so they're not a background service that's running, just something that might be avalable06:41
mithroor again run them in a seperate thread06:41
Patrick`how was the python zeroconf changed for this project?06:41
Patrick`I know, I know, diff :)06:41
mithroPatrick`: well, they are a background service that's running which is available some of the time06:41
mithroPatrick`: it was changed to not need 8 threads06:41
Patrick`ah.06:41
Patrick`good shout!06:41
mithroonly one06:41
mithroI probably broke it in a number of ways06:45
mithrobut it works for what it needs to06:45
Patrick`if it works it works :)06:45
Patrick`ok, I went to linux to clarify the client/server networking code thing in my proposal06:53
Patrick`and of course my parents remind me I said I'd visit today06:53
Patrick`ok, so before I go, can you tell me briefly about the client's dependance on netlib? I have the code handy.07:01
Patrick`I imagine it must use threading, since wx has its own event loop07:01
Patrick`it's not essential for me to know, it can just stay as it is with its own version of libtpproto-py07:03
mithroPatrick`: so in your changes you need to keep the old method working07:04
Patrick`yep07:04
mithroPatrick`: the clients *can not* depend on twisted07:04
Patrick`yeah, I get that07:04
mithroyou can check the client.py file in libtpproto-py07:05
mithromost of the clients use libtpclient-py which wraps libtpproto-py07:05
Patrick`ok, now I could just keep that as it is, maybe have it folded back into the client rather than a library07:08
Patrick`and tpserver-py would have the changed proto folded back into it07:08
mithroPatrick`: hrm?07:09
Patrick`not that I wouldn't get into it I don't want to get bogged down into the inner workings of the client07:09
mithroas you can see from twist.py, it's quite possibly to impliment twisted support without breaking libtpproto-py for other uses07:09
Patrick`oh dur, and I can't just fork it because of all the object descriptions and protocol and so on07:10
*** skiffcz-gsoc has joined #tp07:10
Patrick`so I have to have "old-mode" and "twist-mode" come from the same codebase07:10
mithroPatrick`: yes07:10
mithrothe socket handling side of libtpproto-py is just a small part07:11
Patrick`now to clarify07:11
Patrick`does the client actually use the socket handling stuff?07:11
Patrick`I've been grepping for the relevant imports07:11
Patrick`but I might have to chase dependancies07:12
Patrick`or is it more a case that just can't modify certain things07:12
mithrookay there is the following07:13
mithrocommon.py - provides a common interface to low level socket stuff - common between server.py and client.py07:13
mithroclient.py - the "client" interface07:14
mithroserver.py - the "server" interface07:14
mithroI expect you to add a "twist.py" - the "twisted server" interface07:14
*** calidus has joined #tp07:14
mithrohi calidus07:14
Patrick`ok, that's eas(ier) than hacking into the client stuff07:15
Patrick`phew :)07:15
calidushi07:15
mithroPatrick`: the doco for libtpproto-py is quite good07:15
mithrotry running epydoc over it07:15
llnzor read the epydoc online07:16
Patrick`ok, that should quell the flow of obvious questions for a few months07:16
mithrollnz: does it work/is it up to date?07:17
llnzmithro: should be07:17
* llnz checks07:17
llnzGenerated by Epydoc 2.1 on Sun Mar 29 07:15:33 200907:18
llnzlooks good to me (in EDT)07:18
llnzumm... maybe EST (eastern time, USA)07:19
mithrookay cool07:19
mithroI have not touched them in quite a while07:20
llnzand the twisted server probably should be in a file called twistserver.py, incase a twistclient.py is ever written07:20
Patrick`ah.07:21
* llnz wanders off07:21
llnzlater all07:21
*** llnz has quit IRC07:21
mithrollnz: true07:22
mithrobut a git-mv is not that expensive :)07:22
Patrick`ok, I gotta run07:22
Patrick`I still have some questions but I can dig around in the code when I have time07:23
*** skiffcz-gsoc1 has joined #tp07:23
Patrick`(what if the SQL and zeroconf get used by the client)07:23
*** skiffcz-gsoc has quit IRC07:23
*** skiffcz-gsoc1 has quit IRC07:30
*** skiffcz has joined #tp07:31
Patrick`ok, I've updated my proposal and posted on the thread07:32
Patrick`I'm off07:32
mithroPatrick`: sql == unlikely07:36
mithroPatrick`: zeroconf == likely07:36
BCarlyon|HomeHow do you generate \x00\x00\x00\x01 from 1 in php?07:43
BCarlyon|HomeThats where I'm stuck at the moment.07:43
BCarlyon|HomeI think I've missed something obvious here.07:44
Epyonmithro, pong07:47
mithroEpyon: going to update your application07:48
Epyon?07:49
mithroEpyon: i made some comments on it last night07:50
Epyonah, ok, I'll take a look :)07:51
mithroyou mentioned that you would update it last night07:54
mithrobut I didn't see any updates07:54
Epyonbrb07:59
*** Epyon has quit IRC07:59
*** epyon has joined #tp08:09
*** verhoevenv has joined #tp08:10
calidushi,first and foremost, i am new to this community and hope you guys dont mind if i make a mistake or two08:13
calidusi have taken a look at the ideas for the projects in GSoC for some time now, and has decided to focus on the idea given by thousand parsec which is the universe in an hour08:13
calidusby means of design, we are suppose to create a new set of rules regarding the gameplay and objects inside it together with their respective classes, am I right?08:14
calidusis there any more stuffs that we need to take care other than the above08:15
mithrocalidus: if you have any questions, feel free to ask08:16
calidusmithro: thanks08:18
epyonmithro, I can update the wiki page on monday, my wierd router has some problem with POST requests, and I can't do it from home. However, if you want any questions answered live, I'll be more than happy08:18
calidusmithro, about the designing of rulesets, does it have to based on previously implemented ruleset?08:19
mithrocalidus: no08:19
mithroepyon: you missing a capital :)08:20
mithroepyon: looks weird :)08:20
mithroepyon: Im happy for you to answer live and let me help direct the answers08:20
calidusmithro: do you mind giving me more explanation or maybe a rough idea on what is expected in our design of ruleset?08:24
mithrocalidus: have you played many 4x or board games?08:25
calidusmithro: yup08:25
mithrocalidus: so you need to do a couple of things08:27
mithroa) figure out the "rules of the game"08:27
calidusmithro: i have some idea on the "rules of the game", but not too sure whether i am going to the right direction08:27
mithrocalidus: so I would recommend playing a couple of it08:28
mithros/couple/a couple of rounds/08:28
mithroand document how you would expect it would work08:28
calidusmithro: ok. thanks :)08:35
mithrocalidus: actually take a look  at the wiki08:35
mithroEpyon's application has an example of what I mean08:36
mithrohttp://www.thousandparsec.net/wiki/User:Epyon/2009/EliteSec#Gameplay_example08:38
tpb<http://ln-s.net/34b3> (at www.thousandparsec.net)08:38
epyonAbout Single Player support -- yes, I'd really love that ( as I have written in the app )08:38
epyonWhat is "scenario idea"?08:39
mithrothe idea that you can load up a "preconfigured" scenario08:39
mithrokinda like you know those pre-setup chess boards08:39
epyonIt's hard to talk about preconfigured in a random universe -- but a preconfigured set of missions would be tempting08:40
epyonabout market and missions -- yes, that would need an update of a client08:40
epyonattack on orbit is either a players choice ( if a player attacks, or the servers choice -- if pirates attack )08:40
mithroIE Caption Epyon is in debt 10,000 credits to a loan shark, can you raise enough money before they blow you out of the sky08:41
epyonlol :D08:41
mithroepyon: I think you need to be more clear08:41
epyonthat is?08:41
mithroepyon: IE "Player A" sets an order to "attack players who enter this system"08:41
mithroPlayer B then enters the system08:41
mithroetc08:41
epyonNope, that's not the way it works :)08:41
mithroepyon: see - you need more description on that08:42
epyonAttack can be made only when both players are on orbit in the same turn08:42
epyonyou select a player in orbit and do attack08:42
mithroepyon: I'm worried these "customise the client" things08:42
epyonthis way it works asynchronously08:42
mithroepyon: i want you to think in terms of the current UI08:42
mithrorather then how an idealised UI would work08:43
epyonin case of missions I need a list with yes/no08:43
epyonin case of market i need a list with number input08:43
mithroepyon: you could probably make the market part of the "resource" bits08:43
epyonassessing the validity of the input can be even moved to the server08:43
mithroand use the message board for the misson stuff08:44
mithro(we actually support multiple message boards, the client just doesn't show it atm)08:44
epyonyeah, I looked at the specs a little08:44
mithro[23:43] <epyon> assessing the validity of the input can be even moved to the server <- that comment scares me a little bit08:44
mithroepyon: what do you mean by that?08:44
mithroezod: debian packages for tpserver-cpp 0.6.1 have been uploaded to packages.thousandparsec.net08:45
mithrosupports debian etch and lenny - ubuntu gutsy, hardy and intrepid08:46
mithroezod: how do I update the xml file>08:46
mithros/>/?/08:46
epyonI mean that even if you give an order to buy 1000 tons of medicine, the next turn you'll get info that you bought 608:48
mithroepyon: well that would be required08:48
mithroepyon: remeber you can't trust anything the client says :)08:48
mithrotp04 support would obviously make this a whole lot easier08:48
epyonsure, it would be more cool if the check would be also on client side, but it isn't required because the serever needs to check it anyway08:48
mithrobut that may or may not happen this GSoC, and you depending on that would be *bad*08:48
epyonI'd like to know what are the statuses of all the clients at the moment...08:49
mithroepyon: the less you depend on client changes, the happier I would be with this application08:50
mithroepyon: have you tried playing a game of minisec or risk locally?08:50
mithro(using the singleplayer stuff?)08:51
mithroepyon: the tpclient-pywx is slightly more advanced, as it has singleplayer support08:51
mithroI wish someone would propose a task to add singleplayer support to tpclient-pyogre08:52
mithroepyon: but actual play wise, they are fairly comparable08:52
mithroepyon: I'm also assuming you are planning to use the C++ server?08:53
*** cd_d has quit IRC09:06
CIA-1mithro web * rf153b93f326d / (bits/instructions-linux-ubuntu.inc downloads/downloads.xml): Update the download instructions.09:13
epyonyes, as I am very proficient with C++ so I'm sure to be able to do modifications if needed09:20
mithrothat makes me sad, but llnz happy ;)09:20
mithrodunno what it will make nash09:20
mithronot sure if he prefers python or C++09:20
epyonoh don't worry :)09:21
epyonyou'll get the art xD09:21
verhoevenvmithro: I'm not getting the minisec colonise order to work in daneel09:24
verhoevenvmithro: And I think it has something to do with the fact it doesn't have arguments09:25
mithroVadtec: on tpserver-cpp?09:25
mithroopps09:25
mithroi mean verhoevenv09:25
verhoevenvYeah.09:25
mithroezod: poke poke09:25
verhoevenvIt works perfectly with tpclient-pywx though.09:25
verhoevenvSo 'm probably just doing something wrong with that whole cache stuff.09:25
mithroso it's a daneel-ai bug then?09:25
verhoevenvYeah.09:26
verhoevenvI'm currently not handling orders very elegantly.09:26
mithrohow are you building the order?09:26
verhoevenv        order = findOrderDesc("Colonise")09:26
verhoevenv        args = [0, objid, -1, order.subtype, 0, []]09:26
verhoevenv        o = order(*args)09:26
verhoevenv        evt = cache.apply("orders","create after",objid,cache.orders[objid].head,o)09:26
mithroprint out type(o)09:27
mithroverhoevenv: btw - you know you can turn on debugging and then it prints all the to/fro the server to the console09:27
verhoevenvmithro: Debuggin on the server side? Or where?09:28
mithroon the client side09:28
*** calidus has quit IRC09:28
verhoevenvHmm, how?09:28
mithrogive debug to the connection09:28
mithroclass ClientConnection(Connection):09:29
mithro"""\09:29
mithroClass for a connection from the client side.09:29
mithro"""09:29
mithrodef __init__(self, host=None, port=None, nb=0, debug=False):09:29
mithroor on setup09:29
verhoevenvHmm, okay, will take a look at that.09:29
mithrodef setup(self, host, port=None, nb=0, debug=False, proxy=None):09:29
mithroyou can turn it on in tpclient-pywx in the preferences09:30
mithroyou can then compare the raw stuff it's sending09:30
*** bddebian has quit IRC09:31
verhoevenvWhat the hell. I didn't change anything except for run it in debug mode, and now it works? It failed 10 times before. :/09:33
mithro:(09:37
mithrodoes changing it back to non-debug mode make it stop working?09:37
mithrothey should be identical09:37
mithroanyway I should be in bed09:38
mithrognight!09:38
verhoevenvIt works, and I didn't change a thing except talk to you. I think.09:39
verhoevenvMagic.09:39
verhoevenvI suppose it'll break when you're in bed.09:39
verhoevenvYep, it doesn't work anymore. What. The. Fuck.09:46
*** BCarlyon|Home has quit IRC09:59
JLPahoy everyone10:01
*** konr has joined #tp10:02
*** konr has quit IRC10:16
BCarlyonAhoy10:20
*** konr has joined #tp10:41
verhoevenvOh. Damn.10:42
verhoevenvHow obvious.10:42
verhoevenvYou can't colonize with a scout.10:42
verhoevenvAnd because I produced scouts and they are faster, it *usually* meant they found a planet quicker.10:43
verhoevenvSo the colonize would fail.10:43
verhoevenvGrmf.10:43
BCarlyonfail10:52
*** Greg___ has joined #tp10:53
verhoevenvAnyway, gotta go.10:54
*** verhoevenv has quit IRC10:54
*** black_pearl has joined #tp11:05
* JLP sees the very first parsek application, yay11:12
Vadtecmornin11:14
black_pearlgud morning all11:16
JLPmorning Vadtec and black_pearl11:19
Vadtechow goes its JLP11:19
JLPVadtec: i'm looking for one hour i've lost somwhere, and reading the application for parsek11:21
Vadtecheh, time jumped forward on ya did it?11:22
Vadtecyou will find your lost hour in about 6 months or so11:22
Vadtec:P11:22
JLPyeah, someone stole it :)11:22
JLP:)11:22
Vadtec:P11:23
Vadtecim reading up on git11:23
Vadtecgonna try to setup my own git repo for the work i do11:23
Vadtecthen im gonna start going over the libtpprotocol-cpp code and figure out exactly what I can unit test and what i cant11:24
Landonheh11:39
Landonleft ogre compiling all night to wake up to a missing libgl error11:39
Landonwas very near the end though :)11:39
Vadtecdont ya just love that11:55
*** black_pearl has quit IRC11:58
LandonVadtec: not as much as realizing it looks like the ogre-python package builds ogre for ya :P12:08
jmtanhi Landon, are you compiling python-ogre for linux?12:10
Landonyeah12:10
jmtandid you try the deb packages on the thousandparsec repo?12:11
Landonhoping to test out the 3d client a bit before writing an application   involving it :)12:11
Landonthey're behind, the client is requiring python-ogre 1.6.112:11
jmtanthat's cool :)12:11
jmtanhmm did you try changing requirements.py to 1.6.0 and giving it a shot?12:12
Landon>.> not before I uninstalled ogre12:13
Landon:P12:13
jmtanoh, yeah i just guessed as much12:13
jmtanhave you hit any snags with compiling python-ogre so far?12:14
Vadtecif i didnt have an ATI video card id mess around with the 3d stuff, but i aint gonna replace my xorg core with whatever the ati compatible core is, cause it totally hoses KDE12:14
Landonstill retrieving libraries12:14
Landonthank god for fat university pipes, but still going slow12:15
jmtani've spent quite a bit of time compiling python-ogre, so i may be able to help with issues12:16
LandonI'm actually going out to airsoft today, so I won't be back for ~7 hours12:17
jmtanthe linux build should be pretty smooth though12:17
LandonI'll let you know if anything pops up, I just queued some commands from the wiki12:17
Landon:p12:17
jmtansure12:18
*** skiffcz has quit IRC12:25
Vadtecgit...is weird...and its remote stuff is even weirder12:29
Vadtechow it keeps track of what is master and what is HEAD im not sure, but it apparently is12:30
*** black_pearl has joined #tp12:39
ezodVadtec: .git/config12:54
Vadtecezod: i know that much, im just saying, the way all these articles describe it...makes my head spin13:00
Vadtecim sure as i use it more ill come to understand and like it13:01
*** black_pearl has quit IRC13:02
Vadtecive just been a subversion guy for so long that its all confusing13:02
*** jmtan has quit IRC13:20
Vadtechrm13:22
*** konr has quit IRC13:30
Vadtecezod: do you have experience setting up gitosis?13:31
Vadtecezod: specifically, gitosis on redhat based distros13:32
Vadtecsuch as CentOS (which is what my server runs)13:32
BCarlyoncentos boo13:36
BCarlyonhiss13:36
alanpHello13:37
BCarlyongit cant be that bad its written by linus13:37
BCarlyongit is my primary repo13:38
mithroezod: poke?13:42
ezodVadtec: i have experience with neither gitosis nor redhat :(13:42
ezodmithro: hi13:42
ezoddaneel-ai.exe works, in case you didn't catch that13:43
mithroezod: yes I did13:44
ezodtesting all on win32 today13:44
mithroso are we ready for a release?13:44
ezodshould be as soon as i verify it all together13:45
ezodwe'll need to do a bunch of stuff yet - like adding proper cwd's in the xml files and some registry stuff13:45
ezodwe can do that in the installer i imagine13:46
*** epyon is now known as epyon|afk13:47
mithroI have not updated the nsis install script yet13:47
Vadtecezod: oh, thanks13:48
VadtecBCarlyon: haha, lets not start an os flaming session :P13:48
BCarlyonIt was meant in jest13:49
Vadteci know13:49
Vadtecthats why i said :P13:49
BCarlyonBut I am not a fan of CentOS I've had too many problems with it13:49
Vadtecyou were using 64bit or you expected each release to have the latest and greatest, which isnt what centos is about13:50
alanpwhat's wrong with 64-bit centos?13:50
BCarlyonneother the people running it were iditios13:50
BCarlyonidiots*13:50
alanpi've been running it for years13:50
BCarlyonalanp: and you have full control of it?13:51
alanpbcarlyon, yes13:51
alanpalthough13:51
BCarlyonthe centos I have access to, has no ability for me to setup cron13:51
Vadtecalanp: you run into A LOT of lib issues when you get into compiling custom packages13:51
BCarlyonI have to email the server admins, and then they set it up13:52
alanpVadtec, you're not really supposed to compile custom packages, it's binary based13:52
BCarlyonfsking register one13:52
alanpand if you do, you should be making packages13:52
Vadtecalanp: kinda hard to make packages when you cant get the build system to find the right libs or compile them due to weird requirements13:52
BCarlyonDreamhost ftw13:52
BCarlyonOr your own server of course.13:52
Vadtecalanp: VLC media player is a perfect example13:52
BCarlyonI should probably get the home server alive again and stop living in the radio office13:53
alanpwhy do you need VLC on an enterprise operating system?13:53
alanpyou shouldn't even have X Windows13:53
Vadtecalanp: streaming is what i was going to use VLC for13:53
alanpahh13:53
Vadtecalanp: and i run CentOS on my home box, which does have X+KDE on it13:53
Vadtecbut not 64bit13:53
alanpcentos isn't meant for a home machine13:54
Vadtectrue13:54
alanpso, really you're out of their scope13:54
Vadtecbut its better than fedora13:54
alanpyou probably want something like gentoo or ubuntuy13:54
Vadteci dont mind most of the hassels of running enterprise grade at home13:54
Vadtecim just now to the point where i have to install custom packages like newer pythons or glibs13:55
Vadtecim gonna be switching to debian at some point, not a big fan of gentoo or ubuntu13:55
mithroezod: so have you started a singleplayer game with daneel-ai exe and tpserver-cpp exe on windows?13:56
ezodmithro: not with the exe yet, i am just about to though13:57
mithrookay13:57
mithroLandon: you about?13:59
Vadtecugh, this is pissin me off....why is git not using the publickey auth first....the git user doesnt have a pass...14:01
Vadtecand i know my key is good, cause i use it all over the place14:02
Vadtechrm14:07
Vadtecgrumble14:07
Vadtecand other words of foulness14:07
*** epyon|afk has quit IRC14:07
*** epyon|afk has joined #tp14:07
Vadtecblah14:15
Vadtecat least its working14:15
Vadteci think the silly ssh key wasnt getting updated for some reason14:15
Vadtecah, i know what happened, encoding was different on my vps, forgot to force it into UTF-814:18
CIA-1mavrinac libtprl * r58576992a1a9 /.gitignore: Add more files to .gitignore.14:25
CIA-1mavrinac libtprl * r7160ca625f82 /tprl/console.cpp:14:25
CIA-1Properly tokenize commands before sending and for help. (Per alanp,14:25
CIA-1fixes issue #74)14:25
CIA-1http://code.google.com/p/thousandparsec/issues/detail?id=7414:25
tpb<http://ln-s.net/34do> (at code.google.com)14:25
ezodalanp: thanks14:25
mithroezod: I have noticed that there is no good way to filter mail from the issue tracker :(14:28
ezodwhat do you mean?14:30
ezodafter we get this release stuff taken care of i was going to look at integrating it on the website, maybe tpb could announce bugs, improve auto-assignment, etc.14:31
ezodit seems pretty flexible14:31
ezodworst case i could always write custom scripts that do whatever we want from the csv export14:31
ezodmithro: has the tracker been spamming you or something though?14:32
mithroezod: I mean the email that code.google.com doesn't have any useful headers for filtering14:33
mithroezod: yeah14:33
ezodoh like in gmail?14:33
ezodi see14:33
Vadtecoops...nothing like hosing a repo to give you fits, good thing its not a production repo... :P14:33
ezodi'll keep that in mind, there may be ways to mitigate that from the tracker end14:34
mithroezod: I might poke the guys internally14:34
ezodgood to have an inside man ;)14:34
mithroyay working for Google! :)14:34
mithroalthough it would be better if it was "yay the code is OSS"14:35
ezodyup :)14:35
mithroezod: so... you working on this release right now, right?14:35
ezodmithro: ya, although currently i keep getting off track14:35
mithroezod: that is why I was poking you :P14:36
ezodok, very close, will keep reporting14:36
alanphmm14:37
* mithro is only up because he has indigestion14:38
ezodwas wondering14:38
mithroezod: if we could get this release out the door today, it would be great14:38
ezodmithro: that's my hope and expectation14:38
mithroezod: great!14:39
ezodbut possibly my today, not your today :P14:39
ezod2:38pm here14:39
mithroha14:40
* ezod away, working on win32 stuff14:40
mithroit's 5:40am here14:40
alanpezod, has that diff been committed?14:41
ezodalanp: yep14:45
*** Greg___ has quit IRC14:45
alanpsweet14:48
mithroezod: I think I'm going to head back to bed for a couple of hours15:02
mithroezod: is there anything you need from me before then15:02
mithro?15:02
Vadtechrm....im getting a weird error from my gitosis....*keeps reading*15:04
ezodmithro: not that i can think of15:08
ezodmithro: tpserver-cpp win32 build on ~tim is up to 0.6.1 tag?15:08
ezod(or close enought that it won't matter for testing?)15:08
mithroezod: not quite, I think it's missing a few patches - but it should be good enough15:08
ezodk15:09
mithro* master        47b2170 More windows support.15:09
mithrothat was what it was built from15:09
*** llnz has joined #tp15:10
llnzmorning all15:10
alanpmorning15:11
Vadtecafternoon15:11
ezodmithro: alright, i'll test from that, but there are risk-related commits between there and 0.6.1 tag so i'll re-test the final release build15:11
ezodalso there are turn timer commits after 0.6.115:12
ezodhi llnz15:12
llnzhi ezod15:12
mithrohey llnz15:12
*** skiffcz has joined #tp15:12
mithroezod: there are?15:13
llnzhi mithro ,. you're up early15:13
mithroI'm rebuilding now15:13
llnz0.6.2 isn't far off15:13
mithrollnz: the turn-timer changes the big different?15:13
llnzi think if the bugs in the ruleset can be fixed soon, then i'd release15:13
llnznot a big change15:14
llnzjust made it clearer15:14
ezodmithro: ok, thanks - 0.6.1 tag or latest head?15:14
mithro0.6.1 tag15:14
ezodcool15:15
Vadtecoi....iforgot to git push the additions for the test repo i setup15:19
Vadtecsomeone slap me15:19
Vadtecplease15:19
mithroseem to build fine15:21
mithrouploading zip now15:21
Vadtecok, i have a general question about git with regards to repos15:21
Vadteci have a test git server up right now15:21
Vadtecif I create multiple repos on it15:22
Vadtecdoes each repo increase a global versioning counter like in subversion15:22
Vadtecor is each repo a repo in its own right?15:22
*** jnengland77 has joined #tp15:22
Vadtecand they dont interfere with each other15:22
mithroVadtec: they are seperate in their own right15:24
mithrojust like seperate subversion repos15:24
Vadtecmithro: cool, i was referring to subversion repos that have like /proj1/trunk /proj2/trunk etc, which makes sense for git as well, if its in the same repo it will deal with that repo15:26
mithroezod: 3 minutes and it should be up15:26
Vadtecthe difference is15:26
mithroVadtec: project1/project2 are in different repos then they are in different repos15:26
Vadtecwell15:26
mithroif they are in the same repo, then they are in the same repo15:27
Vadtecoriginally, i had one subversion server running and wanted to host two projects on it15:27
Vadtecso i created two dirs15:27
Vadteci really didnt want to run two server procs at the time15:27
Vadtecdue to process limitations15:27
Vadtecbut with gitosis15:28
mithroVadtec: subversion supports having two seperate repos served out of the one server15:28
Vadtecmithro: i never got that to work for some reason15:28
Vadteccould just be me15:28
jnengland77Or use xinetd.15:28
mithroVadtec: otherwise sf.net would be screwed :)15:28
Vadtecbut so far git seems to be more inteligent that svn15:28
Vadtecso im gonna give it a serious go15:28
mithrogit is like in a totally different league to subversion15:29
mithroezod: http://www.thousandparsec.net/~tim/tpserver-cpp-win32-0-6-1.zip15:30
tpb<http://ln-s.net/34eS> (at www.thousandparsec.net)15:30
mithrogive that a whirl15:30
mithroezod: we should get you setup with a cross-compile environement sometime in the near future too15:30
alanpcross compile for which os?15:41
mithroalanp: cross compile for windows on Linux15:45
* Patrick` gets back15:46
Patrick`only one massive bag of crap they were going to throw out, this time15:46
mithroPatrick`: all your childhood memories, ha? :)15:48
Patrick`actually yes, this time15:48
Patrick`a wonderful little book I had called "germ wars"15:48
Patrick`a cartoon depicion of the immune system15:48
Patrick`now that I'm studying some molecular biology and pharmacology I can actually understand it :D15:49
mithroha15:50
Patrick`I guess it's kind of like going back and reading Godel, Escher, Bach or The Magical Maze15:50
Patrick`(which were the genesis of my computer science side)15:50
Patrick`only got about 40 pratchett's to move out and I'm basically done but I've run out of bookshelves in my flat!15:50
mithroPatrick`: wow, your really living up in the world if you have bookshelves - I just have bricks and planks ;)15:51
Patrick`they're most probably cheaper than yours :P15:52
Patrick`they're 6 months old and the laminate's coming off15:52
mithroPatrick`: I didn't pay for any of the materials :)15:52
*** zerotruths has joined #tp15:52
Patrick`I got ripped off, bricks and planks would work better15:52
Patrick`I guess it's a symptom of giving students free money in the UK ... we go and blow it15:53
BCarlyonSomeone mention free money?15:53
Patrick`mumble mumble15:53
Patrick`come to britain and do an art history degree15:53
BCarlyonI live in britain and do a New Media degree15:54
Patrick`free money!16:01
BCarlyonwhere?16:03
Vadtec*grumble* *annoyance* *more annoyance* why isnt this silly thing letting me clone my repo....16:03
BCarlyongit clone *git url*16:04
Vadtecyeah I am16:04
BCarlyonwhats up with it?16:04
Vadtecsec16:04
Vadtechttp://vadtec.pastebin.com/d7f022d6716:05
tpbTitle: vadtec private pastebin - collaborative debugging tool (at vadtec.pastebin.com)16:05
Vadtecive checked the perm in the dir16:05
BCarlyonIs the git deamon on, listening, and the relevant port open onn the dir?16:05
BCarlyonDid you add the required thing to the git config file in the .git folder?16:06
mithroVadtec: you need to give it the complete URL normally16:06
BCarlyonThats just off the top of my head16:06
Vadtecgit 15743 0.0 0.1 5388 668 ? Ss 22:55 0:00 git-daemon --base-path=/home/git/repositories/ --verbose --detach16:06
mithroI doubt you have a16:06
mithro /libtpproto-cpp.git16:06
mithrooh16:06
mithroyou also need to set16:06
Vadtechttp://vadtec.pastebin.com/d4bc7aa5316:07
tpbTitle: vadtec private pastebin - collaborative debugging tool (at vadtec.pastebin.com)16:07
BCarlyonBeen a while since I ran my own deamon16:07
Vadteci have a libtpproto-cpp.git dir16:07
mithrogit-daemon-export-ok16:07
Vadtecand yes, that repo has git-daemon-export-ok in it16:07
mithroah16:07
Vadtecwell16:07
Vadtecwait16:07
mithropermissions?16:07
Vadtecdoe that file go *in* the repo16:07
BCarlyonYou cant checkout a file you have to checkout a branch dont you16:07
Vadtecor the repo dir on the repo server?16:07
Vadtecblah16:08
Vadtecnvm16:08
Vadtec"To export only some repositories and not others, you need to touch git-daemon-export-ok inside the root directory (e.g. /home/git/repositories/free_monkey.git) of each repo that you want public. Then remove "--export-all" from the git-daemon command above. "16:08
BCarlyonyeah its git://url/name_of_git_repo16:08
Vadteci added it to the wrong place16:08
BCarlyonrofl16:08
Vadtecgit clone git://vps2.vadtec.net/libtpproto-cpp.git16:09
Vadtecsomeone give that a whirl and see if it works please16:09
mithroVadtec: worked fine here16:09
mithrobut you have a16:09
Vadtec*sigh* someone slap me, i need it16:09
mithrogit-daemon-export-ok file in your repository16:10
Vadtecyeah16:10
Vadteccause i thought thats where it went16:10
BCarlyongood for me too16:10
mithroand you probably want to clone the existing libtpproto-cpp?16:10
Vadtecuntil i re-read what it said to do16:10
Vadtecmithro: yes I will16:10
Vadteci already have16:10
Vadtecim just setting up a private repo for it16:10
mithroVadtec: okay cool16:10
Vadtecso i dont clobber the main one while im working on unit testing it16:10
mithroanyway I'm off to bed for an hour or two16:11
mithroVadtec: that is called a "branch" :)16:11
Vadteci know16:11
mithroVadtec: you shouldn't be scared of them16:11
mithroVadtec: they aren't broken like in subverison :)16:11
Vadtecheh16:11
mithrosvn makes people think branches are all painful and horrible16:11
Vadtecim not scared of them per se16:11
mithrobut in reality they are all great and awesome16:12
mithroanyway...16:12
Vadtecjust dont wanna clobber anything on the core repo16:12
Vadtecnow16:12
Vadtecgo sleep16:12
Vadteci needed to setup a git repo on my own anyways so i could learn16:12
Vadtec:P16:12
Patrick`hey, I used a subversion branch and it was fine16:13
Patrick`and by fine I mean I've never actually merged a branch back into trunk before16:13
Vadtechaha16:13
VadtecPatrick`16:13
Vadtecyou havent lived until you try to merge back into trunk/16:14
Vadteci personally dont merge into trunk/ until im ready for a code release16:14
Vadtecso trunk/ is never up to date with the current development track16:14
Vadtecsaves ALOT of head aches16:14
Patrick`I'm gonna be doing that in a day or two I think16:14
Vadteci use trunk/ kinda like a holding place for copies of the code that are guaranteed to compile and run16:15
Vadtecbut arent exactly "releases"16:16
Vadtecif some schmoe wants the raw development branch, they can get it on their own16:16
Vadtecsaves me having to deal with people asking "why wont it compile?"16:16
Vadtecin theory at least16:16
BCarlyonWhy won't it compile?16:17
Vadteclol16:17
* BCarlyon beats the wont compile drum16:17
BCarlyonWhy won't it compile?16:17
BCarlyonHelp me!16:17
*** konr has joined #tp16:17
Vadtecso now....i wonder if its possible to clone the TP libtpproto-cpp repo directly into my libtpproto-cpp repo....16:19
Vadtecill probably have better luck just cloning it and removing .git/ and adding the source to my own repo16:19
BCarlyonyip16:20
BCarlyonIts also more logical16:22
Patrick`aah, good ol' python16:29
Patrick`compile? what is this mysterious wod16:30
Patrick`*compile? what is this mysterious word16:30
Vadtecrandom question, does git have keywords like svn does such as $Id$ or $URL$ ?16:34
Vadtecim looking in the manual but not seeing anything16:35
llnzVadtec: i believe it does16:35
Vadteckk, ill look for it later16:36
*** zerotruths has left #tp16:37
Vadtecright now im gonna get to work looking at the source to see what i can unit test16:38
Vadtecor well, at least try to16:38
* ezod thinks all win32 pieces are working!16:43
llnzezod: cool16:43
llnzVadtec: if you are going to start now, I am happy to create a unittesting branch in the main repository16:43
Vadtecim ok with that16:44
Vadtecjust so long as im not screwing with a core branch :P we dont need me screwing things up16:45
Vadtecllnz: however you want to do it is fine with me, i just got my own git repo setup so I can use it (which i probably will anyways so if i screw it up it only affects me) or i can use the TP repos and work on a branch, makes no difference to me16:48
llnzVadtec: work on your own repo for now, easy enough to move it across later16:49
Vadtecwill do16:49
Patrick`Vadtec: you doing testing of the c++ network stuff?16:51
VadtecPatrick`: i plan to yes, its what i would like to do for gsoc, but ill stick around and do what i can anyways16:52
Patrick`Vadtec: well, if we both end up doing it it would be great to work vaguely in parallel16:52
Patrick`my proposal covers unit testing of the python protocol :)16:53
Vadteccool16:53
Vadtecwow its been forever since i did C++....i need a tetnus (sp?) shot im so rusty17:17
Vadtechmmmm, i better brush up on my C++ before i get to involved with my train of thought...17:31
Patrick`hah, my mate was really into trains17:33
Patrick`he was really enthusiastic about that community project to build a new steam train17:33
Patrick`put loads of money into it but never visited17:33
Patrick`it finally got built and went on tour17:33
Patrick`queued up at his local station - but with all the jostling, he fell right in front of it as he got his first look at it17:33
Patrick`chuffed to bits17:34
BCarlyono.017:34
Patrick`badum-tish17:34
Vadtecummm what17:34
Patrick`I'll be here all nighyt17:34
Vadteck then...17:35
VadtecO.o17:35
Vadtec:P17:35
Patrick`I made it a bit long-winded, sorry17:35
Patrick`I can take any joke and ruin it with bad delivery17:36
Vadtecit was more the randomness of it that i was referring to17:37
* ezod has never previously heard the term 'chuffed'17:38
Patrick`ezod: I think it's an englishism17:39
ezodso says the all-knowing google17:39
Vadteci just heard a scary prospect....using svn AND git on the same directory....should i internet-stab the "person" who suggested it?17:45
Patrick`yes.17:46
Patrick`git would track .svn and svn would track .git :)17:46
Vadtecyeah17:46
Vadtecnot exactly my idea of a good thing17:46
ezodhttp://www.kernel.org/pub/software/scm/git/docs/git-svn.html17:46
tpb<http://ln-s.net/2BYx> (at www.kernel.org)17:46
VadtecO.o17:47
ezodi've never used it but i hear it works remarkably well17:48
BCarlyonunless you tell eachto ignore the other17:48
BCarlyonI did that last year17:48
Vadtecwell i can see the advantages of keeping both a git and svn repo up-to-date off the same source base17:49
Vadtecsvn:ignore and .gitignore could take care of the rest17:49
BCarlyonlast years project has a svn repo (or was it cvs?) I used git locally17:49
Vadtecso git-svn is a svn frontend...but done git style?17:50
ezodit's bidirectional17:50
Vadteci think for my own sanity ill stick to one or the other17:51
BCarlyonor all three?17:53
BCarlyonI mean four.17:53
*** cherez has joined #tp17:55
*** nash has joined #tp18:01
epyon|afknash !18:58
*** epyon|afk is now known as epyon18:58
nashepyon: !18:58
epyonnash, have you seen the EliteSec proposal?18:58
nashnot yet...18:59
nashgot a link?18:59
epyonhttp://www.thousandparsec.net/wiki/User:Epyon/2009/EliteSec19:00
tpb<http://ln-s.net/34K:> (at www.thousandparsec.net)19:00
nashepyon: I'll look shortly (sorry - busy morning at work)19:00
nashepyon: Brief look... looks fairly comprehensive ;-)19:01
epyonI hope I adressed at least some of your doubts :)19:02
nashI hope so too ;-)19:03
epyon:D19:04
Patrick`so does tp currently have any concept of sensors, planetary coverage etc19:08
JLPPatrick`: it can be implemented as a resource, but the client wouldn't know how to show it in a specific way on the starmap, just together with other resources as plain data, unless you would teach the clients about it19:18
Patrick`more to the point, you could just not give data on certain planets :)19:18
Patrick`not for a proposal, this is just Yet Another Thing Stars! Would Need19:18
JLPyeah that would be for tp04 as part of history and visibility support19:19
Patrick`ooh, it's got a name :)19:20
Patrick`so you could do things like "gravity 1.2, average temperature -8, 13% of earth normal levels of bauxite, this data was collected 20 years ago"19:21
Patrick`found the wiki page19:23
Patrick`mmm19:23
llnzPatrick`: yes you can do that19:23
CIA-1mavrinac libtpclient-py-stable * r5bedba546bf8 /tp/client/SinglePlayer.py: Also search in environment path when verifying command existence.19:30
ezodgit submodules are ridiculous19:36
mithroezod: which part?19:43
ezodmithro: just venting frustration19:44
ezodi do git pull in libtpclient-py module dir19:44
mithroezod: out of interest, what is causing you the most pain?19:44
ezodi have to checkout stable branch because it doesn't know what branch i want19:44
ezodso now my latest commit is in SinglePlayer.py19:45
ezodi run tpclient-pywx - and it behaves as previous to the commit19:45
ezodso i check SinglePlayer.py again, and it somehow reverted to the old head19:45
ezodi figure, ok, so i'll do yet another commit to update the head reference19:46
ezodand somehow git tried to add all of libtpclient-py to the tpclient-pywx repo instead19:46
ezod:(19:46
ezodso round 2 with a fresh clone19:46
mithro:/19:46
mithroI'll be back in 15-20 minutes19:46
*** peres has joined #tp19:47
ezodi don't understand why the act of running tpclient-pywx is reverting the submodule19:49
ezodbut that seems to be what is happening19:49
mithrorequirements.py does a "git submodule update"19:50
ezodhah, yeah, just noticed that19:50
mithroanyway - I'm runing late now19:50
mithroseeya!19:50
ezodlater19:51
CIA-1mavrinac tpclient-pywx-stable * r8dce693418b0 /libtpclient-py: Update libtpclient-py submodule to latest head.19:53
Landonmithro: I'm about now20:02
*** mithro_ has joined #tp20:06
mithro_back20:07
tansell__laptopLandon, so I notice your currently trying to get python-ogre to build?20:08
ezodmithro_: there appears to be a problem with the windows build of tpserver-cpp - when i start risk, i own no objects - win32 only20:08
ezodi shot a status update to the mailing list20:09
Landontansell__laptop: yup20:09
*** mithro has quit IRC20:09
tansell__laptopLandon, how are you going so far?20:09
Landonnothing unsurmountable yet20:10
*** tansell__laptop is now known as tansell-laptop20:10
tansell-laptopLandon, great to here20:10
tansell-laptopthere are some packages I have yet to release avaliable20:10
tansell-laptopthey don't work on intrepid yet however20:10
Landonhad to manually download a few libraries, but the python script is chugging through now20:10
*** mithro_ is now known as mithro20:10
tansellLandon, a good 3d proposal would be killer20:12
tansellwe have yet to get one20:12
Landonwell I don't know how "good" mine will be20:14
LandonI'm still thinking about doing tech trees :)20:15
Landonat least, that's what my first app will be20:15
Landononce I get some time to play witht he client (didn't realize it would take me 2 days to get everything installed! :P for once, windows has it easier )20:16
tansellLandon, sadly this is the case for python-ogre20:17
tansellI'm trying to fix the problem but I'm currently waiting on a fix from upstream20:17
Landonany idea what's keeping it from being included in debian repos in the first place?20:17
ezodupstream doesn't really do releases20:18
ezodor source ones anyway20:19
ezodso gentoo ebuilds are a huge headache, which is why i still don't have a working ebuild for the 3d client20:19
Landonahh, gotta go move my car before I get a ticket20:22
tansellezod, oh?20:22
tansellezod, I thought someone else was doing the overlays for gentoo20:23
tansella lot of stuff has stabilised20:23
tansellezod, wanna try get that working next? :P20:23
tansellezod, re the risk bug20:25
tansellezod, does minisec wrok?20:26
tansellbah, I can't type today20:26
tansellezod, poke poke20:35
* mithro feels a bit like he is talking to himself :)20:39
skiffczJLP, llnz, mithro, ping?20:48
skiffczhttp://www.thousandparsec.net/wiki/Google_Summer_of_Code_2009_-_Skiff%27s_Proposal20:48
tpb<http://ln-s.net/34hH> (at www.thousandparsec.net)20:48
shenkimithro,tansell , tansell-laptop, identiy-crisis-tim: how about an application to port tpserver-cpp to windows?20:48
shenkiwould that be interesting to you guys?20:48
tansellshenki, I wouldn't put it as a high priority20:49
shenkitansell: lower than the website stuff?20:50
tansellshenki, yes20:50
tansellshenki, we have an almost working solution for win3220:51
skiffczhuh, well, I have finished it, at last ... now I have to go to a deep cave andd die to exhaustion, because today was a hard day :) Good night all, and Im looking forward some comments soon. Thanks in advance ;)20:51
tansellskiffcz, I'm a little worried about the choice of using C++ for this project20:51
tansellthe Python libraries are significantly more mature (having 3 clients use them)20:52
shenkiskiffcz: perhaps you could have a section where you argue for using c++. then tansell can tear it apart if he still disagrees, but atleast then everyone can see why the language was chosen20:54
tansellwhile I'm definitely interested in having an AI which competes against daneel-ai - I'm not sure how high a priority it would be20:55
tansellskiffcz, otherwise your proposal is very good20:58
tansellI like the agent idea20:58
ezodtansell: re pyogre ebuilds, i haven't heard back from either of the 2 people who said they'd take a look, so yeah, that's on the list21:01
cherez'lo folks; how are things?21:01
tansellhey cherez21:01
ezodtansell: re minisec, not sure, will check right now21:01
tansellcherez, get my email?21:02
cherezYeah.21:02
cherezI've not had internet the last few days.21:02
ezodtansell: same problem with minisec21:03
tansellezod, hrm21:04
tansellezod, the person doesn't control anything when they connect?21:04
cherezI need to figure out my proposal and find time to write it....21:04
tansellwhat happens if you start tpserver-cpp manually?21:04
jnengland77cherez: Me, too.21:04
tansellthe sooner you put in a draft21:05
tansellthe sooner we can give you feedback21:05
cherezYeah, I know, but I've got 3 fairly major projects demanding I have no time.21:05
shenkitansell: do you have advice on the timeline for my proposal?21:06
shenkias in, how to scheudle the work?21:06
cherezDoes the TP protocol have any explicit design document?21:06
llnzcherez: yes21:06
llnzcherez: TP03 doc: http://www.thousandparsec.net/tp/dev/documents/protocol3.php21:07
tpb<http://ln-s.net/Ix2> (at www.thousandparsec.net)21:07
llnzcherez: TP04 doc: http://www.thousandparsec.net/tp/dev/documents/protocolxml.php21:07
tpb<http://ln-s.net/JXj> (at www.thousandparsec.net)21:07
tansellprotocolxml.php could use some more love21:07
tansellprotocol4.php shows our thoughts about moving to the next protocol version21:07
cherezI presume we want libtpproto2-py to be moved to TP04?21:08
ezodtansell: well that's strange - started manually with same parameters, works fine21:09
tansellcherez, it should support both tp03 and tp04 depending on which xml file it loads21:10
ezodtansell: i haven't looked into this very thoroughly yet, btw, if you couldn't already tell :\21:10
tansellezod, sounds kinda like a problem shenki was having a while ago21:10
tansellezod, how fast is your machine?21:11
cherezSo that protocol.xml defines everything that matters between versions?21:11
cherezThat's handy.21:11
tansellit defines the *whole* protocol21:11
tansellthere is a protocol.xml for tp04 and tp0321:11
tansellsee the documents.git repository21:11
tansellprobably needs some more work21:11
cherezAhh, terrific.21:11
cherezSo where is it lacking, then?21:12
*** Greywhind_ has quit IRC21:13
Landontansell: whoo, running ogre demos21:13
LandonI think I've got a step or two21:13
shenkiare there tests for the protocol?21:13
Landonleft21:13
shenkias in, if i decide to write a tp-forth implimentation tomororw, can i verifiy that my implimentation is correct?21:13
tansellshenki, nope!21:13
shenkioh oh oh. there's a SoC project.21:14
tansellwell, you can just test against the server21:14
tansellshenki, it's quite hard to do21:14
cherezRight, testing of course.21:14
shenkitansell: that's why you get a SoC peon to do it :)21:14
shenkisave you the hard work21:14
ezodtansell: the win32 test box isslow - 1ghz pentium 321:15
tansellezod, hrm....21:15
tansellezod, did you start it up manually with daneel-ai too?21:15
cherezAre there any things that haven't been implemented at all, or am I looking at primarily coming up with test and integrating it?21:16
tansellcherez, there is quite a bit which hasn't been implimented in libtpproto-py21:16
tansellbut a lot of it has semiworking stubs21:16
tanselland protocol.xml is missing some stuff21:19
tansellbut not a huge amount21:19
ezodtansell: i think you're getting at the right idea21:19
tansellwe also have not defined things like research21:19
tansellezod, oh21:19
tansell?21:19
tansellI don't have a windows computer here21:19
tansellmine is at home21:19
ezodtansell: i just tried single player {minisec, risk, risk+daneel} and all worked21:19
tansellezod, I think it's got to do with the tpserver-cpp turn stuff21:20
ezodk, so this isn't necessarily a win32 thing21:20
ezodthat's good to know21:20
tansellbasically daneel-ai takes over the universe before tpclient-pywx has even joined21:20
tansellor something like that21:20
ezodmy observations seem to corroborate that21:21
Landontansell: now I'm having problems here, I get consistent Task Failed on step 5 of http://wiki.python-ogre.org/index.php/LinuxBuildV221:21
tpb<http://ln-s.net/2VC0> (at wiki.python-ogre.org)21:21
*** bddebian has joined #tp21:21
cherezDo the other implementations have any pluggable protocol specifications?21:21
ezodtansell: alright, then it seems once i get this subprocess killing crap taken care of we're as good to go on win32 as on linux21:21
tansellthe problem is that the next turn stuff should only work when all the players have joined the game21:22
tansellcherez, the java library21:22
tanselland the ruby library21:22
tanselljava library is the most advanced21:22
tansellanyway have to go to lunch21:22
tansellbe back in an hour21:22
cherezAlright.21:22
bddebiantansell? What happened to mithro? :)21:23
shenkihe's having an identity crisis21:23
shenkithere's 3 of him today21:23
shenkijust wait for tomorrow, there will be more21:23
bddebianheh21:24
bddebiantansell: Yes, that means tpclient-pywx is in now.  Everything should be there now.  I will check on the new cpp server version.21:27
*** Greywhind has quit IRC21:29
llnzskiffcz: i quite like your proposal21:30
Landonthat's the damnedest thing21:41
Landonpython-ogre downloads and installs boost 1.3621:42
Landonbut later on it requires 1.3721:42
Vadtecso, now that im mostly done dealing with a nasty server attack, im gonna get back to looking at the protocol code and figure out what i can unit test in it21:53
Vadtecmost i can do right now anyways is wait for the bots to stop attacking21:54
Vadtecdamn kiddies, they all need a good butt kicking21:54
Vadtecand their parents too21:54
*** xenozephyr has joined #tp21:58
xenozephyrhi, so I'm interested in doing a web based client for GSoC, who can I talk to about that?22:02
ezodxenozephyr: depends, what language(s) would you be using?22:02
xenozephyrwell Ive done a bit of work in PHP but I can also do Flash (Actionscript/Flex moreso)..22:03
ezodwe do have a protocol library for php, not sure how mature it is22:05
ezodincidentally there are ones for ruby and python as well that could apply to web clients22:05
xenozephyrbut whatever you guys would have preference in, I can do that too.. lol im a fast learner. I was also considering using dojo/javascript22:06
ezodwell the front end is really your call22:06
ezodpython protocol library is the most mature22:07
* ezod bbl22:08
tansellback22:29
tansellLandon, actually - that is probably me breaking things22:30
*** konr has quit IRC22:31
*** konr has joined #tp22:32
tansellezod, poke22:32
ezodtansell: hey22:35
ezodso it looks like Popen.kill() basically doesn't work in windows22:35
tansellezod, on windows you have to do something special?22:36
ezodabout to google around about that22:36
ezodyou know what it is?22:36
Vadtecis there any documentation for the protocol as to what happens in what order? say for example, login and logout22:37
tansellezod, are you running tpclient-pywx using the windows python or the cygwin python?22:40
llnzVadtec: there is an example at the bottom of the protocol3.php page22:40
ezodwindows python22:40
ezodi'm doing everything with windows python in fact, on purpose22:41
Vadtecllnz: thanks, i guess i just missed it22:41
llnzVadtec: it's pretty simple though, doesn't show everything22:41
Vadtecllnz: just so long as it gives me an idea of what I can try and unit test and go from there its all good22:42
ezodtansell: subprocess.py from 2.6 imports all sorts of win32 stuff, and i can't find any evidence that what i'm doing shouldn't work22:44
tansellezod, great - just wanted to make sure22:44
tansellezod, try it outside the confines of the singleplayer stuff?22:45
ezodtansell: did - opened python, from subprocess import Popen, Popen("notepad.exe", shell=True), Popen.kill(), notepad stays open22:46
ezodi'll try without shell=True22:46
ezod...and there we go22:47
tansellyou are probably just killing the shell22:47
ezodevidently22:48
tansellezod, why do we have shell=True?22:48
ezodtbh, i forget22:48
ezodi'm going to try taking it out22:48
Vadtecllnz: http://www.thousandparsec.net/tp/dev/documents/protocol3.php <--- this link correct? sorry, im dealing with a ddos attack on a server i admin, try to negate it22:50
tpb<http://ln-s.net/Ix2> (at www.thousandparsec.net)22:50
llnzVadtec: y22:50
llnzyes22:50
chereztansell: Is there any existing method to test protocol compatibility?22:50
Vadtecthanks22:51
tansellcherez, no22:52
tansellcherez, but there are some start of unittests in libtpproto2-py22:52
tansellthat is quite different to protocol compatibility22:52
shenkicherez: if you're interested in testing, you could finish off the unit tests in libtpproto2-py22:59
*** tansell_laptop has joined #tp23:01
Landontansell: I got this straight from the python-ogre site, not the TP repos23:04
Landonbut it's fixed23:04
Landonand now ogre is all compiuling23:04
tansell_laptopLandon, yes it takes forever :(23:05
tansell_laptopLandon, hence why I'm trying to get the debs working23:05
cherezWould it be feasible to try to make a mechanism for testing protocol compatibility, or would that call for a new protocol version?23:05
tansell_laptopcherez, it would be a hard problem to solve23:06
tansell_laptopcherez, you would be better concentrating on making libtpproto2-py work23:06
tansell_laptopand creating tests which work23:06
shenkicherez: with the unit tests as an introduction to the protocol, you could then stick around and help write a tpproto-test package. but it's important not to attempt something huge as a first step23:08
*** tansell-laptop has quit IRC23:15
*** tansell_laptop is now known as tansell-laptop23:17
cherezI'm just not too sure what I'll be writing in the proposal.23:19
tansell-laptopcherez, oh?23:20
Vadtecok finally, ddos over, maybe now i can get something done23:22
tansell-laptopcherez, your proposal is "make libtpproto2-py as usable as libtpproto-py" while including good unittest coverage23:22
tansell-laptopto prove it works23:23
Landonhm23:27
Landonhaving no luck with python-ogre at all, it's trying to use boost 1.36 now23:28
tansell-laptopLandon, what is your exact error?23:30
Landon03-29 22:27 PythonOgre.BuildModule DEBUG    /usr/bin/ld: cannot find -lboost_python-gcc43-mt-1_36-py2523:30
tansell-laptopLandon, did you build boost 1.36 in the earlier steps?23:32
llnzbbl23:32
*** llnz has quit IRC23:32
Landonno, when I did that it complained that I didn't have boost 1.3723:32
Landonso I had it install that23:33
tansell-laptopLandon, that would be why it can't find 1.3623:34
Landonah there's a boost_python_index to build too23:36
Landongoing to see if rebuilding that helps anything23:36
Landonnope :\23:38
tansell-laptopwhat did you modify to get it to build boost 1.36?23:39
Vadtecllnz: when im looking at the protocol specs, if i see <4>, does that mean its 4 32bit integers?23:40
Vadtecoh23:40
Vadtecherm23:40
Vadtechrm23:40
Vadteche left23:41
Landontansell-laptop: nothing, I had it rebuild the python_boost bit, since it seemed that was giving it trouble23:41
tansell-laptopyou ran this command23:42
tansell-laptopright23:42
tansell-laptoppython python-ogre/BuildModule.py -b cmake gccxml boost boost_python_index pygccxml cg scons freeimage zziplib pyplusplus23:42
tansell-laptopStep 223:42
Landonyeah23:42
tansell-laptopand they all succeeded23:44
tansell-laptop?23:44
Landonyup23:45
tansell-laptopso you should have boost 1.36 in the root directory23:46
Landonlibboost_python-gcc43-mt-1_37.so and libboost_python_index-gcc43-1_36.so are present23:47
Landonso, both foso, bothr python23:47
Landonhm23:48
Landonthere is a broken link in my lib directory23:48
tansell-laptophow did you get a libboost_python-gcc43-mt-1_37.so23:49
LandonI redid step 2 for boost with boost 1.3723:49
LandonI just renamed the directory to boost_1_36 and all installed fine23:49
Landonok, so this is after rerunning the step ot install boost (and installing 1.36 with it this time, not 1.37)23:50
Landoncannot find -lboost_python-gcc43-mt-1_36-py2523:50
tansell-laptophow did you "redo step 2 with boost 1.37" ?23:50
Landondownloaded the tarball for boost 1.37, extracted it23:51
Landonand renamed the folder23:51
tansell-laptopbad Landon23:51
tansell-laptopthat is a "probably have to start from step 1 again" type thing :(23:52
Landonyeah, I'm redoing it now23:52
Landonor at least the parts that I diverged from23:52
tansell-laptoplooks like a simple change needed in PythonOgreConfig_posix.py23:53
tansell-laptop ## path to the root of your boost dir, should have 'libs', 'boost' subdirectories23:53
tansell-laptop-PATH_Boost = os.path.join(LOCAL_INCLUDE, 'boost-1_38')23:53
tansell-laptop+PATH_Boost = os.path.join(LOCAL_INCLUDE, 'boost-1_36')23:53
tansell-laptop ## Path to your boost_pythonxxxx lib file23:53
tansell-laptop PATH_LIB_Boost = LOCAL_LIB23:53
tansell-laptop ## and the name of the boost python library23:53
tansell-laptop-LIB_Boost = 'libboost_python-gcc42-mt-1_37'23:53
tansell-laptop+LIB_Boost = 'libboost_python-gcc42-mt-1_36'23:53
Landonyeah, I'll change it after this compile fails23:55
tansell-laptopLandon, you might have to remove your root directory and start from step 2 again23:56
Landonhopefully I didn't screw everything up that bad :p23:57
ezodtansell-laptop: ah, shell=True was for arguments23:57
tansell-laptopezod, what was wrong with the arguments?23:57
ezodtansell-laptop: well, originally, they were all just tacked into the command string23:58
ezodbut no worries, i got it23:58
ezodthe easy way to do this is to just do commandstring.split() right in the Popen constructor ;)23:59

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!