| *** tpb has joined #tp | 00:00 | |
| *** ChanServ sets mode: +o tpb | 00:00 | |
| epyon-kitsune | llnz: what non-tpframe network connections happen to a tpserver? | 00:03 |
|---|---|---|
| epyon-kitsune | (disregarding http wrapping) | 00:03 |
| llnz | incoming to the server: none (a separate web server will host some content though) | 00:04 |
| llnz | outgoing from the server: http to the metaserver, dbus to dbus for avahi (if needed) | 00:05 |
| epyon-kitsune | great | 00:06 |
| llnz | epyon-kitsune: any particular reason for copying files and not just updating the existing ones? | 00:09 |
| epyon-kitsune | I think the reason will become apparent soon enough. | 00:10 |
| epyon-kitsune | I'm just writing the frame class from scratch | 00:10 |
| epyon-kitsune | If I did that in the repo, it'd just become a big mess | 00:10 |
| llnz | it wouldn't hurt if the new frame class implemented outputframe and inputframe | 00:10 |
| llnz | at least for now | 00:10 |
| llnz | or they became wrappers around the new frame class | 00:11 |
| epyon-kitsune | actually I wonder if a FrameFactory wouldn't be better... but I'll leave that for later -- I need to get this stuff testable ASAP | 00:11 |
| llnz | if you can get the byte buffers to and from the network, you should be ok | 00:12 |
| epyon-kitsune | yeah, and that's where I'm heading right now. | 00:12 |
| epyon-kitsune | llnz: I'll take the liberty of doing some hacks to speed it up. We'll be able to clean them up later. | 00:18 |
| epyon-kitsune | (speed up getting the server working that is) | 00:19 |
| llnz | i suggest you don't break too many things, processing the frames is one of those | 00:19 |
| epyon-kitsune | processing? | 00:20 |
| llnz | ie, everything inside playeragent and probably some in playerconnection | 00:23 |
| epyon-kitsune | llnz: yeah, I'll try to keep those untouched. | 00:23 |
| llnz | ok, cool | 00:23 |
| *** nash has joined #tp | 01:19 | |
| *** ezod has quit IRC | 02:42 | |
| *** ezod has joined #tp | 03:07 | |
| *** glew has quit IRC | 03:27 | |
| alanp | *: ping | 03:37 |
| llnz | hi alanp | 03:43 |
| alanp | hey | 03:45 |
| alanp | was looking for cab fare estimate ejc -> mtv | 03:46 |
| alanp | looks like about $30-40 :( | 03:46 |
| alanp | but cab: 20 mins transit: 2+ hrs | 03:46 |
| alanp | might be worth it | 03:46 |
| *** verhoevenv_ has quit IRC | 03:49 | |
| llnz | alanp: it appears to be about an hour by public transport | 03:51 |
| alanp | ? | 03:51 |
| alanp | google maps shows me 2+ | 03:51 |
| llnz | http://maps.google.co.nz/maps?f=d&source=s_d&saddr=SJC,+San+Jose,+CA,+United+States&daddr=1786+West+El+Camino+Real,+Mountain+View,+CA,+United+States&hl=en&geocode=FacrOgId7o67-Cnty17wvsuPgDGQxhv1NnD4xQ%3B&mra=ls&dirflg=r&ttype=dep&date=07%2F08%2F10&time=09:00&noexp=0&noal=0&sort=&sll=37.392383,-122.092674&sspn=0.005029,0.011362&ie=UTF8&ll=37.402074,-122.048149&spn=0.080456,0.181789&z=13&start=0 | 03:52 |
| alanp | oh fail to look | 03:52 |
| tpb | <http://ln-s.net/7PdB> (at maps.google.co.nz) | 03:52 |
| alanp | wtf! | 03:52 |
| alanp | why do the list the 2 hour trip time first | 03:52 |
| llnz | timing? | 03:52 |
| alanp | i guess :( | 03:53 |
| llnz | also, if you pay a little more, catch the 522 express bus - it's faster (you might have to walk a bit further because it only stops at some places) | 03:55 |
| alanp | ah | 03:56 |
| alanp | go to sf much? | 03:56 |
| llnz | i have visited three times, once in 1999 with family (mostly around SF), twice visited Google ('07 and '08) | 03:58 |
| alanp | ah cool | 03:58 |
| llnz | the south bay area has a very good bus system, and the train to SF is quite usable | 03:59 |
| *** DTRemenak has quit IRC | 04:32 | |
| llnz | bbl | 04:34 |
| *** llnz has quit IRC | 04:34 | |
| *** nash has quit IRC | 04:37 | |
| *** DTRemenak has joined #tp | 05:01 | |
| *** llnz has joined #tp | 06:24 | |
| *** cahirwpz has joined #tp | 06:53 | |
| *** DTRemenak has quit IRC | 07:10 | |
| *** DTRemenak has joined #tp | 07:13 | |
| *** bisc has quit IRC | 08:14 | |
| *** tote has quit IRC | 09:28 | |
| *** tote has joined #tp | 09:37 | |
| * llnz wanders off | 09:54 | |
| llnz | later all | 09:54 |
| *** llnz has quit IRC | 09:54 | |
| *** StupidIncarnate has quit IRC | 11:09 | |
| *** epyon-kitsune has quit IRC | 12:35 | |
| *** epyon-kitsune has joined #tp | 12:35 | |
| *** Greywhind has joined #tp | 13:18 | |
| *** glew has joined #tp | 14:21 | |
| ezod | glew: ping | 14:53 |
| glew | ezod: pong | 14:54 |
| ezod | hi | 14:54 |
| ezod | anything new? | 14:54 |
| ezod | i guess from your commits you're testing saving atm? | 14:55 |
| glew | yeah I tested some saving | 14:56 |
| glew | I also played a little MTsec so i could see the sqlite database it made | 14:56 |
| glew | I was going to try modifiying information in it | 14:57 |
| *** PrimusInterPares is now known as Linkadmin | 14:57 | |
| glew | and I also noticed my addslashes() function probably isnt working correctly because the fields that use it are chopped to just the first word | 14:57 |
| ezod | can you load yet? | 15:03 |
| ezod | (i haven't tried) | 15:03 |
| *** verhoevenv has joined #tp | 15:05 | |
| glew | ezod: I think loading works, although I' | 15:06 |
| glew | m not sure if the ai loading is fine | 15:06 |
| glew | it should be cause it just starts the client right now | 15:06 |
| ezod | ok | 15:09 |
| *** null_000 has joined #tp | 15:12 | |
| null_000 | hi all | 15:18 |
| *** null_000 has quit IRC | 15:33 | |
| *** null_000 has joined #tp | 19:09 | |
| null_000 | hey all | 19:10 |
| *** glew has quit IRC | 19:11 | |
| *** glew has joined #tp | 19:11 | |
| *** tpb has joined #tp | 19:19 | |
| *** ChanServ sets mode: +o tpb | 19:19 | |
| *** llnz has joined #tp | 20:10 | |
| llnz | morning all | 20:11 |
| null_000 | morning llnz | 20:26 |
| llnz | hi null_000 | 20:27 |
| null_000 | I saw in the logs there were some questions about the AI | 20:27 |
| null_000 | do you know anything about that? | 20:27 |
| llnz | nope | 20:28 |
| * llnz opens the irc logs | 20:28 | |
| llnz | http://www.thousandparsec.net/~irc/logs/%23tp.2010-08-03.log.html#t2010-08-03T17:51:08 | 20:29 |
| tpb | <http://ln-s.net/7Q35> (at www.thousandparsec.net) | 20:30 |
| llnz | ? | 20:30 |
| null_000 | yes this one | 20:31 |
| null_000 | if you don't know anything about that I'll have to find glew | 20:32 |
| llnz | i don't know about it | 20:35 |
| verhoevenv | null_000: I don't know the exact details, but glew wants to be able to save and load AI state from a database for single player save game support. | 20:37 |
| verhoevenv | Which I believe not to be very hard in the original daneel-ai, but I have no idea how easy or hard it would be in your AI. | 20:37 |
| null_000 | oh right... we talked about that a bit | 20:38 |
| null_000 | verhoevenv: if I remember correctly your AI is stateless (no need to save any data) | 20:38 |
| verhoevenv | I think there is support for turn-transcending state in the form of the don't-remove-these constraints. | 20:39 |
| verhoevenv | Marked with *, I think. | 20:39 |
| verhoevenv | Apart from that, the rest is rebuild each turn from the downloaded information. | 20:39 |
| ezod | the easy thing to do is to add some function that outputs a pickled string of a dict containing all stateful objects in the AI | 20:40 |
| null_000 | this can be easily done for MTSec | 20:41 |
| ezod | if other AIs in other languages can similarly "pickle" and "unpickle" their state, then it's a viable general process | 20:41 |
| null_000 | but I'm not sure where the constraints are stored | 20:41 |
| * null_000 has to go | 20:45 | |
| null_000 | bye all.. | 20:45 |
| *** null_000 has quit IRC | 20:45 | |
| *** llnz2 has joined #tp | 21:00 | |
| *** llnz has quit IRC | 21:00 | |
| *** llnz2 is now known as llnz | 21:00 | |
| *** Greywhind has quit IRC | 21:14 | |
| *** null_000 has joined #tp | 21:15 | |
| *** null_000 has quit IRC | 21:53 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!