Friday, 2007-03-02

*** tpb has joined #tp00:00
*** ChanServ sets mode: +o tpb00:00
mithronash: btw can you remind me to submit a SoC mentor application? I had another chat with Leslie and she said we should apply anyway00:01
nashWhen does it need to be in by?00:01
mithrowell it opens 5th of march00:02
mithroMonday next week00:02
* nash also knows Chris Dibona which may help00:02
mithrocan closes Thursday - but the earlier the better00:02
nashokay - so Sunday basically00:03
nashmithro: What else is on your todo list TP wise then - I can list htem too ;-)00:04
mithronash: Take over the universe? :)00:04
nashTP related00:05
mithronash: intergrate jotham's battleviewer into tpclient-pywx00:05
mithronash: intergrate tpserver-py into tpclient-pywx00:05
mithro(for quick games)00:05
mithronash: close long outstanding bugs with tpclient-pywx00:05
nashBoth seem to lower prio then TP04 and MTSec definition to me00:06
nashmithro: Actually there are a couple of bugs which are Todo items - so they shoudl be closed as such00:06
mithronash: reorganise the website to be easier to find relavent information00:06
mithronash: get nice looking and working well00:07
mithronash: get tp into the top 20 projects on sf :)00:07
mithroincrease the websites page rank00:07
mithronash: get tpclient-pywx so you can play a game out of the box00:08
mithronash: managed the students doing the java and midp client stuff00:09
mithronash: get the intro movie to play instead of the static splash screen00:09
nashmithro: Feel free to list more... but none of them are important.00:11
nashBiggest hurdle for TP is lack of man-power (like 99% of porjects really)00:11
mithronash: well they are of varying importance00:12
mithronash: tp04 is important - however users don't really "care" about tp0400:12
nashNo, but it allows you to build a better game experience00:12
nashMaybe a TP04 which just addresses the main issues, with TP06 being the XML reboot00:12
mithronash: most of the "game experiance" problems aren't caused by the protocol - they are caused because I suck at GUI programming :)00:13
mithronash: the biggest "game experiance" thing that tp04 fixes is the EOT thing00:14
mithroand most of these game experiance problems will be directly applicable to MTSec as well00:15
nashHence the suggestion for a TP04 being TP03 + obj del & turns00:18
nashMy point is you want to make the project attractive00:18
nashA game which people can play will inspire people to get involved00:18
nashYou also need to deal with game start up - how can I start a game in a sane way00:19
mithronash: thats the part i'm trying to deal with by intergrating the tpserver-py00:20
nashWrong solution00:20
nashThat doesn't work in the general case00:20
mithrowhat general case?00:20
mithropeople want to start a game and play it?00:21
mithrowith an intergrated tpserver-py people just click a "New Game" button on the client and it uses a wizard to set it up00:22
nashYes but that works for one player using only tpserver00:22
nashIt doesn't work if I want to set up with my friends using the java or tpe clients00:23
nashThere needs to be protocol to tell the server "set up this game"00:23
nashWith these players00:23
mithronash: you just give the URL (which the wizard will end with) to your friends and they use it (or use browser the metaserver)?00:25
nashWell metaserver is no good, I want a private game00:26
nashmetaserver tells teh whole world00:27
nashBut my point is - if it is generic protocol it is a long term solution, not mulitple ad-hoc client solutions00:27
mithronash: well put forward a proposal on how to do it :)00:28
mithronash: still doesn't solve the problem where you don't want to run your game on a public server - have to start up the server somehow00:28
nashif (_unix){ system("tpserver-cpp &"); }00:29
mithroshit I ave to run00:29
*** mithro has quit IRC01:18
*** nash has quit IRC01:33
*** nash has joined #tp01:33
*** mithro has joined #tp01:58
CIA-20[email protected] * tpclient-pywx-dev/windows/ (11 files):01:59
CIA-20Implimented [ 1333188 ] Connect window remebers.:01:59
CIA-20 Also fixed a bunch of autoconnect bugs.01:59
mithrowelp, another long standing bug down02:03
nashWell just gave you one to compensate02:08
mithronash: o?02:10
nashBasically I fired of a list of objects to get (100+ items) and the server seems to have regened the turn before it finished responding - so it game me a Fail on one of them02:11
nashBut there is no easy way of knowing which one failed02:11
mithronash: well - firstly you should log a bug as the server shouldn't be regenerating in the middle of a pending request?02:12
mithroor was the regen between getting the list and sending the get orders?02:12
nashIt may have been riming related02:12
nashI am using 10s turns over a slow connection02:12
mithroi'm pretty sure tpserver-cpp has timing issues :)02:13
nashso the server may not have got the request until after it regened02:13
mithronash: ahh okay02:13
mithroi keep a list of what I requested as you suggested02:13
mithrobut the fail frame should most probably have a reference to what actually failed02:14
nashYeah, but the protocol should handle it - that way the server can send objects back out of order for instnace02:14
mithroif the fail frame had something like a GRS stuff, would that be okay?02:15
CIA-20[email protected] * libtpclient-py-dev/tp/client/ (10 files): Better documentation on what the Threads do.02:15
nashCould be.02:15
mithrothe GRS needs a little work - but it's pretty good02:15
nashYeah - I should actually implement it in the client02:16
nashWill be done soon02:16
nashFor user UI you want it02:16
nashSo... are you working towards MTSec release02:16
mithronash: i cleaning up a few smaller things at the moment, don't have the patiences at the moment to work on big tasks like my honors project or the starmap code02:17
nashDo you know the state of MTsec in the server(s)?02:18
mithroyou see Lee's post?02:18
mithroSubject: Re: [tp-devel] nash's requested status of MTSec and tp0402:19
mithroDate: Wed, 28 Feb 2007 01:23:37 +1300  (Tue, 22:53 CST)02:19
nashYes, /me had forgotten about it there02:19
mithrofor tp04 I would like to move to using libtpproto2-py, which should be ~API compatible with libtpproto-py but a lot cleaner and a huge amount less code (making it easier to maintain and update)02:21
nashmithro: Seriously - your #1 priority should be getting someway of playing MTSec02:23
mithronash: well, I'm dependent on the server really02:23
nashBut we agree that TP04, or at least TP03 is practically necessary for mtsec (for real games)02:25
mithrothe client is actually quite a long way towards supporting MTSec (the client already supports creating and updating designs)02:25
mithrothe client even calculates design properties locally02:27
mithroit also supports all the order arguments needed by MTSec (that I can see)02:29
mithroso basically somebody needs to hurry up and impliment it in the server02:29
nashBut has it been tested or anything?02:30
mithronash: can you see anything missing? (apart from the obvious UI problems)02:30
nashNothing obvious02:30
mithronash: the design stuff was being tested - but it hasn't been recently (as it's not needed for MiniSec) and nobody is working on MTSec in the server at the momnet02:31
nashOkay - so it's llnz who needs to be kicked?  Or is there multiple servers?02:31
mithrothere are multiple server, i wrote tpserver-py because Lee was busy with his studies02:32
nashSo how is tpserver-py in regards to mtsec?02:32
nashBTW: Is there any plans to MTSec ++?02:37
nashIe the next game after MTSec?02:37
nashmithro: Anywyay I'm about to have lunch...02:46
nashHave fun... ;-)02:46
*** mithro has quit IRC02:51
*** mithro has joined #tp02:51
CIA-20[email protected] * tpclient-pywx-dev/ (16 files in 2 dirs): Ask to update at end of turn.04:09
CIA-20[email protected] * tpclient-pywx-dev/windows/ (
CIA-20Use Yes/No instead of Ok/Cancel.:04:09
CIA-20 With Ok/Cancel the Cancel button is always the default. With Yes/No I can set yes to be the default.04:09
* mithro closes04:11
tpbBug #1333653: Properly support EOT04:11
mithrothat was a good bug to fix - kind of makes 10 second turns a bit annoying however04:24
nash10s are great for testing AIs.04:24
nashHowever no good for humans04:24
* nash notes "I'm finished is even better"04:24
mithronash: thats pending a protocol update :)04:25
* mithro ponders, whats a sty file04:26
mithroooo, somebody used a sony playstation portable to access our website04:28
mithroanyway got to run to a lecture04:29
nashhave fun04:29
*** mithro has quit IRC04:35
*** mithro has joined #tp04:36
* mithro ponders04:37
* mithro goes offline to concentrate on his lecture for once05:05
*** mithro has quit IRC05:24
*** jotham is now known as jewishbanker05:48
*** jewishbanker has left #tp05:59
*** tpb has joined #tp06:00
*** ChanServ sets mode: +o tpb06:00
*** nash has quit IRC06:33
*** JLP has quit IRC07:42
*** JLP has joined #tp07:46
*** llnz has joined #tp10:04
* llnz ponders10:06
*** llnz2 has joined #tp10:27
*** llnz has quit IRC10:27
*** llnz2 is now known as llnz10:27
*** JLP has quit IRC10:37
*** llnz2 has joined #tp10:43
*** llnz has quit IRC10:46
*** llnz2 is now known as llnz10:46
llnzblog updated10:47
llnzlater all10:47
*** llnz has quit IRC10:47
*** JLP has joined #tp11:33
*** JLP has quit IRC11:50
*** JLP has joined #tp12:14
*** JLP has quit IRC12:34
*** tuna-fish has quit IRC13:11
*** tuna-fish has joined #tp14:14
*** JLP has joined #tp14:57
CIA-20Jure Repinc <[email protected]> * web/news/ ( Links for C++ Server 0.4.1 news16:37
CIA-20Jure Repinc <[email protected]> * web/links.php : Two more Thousand Parsec links16:37
CIA-20Jure Repinc <[email protected]> * web/ (faq.php faq.php): New FAQ entry: Nothing happens when a new turn starts17:19
*** zipola has joined #tp17:31
*** zipola has quit IRC20:54
*** llnz has joined #tp22:40
*** nash has joined #tp23:00
* nash waves23:02
llnzhi nash23:02
nashhow is life?23:03
tpbI guess you are new, llnz23:05
nashI'm pretty cool.  tpb is an idiot ;-)23:12
nashSorry again about that bug23:12
llnzno problem, don't worry about it23:13
nashSo what are you up to?  Any idea when the server will be able to do MTSec23:14
nashYou seem to be the critical path for that...23:15
nashAs I was saying to mithro yesterday, minisec is a nice tech demo, MTSec however seems like quite a playable game23:18
llnzI intend to work on it soon23:20
nashCool ;-)23:21
* nash will be moving his client to support it too ;-)23:21
nashI'm currently improving my client to make it a lot more efficient23:30
llnzon the server side there are a few changes that need to be made, so it will take a while23:30
nashYou have your own dev branch I assume?23:30
llnzno, not really23:32
llnzi'm thinking about one though, until the AI comp is closed23:32
nashDoes that make it hard to do non-trivial changes then?23:34
llnzit's harder to keep everything in sync, and being able to commit the dev branch to the stable branch23:38
llnzmithro can't do that with his two dev braches because of some changes made to the stable branches23:39
nashBut the mithro has a long time between syncs23:39
* nash currently has 4 branches for his client dev23:39
llnzwell, i guess I have a local dev branch, and the repo on is the stable one23:39
nashTo be fair - one is the online source - so I really have 3 branches23:40
nashSo you do have a dev branch then ;-)23:40
nashSo you can commit a partial unworing code without breaking the main server23:40

Generated by 2.5 by Marius Gedminas - find it at!