| 00:15.53 | mrapple | mdskpr_: if your bored, an old bug I never looked in to https://sourceforge.net/tracker/index.php?func=detail&aid=3114569&group_id=3248&atid=103248 |
| 00:16.19 | mrapple | here's a helpful backtrace http://pastebin.com/fRkLWSRn |
| 00:16.23 | mdskpr_ | is never bored |
| 00:16.26 | mdskpr_ | but sure |
| 00:17.04 | meem1029 | Anyone have ideas as to why 2.0.16 would segfault on a specific server when it never has before now? |
| 00:17.04 | blast007 | JeffM: I wonder if the 'localhost' lookup is a limitation in my router |
| 00:17.16 | JeffM | could be |
| 00:17.24 | blast007 | I'm using my router as my DNS resolver (which is supposed to forward it on) |
| 00:17.32 | blast007 | an nslookup for localhost doesn't work |
| 00:17.39 | JeffM | mrapple, did you see that I wrote a website spec? |
| 00:17.44 | JeffM | or at least the start |
| 00:17.53 | mrapple | JeffM: when did you write that? |
| 00:17.58 | JeffM | that day |
| 00:18.01 | JeffM | when I was at lunch |
| 00:18.15 | mrapple | yes i saw that, the same day :P |
| 00:18.21 | mdskpr_ | mrapple: that bug has been fixed |
| 00:18.26 | mrapple | oh really? |
| 00:18.30 | mdskpr_ | look at the comments |
| 00:19.06 | mrapple | oop just kidding |
| 00:19.06 | blast007 | meem1029: what OS are you on? |
| 00:19.09 | mrapple | i forgot BC fixed it |
| 00:19.14 | meem1029 | ubuntu 10.10 |
| 00:19.55 | blast007 | meem1029: okay, as a test, navigate into ~/.bzf/ and rename the cache directory to cache.old |
| 00:19.58 | blast007 | then try joining again |
| 00:20.06 | blast007 | might be a corrupted cache file |
| 00:20.24 | meem1029 | Ah, I will try that soon. I just decided to try uninstalling and reinstalling via apt. |
| 00:23.43 | meem1029 | yep, that worked |
| 00:23.53 | blast007 | k, cool |
| 00:24.05 | meem1029 | thanks! |
| 00:24.31 | blast007 | now if you'd want, you can copy the good cache file back into your old directory (overwriting the bad file) so you don't have redownload every map/texture again |
| 00:40.33 | *** join/#bzflag moiree_ (~moiree@dslb-084-059-052-228.pools.arcor-ip.net) |
| 00:40.48 | blast007 | JeffM: ah, interesting, when I ping localhost, I get the IPv6 loopback (::1). Wonder if ares is getting *that* and then the game is choking |
| 00:41.06 | JeffM | would not suprise me |
| 00:43.28 | *** join/#bzflag ahs3- (~ahs3-@adsl-065-005-193-158.sip.rdu.bellsouth.net) |
| 00:46.19 | *** join/#bzflag ahs3- (~ahs3-@adsl-065-005-193-158.sip.rdu.bellsouth.net) |
| 00:49.28 | CIA-114 | BZFlag: 03dtremenak * r21523 10/trunk/bzflag/ (27 files in 26 dirs): Fix build for directories with spaces in their names. Set some dependencies correctly. Move plugin projects in fullbuild into their own solution-folder. |
| 00:54.27 | *** join/#bzflag ahs3- (~ahs3-@adsl-065-005-193-158.sip.rdu.bellsouth.net) |
| 00:55.45 | *** join/#bzflag ahs3- (~ahs3-@adsl-065-005-193-158.sip.rdu.bellsouth.net) |
| 01:02.22 | *** join/#bzflag Guest764 (ca59b013@gateway/web/freenode/ip.202.89.176.19) |
| 01:02.30 | Guest764 | delusional |
| 01:02.32 | Guest764 | .... |
| 01:03.43 | delusional | thats me guest784 |
| 01:03.50 | delusional | what's up? |
| 01:05.23 | delusional | guest 764 |
| 01:05.33 | mdskpr_ | guest764 |
| 01:07.11 | CIA-114 | BZFlag: 03JeffM2501 * r21524 10/trunk/bzflag/plugins/python/: remove python plug-in, it was never finished and will be a pain to update |
| 01:07.41 | CIA-114 | BZFlag: 03JeffM2501 * r21525 10/trunk/bzflag/plugins/Makefile.am: remove python plug-in, it was never finished and will be a pain to update |
| 01:13.42 | *** join/#bzflag ahs3- (~ahs3-@adsl-065-005-193-158.sip.rdu.bellsouth.net) |
| 01:21.52 | mdskpr_ | JeffM: do you want to look over McSpider's openffa patfch before i submit it? |
| 01:22.01 | JeffM | not really |
| 01:22.17 | JeffM | did he add the new game type to the API? |
| 01:22.31 | mdskpr_ | idk |
| 01:22.35 | JeffM | check |
| 01:22.43 | JeffM | look for game type |
| 01:22.58 | mdskpr_ | yes he did |
| 01:23.06 | mdskpr_ | but just changed the names |
| 01:23.37 | mdskpr_ | from TeamFlagGameStyle to ClassicCTF |
| 01:23.53 | JeffM | he should have added one cus there are 2 FFAs now |
| 01:24.21 | mdskpr_ | hmm ok |
| 01:24.28 | mdskpr_ | ill see if i can add it |
| 01:24.52 | JeffM | i'll just be another enum |
| 01:25.02 | mdskpr_ | ok |
| 01:25.06 | JeffM | and then just be sure to set it right |
| 01:25.12 | JeffM | breaks the world |
| 01:25.52 | Thumper_ | What class executes when you hit Connect from the bzflag join menu? |
| 01:26.07 | JeffM | on the client or the server? |
| 01:26.13 | Thumper_ | client |
| 01:26.17 | CIA-114 | BZFlag: 03JeffM2501 * r21526 10/trunk/bzflag/ (58 files in 21 dirs): |
| 01:26.18 | CIA-114 | BZFlag: change the way the API works |
| 01:26.18 | CIA-114 | BZFlag: update most of the plug-ins ( I'll finish the rest soon ) |
| 01:26.23 | JeffM | ServerLink |
| 01:26.24 | Thumper_ | there's a bzflag join menu on the server? :) |
| 01:26.25 | Thumper_ | thanks |
| 01:26.35 | JeffM | it's constructor connects to the server |
| 01:26.44 | JeffM | so yeah.. that breaks like all the plugins |
| 01:26.55 | JeffM | I've updated all but 6 of them and I'll finish them after I eat |
| 01:27.09 | JeffM | mdskpr_, you'll have to merge your changes in, sorry |
| 01:27.27 | JeffM | then I'll spend the rest of the night updating the docs |
| 01:27.32 | JeffM | but the API is super cool now :) |
| 01:27.34 | JeffM | bbl |
| 01:28.25 | delusional | peachy keen |
| 01:29.51 | McSpider | what did I forget? |
| 01:30.04 | mdskpr_ | an API change |
| 01:45.45 | CIA-114 | BZFlag: 03dtremenak * r21527 10/trunk/bzflag/MSVC/build/fullbuild.sln: more dependencies |
| 01:52.36 | CIA-114 | BZFlag: 03trepan * r21528 10/trunk/bzflag/include/bzfsAPI.h: * unused vars make --enable-debug unhappy |
| 02:01.23 | *** join/#bzflag PrezKennedy (MK@whitecalf.net) |
| 02:26.36 | *** join/#bzflag DTRemenak (~DTRemenak@about/essy/CrazyCoder/DTRemenak) |
| 02:26.37 | *** mode/#bzflag [+v DTRemenak] by ChanServ |
| 02:36.21 | *** join/#bzflag JeffM (~Jeff@unaffiliated/jeffm2501) |
| 02:36.21 | *** mode/#bzflag [+v JeffM] by ChanServ |
| 02:58.29 | CIA-114 | BZFlag: 03bthansen * r21529 10/trunk/bzflag/configure.ac: Drop python plugin from the build |
| 03:35.06 | Constitution | ~jeffm++ |
| 03:37.13 | JeffM | ? |
| 03:38.42 | JeffM | some of these plugins are soooooo bad :) |
| 03:39.51 | Constitution | ever so modest ;) |
| 03:40.54 | JeffM | they use a seperate event handler class for each event type |
| 03:41.04 | JeffM | helloo.... switch statement anyone? |
| 03:46.30 | trepan | if you want cleanliness, enforce it (see 2.99.x / EventHandler.h, which could be done C style) |
| 03:48.24 | trepan | guess you could also go with function callbacks, and avoid classes entirely |
| 03:51.09 | JeffM | oh I just find it funny |
| 04:05.23 | CIA-114 | BZFlag: 03jeffm2501 * r21530 10/trunk/bzflag/plugins/ (15 files in 6 dirs): update the last few plugins to new API style |
| 04:06.52 | trepan | JeffM: is eInBuilding new? warning for a missing switch value in bzfs.cxx:6107 |
| 04:07.35 | JeffM | did I miss taking out one? I saw that we didn't have zoned and some others |
| 04:08.23 | trepan | ah, looks like some backporting, I see that ePaused is a fallthrough |
| 04:08.43 | trepan | can probably take that FIXME out as well |
| 04:09.52 | JeffM | ahh look we do have pause |
| 04:10.06 | JeffM | I don't see in building in playerstate |
| 04:10.16 | trepan | only exists in 2.99.x |
| 04:10.22 | JeffM | k |
| 04:10.24 | JeffM | then I missed it |
| 04:10.57 | trepan | InBuilding, PhantomZoned, and BackedOff look like they're only in 2.99.x |
| 04:11.28 | JeffM | yeah |
| 04:18.27 | trepan | getPluginMinWaitTime() isn't defined when _USE_BZ_API is off (bzfs.cxx:5371) |
| 04:18.53 | trepan | is there a reason for that to be a build option? |
| 04:19.11 | JeffM | why do we need to turn off plugins? |
| 04:19.27 | trepan | is there a reason for that to be a build option? <-- same question? |
| 04:19.49 | JeffM | I don't think we need it as an option anymore |
| 04:19.49 | JeffM | it's kinda trusted |
| 04:19.59 | JeffM | windows dosn't have a way of turning it off |
| 04:20.11 | JeffM | and IIRC 2.99 had em on by default |
| 04:20.31 | trepan | ya, I don't remember having to use --enabled-shared |
| 04:20.50 | Thumper_ | The 2.99 build doesn't need --enabled-shared but we don't have that in 2.3 yet |
| 04:21.14 | JeffM | yeah but can't you still turn it off in 2.99? |
| 04:21.23 | JeffM | we are saying why do we even need that? |
| 04:21.27 | Thumper_ | I dunno - never tried :) |
| 04:21.36 | JeffM | what advantage would you have for building a bzfs that dosn't support plugins? |
| 04:22.11 | trepan | do all "supported" systems have .so support? |
| 04:22.23 | trepan | or .dll, as you OS dictates |
| 04:22.23 | JeffM | all the ones we care about |
| 04:22.41 | JeffM | do we really care about servers on irix ? |
| 04:22.52 | trepan | plural? :) |
| 04:22.58 | JeffM | heh |
| 04:23.06 | JeffM | yeah that guy can install linux ;) |
| 04:23.22 | JeffM | or figure something out and submit a patch |
| 04:23.42 | JeffM | bah.. need directX before I can do a fullbuild on this machine :( |
| 04:34.18 | CIA-114 | BZFlag: 03bthansen * r21531 10/trunk/bzflag/plugins/ (6 files in 6 dirs): |
| 04:34.19 | CIA-114 | BZFlag: Fix plugins for --enable-debug compiles |
| 04:34.19 | CIA-114 | BZFlag: - Fix event data type name to conform to the new API |
| 04:34.19 | CIA-114 | BZFlag: - Comment out names for unused parameters |
| 04:34.19 | CIA-114 | BZFlag: - Fix incorrect assignment in if test |
| 04:34.19 | CIA-114 | BZFlag: - Add default case for switch statements |
| 04:34.19 | CIA-114 | BZFlag: - Don't return a value from a void function |
| 04:35.12 | JeffM | aww you screwed me dude |
| 04:35.15 | JeffM | I was in the middle of that |
| 04:35.23 | Thumper_ | oops sorry |
| 04:35.38 | JeffM | lets see if I can merge :) |
| 04:35.43 | Thumper_ | thought you were done with your last commit comment |
| 04:35.56 | JeffM | no that was just the bassics |
| 04:36.04 | JeffM | I'm doing a full build to see what I missed |
| 04:36.18 | Thumper_ | it builds on linux now |
| 04:36.42 | CIA-114 | BZFlag: 03trepan * r21532 10/trunk/bzflag/src/bzfs/ (bzfs.cxx bzfsAPI.cxx): |
| 04:36.43 | CIA-114 | BZFlag: * make it compile when _USE_BZ_API is not defined |
| 04:36.43 | CIA-114 | BZFlag: * compensated for the missing eInBuilding case value by adding a default: |
| 04:36.44 | trepan | and now it builds without plugins |
| 04:37.33 | JeffM | yay more conflicts |
| 04:37.37 | Thumper_ | :( |
| 04:38.00 | Thumper_ | I'll ask next time :) |
| 04:38.41 | JeffM | no worries |
| 04:38.44 | JeffM | it merged ok |
| 04:39.31 | JeffM | just had to make sure I didn't blow out any of your changes |
| 04:41.24 | CIA-114 | BZFlag: 03jeffm2501 * r21533 10/trunk/bzflag/ (19 files in 16 dirs): More missed API fixes. |
| 04:43.40 | Thumper_ | waits to see if the test trunk servers build now - they're stuck at r21523 since the build was broken |
| 04:49.24 | Thumper_ | nope |
| 04:49.46 | Thumper_ | JeffM: unused param in SAMPLE_PLUGIN - do you want me to fix it? |
| 04:53.28 | JeffM | sure |
| 04:54.07 | CIA-114 | BZFlag: 03bthansen * r21534 10/trunk/bzflag/plugins/SAMPLE_PLUGIN/SAMPLE_PLUGIN.cpp: Unused parameter breaks the --enable-debug build on linux |
| 04:59.18 | meem1029 | So I've learned 2 things from my problems getting things to run over the last couple of days. |
| 04:59.30 | meem1029 | 1: apt-get is good. |
| 04:59.51 | meem1029 | 2: When getting a package, try the ubuntu repos before downloading source and compiling it. |
| 05:00.39 | BulletCatcher | And you owe it all to BZFlag. :-) |
| 05:01.00 | meem1029 | Yep! |
| 05:01.50 | meem1029 | Ya, for some reason I clicked on the link given in configure to install c-ares. I just now decided it was probably something I messed up (ok, I've known that for a while) and tried to fix it. My first fix was apt-get and it worked! |
| 05:02.14 | Thumper_ | Does anyone have a windows client that can connect to the bzflag3 trunk test servers? (any server) |
| 05:02.33 | Thumper_ | There still seems to be something screwy with mine - I can connect to a localhost server no problem but not the public ones |
| 05:04.02 | McSpider | Same on Mac OSX |
| 05:04.24 | Thumper_ | I think the linux connection code still has some issues |
| 05:04.26 | trepan | ares related? |
| 05:04.32 | meem1029 | I'll try r 21534 and see if there are problems. |
| 05:04.54 | Thumper_ | not sure - I'm using a system installed c-ares-dev pkg |
| 05:05.21 | Thumper_ | I can see initial communication with teh server in tcpdump where it sends BZFLAG and gets the server version back |
| 05:05.30 | Thumper_ | but something goes wrong after that - haven't figured it out yet |
| 05:06.06 | meem1029 | I think 21534 broke something. It was working before for me but isn't now. |
| 05:06.13 | meem1029 | Although I can't even see any servers up. |
| 05:06.52 | Thumper_ | they're restarting |
| 05:06.57 | Thumper_ | build just finished |
| 05:07.10 | meem1029 | Ok. That would explain why I can't see any. |
| 05:07.17 | Thumper_ | there's a 20s pause between shutdown and restart so if it goes into an infinite loop it doesn't kill the server |
| 05:07.51 | Thumper_ | or not - they're all maked 'Dead' :) |
| 05:07.58 | Thumper_ | marked* |
| 05:08.07 | meem1029 | Hmm, that sounds problematic. |
| 05:08.28 | Thumper_ | Segmentation fault (core dumped) ${BZFS} |
| 05:08.42 | Thumper_ | guess the new plugin API changes don't work ;) |
| 05:08.50 | Thumper_ | these servers use logDetail and serverControl plugins |
| 05:09.38 | meem1029 | Hmm, bzfs is starting for me. Not trying it with plugins though (I don't really know how to do server stuff). |
| 05:09.52 | Thumper_ | the logDetail plugin crashed teh server |
| 05:10.15 | JeffM | I'm still testing the changes |
| 05:10.33 | Thumper_ | I have a backtrace for you if you want it |
| 05:10.49 | JeffM | I'm debuging here |
| 05:11.05 | JeffM | I jus wanted the major changes in so people can work around them |
| 05:11.16 | JeffM | did other plugins load ok? |
| 05:11.54 | Thumper_ | they all use the same two plugins and died -not sure which one I load first |
| 05:11.58 | Thumper_ | I can try them on my workstation |
| 05:12.07 | JeffM | that's the only 2 y ou load? |
| 05:12.11 | Thumper_ | yes |
| 05:12.12 | JeffM | k |
| 05:12.15 | Thumper_ | (currently) |
| 05:12.17 | JeffM | should be simple to dupe then |
| 05:12.21 | Thumper_ | backtrace -> http://norang.ca/tmp/r21534.html |
| 05:13.13 | Thumper_ | yup crashes immediately on loading the logDetail plugin |
| 05:14.00 | Thumper_ | serverControl seems to load okay |
| 05:14.25 | JeffM | ahh yeah I made a bad assumption |
| 05:14.57 | JeffM | one sec |
| 05:18.58 | CIA-114 | BZFlag: 03jeffm2501 * r21535 10/trunk/bzflag/plugins/logDetail/logDetail.cpp: don't assume the passed in event is valid, it's called in init with null |
| 05:19.01 | JeffM | should be good |
| 05:19.45 | Thumper_ | looks good to me - servers should be back shortly after 1:30 EST |
| 05:19.55 | Thumper_ | they try to build every 15 minutes if there are new commits |
| 05:26.31 | CIA-114 | BZFlag: 03bullet_catcher * r21536 10/trunk/bzflag/src/bzfs/bzfsPlugins.cxx: |
| 05:26.32 | CIA-114 | BZFlag: Use valid syntax in debug message format strings. |
| 05:26.32 | CIA-114 | BZFlag: Use consistent style in debug messages. |
| 05:28.32 | JeffM | thanks |
| 05:28.52 | JeffM | I want to go remove all the debug messages from the plugins at some point so they are consistent too |
| 05:29.26 | CIA-114 | BZFlag: 03trepan * r21537 10/trunk/bzflag/ (include/TextUtils.h src/bzflag/CommandsImplementation.cxx): * backported the client /debug [level] command |
| 05:31.30 | BulletCatcher | Some of the plugin debug messages are useful. Others could just have their numeric levels increased so they are less noisy. |
| 05:32.19 | JeffM | the ones I want to pull are the startup and shutdown ones |
| 05:32.26 | JeffM | cus the server will send those anyway |
| 05:32.46 | JeffM | I don't want to remove the other ones that are unique to specific plugins |
| 05:33.23 | BulletCatcher | It adds an extra layer of confidence to have each plugin announce itself. |
| 05:33.25 | JeffM | or at least put them in the base class |
| 05:33.43 | JeffM | I just don't like duped code |
| 05:34.12 | JeffM | I can move them into the base class, they will still be called by the plugin but the code dosn't have to be in the plugins cpp |
| 05:34.52 | JeffM | what level would you want that as? |
| 05:34.55 | JeffM | 4? |
| 05:35.41 | BulletCatcher | I run my servers with -dd, and it is comforting to have plugin loading be announced. |
| 05:36.01 | JeffM | the message you get now are the level 1 messages I did not remove |
| 05:36.08 | JeffM | that's done by bzfs not the plugin |
| 05:36.14 | JeffM | there is a level 4 message buried in the plugin |
| 05:36.19 | JeffM | well some of them |
| 05:36.57 | JeffM | it's that level 4 one I want to comonize |
| 05:37.04 | Thumper_ | 4 means you get all of the server network messages to which tends to flood the output channel - are they useful on their own other than for occasional debugging? |
| 05:37.11 | Thumper_ | to-> too |
| 05:37.22 | JeffM | yeah I think they are only usefull during plugin dev |
| 05:37.30 | Thumper_ | then 4 is fine |
| 05:37.33 | JeffM | thats why I don't want the code in there |
| 05:37.50 | Thumper_ | you do -dddd to check it loads and then you move on |
| 05:37.53 | JeffM | so actualy I added a message for BulletCatcher you will now get an uload message |
| 05:37.59 | JeffM | at level 1 |
| 05:38.08 | Thumper_ | ok |
| 05:38.14 | JeffM | not that many people unload at runtime |
| 05:39.59 | Thumper_ | yay the servers are magically up again. I've gotta sleep - catcha later |
| 05:40.05 | JeffM | see ya |
| 05:43.30 | trepan | (une |
| 05:43.34 | CIA-114 | BZFlag: 03trepan * r21538 10/trunk/bzflag/src/ (bzflag/playing.cxx common/global.cxx): * backported _forbidIdentify and _forbidHunting |
| 05:43.38 | trepan | (useful for no-radar servers) |
| 06:13.57 | *** join/#bzflag Marzipan- (~Marzipan@p5B2253B1.dip.t-dialin.net) |
| 06:15.30 | *** join/#bzflag Marzipan (~Marzipan@bzflag/player/Marzipan) |
| 06:17.39 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7658 10/w/BZFS_API_2.4_Upgrade: Initial notes on upgrading plug-ins from 2.0.4 to 2.3.4 |
| 06:21.32 | *** join/#bzflag fahadsadah (fahad@unaffiliated/fahadsadah) |
| 06:49.04 | BulletCatcher | Good timing. I have some 3rd party plugins I must update now. |
| 07:06.42 | *** join/#bzflag trepan_ (~trepan@204.237.8.232) |
| 07:47.25 | Pimpinella | hmm, todo: disable endshot detection when you stole GM |
| 10:11.01 | Constitution | woot http://www.youtube.com/watch?v=M-lwleibmQA |
| 10:11.13 | Constitution | basic, but functional :) |
| 10:12.30 | Manu | Pimpinella: IIRC it was working fine with GM of course there is a min. secure value for endshot detection |
| 10:30.24 | *** join/#bzflag Zac_ (ca59b013@gateway/web/freenode/ip.202.89.176.19) |
| 11:16.26 | Zac_ | can anyone tell me the worldsize of missile war? |
| 11:42.17 | Zac_ | anyone here a linux user? |
| 12:05.20 | blast007 | Zac_: join the server and do /diff or /set _world* |
| 12:05.32 | blast007 | and yes, there are linux users. If you have a real question, just ask it. |
| 12:09.14 | blast007 | figures. |
| 12:39.26 | delusional | in bzflag CommandsImplementation.cxx:160: error: extra â;â |
| 12:40.27 | *** join/#bzflag trepan (~trepan@unaffiliated/trepan) |
| 12:40.27 | *** mode/#bzflag [+v trepan] by ChanServ |
| 12:41.56 | *** join/#bzflag kierra (~jolie@unaffiliated/kierra) |
| 12:41.58 | CIA-114 | BZFlag: 03trepan * r21539 10/trunk/bzflag/src/bzflag/CommandsImplementation.cxx: * remove the extra ';', thanks delusional |
| 12:43.13 | CIA-114 | BZFlag: 03trepan * r21540 10/trunk/bzflag/src/bzflag/playing.cxx: * protect client-side BZDB variables |
| 13:38.27 | mrapple | w/win 19 |
| 14:11.39 | *** join/#bzflag bier|tp (~bier@p54A5BE57.dip.t-dialin.net) |
| 14:22.08 | *** join/#bzflag Upsetter (~er@i59F6DC39.versanet.de) |
| 14:44.52 | *** join/#bzflag spldart (~spldart2@bzflag/contributor/spldart) |
| 14:44.52 | *** mode/#bzflag [+v spldart] by ChanServ |
| 15:01.06 | *** join/#bzflag eTangenT (~lamer@c-98-245-170-238.hsd1.co.comcast.net) |
| 15:04.42 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 15:04.42 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 15:07.55 | spldart | prepares for his first build of bz in squeeze. Hopefully I have more luck than I've had in windows. heh |
| 15:16.10 | blast007 | spldart: you never gave me what error you had in Windows.. ;) |
| 15:21.18 | meem1029 | Oh, by the way, I just decided to try the newest version and it hangs whenever I try to connect to a server. In about 30 seconds it says Error Connecting to server. |
| 15:21.32 | meem1029 | I haven't tried local server yet though since I have to go to class. |
| 15:30.04 | spldart | yes I did.. check your logs |
| 15:30.15 | spldart | I pastebin' |
| 15:30.19 | spldart | just the errors |
| 15:30.43 | spldart | what debian package gets me that pesky gl.h package? |
| 15:31.01 | spldart | thought it was that mesa package and nvidia-glx-dev |
| 15:53.37 | *** join/#bzflag temporalD (~a_temp_di@bzflag/serverop/TemporalDistraction) |
| 15:53.39 | BulletCatcher | My notes list libgl1-mesa-dev, libglu1-mesa-dev, libsdl1.2-dev and/or libsdl-dev, libcurl3-dev, libsdl-sound1.2-dev . |
| 15:54.05 | Thumper_ | BulletCatcher: do you have a public trunk server available for a quick test? |
| 15:55.03 | BulletCatcher | No. |
| 15:55.05 | BulletCatcher | I have some plugins that need to be updated first. |
| 15:55.12 | BulletCatcher | But I can't connect to your servers, either. |
| 15:55.33 | Thumper_ | ok I'll build on out on bzflag2.norang.ca which has an updated c-ares to rule that out as the cause of my connection issues in linux |
| 15:55.34 | blast007 | spldart: ah, missed that. However, those error don't help much. When you get a chance, post the entire build log from the Output window (you'll probably have to do a full rebuild to get the post-build steps to run again) |
| 15:55.43 | Thumper_ | s/on/one/ |
| 15:56.14 | Thumper_ | localhost connections work for me (the last time I tried it) |
| 15:56.19 | BulletCatcher | Same here. |
| 15:57.58 | blast007 | I'm building an updated bzfs on my server as well |
| 15:58.11 | Thumper_ | the race is on :) |
| 15:58.13 | blast007 | I'll throw up decks ;) |
| 16:02.06 | blast007 | there, mine is up. |
| 16:02.38 | spldart | oki doki blast |
| 16:02.42 | BulletCatcher | L4m3r: Is there any chance I can convince you to contribute your fairCTF plugin to the BZFlag project? |
| 16:04.00 | spldart | thanks bulletcatcher. I'm goin through your notes now to see if I'm missing any |
| 16:04.49 | BulletCatcher | Same connection problem to your decks server, blast007. |
| 16:05.23 | blast007 | what happens when you try? |
| 16:05.51 | BulletCatcher | After about 30 seconds it says error connecting to server. |
| 16:05.56 | blast007 | hmm |
| 16:06.07 | spldart | looks like i was missing the libsdl1.2-dev |
| 16:06.26 | BulletCatcher | I think it makes the TCP connection, but there is a protocol bug. |
| 16:06.27 | blast007 | BulletCatcher: try connecting to port 5000 (won't show on the list as it's not a bzfs, just a netcat) |
| 16:06.30 | spldart | retrying ./configure |
| 16:06.30 | Thumper_ | ditto - can't connect here either |
| 16:06.50 | spldart | crap.. fail... something with squeeze is a little different than lenny. |
| 16:06.52 | blast007 | (note I shut down the server as well) :) |
| 16:07.01 | spldart | I'll need to get squeeze on my main rig at home asap |
| 16:07.18 | Thumper_ | spldart: I'm on squeeze |
| 16:07.23 | blast007 | BulletCatcher: okay, I see the BZFLAG header get passed |
| 16:07.36 | blast007 | so it's at least getting that far |
| 16:08.01 | BulletCatcher | port 5000 gives the same behavior on the client |
| 16:08.05 | blast007 | does bzadmin work? (I have my 15154 decks back up) |
| 16:09.19 | BulletCatcher | bzadmin does the same thing. |
| 16:09.23 | blast007 | hmm |
| 16:09.27 | BulletCatcher | to port 15154 |
| 16:09.45 | blast007 | guess my bzadmin doesn't even connect locally |
| 16:09.52 | spldart | Thumper_ You wouldn't by any chance have a list of dependencies for bzflag build would you? |
| 16:09.55 | blast007 | that was working sometime yesterday though |
| 16:10.20 | blast007 | spldart: it might be in the README.Linux (or similar named file) |
| 16:10.24 | Thumper_ | spldart: not but I have a list of what I have installed that works for me |
| 16:10.56 | spldart | I was poking through that readme |
| 16:12.07 | blast007 | at a minumum, you'll need the ares dev package, sdl and sdl audio dev package, ncurses dev package, and then the usual make, automake, autoconf, libtool, g++ |
| 16:12.19 | blast007 | probably want the glew dev package too |
| 16:12.50 | spldart | gets back to work. CRAP! G++ |
| 16:12.53 | spldart | I forgot to install that |
| 16:12.55 | spldart | gah! |
| 16:12.58 | Thumper_ | heh |
| 16:13.34 | spldart | I've made that mistake before.. I feel so stupid :( |
| 16:13.46 | spldart | and configure is happy now |
| 16:14.28 | blast007 | spldart: I made that mistake two days ago :P |
| 16:14.42 | blast007 | "why is gl.h missing? ooooo" |
| 16:14.48 | spldart | lol |
| 16:14.51 | Chestal | you can also try apt-get build-dep bzflag |
| 16:15.08 | spldart | whazzat doo? |
| 16:15.23 | Chestal | should install all packages needd to build bzflag deb package |
| 16:15.30 | spldart | ah |
| 16:15.34 | spldart | yeah.. just googled |
| 16:15.45 | Chestal | doesn't know about the dev branch obviously :-) |
| 16:15.49 | spldart | makes a note of that for when he ups the main machine |
| 16:16.24 | BulletCatcher | It would be really helpful if someone would add the Debian/Ubuntu package requirements to the README.Linux file. |
| 16:16.28 | Thumper_ | blast007: according to tcpdump I'm sending 10 bytes for BZFLAG\r\n\r\n but getting length 0 back in the reply |
| 16:16.42 | *** join/#bzflag temporalD (~a_temp_di@bzflag/serverop/TemporalDistraction) |
| 16:17.44 | spldart | ooOOoo errors |
| 16:18.00 | Thumper_ | blast007: http://norang.ca/tmp/r21540.html |
| 16:18.25 | spldart | http://bzflag.pastebin.ca/2063701 |
| 16:18.50 | blast007 | hmm |
| 16:19.16 | BulletCatcher | spldart: You need to re-run configure after installing development packages. |
| 16:19.33 | spldart | I did... lemme clean though and start over |
| 16:19.43 | blast007 | also, make sure you installed the ares dev package |
| 16:20.00 | BulletCatcher | To be really thorough, do a "make maintainer-clean" and start over with autogen.sh. |
| 16:20.12 | blast007 | spldart: specifically, libc-ares-dev |
| 16:20.27 | blast007 | (not libares-dev) |
| 16:20.42 | blast007 | so I guess I should call it what it really is.. c-ares ;) |
| 16:21.35 | Thumper_ | goes for lunch |
| 16:22.48 | *** join/#bzflag JeffM (~JeffM@69.36.85.130) |
| 16:22.49 | *** join/#bzflag JeffM (~JeffM@unaffiliated/jeffm2501) |
| 16:22.49 | *** mode/#bzflag [+v JeffM] by ChanServ |
| 16:33.51 | *** join/#bzflag JeffM (~JeffM@69.36.85.130) |
| 16:33.51 | *** join/#bzflag JeffM (~JeffM@unaffiliated/jeffm2501) |
| 16:33.51 | *** mode/#bzflag [+v JeffM] by ChanServ |
| 16:38.11 | spldart | Should show up as ver 2.0.17 correct? |
| 16:38.42 | blast007 | no |
| 16:39.04 | blast007 | blast@porteighty:~$ bzfs23 -version |
| 16:39.05 | blast007 | BZFlag server 2.3.4.20110518-DEVEL-linux-gnu (protocol 0204) http://BZFlag.org/ |
| 16:39.33 | blast007 | make sure you're running the right binary (in the event that you also have bzflag/bzfs installed from the debian repo) |
| 16:55.22 | spldart | When I do './src/bzflag/bzflag --version' I get this 'BZFlag client 2.3.4.20110518-DEVEL-linux-gnu-SDL (protocol 0204) http://BZFlag.org/' but when I run ./src/bzflag/bzflag I see 2.0.17 in the version id... But I only see like 10 servers |
| 16:55.25 | spldart | bizarre |
| 16:55.33 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 16:55.33 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 16:56.26 | spldart | is it just that the 2.0.17 is in that picture in the splash picture at above the menu? |
| 16:57.08 | spldart | Yep.. that's it.. |
| 16:57.14 | blast007 | oh, that |
| 16:57.14 | spldart | I've got the right one |
| 16:57.18 | blast007 | k |
| 16:57.26 | spldart | one bzexcess and the rest norang servers |
| 16:57.27 | spldart | i'm good |
| 16:57.50 | spldart | Now to figure out why winderze wont build. meh |
| 16:57.51 | spldart | hehe |
| 16:59.54 | blast007 | the windows one *did* build, it just failed running post-build steps (which are just copying the binaries into another directory) |
| 17:05.01 | BulletCatcher | JeffM: The API upgrade is going to be quite a burden to server operators who use 3rd party plugins. |
| 17:05.02 | BulletCatcher | The new wiki page about it is a start, but lacks the step-by-step plugin upgrade procedure that will be needed. |
| 17:05.19 | JeffM | BulletCatcher, that's because I had to sleep |
| 17:05.42 | JeffM | I am in the process of writing , exactly that ;) |
| 17:05.47 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7659 10/w/BZFS_API_2.4_Upgrade: |
| 17:05.49 | *** part/#bzflag kierra (~jolie@unaffiliated/kierra) |
| 17:05.51 | BulletCatcher | yay :-) |
| 17:06.22 | JeffM | mostly this page is me writing it out, then when I am sure it's all good, I'll incorporate it into the main wiki |
| 17:08.32 | BulletCatcher | You have already mentioned most of the to-do items I encountered. |
| 17:08.33 | BulletCatcher | I'll be patient. |
| 17:08.41 | JeffM | :) |
| 17:08.53 | JeffM | I built up a list of steps as I was doing all of them |
| 17:09.05 | JeffM | it wasn't too hard once you got a rythm going |
| 17:10.24 | JeffM | and I will put notes on there about how we'll help people upgrade of they come on IRC |
| 17:21.29 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7660 10/w/BZFS_API_2.4_Upgrade: step by step |
| 17:23.49 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7661 10/w/BZFS_API_2.4_Upgrade: /* Assistance */ |
| 17:24.14 | JeffM | that should be the bassics |
| 17:25.25 | Thumper_ | JeffM: do we need some info about -publickey in there? |
| 17:25.50 | JeffM | in the API docs? |
| 17:25.53 | Thumper_ | nevermind - missed the API in the title |
| 17:25.53 | Thumper_ | :) |
| 17:25.59 | JeffM | the public key dosn't have anything to do with the API |
| 17:26.08 | JeffM | but yes we sould mention it somewhere |
| 17:27.09 | trepan | it's in the manpage, and it gives a clear warning and a URL if you do not provide it |
| 17:28.29 | spldart | blast007: Really? I could have ran that bzflag client? |
| 17:28.34 | spldart | will have to try when he gets home |
| 17:28.43 | blast007 | spldart: probably |
| 17:28.55 | blast007 | the full build log would give what the real problem is |
| 17:29.02 | spldart | oki doki |
| 17:29.23 | spldart | I have to work on a girls laptop first but then I'll check that win bz build tonight |
| 17:29.45 | trepan | JeffM: then again, you oft need a REALLY BIG HAMMER to get through to users |
| 17:30.02 | JeffM | air hammers work best |
| 17:30.44 | trepan | first thought, vapourware ... need more coffee |
| 17:32.07 | smoooth | is the new bzflag going to implement some kind of -publickey infrastructure? |
| 17:32.24 | Thumper_ | yes |
| 17:32.28 | smoooth | sweet :) |
| 17:32.29 | JeffM | it's already done |
| 17:32.32 | JeffM | it was backported |
| 17:33.19 | JeffM | marconi plays the mamba... listen to the radio. |
| 17:34.14 | BulletCatcher | smoooth: The -publickey option is used to identify the owner of a server to the global list server. |
| 17:34.15 | BulletCatcher | There is no RSA-style private/public key system. |
| 17:35.16 | smoooth | ohh k, got it. That's what I was thinking. So just for server/global comm. |
| 17:35.39 | JeffM | man in the middle attacks are not a big problem for BZ |
| 17:35.47 | JeffM | so crypto of any channel is overkill |
| 17:35.59 | BulletCatcher | The key is used for authentication, not for encryption. |
| 17:36.03 | JeffM | yeah |
| 17:36.16 | smoooth | True but keeping identities is a problem.. it might could help there |
| 17:36.45 | JeffM | all our transactions are atomic, and the key is tied to the domain name, it means we know who each request comes from |
| 17:36.53 | JeffM | we don't need to keep a session or any kind of channel open |
| 17:36.58 | BulletCatcher | And although the name of the option is -publickey, the key itself is a shared secret. |
| 17:37.16 | JeffM | and -publicsecret is a silly name ;) |
| 17:37.35 | BulletCatcher | -serverkey might be a better name |
| 17:37.40 | JeffM | true |
| 17:37.46 | JeffM | tho it's only need for public servers |
| 17:37.53 | JeffM | thats what the public part is for |
| 17:37.56 | JeffM | -listserverkey ? |
| 17:38.18 | BulletCatcher | I would interpret that to be a key that identifies the list server. |
| 17:38.30 | JeffM | hmm |
| 17:38.53 | JeffM | -publicserverkeythatyouhavetouse |
| 17:38.59 | smoooth | i was thinking if there was a public key and users had their own private keys.. you could authorize people that way.. like an ssh session? or is that not a useful concept |
| 17:39.46 | BulletCatcher | -serverkey, along with documentation that it is only required for public servers, ought to be the least confusing. |
| 17:39.50 | JeffM | that's a replacement for global auth and outside the scope of this release |
| 17:39.57 | smoooth | gotcha |
| 17:40.13 | JeffM | ssh does that so that it can crypto the channel |
| 17:40.26 | JeffM | cus once you auth you need a secured channel |
| 17:40.31 | JeffM | we don't realy need a secured channel |
| 17:41.08 | JeffM | I did some experimentation with encrypted communiction from server to client and it works it just has some overhead |
| 17:41.17 | smoooth | good point -- would be useful for general authentication but not need to do channel encryption |
| 17:41.47 | JeffM | all it relay prevented was man in the middle attacks |
| 17:41.52 | JeffM | really even :) |
| 17:42.09 | BulletCatcher | It might be nice to encrypt chat messages, especially private chat, but that won't happen in this release. |
| 17:42.30 | JeffM | I don't think we need to make bzflag into a secure chat system ;) |
| 17:43.11 | BulletCatcher | You underestimate the competitiveness of some GU players. |
| 17:43.11 | smoooth | i agree, encryption is too much.. i just like the keys idea for authentication.. that way we can move away from ip address stuff in the leagues |
| 17:43.38 | JeffM | if bz required a registered username then you'd get the same thing |
| 17:43.49 | JeffM | your average player is just going to want to type in a password |
| 17:43.53 | JeffM | not get and install a key |
| 17:44.01 | blast007 | smoooth: how is a public/private key system any better than our current system? |
| 17:44.16 | JeffM | the token is efectivly a temporary key |
| 17:45.19 | smoooth | i dont know if its better. just i know we have problems with ip address.. banning players.. just user control.. if each was given a private key it could possibly help admin/user system with identification.. |
| 17:45.30 | smoooth | or maybe not.. just seems like a novel idea that could be useful |
| 17:45.42 | JeffM | you can ban them by BZID |
| 17:45.45 | blast007 | smoooth: like jeff said - require auth and you get the same thing |
| 17:45.51 | JeffM | so I don't know why you need IPs |
| 17:46.03 | blast007 | if someone wanted to get past a key ban, they'd just get a new key :P |
| 17:46.06 | blast007 | same difference |
| 17:46.25 | blast007 | so it just makes it more complicated without solving any problem |
| 17:46.32 | smoooth | you're probably right |
| 17:46.39 | smoooth | or *are |
| 17:46.52 | JeffM | it is a better solution if you need crypto |
| 17:46.59 | JeffM | that's why it's used for ssh and ssl |
| 17:47.50 | blast007 | "First 100 registrations get their BZSmartCard for free! (A $19.95 value!)" |
| 17:48.01 | smoooth | hah |
| 17:48.45 | trepan | the big benefit to PKE is that it's safe even if someone is monitoring the link (as long as different tokens are encrypted for different sessions) |
| 17:50.30 | blast007 | trepan: but using SSL to the list would largely equate to that feature, no? Our tokens are one-time use and expire after 5 (iirc) minutes, and are tied to the IP that requested them. |
| 17:51.42 | Thumper_ | -publicserverkey :-P |
| 17:52.48 | trepan | blast007: i'm still thinking about -publickey |
| 17:53.03 | blast007 | k |
| 17:53.15 | blast007 | eww, and I guess it's 60 minutes. |
| 17:53.21 | trepan | but ya, one time keys are safe (albeit that there's a period where someone monitoring could spoof) |
| 17:54.20 | trepan | an inordinate amount of effort to join e bzflag server using someone's account, probably just as easy to find a stack overflow exploit :) |
| 17:54.34 | blast007 | they'd have to spoof the IP as well |
| 17:56.01 | trepan | for one-time tokens? |
| 17:56.20 | blast007 | yeah, the IP is stored when a token is requested |
| 17:56.36 | blast007 | and bzfs sends the IP of whoever gave it the token to the list |
| 17:56.41 | blast007 | if it doesn't match, BADTOK |
| 17:57.23 | blast007 | same is done for webauth (though it's up to the remote site to decide if they want to pass the IP to the list) |
| 17:59.33 | CIA-114 | BZFlag: 03JeffM2501 * r21541 10/trunk/bzflag/ (24 files in 24 dirs): |
| 17:59.33 | CIA-114 | BZFlag: add plug-in load /unload events |
| 17:59.33 | CIA-114 | BZFlag: comonize plugin load/unload debug comments. |
| 18:00.06 | trepan | blast007: looking for where bzfs sends the IP along with the token... |
| 18:01.16 | trepan | got it, cool |
| 18:01.47 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7662 10/w/BZFS_API_2.4_Upgrade: |
| 18:02.36 | blast007 | exception is someone on the LAN that the server is on, but then there are easier ways to impersonate someone ;) |
| 18:08.20 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7663 10/w/BZFS_API_2.4_Upgrade: /* Upgrade */ more of a set of guidelines |
| 18:09.26 | CIA-114 | BZFlag: 03JeffM2501 07http://my.bzflag.org * r7664 10/w/BZFS_API_2.4_Upgrade: /* Upgrade */ |
| 18:10.49 | *** join/#bzflag jh` (~jh_@212-149-234-105.bb.dnainternet.fi) |
| 18:16.12 | *** join/#bzflag Erroneous (~DTRemenak@69.36.85.130) |
| 18:16.12 | *** join/#bzflag Erroneous (~DTRemenak@about/essy/CrazyCoder/DTRemenak) |
| 18:16.12 | *** mode/#bzflag [+v Erroneous] by ChanServ |
| 18:21.41 | *** join/#bzflag jh`` (~jh_@212-149-234-105.bb.dnainternet.fi) |
| 18:38.55 | Thumper_ | When I set a breakpoint on bzfs.cxx:4814 I can connect to the server everytime - but without the breakpoing connection fails |
| 18:55.27 | *** part/#bzflag Upsetter (~er@i59F6DC39.versanet.de) |
| 18:55.39 | blast007 | Thumper_: is it bailing out of that function somewhere? |
| 18:56.08 | Thumper_ | not that I'm aware of - maybe it doesn't have all of the header data from the client yet - still looking at it |
| 18:56.41 | blast007 | line 4826 checks the receive status |
| 18:56.56 | Thumper_ | if I break on 4826 it fails |
| 18:57.14 | Thumper_ | so it is somehow timing related |
| 18:58.26 | Thumper_ | Got 0 bytes of header data |
| 19:26.33 | *** join/#bzflag bryjen (~bryjen@76.92.85.169) |
| 19:26.33 | *** mode/#bzflag [+v bryjen] by ChanServ |
| 19:31.51 | JeffM | Thumper_, try adding if (readSize == 0) return; |
| 19:31.51 | JeffM | <PROTECTED> |
| 19:32.18 | JeffM | make it go around the horn again |
| 19:33.15 | Thumper_ | ok |
| 19:48.09 | Thumper_ | added that - still behaves the same |
| 19:52.10 | JeffM | k |
| 19:52.14 | JeffM | was just an idea |
| 19:52.49 | Thumper_ | so if I stop and wait before the netHandler->receive call (with the debug breakpoint) I get 10 bytes everytime |
| 19:52.56 | Thumper_ | but if I let it fly past that I get 0 bytes |
| 19:53.10 | Thumper_ | and it goes south from there |
| 19:53.36 | Thumper_ | throws in a 5s sleep just to see what happens |
| 19:54.00 | JeffM | it should just skip it till later if it gets 0 |
| 19:54.21 | JeffM | so it returns readpart and size is 0? |
| 19:54.27 | *** join/#bzflag mdskpr_ (~mdskpr@108.25.108.121) |
| 19:54.29 | Thumper_ | works with the sleep |
| 19:54.38 | Thumper_ | it returns readall and 0 I think |
| 19:54.40 | JeffM | we should not have to sleep |
| 19:54.47 | JeffM | readall means that it reads 10 |
| 19:54.47 | Thumper_ | I know - it was just to test a theory :) |
| 19:54.48 | JeffM | and that is wrong |
| 19:54.56 | JeffM | cus I ask for 10 |
| 19:55.03 | Thumper_ | I'll double check that |
| 19:56.18 | mdskpr_ | JeffM: it turned out that McSpider did add openFFA to bzfsAPI, but to the header? should it be in bzfsAPI.cxx also? |
| 19:56.38 | JeffM | yeah it needs to be set properly in the API |
| 19:56.49 | JeffM | just adding the enum dosn't mean that it will be set correctly |
| 19:57.12 | mdskpr_ | ah ok, so it should be set the same places RabbitChase is set? |
| 19:57.50 | JeffM | there is a function that returns that enum |
| 19:57.50 | Thumper_ | JeffM: it's readPart and 0 this time |
| 19:57.56 | Thumper_ | maybe I remember it wrong |
| 19:58.10 | JeffM | Thumper_, ok, then it should skip it and go into the normal thread sleep and come back around |
| 19:58.15 | JeffM | where does it go after that? |
| 19:58.30 | JeffM | does it ever get back into the process function? |
| 19:59.02 | Thumper_ | it only hits the breakpoint at bzfs.cxx:4826 once |
| 19:59.21 | Thumper_ | line after RxStatus e = netHandler->receive(strlen(BZ_CONNECT_HEADER)); |
| 20:00.14 | JeffM | step along and see where it goes |
| 20:00.45 | JeffM | readSize = netHandler->getTcpReadSize(); returns 0? |
| 20:01.21 | JeffM | so that return you added should kik you out of that function |
| 20:01.54 | JeffM | then it should pass all the kill code in bzfs.cxx and come around again |
| 20:02.16 | Thumper_ | ack I took that out again :-P |
| 20:02.18 | Thumper_ | puts it back |
| 20:02.41 | JeffM | if readpart can return 0 for a buffer size then I think that it needs to be there |
| 20:02.59 | JeffM | I made the assumption that readpart would be a number from 1 to 10, not 0 |
| 20:03.20 | JeffM | cus that's kinda what it's suposed to do :) |
| 20:08.18 | Thumper_ | first time throught I get ReadPart and readSize of 0 -> so I return |
| 20:08.22 | Thumper_ | then I never hit that breakpoint again |
| 20:08.39 | JeffM | step out and see why it never goes in there |
| 20:08.47 | JeffM | does something delete the peer? |
| 20:09.02 | Thumper_ | I usually run into my 30s timeout before I get done so the client bails |
| 20:10.42 | JeffM | it needs to be debuged |
| 20:10.49 | JeffM | increase the timeouts if you must |
| 20:10.52 | Thumper_ | working on it (slowly) |
| 20:10.56 | JeffM | but something is getting rid of it peer |
| 20:27.28 | *** join/#bzflag Anxuiz (~Anxuiz@unaffiliated/anxuiz) |
| 20:29.42 | Thumper_ | JeffM: it looks like it just goes around the main loop and waits on the select call |
| 20:29.59 | JeffM | so then it thinks it has no more data? |
| 20:30.01 | Thumper_ | I can see that time out once or twice before the client bails |
| 20:30.07 | Thumper_ | that's what it looks like |
| 20:30.08 | JeffM | then why did the data size return 0 |
| 20:30.28 | JeffM | something is wrong on the linux networking code if it's saying it has data but that data is 0 |
| 20:30.30 | Thumper_ | it works every 2nd time I run it in debug |
| 20:30.31 | JeffM | that's where I think the problem is |
| 20:30.58 | JeffM | step into the first select and the nethandler stuff, see how much it actualy gets |
| 20:33.52 | Thumper_ | part of the problem is stepping makes it slow enough to work everytime :-P |
| 20:34.11 | Thumper_ | keeps looking at it |
| 20:35.08 | JeffM | I don't think the size should be 0 when it does a readpart |
| 20:35.18 | *** join/#bzflag TD-Linux (~thomas@68-115-86-128.dhcp.roch.mn.charter.com) |
| 20:35.18 | *** join/#bzflag TD-Linux (~thomas@about/essy/indecisive/TD-Linux) |
| 20:35.24 | JeffM | ether that or I am totaly not understanding how networking works |
| 20:35.47 | JeffM | when select says you have data, you have friggin data ;) |
| 20:47.30 | JeffM | you should step into NetHandler::receive |
| 20:47.47 | JeffM | and see what int size = recv(fd, tcpmsg + tcplen, (int)length - tcplen, 0); returns |
| 20:47.57 | JeffM | and if tcpLen is set allready |
| 20:48.19 | mdskpr_ | right there IIRC, tcplen=0 |
| 20:48.33 | JeffM | ok good, what does size return from recv |
| 20:51.54 | Thumper_ | I think you can get returned sized of 0 when you signals and things interrupted the system call |
| 20:52.10 | Thumper_ | s/sized/sizes/ |
| 20:52.21 | Thumper_ | let's try that again :) |
| 20:52.28 | JeffM | if recv returns 0 then that code treats it as a discoonect |
| 20:52.37 | Thumper_ | signals can interrupt the read system call so it returns 0 |
| 20:52.53 | JeffM | that means our networking has never worked in that case |
| 20:53.06 | JeffM | I did not change the code in nethandler::read |
| 20:53.24 | JeffM | cus if that did read data but returns 0 then we have lost data |
| 20:53.27 | JeffM | and that ain't good |
| 20:53.55 | Thumper_ | there's no wait on a local server in linux - the socket data can be immediately read by the client so that explains the difference between the remote public server and the local one I think |
| 20:54.31 | JeffM | http://pubs.opengroup.org/onlinepubs/009695399/functions/recv.html |
| 20:54.55 | JeffM | recv is defined that the return of 0 means a shutdown |
| 20:55.24 | JeffM | so if your linux system is causing that to return 0 when it is not a discconect case then you do not have a standards compliant sockets system |
| 20:55.31 | JeffM | so I doubt that is the case |
| 20:56.05 | Thumper_ | then i must be thinking of the -1 case for the interrupt issue |
| 20:56.13 | Thumper_ | didn't mean to make us _more_ confused ;) |
| 20:56.29 | JeffM | less then 0 is an error case, and will not return ReadPart |
| 20:56.30 | Chestal | yes, it should return -1 if it is interrupted before reading data (EINTR) |
| 20:56.59 | JeffM | interesting, we call that "read part" |
| 20:57.09 | JeffM | ok I think I know what the issue is |
| 20:57.15 | JeffM | I need to make a change then |
| 20:57.24 | JeffM | I have to try again untill it's readall |
| 20:57.42 | Thumper_ | I think you need to keep collecting until it's readall |
| 20:58.31 | JeffM | yes |
| 20:58.50 | JeffM | add these 2 lines before // the dude has sent SOME data |
| 20:58.51 | JeffM | <PROTECTED> |
| 20:58.52 | JeffM | <PROTECTED> |
| 20:58.53 | Thumper_ | what's ReadPart mean - you asked for 10 bytes but got less? |
| 20:58.58 | JeffM | yes |
| 20:59.25 | *** join/#bzflag bier|tp (~bier@p54A5BE57.dip.t-dialin.net) |
| 20:59.43 | Thumper_ | ok gimme a minute |
| 20:59.48 | JeffM | sorry, e = netHandler->receive(strlen(BZ_CONNECT_HEADER)); |
| 20:59.55 | JeffM | I forgot the e in that first paste |
| 21:01.28 | JeffM | the problem is that we don't indicate that it was an "again" case |
| 21:03.03 | JeffM | hmm, I wonder if I want to use tcpReceive instead |
| 21:03.16 | JeffM | no that's for bz packets |
| 21:03.40 | Thumper_ | `I think that works |
| 21:13.20 | JeffM | test the piss out of it :) |
| 21:15.28 | Thumper_ | bah - it worked for my simple bzfs -dddd -p 10000 test, but not when I run with the full server options |
| 21:15.47 | Thumper_ | I think that while loop pegs the cpu at 100% for some cases |
| 21:18.17 | mdskpr_ | JeffM: you said I had to 'merge' the openFFA changes in? whats the difference between that and commit? |
| 21:32.16 | JeffM | mdskpr_, have you gotten latest after my API changes? |
| 21:32.28 | mdskpr_ | yes |
| 21:32.33 | JeffM | then you have merged |
| 21:33.17 | JeffM | it was possible tht some of the things I changed overlaped the things you changed so there could have been conficts when you got latest |
| 21:33.24 | JeffM | that is the merge process |
| 21:33.42 | mdskpr_ | k, I applied that patches AFTER i updated so it should be fine |
| 21:33.54 | JeffM | Thumper_, yeah I'm not sure that is the correct fix, I must ponder |
| 21:34.37 | JeffM | what confuses me is why select hangs for you and it dosn't get around back into the process method |
| 21:34.45 | JeffM | does someone delete the peer? |
| 21:35.14 | JeffM | mdskpr_, ahh I figured you had the patch applied before you got latest |
| 21:35.21 | JeffM | that's what I'd do |
| 21:35.28 | JeffM | cus you were ready to check in |
| 21:35.46 | mdskpr_ | i did, but McSpider sent another patch so i just ersed my tree and re checkouted |
| 21:35.50 | JeffM | ahh |
| 21:36.05 | mdskpr_ | is there a way to reverse a patch? |
| 21:36.09 | JeffM | revert |
| 21:36.24 | JeffM | it will replace your changes with what is in the repo |
| 21:36.52 | Thumper_ | JeffM: what I have is very broken. I'm going to go read the NetHandler code and try to understand what it does |
| 21:36.55 | *** join/#bzflag bier (~bier@p54A5BE57.dip.t-dialin.net) |
| 21:37.21 | JeffM | I understand why recieve could return ReadPart and 0 now |
| 21:37.36 | JeffM | but what I don't understand is why we never get back there on the next pass |
| 21:37.45 | JeffM | we should try to read every pass |
| 21:37.53 | JeffM | one of them should return data |
| 21:38.12 | JeffM | unless select clears out the data in a socket |
| 21:39.54 | CIA-114 | BZFlag: 03mudskipper * r21542 10/trunk/bzflag/ (26 files in 7 dirs): Accepted patch #3303614 from McSpider, adding OpenFFA, fixed robot targeting of robots, bumped PROTO->205 |
| 21:43.03 | JeffM | Thumper_, does deleteMe get set for your peer in that first run? |
| 21:44.03 | JeffM | you can also put a breakpoint in NetHandler::flushData to see if anyone calls that |
| 22:46.06 | mdskpr_ | JeffM: for the image downloading in bzflag, you want to force ONLY images.bzflag.org? do you want the player to still be able to choose others? |
| 22:47.05 | JeffM | they should always be allowed to set there own |
| 22:47.17 | JeffM | the default now is *.bzflag. org and *.bzflag.bz |
| 22:47.25 | JeffM | I want *.images.bzflag.org as the only default |
| 22:47.37 | mdskpr_ | oh ok |
| 22:47.59 | JeffM | at some point we will want to make a dialog that says "this map uses images from . XXX.YYY.ZZZ do you want to allow it?" |
| 22:48.02 | JeffM | but not now |
| 22:48.25 | mdskpr_ | you sure? that wouldn't be too difficult or invasive |
| 22:48.35 | JeffM | really? how do you make a dialog in bzflag? |
| 22:48.38 | JeffM | ;) |
| 22:49.05 | JeffM | we have no "pop up dialog" equivalent |
| 22:49.17 | mdskpr_ | ic |
| 22:49.23 | JeffM | and you don't want to use the chatbox that'd be silly |
| 22:49.33 | JeffM | Erroneous was working on diaogs and stuch for the GUI |
| 22:49.33 | mdskpr_ | right |
| 22:50.51 | Erroneous | hm, I need to migrate the navqueue stuff in. then dialogs should be pretty straightforward. |
| 22:51.07 | Erroneous | not sure the server list is going to work though; I think it depends pretty heavily on scalable fonts |
| 22:51.23 | JeffM | we can render more fonts if need be |
| 22:51.32 | Erroneous | yeah |
| 22:52.15 | JeffM | serverside scoring got put in? |
| 22:55.46 | mrapple | JeffM: i could work on that little dialog asking for image approval |
| 22:56.01 | JeffM | mrapple, we don't have a dialog system in the game yet |
| 22:56.03 | mrapple | "Press Y to continue or N to abort" |
| 22:56.07 | JeffM | so what would you "work on" |
| 22:56.20 | mrapple | i imagine it as something replacing the connecting GUI |
| 22:56.26 | JeffM | no |
| 22:56.32 | mrapple | you want it to pop up while connecting? |
| 22:56.38 | JeffM | yes |
| 22:56.45 | mrapple | it could go where 'start server' went |
| 22:57.01 | JeffM | Erroneous has code for a generic diallog system for notifcation and allersts |
| 22:57.05 | JeffM | we want something generic |
| 22:57.07 | JeffM | not a hack |
| 22:57.19 | mrapple | he has them done? |
| 22:57.37 | JeffM | in 2.99 IRC |
| 22:57.40 | JeffM | or the guts for them |
| 22:57.47 | mrapple | ok ill look into it |
| 22:57.59 | Erroneous | it's closer |
| 22:58.05 | JeffM | by that you mean first ask him if he's going to do it? |
| 22:58.06 | JeffM | ;) |
| 22:58.29 | *** join/#bzflag Pimpinella (~frank@gondolin.pimpi.org) |
| 22:58.31 | mrapple | going to do what? write the image thing or write the dialog thing |
| 22:58.58 | JeffM | the dialog thing |
| 22:59.05 | JeffM | gun... do not jump it |
| 22:59.26 | mrapple | i was just curiuos, its not like i started writing something... |
| 22:59.34 | JeffM | I know |
| 22:59.42 | JeffM | you just get hella excited then go off on your own ;) |
| 22:59.43 | Erroneous | please don't add dialogs until I get at least the navqueue stuff moved down. it will be harder for you AND harder for me :) |
| 22:59.52 | mrapple | will do |
| 23:00.05 | JeffM | you are more then welcome to help, but work with Erroneous on it |
| 23:00.22 | mrapple | JeffM: im curiuos, i wouldnt say excited, and im currently busy with other things, hence curious |
| 23:00.33 | JeffM | it sounds like you are exiced |
| 23:00.53 | mrapple | i guess its hard to convey tones of voice with text chat then |
| 23:01.01 | JeffM | mostly the throwing out ideas before asking about the history of the featue :) |
| 23:01.05 | JeffM | it is indeed |
| 23:07.12 | JeffM | if server side scoring got put in, where is it set? I want to add an event |
| 23:09.39 | blast007 | JeffM: do we also resources.bzflag.org? |
| 23:09.44 | JeffM | ok |
| 23:09.48 | blast007 | (in the allow list) |
| 23:09.49 | JeffM | those 2 URLs |
| 23:09.53 | JeffM | that's fine |
| 23:10.02 | JeffM | as long as it's not just a TLD |
| 23:10.29 | blast007 | ooo, we can just move everything under my.bzflag.org! that makes everything easier. |
| 23:10.33 | blast007 | ;) |
| 23:10.41 | JeffM | rrrrrriiiiiigggghhhhttttttttttt...... |
| 23:10.52 | JeffM | cus IE can't handle a site cookie |
| 23:11.21 | blast007 | yeah, I wonder how far back in versions you have to go to find one that doesn't |
| 23:11.27 | JeffM | 3 |
| 23:11.31 | blast007 | hehe |
| 23:23.33 | CIA-114 | BZFlag: 03JeffM2501 * r21543 10/trunk/bzflag/ (10 files in 4 dirs): add a subset of ServerSidePlayers to the API, mostly for BZFCron |
| 23:24.24 | JeffM | hmm I should make a ssp that just says quotes from the fact sphere |
| 23:39.24 | *** join/#bzflag bier|tp (~bier@p54A5BE57.dip.t-dialin.net) |
| 23:53.53 | CIA-114 | BZFlag: 03trepan * r21544 10/trunk/bzflag/ (3 files in 2 dirs): |
| 23:53.54 | CIA-114 | BZFlag: * added the UNUSED(var) macro to common.h |
| 23:53.54 | CIA-114 | BZFlag: * fixed the unused var warnings |