Wednesday, 2007-05-23

*** tpb has joined #tp00:00
*** ChanServ sets mode: +o tpb00:00
mithronash: ping?00:01
mithrodmpayton: ping?00:01
dmpaytonmithro: pong00:01
mithrohey dmpayton00:01
mithroso how goes everything?00:01
Epyonping all00:01
EpyonxP00:01
mithrohey Epyon00:01
mithrowhat is the time where you are?00:01
dmpaytonmithro: Everything's alright. Finally gonna get health insurance. :D00:02
Epyon6 in the morning. High time to get some sleep :)00:02
mithrodmpayton: oh? Health Insurance is so weird in the US00:02
dmpaytonmithro: Seriously. Atleast our plan is gonna be relatively cheap. $200/mo for my fiance and I.00:03
mithrodmpayton: I seem to remeber most people get health insurance through their job?00:03
* mithro 's health insurance means it only costs me $100 for new glasses each year00:04
dmpaytonmithro: Not all jobs offer health coverage...depends on if the company can afford it, which, in the case of my current employer, they can't00:05
nashmithro: pong00:09
mithronash: I got disconnected when I was at uni00:09
mithrodid you manage to get anywhere?00:10
nashmithro: No, as I said - are you sure it is not a compilation issue?00:10
mithrodmpayton: so when will we start seeing code?00:10
mithrocompilation issue?00:10
nashYes - or magic variable iussues00:12
mithropossibly00:12
mithroactually I just did a diff on the binaries and think I know what the issue is now00:13
nashokay00:15
mithronash: and could you see how I could can the execv to execv_git_cmd ?00:16
nashi assume you meant change execv -> execv_git_command00:18
nashAnd the short answer is... not particularly easily... unless you paste a git- on the front, and let execv_git_command peel it off...00:19
mithroyou mean "git-git-cvsserver"?00:21
mithrodmpayton: your have gone very quiet :)00:24
nashNo, when execv_git_command executes, it turns git-foo into git foo00:32
mithrooh00:33
mithroi don't think that would work very well00:33
mithronash: anyway to get past needing to have the full path in execv?00:34
nashexecvp00:35
nashsearchs $PATH00:35
mithronash: can you check that you can still pull and push?00:41
nashtrying to pull now00:42
nashI can fetch00:43
mithrookay seems cvs is working too00:50
nashsweetness & light00:54
mithrolight?00:55
dmpaytonmithro: Sorry for the lack of reply. Just got hit with some really bad news.00:56
mithrodmpayton: damn :/00:57
mithronash: does the parent bug effect galaxie in anyway?01:22
nashNot directly, but it makes the UI strange01:23
mithrookay - was preventing the AI playing on tpserver-py atm?01:24
nashIssues with orders last I saw - but I was having an issue with tpserver-cpp as well, so I must have broken something I think01:26
mithroorders where completely not working until last nights commit01:26
mithrothe server was just throwing away all the arguments01:27
nashOkay - I'll give it a burl tonight if I can, and see if it is working now01:27
mithrookay - that would be good01:30
* pschulz01 wonders if his 100 battleships have been built yet?01:32
mithroeventually I'm going to need to write some tests01:32
mithroit would be really good if you could do a "tp-server-test" and it would run through and test all the differnet parts01:32
nashI agree... Also need some tpclient tests as well ;-)01:34
mithroi have no idea how to do tests for tpclient-pywx01:34
nashheh01:36
mithroTesting UI's suck :/01:36
nashI can look at some for galaxie.  There are somethings that are easy to do01:36
mithrothen of course you have the fact that libtpproto-py dynamically builds the order classes01:37
nashheh01:41
mithrohopefully, when I get back to libtpproto2-py I can acutally have some tests01:42
*** dmpayton has quit IRC02:02
mithronash: i'm not quite sure but I think this parent bug is a sqlite bug03:15
nashmithro: Okay... sqlite being used by many many people in many applications, tpserver-py being used by 2... I'll agree with that03:16
mithronash: well - I'm getting strange results in my sql query (which is a rather evil query)03:17
mithrowhere = (((c.size+bp_s)*(c.size+bp_s)) >= \03:17
mithro((c.posx-bp_x) * (c.posx-bp_x)) + \03:17
mithro((c.posy-bp_y) * (c.posy-bp_y)) + \03:17
mithro((c.posz-bp_z) * (c.posz-bp_z)))03:17
mithrocan see see a problem with that?03:18
nashAre the types right?03:18
mithroseems so03:18
nashno overflow etc?03:18
nashSo it's supposed to be pythag in 3d?03:19
mithronash: yeah03:19
mithrohave I stuffed up?03:19
mithrohrm - it could be overflowing I guess03:19
nashShould it really be c.size+bp_s or c.size-pb_s, having NFI what03:20
nashc.size and bp_s is03:20
nashAnd yes, these numbers will overflow happily03:20
nashIs there a hypot function?03:20
mithronash: not that I can see03:21
mithrothere is no pow function03:22
mithrootherwise it would be03:22
nashMost objects are either really close, or really far - so use a linear difference formula - delta x + delta y + delta z, and see what happens03:22
mithro(c.size+bp_s)**2 >= (c.posx-bp_x)**2 + (c.posy-bp_y)**2 + (c.posz-bp_z)**203:23
mithrostill occuring with the following03:28
nashWell yeah.03:29
nashAs I said use    (c.size+bp_s) (c.posx-bp_x) +  (c.posy-bp_y) + (c.posz-bp_z)03:29
nashAs I said use    (c.size+bp_s)  >=  (c.posx-bp_x) +  (c.posy-bp_y) + (c.posz-bp_z)03:29
nashNot correct but it will avoid overflow03:30
mithronash: yeah - it seems to work03:30
mithromust be an overflow issue?03:30
nashMost likely03:31
nashWhich is not a bug in sqllite really03:31
nashwhere clauses aren't meant for complex maths03:31
mithronash: i know - but I can't see an easier way to do it?03:34
mithroI guess I could select all of them into a numeric array03:34
nashUse the linear version, then do a second filter when it comes over03:35
nashusing hypotll or the equiv in python03:38
mithronash: I'm pretty sure one exists :P03:41
nashOr turn that into a SQL function03:41
mithrothat isn't portable03:42
nash long double hypotl(long double x, long double y) {03:44
nash03:44
nash long double ax, ay;03:44
nash03:44
nash /* If x or y is a +-Inf, return +Inf. */03:44
nash if (isinf(x) || isinf(y))03:44
nash return (-1.L / zero);03:44
nash03:44
nash /* If x or y is a NaN, return NaN. */03:44
nash if (isnan(x) || isnan(y))03:44
nash return (x*x+x) + (y*y+y);03:44
nash03:44
mithronash: umm?03:44
nash ax = fabsl(x);03:44
nash ay = fabsl(y);03:44
nash03:44
nash /* Special case if x and/or y is zero. */03:44
nash if (ax == zero || ay == zero)03:44
nash return (ay + ax);03:44
nash03:44
nash /* Avoid overflow in sqrtl(x*x + y*y) */03:44
nash if (ax >= ay) {03:44
nash ay /= ax;03:44
nash ax = ax * sqrtl(1.0L + ay * ay);03:44
nash } else {03:44
nash ax /= ay;03:44
nash ax = ay * sqrtl(ax * ax + 1.0L);03:44
nash }03:44
nash03:44
nash return ax;03:45
nashahh03:45
nashbugger03:45
nashif (ax >= ay) {03:45
nash        ay /= ax;03:45
nash        ax = ax * sqrtl(1.0L + ay * ay);03:45
nash} else {03:45
nash        ax /= ay;03:45
nash        ax = ay *  sqrtl(ax * ax + 1.0L);03:45
nash}03:45
nashThat bit there can be done as a where clause03:45
nashwhere (ax > ay  && ...) or (ay < ax && ...)03:45
mithrothat could be done as03:46
mithrowhere (ax > ay && (ax*sqrt(1.0L+(ay/ax)*(ay/ax)) > (s+s)) ....03:48
nashmay help03:49
nashANyway - I'm off home03:49
nashtalk to you later...03:49
*** nash has quit IRC03:49
CIA-3mithro tpserver-py * r10e07b64c902 /tp/server/ (__init__.py bases/SQL.py): Fixes for the changes introduced in the libtpproto-py cleanup.03:55
CIA-3mithro tpserver-py * rc5d64c3fc4e9 /tp/server/db/__init__.py:03:55
CIA-3Better way of doing the DB proxy.03:55
CIA-3Should now support all the different ways you can call execute.03:55
CIA-3mithro tpserver-py * r8cc5cbf0f613 /tp/server/bases/Order.py: No longer need to parse the extra value.03:55
CIA-3mithro tpserver-py * r8eba7a6daed9 /tp/server/bases/Object.py: Use 64 bit Integers. Use a linear calculation for positioning.03:55
*** pschulz01 has quit IRC04:24
*** nash has joined #tp04:38
mithrowb nash05:02
*** llnz has joined #tp05:30
*** pschulz01 has joined #tp05:44
mithrowb pschulz0105:47
mithrohey llnz05:47
pschulz01mithro: I'm discussing packaging in #ubuntu-motu05:48
mithropschulz01: okay cool05:48
* mithro pokes TBBle 05:48
*** llnz2 has joined #tp05:49
mithrowb llnz :P05:50
*** llnz has quit IRC05:50
*** llnz2 is now known as llnz05:50
llnzhi mithro05:50
mithrohow is everything?05:52
*** pschulz01 has quit IRC05:53
*** jotham has quit IRC05:53
TBBleAnime Society computer died. >_<05:54
*** pschulz01 has joined #tp05:54
*** jotham has joined #tp05:54
mithroTBBle: ?05:54
mithroTBBle: pschulz01 has been working on some debs05:55
mithrojotham: ping?05:55
mithronash: ping?05:56
nashmithro: pont?06:15
nashs/t/g06:15
mithropont has a cool sound to it06:16
nashshame it means nothing06:19
mithronash: just trying to get galaxie issuing orders to tpserver-py06:21
*** tuna-fish has quit IRC06:22
nashIt seems to be having grief with order names06:23
nashANywya - I need to go for a while06:26
nashI should have more TP time on the weekend.06:26
pschulz01nash: I'm on my way to pulverise 'nash Star System'06:26
*** tuna-fish has joined #tp06:29
*** tuna-fish has quit IRC06:35
TBBleOh, goody.06:48
pschulz01TBBle: Just figured out some stuff!!! Expect code debian packages for tpserver-cpp shortly.06:49
TBBleExcellent. You're working on the git tree or the tarballs?06:55
pschulz01TBBle: git06:55
pschulz01TBBle: .. but there is no reason that it won't apply to the tarballs.06:56
mithroTBBle has been working on packaging the python stuff06:56
*** Demitar has quit IRC06:56
pschulz01TBBle: If it's in git then guy guys can edit package descriptions etc. :-)06:56
pschulz01you guys06:57
pschulz01Eg.. what should tpserver-cpp-scheme-guile contain?06:57
*** tuna-fish has joined #tp06:59
*** tuna-fish has joined #tp07:02
pschulz01back in a little while..07:04
*** tuna-fish has quit IRC07:23
*** tuna-fish has joined #tp07:23
*** llnz2 has joined #tp07:25
mithrollnz: connection playing07:28
llnz2yeah07:39
*** llnz has quit IRC07:39
*** llnz2 is now known as llnz07:39
mithrohrm....07:57
pschulz01How do I capture a planet?08:00
mithropschulz01: bomb the hell out of it08:01
mithropschulz01: then recolonise08:02
pschulz01so.. 20 battleships's arn't enough?08:02
mithropschulz01: that should be enough08:03
mithroyou need a frigate to colonise however08:03
pschulz01mithro: Oh.08:03
mithrobattleships can't colonise08:03
pschulz01I'll have to stop my '100 battleship build'.. or can I insert a 'frigate build order before it?08:04
pschulz01That seems to hav eworked.08:05
mithropschulz01: you can insert before in theory :P08:05
mithronash: ping?08:05
pschulz01In practice too..08:06
nashmithro: ?08:11
nashI found the orders bug08:11
mithronash: you send order probes right?08:12
nashYeah..08:12
mithrowhere they broken too?08:12
mithroeither there is a problem in my decoding - or your sending something wrong08:13
nashThe probes are working08:13
nashBut I'm getting a problem on the first turn08:13
mithronash: they aren't occuring here08:14
nashAuto fixes on 2nd and later turns... but tpserver-py never sends a 'new turn' message08:14
mithros/occouring/working08:14
mithronash: you need to send it a sigusr108:14
nashThere now...08:15
nashsending a sigusr1 solves the issue.08:15
mithro?08:15
nashI'm about to send a patch up08:15
mithrookay08:16
CIA-3nash galaxie * r2ec6992d6df0 / (tpe_orders.c tpe_sequence.c): Fix sequence bug with UINt32 and UINT64 confusion.08:18
CIA-3nash galaxie * rd6f4d24fa914 /tpe_ship.c: Same bug in ship designs too.08:18
nashTry that08:19
mithroData: '\x00\x00\x00\x02\xff\xff\xff\xff\x00\x00\x00\x02\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00' Structure: [ISI][II] Extra Data found: '\x00\x00\x00\x00\x00\x00\x00\x00'08:20
mithrodo you actually look at OrderDescs or just assume buildfleets are the same?08:22
nashAt this point I just assume08:25
nashHaven't fixed that yet08:25
nashhmm... ** Error: Seq 17: Err 4: No such <class 'tp.server.bases.Order.Order'>.08:25
mithro?08:25
mithronash: when did you do the last pull of tpserver-py and libtpproto-py ?08:26
nashyesterday08:27
mithrocan you try an update?08:27
mithroany luck?08:31
mithrowhat format do you use for BuildFleet?08:34
nashUmm..08:41
nashIt was what was documented in one of the places...08:42
mithroit's documented?08:42
nashWhat I send is:08:42
nash<objid><slot><build id><0><0><0><N items (<type>,<number>)... >>08:43
mithrohrm....08:45
mithronash: seems to work if I add the extra ARG_STRING08:47
nashOne of those zeros is that string...08:48
nashoh... also there is the name at the end08:49
nashmithro: For the AI comp I just got it to work with tpserver-cpp ;-)08:50
nashDynamic order parsing is on the list todo - but I haven't figured out how to do it sanely for things like AIs... there is no semantic meaning in the protocol08:50
CIA-3mithro tpserver-py * r36f3ecd431db /tp/server/rules/minisec/orders/BuildFleet.py: Added name attribute.08:51
mithro:P08:51
nashmithro: I still have  build in the quere however08:52
nashOh... now it is 'Unknown'08:53
mithrohow do I check?08:53
nashClick on the object...08:53
nash(So <x> Solar System, then down the bottom click on the <x> Planet08:54
nashYou will see the order queue there.08:54
nashmithro: Anywya - I'm heading to be shortly08:54
mithrookay08:54
nashTalk to you tomorrow08:54
pschulz01mithro: Which guile packages do I need to build tpserver-cpp guile files?09:02
mithroyou mean mzscheme?09:02
llnzpschulz01: guile-1.6-dev09:02
pschulz01I have mzscheme building OK.09:02
pschulz01.. just need guile now.09:03
llnzor maybe guile-1.8-dev09:03
pschulz01llnz: ta09:03
pschulz01llnz: Just the file(s) in usr/share/tpserver/tpscheme/guile in guile package?09:06
llnznot just them, also usr/lib/tpserver/tpscheme/libguile*09:08
llnzie, library and the static data file09:09
pschulz01is there an equivalent usr/lib/tpserver/tpscheme/libmzscheme*09:10
llnzyes09:11
llnzusr/lib/tpserver/tpscheme/libtpguile*09:11
*** _JLP_ is now known as JLP09:25
JLPahoy people09:26
llnzhi JLP09:26
mithrohey JLP09:28
mithrocan't figure out how to solve this problem09:34
mithroahh huh!09:57
CIA-3mithro libtpproto-py * rf0ceb19064a8 /tp/netlib/xstruct.py: Make the error message a bit more informative.10:06
CIA-3mithro libtpproto-py * rbf7f13f42264 /tp/netlib/objects/OrderDesc.py: Make the error message easier to figure out.10:06
CIA-3mithro libtpproto-py * r78ef55e7c72d /tp/netlib/objects/Description.py: More informative messages.10:06
CIA-3mithro libtpproto-py * r157d5c2f6801 /tp/netlib/objects/Order_Probe.py: Fix the Order Probe message for the new order stuff.10:06
CIA-3mithro libtpproto-py * r533856fdc77e /tp/netlib/objects/OrderDesc.py: Start with an empty substruct.10:06
CIA-3mithro libtpproto-py * r1fc3532cfb32 /tp/netlib/server.py:10:07
CIA-3Use the type given by the number.10:07
CIA-3Fixes problem trying to differentiate probe and insert order packets.10:07
CIA-3mithro libtpproto-py * rc2ca77367d60 /tp/netlib/objects/Order.py: Allow the number of turns to be -1.10:07
CIA-3mithro tpserver-py * r5803ae769484 /tp/server/rules/minisec/orders/BuildFleet.py: Fixed the default value. Unpack the value.10:08
CIA-3mithro tpserver-py * r1b6e82022220 /tp/server/rules/minisec/objects/Fleet.py: Small cosmetic fix.10:08
pschulz01llnz: still around?10:12
pschulz01llnz: See http://git.mawsonlakes.org/cgi-bin/gitweb.cgi?p=tp/tpserver-cpp10:15
tpb<http://ln-s.net/YBr> (at git.mawsonlakes.org)10:15
pschulz01lI need to find out what all the dependancies are as I'm still not building the mzscheme and guile libraries...10:16
pschulz01Good night all.10:16
llnzpschulz01: cool10:20
llnzgood night10:20
* llnz wanders off10:20
llnzlater all10:21
*** llnz has quit IRC10:26
*** Demitar has joined #tp10:31
*** Demitar has quit IRC10:34
CIA-3mithro libtpproto-py * rb4403b352d59 /tp/netlib/xstruct.py: Catch when people do stupid things like give you to many arguments.10:36
CIA-3mithro libtpproto-py * rc3ba1f07710c /tp/netlib/objects/OrderDesc.py:10:36
CIA-3Should set substruct to empty, not struct.10:36
CIA-3Remove extra prints.10:36
CIA-3mithro tpclient-pytext * r3a6b3629ff7c /tpclient-pytext:10:36
CIA-3Fixed up some str's.10:36
CIA-3Move to the new apply python syntax.10:36
CIA-3mithro tpclient-pytext * rce47f7aeff99 /.gitignore: GIT Ignore file.10:36
tpbaloril has quit worldforge (Ping timeout: 620 seconds)11:29
tpbaloril has joined on worldforge11:41
tpbaloril has quit worldforge (Remote host closed the connection)11:46
tpbaloril has joined on worldforge11:51
tpbaloril has quit worldforge (Read error: Connection reset by peer)12:03
tpbaloril has joined on worldforge12:03
*** dmpayton has joined #tp12:23
mithrohey dmpayton12:55
dmpaytonHey mithro14:17
*** dmpayton has quit IRC16:47
*** xdotx has joined #tp16:53
*** Demitar has joined #tp17:08
*** nash has quit IRC17:42
*** nash has joined #tp18:33
*** pschulz01 has quit IRC18:41
*** mithro has quit IRC19:17
*** mithro has joined #tp19:31
mithromorning people19:33
mithro~seen niphree19:33
tpbmithro: niphree was last seen in #tp 2 days, 10 hours, 22 minutes, and 15 seconds ago: <niphree> mithro: brb - another project meeting :]. I'll be back in few hours19:33
mithro:/19:33
mithrohey nash19:52
mithroo so quiet20:32
mithro~seen jotham20:32
tpbmithro: jotham was last seen in #tp 1 week, 5 days, 20 hours, 13 minutes, and 9 seconds ago: <jotham> cool20:32
mithrojotham: ping?20:32
nashheyo mithro20:38
mithrobblr21:07
*** mithro has quit IRC21:27
*** xdotx has quit IRC22:19
nashtpb: seen dystopicfro22:25
tpbnash: dystopicfro was last seen in #tp 3 weeks, 1 day, 20 hours, 46 minutes, and 25 seconds ago: <DystopicFro> nash: I'll email you the location that you can clone from when I get it up22:25

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