| 02:35.15 | *** join/#bzflag JeffM (~JeffM@cpe-23-241-222-73.socal.res.rr.com) |
| 03:09.12 | *** join/#bzflag Pimpinella (~frank@gondolin.pimpi.org) |
| 03:23.30 | *** join/#bzflag drizzo4shizzo (~dan@pool-74-96-96-26.washdc.fios.verizon.net) |
| 03:25.11 | drizzo4shizzo | ping |
| 03:53.39 | drizzo4shizzo | curious about the latest git plans |
| 03:57.59 | drizzo4shizzo | found github containing svn_to_git dir with git-svn scripts - rather than actual code tree I was hoping for #sadbanana |
| 03:58.31 | drizzo4shizzo | well, goodnight self :-P |
| 04:11.30 | JeffM | drizzo4shizzo, there are branches for differnet versions |
| 04:12.32 | drizzo4shizzo | these branches live in SVN currently? |
| 04:13.06 | JeffM | no |
| 04:13.08 | JeffM | <PROTECTED> |
| 04:13.22 | drizzo4shizzo | oh joy. I must have been too hasty in my searching |
| 04:14.15 | JeffM | what code are you looking for |
| 04:14.21 | JeffM | the stuff compatable with 2.4? |
| 04:14.42 | JeffM | because it's in https://github.com/BZFlag-Dev/bzflag-import-3/tree/v2_4_x |
| 04:15.03 | drizzo4shizzo | oh i see thanks! |
| 04:15.10 | JeffM | sure |
| 04:15.33 | JeffM | the svn branches were cleaned up and shuffled around during the import |
| 04:15.48 | drizzo4shizzo | I guess I was confused cos normally github shows you "master" but we don't have one of those which is fine |
| 04:16.00 | JeffM | yeah it's weird |
| 04:16.24 | JeffM | I had it showing the 2.5.x dev stuff but then I got yelled at |
| 04:16.32 | drizzo4shizzo | haha looks good to me |
| 04:17.01 | drizzo4shizzo | so you freq merge from 2_4_x to 2_6_x and perhaps sometimes rebase the other way |
| 04:17.13 | JeffM | no |
| 04:17.21 | JeffM | nothing happens frequently |
| 04:17.28 | drizzo4shizzo | right! silly me |
| 04:17.32 | JeffM | but changes do get merged when they happen |
| 04:18.01 | drizzo4shizzo | from 2.4 to 2.6 u mean sure |
| 04:18.12 | JeffM | yeah |
| 04:18.38 | drizzo4shizzo | I just discovered lolcommits |
| 04:18.55 | drizzo4shizzo | trigger webcam picture when committing to git repo :-P |
| 04:19.02 | JeffM | why? |
| 04:19.38 | drizzo4shizzo | it is pointless. just another fun code trick |
| 04:20.11 | drizzo4shizzo | https://github.com/mroth/lolcommits |
| 04:20.35 | JeffM | no thanks |
| 04:24.54 | drizzo4shizzo | oh wait there is a master branch. what is that then |
| 04:25.42 | JeffM | just the import scripts |
| 04:26.18 | drizzo4shizzo | oh derp. thats why thats what I got when I cloned |
| 04:26.31 | JeffM | thankfully it didn't take too long |
| 04:26.51 | JeffM | for code compatable with the current shipping release use the URL I gave you |
| 04:27.26 | drizzo4shizzo | yeah I'm used to hving one url for the whole shebang |
| 04:27.50 | JeffM | that would be for small proejects that don't have multiple releases with maintenance |
| 04:28.44 | JeffM | we had discussions about how to structure things, this is what we ended up with |
| 04:28.57 | drizzo4shizzo | I'm not second guessing, I've just never seen it b4 |
| 04:29.07 | drizzo4shizzo | I got a 403 with that URL btw |
| 04:29.25 | JeffM | in a browser? |
| 04:29.43 | drizzo4shizzo | no "git clone ..." |
| 04:29.51 | JeffM | yeah that's not the clone URL |
| 04:29.54 | JeffM | read the page |
| 04:30.02 | JeffM | you are cloning a branch |
| 04:30.13 | JeffM | branches arn't like folders in svn |
| 04:30.19 | drizzo4shizzo | I know thanks jeff |
| 04:30.36 | drizzo4shizzo | I just forgot that "master" has completely different files than the other branches in this case |
| 04:31.32 | drizzo4shizzo | just needed to "git checkout --track origin/v2_4_x |
| 04:31.54 | drizzo4shizzo | thanks for helping - night |
| 04:32.42 | JeffM | sure |
| 05:30.12 | *** join/#bzflag I_Died_Once (~I_Died_On@unaffiliated/idiedonce/x-1828535) |
| 05:47.31 | *** join/#bzflag hj (4c631570@gateway/web/freenode/ip.76.99.21.112) |
| 07:56.19 | *** join/#bzflag Not-001 (~notifico@198.199.82.216) |
| 07:56.19 | Not-001 | [02bzflag-import-3] 07JMakey pushed 034 commits to 03v2_4_x [+0/-0/±5] 13http://git.io/zd5kHg |
| 07:56.20 | Not-001 | [02bzflag-import-3] 07JMakey 03a48fc0f - Allow players to join the rabbit and hunter teams only in rabbit chase games. |
| 07:56.21 | Not-001 | [02bzflag-import-3] 07JMakey 033443f97 - Pay full attention when a plugin modifies a player's team. |
| 07:56.22 | Not-001 | [02bzflag-import-3] 07JMakey 038e58524 - Kick players who send MsgNewRabbit when the game is not rabbit chase. |
| 07:56.23 | Not-001 | [02bzflag-import-3] ... and 1 more commits. |
| 10:08.34 | *** join/#bzflag Constitution (~josh@bzflag/developer/Constitution) |
| 11:07.58 | blast007 | BulletCatcher: Actually, the player should never join as the rabbit, I'd think. |
| 11:41.55 | *** join/#bzflag Bro666 (~paul@89.130.112.113) |
| 12:57.53 | WarPig___ | "07<JeffM> 01nothing happens frequently01" <-- Quote of the day :D |
| 13:22.34 | blast007 | WarPig___: black text doesn't show up too well on a black background ;) |
| 13:25.27 | WarPig___ | ?? No black background I know of. Doesn't look black to me. |
| 13:26.13 | WarPig___ | simply a copy/paste one of JeffM's nuggets of wisdom. |
| 13:27.01 | WarPig___ | "<JeffM> nothing happens frequently" <-- Quote of the day :D |
| 13:27.55 | blast007 | WarPig___: well, I (and others) run IRC clients in a terminal ;) |
| 13:29.22 | blast007 | (which doesn't necessarily always mean a black background, but that's rather common) |
| 13:34.54 | WarPig___ | Ah... well no text formatting of any kind was done (intentionally) by me. Just a simple copy/paste. |
| 13:38.07 | blast007 | yep ;) |
| 13:47.24 | *** mode/#bzflag [+o blast007] by ChanServ |
| 13:47.57 | *** mode/#bzflag [+b *!*@186.150.0.0/16] by blast007 |
| 13:48.16 | *** mode/#bzflag [-o blast007] by blast007 |
| 15:31.08 | WarPig___ | blast007: DR hacker boy on your server at the moment. |
| 15:38.55 | kierra | he took down castle war bztank |
| 15:41.53 | hj | of course - apoc's probably next then |
| 15:42.23 | hj | when does this kid sleep?? |
| 15:43.33 | blast007 | the weekends? |
| 15:50.05 | *** join/#bzflag GvzEvxre (~TimRiker@bzflag/projectlead/TimRiker) |
| 15:50.05 | *** mode/#bzflag [+o GvzEvxre] by ChanServ |
| 15:51.28 | kierra | ok, guys |
| 15:51.48 | kierra | since we really havent been able to stop this DR idiot |
| 15:52.18 | kierra | he joins at will any server with proxies and cheats.....is driving players off |
| 15:52.36 | kierra | bans really arent working |
| 15:52.52 | Not-001 | [02bzflag-import-3] 07blast007 pushed 032 commits to 03v2_4_x [+0/-0/±6] 13http://git.io/zbOOEw |
| 15:52.53 | Not-001 | [02bzflag-import-3] 07blast007 0374876f1 - Undo most of a48fc0f and instead fix ec8c0e0 to properly handle rabbit chase games. |
| 15:52.54 | Not-001 | [02bzflag-import-3] 07blast007 03682ea10 - Ignore generated HTML man pages. |
| 15:53.32 | kierra | I've had more than one player ask me: why not let him play with his home ip & reg'd nick.......since he "promises" not to cheat |
| 15:53.59 | kierra | dont shoot the messenger. But something has to be done that works |
| 15:54.04 | blast007 | because he still will |
| 15:54.11 | hj | blast, that may be true |
| 15:54.13 | blast007 | he's having far too much fun |
| 15:54.14 | hj | he probably will |
| 15:54.22 | hj | however, have we been doing anything that can stop him right now? |
| 15:54.38 | blast007 | hj: fixes are being worked on |
| 15:54.41 | hj | would unbanning him be giving him more access to cheating/allow him to hceat more? |
| 15:54.43 | hj | cheat* |
| 15:54.55 | blast007 | sure |
| 15:55.07 | blast007 | then he wouldn't even need proxies to cheat |
| 15:55.13 | hj | the fixes are great - we should still employ them to make the game's infrastructure stronger |
| 15:55.18 | hj | that's true, but he's using proxies with ease |
| 15:55.32 | hj | either way, the end result is the same - he gets on and cheats at whim |
| 15:56.02 | hj | at this point, i don't see how the bans have been effective - he still gets on every server easily and does what he wants |
| 15:56.04 | blast007 | well, I can make proxies of limited use with one fix I'm working on |
| 15:56.32 | hj | that's solid - how long will it take? |
| 15:56.47 | blast007 | dunno, will need some testing |
| 15:56.54 | hj | because every day that this guy can hack servers, we will lose more and more non-committed players |
| 15:56.55 | WarPig___ | the fix for the ability to use up the slots will help a lot. |
| 15:57.01 | blast007 | he can't "hack" servers |
| 15:57.04 | hj | and bzflag has already been declining rapidly |
| 15:57.42 | hj | well he's shortening people's shots and uses wg everywhere (skipping over the bots/crashing servers/etc) |
| 15:58.03 | kierra | and he can get any flag at will, in ctf, caps over and over again |
| 15:58.07 | WarPig___ | A simple firewall rule to limit tcp connections should be SOP for server owners |
| 15:58.10 | blast007 | when has he actually crashed servers? |
| 15:58.26 | hj | by crashed i mean disallow others from joining |
| 15:58.26 | blast007 | so far I think it's just been filling up slots so that it says "team is full" when you try joining |
| 15:58.30 | WarPig___ | never, to my knowledge. |
| 15:58.32 | kierra | he crashed bztank's castle warfare today |
| 15:58.33 | blast007 | yeah, see, that's NOT crashing |
| 15:58.39 | blast007 | don't confuse terms |
| 15:58.43 | hj | effectively, what's the difference? |
| 15:58.58 | hj | people can't play - that's all that matters to pepople who want to play the game, no? |
| 15:58.59 | blast007 | crashing implies that the actual bzfs process is terminating abnormally |
| 15:59.00 | hj | people* |
| 15:59.01 | kierra | ok, filled the slots....results same temporarily |
| 15:59.18 | hj | agree w/ kierra |
| 15:59.22 | kierra | well, crashing is the term Dr uses |
| 15:59.31 | kierra | he denies access :) |
| 15:59.32 | blast007 | is crashing a car and having a car engine stop working the same thing? no. |
| 15:59.44 | kierra | ok, power thru precision |
| 15:59.50 | hj | for the passengers, yes |
| 15:59.54 | WarPig___ | symantics. Rendering a server inoperable is teh same as crashing it, from the players perspective. It's very different from the dev's perspective. |
| 15:59.58 | blast007 | hj: no.. |
| 16:00.09 | hj | fine, one is more easily fixable than the other, heh |
| 16:00.15 | blast007 | WarPig___: and no, it isn't. |
| 16:00.22 | kierra | lets not argue semantics and forget the problem :) |
| 16:00.26 | blast007 | people still on the server can continue playing, but new players cannot join |
| 16:00.37 | blast007 | whereas if the server crashed, everyone would be booted off |
| 16:00.48 | blast007 | so again, not the same at all |
| 16:00.50 | hj | then when players leave, what happens? can more join? |
| 16:00.50 | WarPig___ | in this case, the players all go NR and can't play. |
| 16:01.12 | hj | ^ didn't know that, war |
| 16:01.16 | WarPig___ | ...or in the case I witnesses, at least. |
| 16:01.17 | kierra | something's been occurring on mofo....huge lag/jit, we all go nr |
| 16:01.24 | blast007 | *shrug* |
| 16:01.29 | blast007 | still not a crash, just a DoS |
| 16:01.34 | WarPig___ | agree |
| 16:01.35 | Delusional | hj: declining, yes... rapidly, no |
| 16:01.36 | hj | well, can we agree on it makes the game unplayable for players? |
| 16:01.40 | kierra | okok, so a fix? |
| 16:01.45 | blast007 | my point is that a crash would require a different fix :P |
| 16:01.52 | hj | is a fair point |
| 16:02.20 | blast007 | (and for the people trying to fix it, knowing the actual issue is important) |
| 16:02.27 | hj | but anyway, i'm saying unbanning him seems like a low risk high reward possibility |
| 16:02.57 | hj | i guess i may not fully understand the details, however |
| 16:03.03 | hj | so let's break this down then |
| 16:03.11 | hj | what does he stand to gain by us unbanning him? |
| 16:03.28 | hj | other than intangibles (peace of mind, feeling of superiority, etc.) |
| 16:03.31 | *** join/#bzflag `ls` (~TimRiker@bzflag/projectlead/TimRiker) |
| 16:03.31 | *** mode/#bzflag [+o `ls`] by ChanServ |
| 16:04.12 | blast007 | hj: he'll have "win" |
| 16:04.14 | blast007 | won* |
| 16:04.16 | WarPig___ | I will say that someone else will have to lift *my* bans. I will not. |
| 16:04.37 | hj | fair enough, blast, yes in his head and in our heads he'll have won and we'll have given up |
| 16:04.38 | blast007 | we do not give in to terrorists ;) |
| 16:04.56 | WarPig___ | ~blast007++ |
| 16:04.58 | hj | well, there's a point where we can stick to our ideals and there's a point where we have to sit back and look at things practically |
| 16:05.00 | hj | don't you think? |
| 16:05.15 | blast007 | hj: the practical response is to shut down his points of attack |
| 16:05.26 | blast007 | there will be other asshats, so why not just fix it? |
| 16:05.36 | hj | we should fix it |
| 16:05.42 | hj | but that doesn't mean we can't employ both methods |
| 16:06.14 | hj | i guess i don't know - how long would it take to make all the patches to counteract his every cheat? |
| 16:07.00 | hj | if it takes any longer than it would to unban him and he stops cheating (even if there's a chance of that occurring) |
| 16:07.11 | hj | then i think unbanning him would be the most practical short-term solution |
| 16:07.34 | hj | especially when players have to deal with not being able to play on their favorite servers |
| 16:08.46 | blast007 | my plan is to have both fixes out today |
| 16:08.58 | hj | wow - that's awesome then |
| 16:09.03 | kierra | cool blast! |
| 16:09.04 | hj | if they work, then excellent |
| 16:09.05 | brad | ~lart hj |
| 16:09.05 | infobot | pulls out a ClueBat (tm) and thwaps hj |
| 16:09.14 | hj | haha |
| 16:09.15 | WarPig___ | lol |
| 16:09.18 | hj | brat :p |
| 16:09.21 | hj | go post that news |
| 16:10.14 | hj | but if he continues to cheat - i really think that should be our backup plan |
| 16:11.35 | brad | I disagree hj |
| 16:11.50 | hj | then what would your plan of action be? |
| 16:12.17 | brad | Fix the bugs, keep banning him |
| 16:12.28 | hj | and when admins aren't around to ban him? |
| 16:12.40 | hj | last night, he was running rampant |
| 16:12.47 | brad | Poll ban :p |
| 16:12.55 | hj | actually, on LOPAH frustration, tt banned him and he joined 1 min later |
| 16:13.27 | hj | tried to - there were only 2 registered players unfortunately - that's when i brought tt over hehe |
| 16:13.32 | blast007 | my fixes will require that server owners build from the latest Git v2_4_x code, of course, so this won't help everywhere |
| 16:14.42 | hj | can you give a quick idea of the scope? |
| 16:14.50 | hj | how many servers would this help |
| 16:14.59 | hj | most/half/etc |
| 16:15.01 | blast007 | how many stars are there in the sky? |
| 16:15.12 | hj | not sure, i've never sat back and counted ;p |
| 16:15.20 | blast007 | me neither |
| 16:15.45 | hj | hopefully it works out for most |
| 16:16.11 | blast007 | most server owners don't care about their server |
| 16:17.42 | hj | well after they're used, we can sit back and reevaluate |
| 16:17.52 | *** join/#bzflag Pimpinella (~frank@gondolin.pimpi.org) |
| 16:17.57 | hj | would be nice to finally be rid of this guy |
| 16:18.25 | kierra | he was poll banned earlier...but came back with different proxy |
| 16:18.46 | hj | yeah - it doesn't take him long - but sounds like that's one of the patches that was being talked about? |
| 16:20.14 | brad | all the proxies I've seen him used are on dns blacklists |
| 16:20.23 | brad | use* |
| 16:20.30 | hj | meaning? |
| 16:20.44 | hj | yeah i'm not nearly as tech savvy as the rest of you, heh |
| 16:20.55 | brad | a plug-in could automatically ban him |
| 16:20.56 | blast007 | meaning it could be possible block the majority of his proxies using a DNSBL plugin |
| 16:21.04 | brad | but it could also block innocent users |
| 16:21.06 | blast007 | but that would only affect full joins |
| 16:21.15 | brad | ya |
| 16:21.23 | hj | full joins? |
| 16:21.35 | brad | not his connection flooding trick |
| 16:21.35 | blast007 | as in, not the half-connections he's doing to flood server slots |
| 16:21.44 | hj | ah |
| 16:22.10 | hj | working on the DNSBL plugin? |
| 16:22.14 | blast007 | no |
| 16:22.21 | brad | :Pp |
| 16:22.26 | brad | saved that job for you hj |
| 16:22.31 | hj | sweeeeeeet |
| 16:22.39 | hj | so what buttons do i click |
| 16:22.41 | hj | ;) |
| 16:22.42 | brad | heh |
| 16:23.03 | hj | brad, you posted! |
| 16:23.06 | brad | I could give it a go |
| 16:23.14 | blast007 | testing one fix right now |
| 16:23.24 | brad | but I suck at making plug-ins, allejo always cleans mine up after:p |
| 16:23.44 | hj | yeah - don't mean to get on yall's asses about this - we all have lives outside of bz as well heh |
| 16:24.03 | hj | well, most of us |
| 16:24.04 | hj | cough, brad |
| 16:25.12 | blast007 | well, I took the day off ;) |
| 16:26.07 | BulletCatcher | Today is a holiday for me. I'm making it harder to cheat, too. |
| 16:27.19 | hj | holiday? for what? |
| 16:27.33 | blast007 | BulletCatcher: did my 74876f1 seem fine? |
| 16:27.41 | BulletCatcher | It's Caesar Chavez day in California. |
| 16:28.02 | blast007 | 74876f1 commit, I mean |
| 16:28.23 | BulletCatcher | I was wondering why you replaced my simple and effective change with one that was much more complicated. |
| 16:29.05 | blast007 | because joining as "rabbit" is invalid and should be rejected regardless of if players are on a server or not |
| 16:31.18 | BulletCatcher | My version prevented joining non-rabbit games as rabbit, and joining a rabbit game as rabbit was harmless in my test. |
| 16:32.04 | BulletCatcher | At least I think I tested that. It was late. |
| 16:32.40 | blast007 | did you test joining as a rabbit and not spawning and then having another client join and see if they would become a rabbit upon spawn? |
| 16:33.00 | BulletCatcher | I definitely didn't test that particular scenario. |
| 16:33.11 | blast007 | me neither, but I will in a sec ;) |
| 16:34.44 | WarPig___ | bigtime jitter/lag at mofo atm |
| 16:37.32 | blast007 | hmm, seems we already prevent joining as the rabbit team somewhere |
| 16:38.21 | blast007 | gonna find where that is checked as that may be a better place for more team validations |
| 16:42.01 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 16:42.01 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 16:45.40 | blast007 | seems like the autoTeamSelect method in bzfs.cxx could handle the check |
| 16:46.03 | blast007 | that's where it forces non-observers on a rabbit chase into the hunter team |
| 16:46.55 | BulletCatcher | With my change, joining a rabbit map as rabbit puts me into the hunter team. |
| 16:48.56 | blast007 | without either of our changes, it does the same (tested in a stock 2.4.2) |
| 16:48.57 | BulletCatcher | I assume that is forced in autoTeamSelect(). |
| 16:49.03 | blast007 | yep |
| 16:49.28 | blast007 | looking to see if that's a viable spot to place the anti-hunter/rabbit options for other modes |
| 16:50.20 | blast007 | I'll be revering my change regardless as the edge case I was thinking of was already handled with autoTeamSelect |
| 16:51.35 | BulletCatcher | One reason I like my maxTeam[] method is because its RabbitTeam and HunterTeam are already referenced, but were not initialized before. |
| 16:52.13 | BulletCatcher | I'm not a fan of referencing uninitialized variables. |
| 16:55.52 | BulletCatcher | A completely different change I am testing is to require player.isAlive() in invalidPlayerAction(). |
| 16:55.53 | BulletCatcher | I see no problems with it in simple tests. |
| 16:57.11 | blast007 | I'm reverting my change and going back to your method. Does that affect anything you're doing right now? |
| 16:57.21 | BulletCatcher | Nope. Go for it. |
| 16:57.35 | Not-001 | [02bzflag-import-3] 07blast007 pushed 031 commit to 03v2_4_x [+0/-0/±5] 13http://git.io/G1bTTA |
| 16:57.36 | Not-001 | [02bzflag-import-3] 07blast007 032de12f2 - Revert "Undo most of a48fc0f and instead fix ec8c0e0 to properly handle rabbit chase games." This reverts commit 74876f1fac45a7b4658f00a2c10f231414dc4e2c. |
| 17:03.35 | blast007 | hmm, my connection timeout code doesn't handle long map downloads (and probably not long texture downloads), as I feared |
| 17:04.38 | BulletCatcher | Commit early; commit often. Assuming a successful partial fix. |
| 17:06.26 | blast007 | would certainly prevent new people from joining mofo, though |
| 17:08.24 | blast007 | even 30KB of map data would take ~6 seconds to down with 200ms of lag to the server |
| 17:08.42 | blast007 | (assuming that an HTTP source is not provided, fastmap or otherwise) |
| 17:08.59 | blast007 | download* |
| 17:10.30 | blast007 | BulletCatcher: where had you placed the (playerData._LSAState == GameKeeper::Player::start ...) in your version? |
| 17:11.12 | BulletCatcher | in doStuffOnPlayer() |
| 17:12.11 | BulletCatcher | But I can see how that might kick a merely laggy player with a large map and/or texture download. |
| 17:13.40 | BulletCatcher | iptables connection limits may be the best defense against this type of DoS. |
| 17:33.23 | blast007 | yeah, fixing this part in bzfs might have to wait for 2.6 where we could completely redo the join game protocol |
| 17:35.56 | BulletCatcher | A 5-minute (or so) timeout in doStuffOnPlayer() ought to be reasonable. |
| 17:41.35 | blast007 | but does that really do all that much to help? |
| 17:42.03 | BulletCatcher | That would speed recovery after he stops making new connections. |
| 17:42.50 | BulletCatcher | As it is now, I think only TCP keepalive is what makes dead connections go away. |
| 17:42.56 | blast007 | he probably doesn't stop until he shuts the attack script down completely (which would probably kill all the connections) |
| 17:43.25 | blast007 | hmm |
| 17:43.26 | BulletCatcher | I saw an instance where his IP address was dead. |
| 17:43.36 | blast007 | k |
| 17:44.01 | BulletCatcher | TCP keepalive did its job, but slowly. |
| 17:48.46 | BulletCatcher | We could set TCP_KEEPIDLE on the connection to make it work more quickly than the Linux default of 2 hours. |
| 18:05.36 | Not-001 | [02bzflag-import-3] 07JMakey pushed 032 commits to 03v2_4_x [+0/-0/±2] 13http://git.io/ZUEbRg |
| 18:05.37 | Not-001 | [02bzflag-import-3] 07JMakey 03c3fa0dc - In invalidPlayerAction() kick dead players. |
| 18:05.38 | Not-001 | [02bzflag-import-3] 07JMakey 031c75a3f - Validate MsgKilled and MsgShotEnd with invalidPlayerAction(). |
| 18:07.37 | BulletCatcher | That's all I have for now. |
| 18:31.19 | Not-001 | [02bzflag-import-3] 07blast007 pushed 031 commit to 03v2_4_x [+0/-0/±2] 13http://git.io/xh2S0Q |
| 18:31.20 | Not-001 | [02bzflag-import-3] 07blast007 03df30ca4 - Limit half-joined player connection duration to five minutes. |
| 18:36.39 | blast007 | Seems like the server handling MsgNewRabbit from the client isn't technically necessary. The only time that happens is when the client pauses or if the client is paused while it receives the message. The server could (or does?) track paused tanks. |
| 18:37.48 | BulletCatcher | The server already has isPaused(), so we should be able to drop MsgNewRabbit in 2.6. |
| 18:38.24 | BulletCatcher | I suppose we could even make it a noop in 2.4. |
| 18:38.55 | blast007 | yeah, looks like it already checks the paused state |
| 18:39.22 | blast007 | and for 2.6 we could do proper team switching |
| 18:43.13 | BulletCatcher | Any reason not to do a logDebugMessage(2,...) in rejectPlayer()? |
| 18:43.44 | BulletCatcher | Mostly, there is no logDebugMessage() associated with rejectPlayer() calls. |
| 18:44.24 | blast007 | removePlayer triggers a logDebugMessage(1, ...) |
| 18:45.07 | blast007 | (which gets called by rejectPlayer) |
| 18:45.37 | blast007 | except in cases where playerData isn't valid anymore |
| 18:45.51 | BulletCatcher | Maybe we just need a better description in rejectPlayer()'s call to removePlayer(). |
| 18:46.42 | blast007 | when I was testing my connection duration fix, I was getting this: |
| 18:46.44 | blast007 | Player [0] removed at 2014-03-28 18:23:14: /rejected |
| 18:47.01 | BulletCatcher | Yeah. I'd like to see more than "/rejected". |
| 18:47.29 | blast007 | so you're thinking of showing the 'reason' as well? |
| 18:47.48 | BulletCatcher | Yup, although I worry that it will break someone's log parser. |
| 18:48.00 | blast007 | perhaps |
| 18:48.15 | blast007 | dunno if anyone does much of that anymore |
| 18:48.24 | blast007 | norang was the only one I knew of personally |
| 18:48.51 | blast007 | (and they may have used the output of LogDetail instead) |
| 18:49.07 | blast007 | bzfs' output isn't generally all that parsable :) |
| 18:50.26 | BulletCatcher | I'm going to be bold and have it show "/rejected: " + reason. |
| 18:50.44 | BulletCatcher | We can respond appropriately if someone complains. |
| 18:51.11 | blast007 | "Fix your scripts" |
| 18:51.27 | BulletCatcher | That may, indeed, be an appropriate response. :-) |
| 18:51.33 | blast007 | or maybe do it as "/rejected (reason)" ? |
| 18:53.40 | BulletCatcher | Parentheses are harder to parse. |
| 18:54.04 | blast007 | Okay. "/rejected {{[reason]}}" ;) |
| 18:54.15 | BulletCatcher | Done. :-p |
| 19:01.40 | BulletCatcher | Ouch. The "reason" string has ANSI codes that I don't want in my log file. |
| 19:02.52 | blast007 | heh |
| 19:03.19 | BulletCatcher | I'll leave it as is for now. |
| 19:03.41 | blast007 | there is a stripAnsiCodes method |
| 19:04.14 | blast007 | function* |
| 19:05.23 | BulletCatcher | Much better. Thanks. |
| 19:09.41 | *** join/#bzflag TimRiker (~TimRiker@proxy.lds.org) |
| 19:09.43 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 19:09.43 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 19:12.53 | Not-001 | [02bzflag-import-3] 07JMakey pushed 031 commit to 03v2_4_x [+0/-0/±1] 13http://git.io/WxxVBQ |
| 19:12.54 | Not-001 | [02bzflag-import-3] 07JMakey 03a1d4ed8 - Log a more verbose reason in rejectPlayer(). |
| 19:13.04 | blast007 | hmm.. trying to figure out the best way to update my production source tree.. I have some local modifications to add some plugins to the build system.. |
| 19:14.18 | blast007 | can't 'git pull --rebase' as I have unstaged changes, can't 'git pull' because I have local changes to files that have been modified upstream. So I either have to commit or stash my changes. |
| 19:14.57 | BulletCatcher | I have a private branch that I rebase onto v2_4_x when I want to incorporate upstream changes. |
| 19:15.08 | BulletCatcher | I have to be careful not to push that branch upstream. |
| 19:15.39 | blast007 | yeah, trying to avoid accidentally pushing it upstream, so I'm leaning towards 'git stash' |
| 19:17.45 | BulletCatcher | What I do is have a local mirror of bzflag-import-3, then I have a clone of that (not upstream) that includes my BRL branch. |
| 19:18.26 | blast007 | probably the best option |
| 19:18.44 | BulletCatcher | As long as I only pull (don't push) from the BRL clone, upstream is safe. |
| 19:19.40 | BulletCatcher | Plus, since the BRL clone uses the local mirror as its upstream I have a chance to catch a mistake. |
| 19:19.49 | blast007 | yep |
| 19:20.27 | BulletCatcher | I even go to the extra effort of doing a development clone from my mirror. |
| 19:21.08 | BulletCatcher | It means I have to push wanted changes twice, but I can also easily and thoroughly discard something I don't like. |
| 19:21.58 | BulletCatcher | This extra effort on my part helps keep my mistakes out of the GitHub repo. |
| 19:23.29 | BulletCatcher | "git clone --mirror git@github.com:BZFlag-Dev/bzflag-import-3.git" creates a mirror without a working tree. |
| 19:23.52 | BulletCatcher | That's what I do my development clones from. |
| 19:26.21 | blast007 | sweet, that's the trick to getting it to track all the branchs |
| 19:26.24 | blast007 | branches* |
| 19:27.37 | BulletCatcher | Test with "git push --dry-run". |
| 19:28.17 | BulletCatcher | I haven't found a clean way to stop it from trying to push refs/pull. |
| 19:28.48 | BulletCatcher | Ugly but works: "mv refs/pull .. ; git push ; mv ../pull refs". |
| 19:30.32 | BulletCatcher | I haven't really used Mercurial at all, but I like the way it is possible to work around almost any Git oddity. |
| 19:35.09 | BulletCatcher | I use Git's "submodule" feature to automate importing the OneVsOne and leagueOverSeer plugins into my private branch. |
| 19:36.34 | *** join/#bzflag Delusional (~delusiona@unaffiliated/delusional) |
| 20:06.03 | Constitution | blast007: won't git pull merge uncommitted changes on the local clone? I'm in the same boat |
| 20:06.13 | blast007 | nope |
| 20:06.29 | blast007 | you have to commit or stash |
| 20:07.00 | blast007 | at least if any of the files modified upstream are ones you have modified locally |
| 20:07.26 | blast007 | (for instance, it didn't care about my configure.ac, but cared about my plugins/Makefile.am) |
| 20:08.08 | Constitution | oh |
| 20:13.25 | blast007 | so I'm gonna go with BC's idea and do a local 'git clone --mirror' of upstream, and then clone that local mirror and commit local-only changes there |
| 20:15.14 | blast007 | soon as I get my updater working, then I'll start working on the next countermeasure |
| 20:34.57 | blast007 | BulletCatcher: do you use 'git remote update' in your --mirror'ed clone to grab the new data from upstream? |
| 20:37.13 | BulletCatcher | I use "git fetch --all". |
| 20:59.55 | blast007 | found this as well: https://help.github.com/articles/duplicating-a-repository |
| 21:23.06 | blast007 | hrm, yeah.. can't figure out how to detect changes to the v2_4_x code so I know when to rebuild... |
| 21:24.53 | blast007 | was gonna use 'git rev-parse', but that won't work if I start committing stuff to my production clone |
| 21:25.19 | blast007 | this was so much easier with SVN :P |
| 22:07.51 | kierra | BulletCatcher: he closed BRL |
| 22:12.00 | kierra | i got the msg :server is full, try again later |
| 22:13.39 | kierra | damn, same msg again on rejoin |
| 22:19.47 | kierra | again the same msg |
| 22:39.45 | BulletCatcher | Lots of "Invalid attempt to end a shot while dead" events for legitimate players. |
| 22:39.59 | BulletCatcher | We'll have to relax that check. |
| 22:40.44 | brad | why was it just certain people getting kicked? |
| 22:40.55 | brad | over and over |
| 22:41.18 | brad | kierra: have you been cheating? ;) |
| 22:41.52 | kierra | always do.....skinny bullets, fat tank |
| 22:41.58 | kierra | brad, bzbg98 |
| 22:42.03 | kierra | waiting on you |
| 22:42.12 | BulletCatcher | The "reasonable timeframe" change seemed to work as expected. |
| 22:43.20 | blast007 | MsgShotEnd is UDP, so that could be part of the issue |
| 22:44.38 | BulletCatcher | Yeah, ordering. |
| 22:45.39 | BulletCatcher | Now I have a replay file to analyze. |
| 22:49.01 | BulletCatcher | oooh, kierra, a 12-year ban! |
| 23:12.07 | kierra | that would save me! |
| 23:13.16 | kierra | ty for your time analyzing, BC |
| 23:15.59 | BulletCatcher | One of the proxy bans you did on BRL is for 12 years. |
| 23:16.00 | BulletCatcher | Probably an extra zero in your excitement. |
| 23:16.14 | BulletCatcher | I'm not complaining. :-) |
| 23:50.20 | *** join/#bzflag thrakattak (cz3141@gateway/shell/devio.us/x-eqfopbntpgenxdam) |