Chat log from the meeting on 2014-04-01
From OpenSimulator
(Difference between revisions)
(Created page with "<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word"> [11:05] Marcus Llewellyn: They work...") |
Revision as of 13:31, 1 April 2014
[11:05] Marcus Llewellyn: They worked. That's not the same as now broken. ;) [11:05] Nebadon Izumi: 8 person physical roller coaster on BulletSim [11:05] Arielle.Popstar @bound2.dyndns.org: yes broken now [11:05] Richardus Raymaker: hi marcus [11:05] Robert Adams: Nebadon... I haven't... Cutulala showed me a physical test setup but I haven't seen anything finished [11:05] Connecting to in-world Voice Chat... [11:05] Marcus Llewellyn: Heya Rich [11:05] Connected [11:05] Justin Clark-Casey: hi folks [11:06] Arielle.Popstar @bound2.dyndns.org:8002: Hi Justin [11:06] Eliopod Beaumont: Hi Justin [11:06] Marcus Llewellyn smiles :) [11:06] Sarah Kline: Hi Justin [11:06] Nebadon Izumi is Online [11:06] Richardus Raymaker: i hope to start building next week. testing grtid this week. (and try to figure out where to start on the sim ) lol [11:06] Richardus Raymaker: hi justin [11:06] Nebadon Izumi: we are going to release a new Physics Arena oar soon that has the physical coaster included [11:06] Nebadon Izumi: i just need to check with Cute if its ready yet [11:07] Nebadon Izumi: soon as I see her again [11:07] Robert Adams: oh wow... [11:07] Marcus Llewellyn: Does it have one of those "You must be this high," signs with it? :) [11:07] Richardus Raymaker: i stay with the old pos way [11:07] Robert Adams: that looks great [11:07] Richardus Raymaker: for now [11:08] Richardus Raymaker: You mean lower then this hight marcus :) [11:08] Richardus Raymaker: or better between this signs [11:08] Nebadon Izumi: ya its pretty awesome [11:08] Nebadon Izumi: the left and right arrow keys make it spin around the track [11:08] Nebadon Izumi: and forward arrow makes it move [11:08] Nebadon Izumi: i made another video of less complex version [11:09] Nebadon Izumi: you can see exacly how it works [11:09] Marcus Llewellyn: All it needs is Oculus and sick bags. [11:09] Nebadon Izumi: https://www.youtube.com/watch?v=W6HS15u-Fck&hd=1 [11:09] Nebadon Izumi: its a torus wrapped around elongated cube track [11:09] Nebadon Izumi: its basically just a car script [11:09] Nebadon Izumi: slightly modified [11:09] Andrew Hellershanks is Online [11:09] Marcus Llewellyn: Prim build or mesh? [11:10] Nebadon Izumi: prims [11:10] Nebadon Izumi: well the 1st video the coaster is mesh [11:10] Nebadon Izumi: but we set physics to none on everything but the torus [11:10] Nebadon Izumi: to reduce stress on physics engine [11:10] Richardus Raymaker: cool, works for monorails etc. harder for train. i tried something else but scripts are to slow for that or sensors [11:12] Nebadon Izumi: exciting stuff :) [11:12] Nebadon Izumi: amazing it works so good [11:12] Andrew Hellershanks: yeah, I was quite impressed [11:12] Justin Clark-Casey: hey robert. As you know, I think we need to do a release soon [11:12] Nebadon Izumi: incase anyone just getting here missed the first video link [11:12] Nebadon Izumi: here it is again [11:12] Nebadon Izumi: https://www.youtube.com/watch?v=g7tYfQ1LYoA&hd=1 [11:12] Nebadon Izumi: physical roller coaster on Bulletsim [11:12] Richardus Raymaker: well, start to work with 0.8 more. and its going betetr [11:13] Richardus Raymaker: think a few bugs maby before release [11:13] Robert Adams: justincc... what do you want fixed for the r elease? [11:13] Justin Clark-Casey: robert: I presume variable regions should still be considered experimental? Would it be okay to have it disabled by defauly behind a configuration flag until its prime time? [11:13] Nebadon Izumi: I'm Exploding! [11:13] Arielle.Popstar @bound2.dyndns.org:8002: megas need fixing [11:13] Richardus Raymaker: hmm, think people disappointed if 0.8 dont have var [11:13] Robert Adams: there is not a special flag.... it is disabled if you don't speciffy a size in the Regions.ini [11:14] Nebadon Izumi: how could 0.8 not have var [11:14] Sarah Kline: I think they are just as ready as Megas ever were [11:14] Nebadon Izumi: that would be impossible [11:14] Richardus Raymaker: and terrain edit worrks sofar i tested acceptable. loading is still slow [11:14] Marcus Llewellyn: Ahhh... watching the video explains a lot. Very cool. :) [11:14] Robert Adams: in most cases, I check if the region is non-legacy and, only do the special code if so [11:14] Richardus Raymaker: hi sarah, i think and hope var;s are betetr then mega's already [11:14] Justin Clark-Casey: robert: you're happy that's enough to stop you getting an avalanche of bugs if it's not totally ready? [11:14] Sarah Kline: we converted our regions and are very happy with how its runninng [11:14] Nebadon Izumi: Richardus it would be impossible for 0.8 to not have var [11:14] Nebadon Izumi: its in core now [11:14] Sarah Kline: and visitors like too [11:15] Robert Adams: if people don't specify a larger size... only the normal avalanche of bugs will be expected [11:15] Richardus Raymaker: sarah, starting soon finaly i hoep with what i want for 1+ year [11:15] Nebadon Izumi: var is definitely ready for prim time if you ask me [11:15] Nebadon Izumi: im using it on several projects [11:15] Robert Adams: the only real breakage seems to be around mega-regions [11:15] Justin Clark-Casey: robert: ok, fair enough :) [11:15] Arielle.Popstar @bound2.dyndns.org:8002: not seems [11:15] Sarah Kline: lol people are going to max them of course [11:15] Arielle.Popstar @bound2.dyndns.org:8002: it is broken [11:15] Marcus Llewellyn: I've been running both var and non-var from master for weeks. I'm not much of a data point, but it's been stable for me. [11:16] Justin Clark-Casey: robert: the other bugs I currently have as major stoppers are mantis 7067, which is avatars hovering above groudn [11:16] Richardus Raymaker: hmm, 2048 is slow laoding. 1280 var works fine think its 1 minute before terrain is loaded [11:16] Nebadon Izumi: ya thats a tough one Justin [11:16] Justin Clark-Casey: robert: mantis 7050 where region crossing back and forth whilst sitting crashes the simulator [11:16] Nebadon Izumi: that likely requires new capsule? [11:16] Justin Clark-Casey: robert: and I need to investigate some issues with npcs [11:16] Mircea Kitsune: hi [11:16] Nebadon Izumi: i guess it could be made a bit better, but proper fix is new capsule i think right? [11:16] Arielle.Popstar @bound2.dyndns.org:8002: is var working with Ode now? [11:16] Marcus Llewellyn: Rich, slow loading how? Terrain? [11:16] Justin Clark-Casey: nebadon: personally, I expect it just requires adjustment to the existing cpasule [11:16] Sarah Kline: I have not noticed it recently [11:16] Nebadon Izumi: ya thats a bandaid though [11:16] Nebadon Izumi: which is fine [11:16] Justin Clark-Casey: well, of course Melanie broke it but there's no timeline for fixing it [11:16] Richardus Raymaker: one annoying problem, is the floating avatr. seen it now to default avater hovers around 0.5 [11:17] Justin Clark-Casey: which really isn't good enough imo [11:17] Robert Adams: justincc: there was a bug about NPC and MoveToTarget... I glanced at the code and the problem wasn't obvious [11:17] Richardus Raymaker: marcus, yes var terrain loading is slow [11:17] Justin Clark-Casey: robert: does it have a mantis? [11:17] Marcus Llewellyn: I can say mesh terrain loads very quickly. :) [11:18] Mata Hari: other kind of annoying bug is when initial inventory fetch completely locks up in non-plaza region [11:18] Justin Clark-Casey: yes, ideally that will get fixed for next release but it was also present in 0.7.6 [11:18] Robert Adams: justincc: http://opensimulator.org/mantis/view.php?id=7051 [11:18] Richardus Raymaker: we talk about terrain [11:18] Billy.Bradshaw @hg.viewtwo.net:8600: @Mata Hari, Seems a good result to Mantis 0007038, have you managed more testing? [11:18] Marcus Llewellyn: BUt yeah... I've seen the terrain loading slow on my 2048. It's pretty bad. I'm guessing that's some sort of UDP thing? [11:18] Justin Clark-Casey: also, the fix is probably not going to be simple as I think it invovles telling the viewer when the request times out [11:18] Justin Clark-Casey: from its perspective [11:19] Justin Clark-Casey: reobrt: thanks, shall I try and take a look? [11:19] Robert Adams: terrain loading is currently one patche per UDP packet.... can take a while [11:19] Mircea Kitsune: Had to re-log because some attachments wouldn't work any more. This seems to happen after teleporting to another region before all your attachments rez (non-hypergrid) [11:19] Sarah Kline: you can minimise the drawing effect by having objects drawing in front of the scene [11:19] Nebadon Izumi: one thing I have noticed about the avatar floating is there seems to be a difference between prims and terrain [11:19] Robert Adams: there is a fix, but it needs some testing [11:20] Sarah Kline: ie putting stuff on your vars lol [11:20] Mircea Kitsune: So from what I'm seeing, if you log in then teleport to another region before all your attachments are rezzed, those attachments will no longer work till you re-log [11:20] Mata Hari: I've been watching it somewhat and it *seems* like it usually gets stock when fetching the links for built-in AO [11:20] Justin Clark-Casey: there is an issue with attachment scripts restarting I think. It might have been pre-existing or got worse [11:20] Mata Hari: *stuck [11:20] Justin Clark-Casey: some kind of race condition [11:20] Richardus Raymaker: but why is a normal 0.7 sim terrain laoding much faster then 0.8 ? [11:20] Richardus Raymaker: unless something hugfh changed [11:20] Nebadon Izumi: because of varregions Richardus [11:20] Nebadon Izumi: something huge did change [11:20] Sarah Kline: on same 256 parcel? [11:20] Mata Hari: most often it seems it's my FS built-in AO that doesn't (ever0 activate [11:21] Marcus Llewellyn: 0.8 terrain load fine on a 256 size sim. a 2048 is *huge* rich. Even a 1024 took me 4 minutes to fly across. That's a lot of terrain. [11:21] Nebadon Izumi: melanie explained the badness of the built in AOs once [11:21] Sarah Kline: zackly [11:21] Richardus Raymaker: the pimple loading in 0.8 slow. not seen it with 0.7.6 [11:21] Richardus Raymaker: no terrain load slow on every type. on 0.8 [11:22] Richardus Raymaker: compared to 0.7.6 [11:22] Andrew Hellershanks: Viewer AO's are bad things? I thought having viewer side AO's was supposed to be a good thing? [11:22] Richardus Raymaker: nebadon, but build in AO's work at least in opensim [11:22] Mircea Kitsune: Never seen terrain loading slow here. I always run latest GIT master myself [11:22] Mata Hari: I must congratulate Orenh on the fix to that mesh asset issue....I can now tp without any trouble wearing mesh for the first time [11:23] Nebadon Izumi: great Mata [11:23] Justin Clark-Casey: yeah, that was a very nice investigation and fix from him [11:23] Mircea Kitsune: Nice [11:23] Marcus Llewellyn: Yeah, I've only seen slow terrain loading on large vars. [11:23] Nebadon Izumi: ya Melanie explained why the built in AO was bad but now I can not remember [11:23] Mircea Kitsune: Didn't see any problem with mesh. But like I said all ttachments break for that login session if you teleport too quickly after login [11:23] Nebadon Izumi: something about how its triggered or something [11:23] Mata Hari: something to do with the links it generates? [11:23] Sarah Kline: me too and ones with complex terraforming [11:23] Andrew Hellershanks: nebadon, ok [11:24] Nebadon Izumi: im sure if you ask her she could explain it again [11:24] Andrew Hellershanks: I was doing some work on an inworld AO recently. I'm trying to see which one(s) of the dozen that I have will work reliably. [11:24] Richardus Raymaker: well, always staying a few seconds on invisible normal pimpel terrain with 0.8 at least on my own grid. think standalone where the same [11:24] Nebadon Izumi: it is a shame that LL didnt design the AO to be viewer side from day one [11:25] Dahlia Trimble is Online [11:25] Nebadon Izumi: would have made for a much more consistent experience [11:25] Andrew Hellershanks: I know a lot of inworld AO's use fast timers but that isn't needed any more with OpenSim. [11:25] Arielle.Popstar @bound2.dyndns.org:8002: Justin, what about the broken mega's and Ode with Vars? will they be fixed before release/? [11:25] Marcus Llewellyn: Didn't LL ad server side scripting support for AOs? [11:25] Mata Hari: trouble with that, Andrew, is regions where scripts are disabled [11:25] Richardus Raymaker: nebadon, ionsteads LL moved soem AO stuff to the server side with lsl scripts [11:25] Nebadon Izumi: maybe but if you go some where that has scripts disabled that is problematic [11:26] Andrew Hellershanks: Mata, true. It may be why some viewers added AO support. [11:26] Mircea Kitsune: Sucks the official SL viewer doesn't have AO support like most custom viewers. Would have been better if AO's were a special wearable item like clothes are [11:26] Andrew Hellershanks: Mata, I have heard of scripts that can still run in no script regions. No idea how they manage to do that. [11:26] Justin Clark-Casey: there will very likely be no ode on var before release. I don't know what's broken with megas. I believe robert is looking at at least one issue [11:26] Mircea Kitsune: Since scripted AO's tend to suck. EG: They don't work in script-disabled regions... at least last time I noticed [11:27] Robert Adams: varregions are not planned to work with ODE [11:27] Arielle.Popstar @bound2.dyndns.org:8002: http://opensimulator.org/mantis/view.php?id=7084 [11:27] Mircea Kitsune: Erm, parcels [11:27] Nebadon Izumi: Justin were you planning on implementing VAR on ODE? [11:27] Justin Clark-Casey: I was thinking of looking at it, but certainly not 0.8 timeframe [11:27] Marcus Llewellyn: Just curiosity... does *any* other physics engine work on vars? POS or whatever? [11:27] Nebadon Izumi: we should change over to the Aurora ODE [11:27] Robert Adams: I need to fix TP and megas [11:27] Arielle.Popstar @bound2.dyndns.org:8002: there was talk from both Justn and robert about doing so in past [11:27] Richardus Raymaker: not sure why people still want to use ode. btw. think bullet still can be tuned a bit to fly smooth. but need to use it more to confirm it [11:27] Nebadon Izumi: that might not be as easy as it sounds though [11:27] Justin Clark-Casey: marcus: there are no other operational phys engines, pos is little more than a stub [11:27] Mircea Kitsune: I assume BulletX and of course POS will work. Since Bullet is the default [11:28] Nebadon Izumi: it does perform much better though is my understanding [11:28] Dahlia Trimble: hi [11:28] Mircea Kitsune: hi [11:28] Robert Adams: Aurora ODE changed the physics API a bunch.... would take come creative porting [11:28] Marcus Llewellyn: I haven't enabled POS in ages, so wasn't sure. [11:28] Justin Clark-Casey: hi dahlia [11:28] Nebadon Izumi: ya that should be asssumed 100% of the time Robert with anything from Aurora [11:28] Nebadon Izumi: lol [11:28] Richardus Raymaker: whats POS ? [11:28] Nebadon Izumi: Piece of shit physics [11:28] Nebadon Izumi: lol [11:28] Mircea Kitsune: Basic physics engine. Like the most basic [11:28] Justin Clark-Casey: piece of... [11:28] Nebadon Izumi: jk [11:28] Justin Clark-Casey: ha [11:28] Arielle.Popstar @bound2.dyndns.org:8002: well the Tp's in megas would be good or it will cause a lot of probs for people [11:29] Marcus Llewellyn: Basically, you dont go throgh stuff. That's it. That's all it did. [11:29] Mircea Kitsune: Haha... didn't know that's what the name standed for [11:29] Nebadon Izumi: no i t doesnt [11:29] Richardus Raymaker: uhmm. lol cant remember when i used that last time [11:29] Justin Clark-Casey: calling pos a physics engine is like calling a stone a space shuttle [11:29] Mircea Kitsune: right [11:29] Dahlia Trimble: Physics of Simplicity [11:29] Nebadon Izumi: ya thats it [11:29] Mircea Kitsune: More like something that lets avatars move without a physics engine existing [11:29] Nebadon Izumi: everything is a cube [11:29] Richardus Raymaker: justin, depends wich direction the need to go [11:29] Nebadon Izumi: and no hollows or cuts [11:29] Marcus Llewellyn: It was blessing back in, like... 0.4. Before that everything was phantom. lol [11:30] Andrew Hellershanks: Dahlia, ok, ty. I was thinking of something else for POS. ;) [11:30] Justin Clark-Casey: robert: you're going to do something on mantis 7067? [11:31] Dahlia Trimble: Andrew, most people do :P [11:31] Richardus Raymaker: i think thats really other one that need to be fixt before release [11:31] Justin Clark-Casey: it's another joke name, like robust ;) [11:31] Marcus Llewellyn: lol [11:32] Dahlia Trimble: I vaguely remember POS was danx0r's attempt to write a physics engine from scratch in c# [11:32] Dahlia Trimble: he didnt get far [11:32] Marcus Llewellyn: It really was better than nohting back in the day. [11:32] Nebadon Izumi: I thought POS came after ODE [11:32] Dahlia Trimble: not sure [11:32] Nebadon Izumi: danx0r did a lot of the early ODE work [11:32] Marcus Llewellyn: I vaguely remeber ODE being super duper experimental back then. [11:32] Nebadon Izumi: ya [11:33] Nebadon Izumi: check this out [11:33] Justin Clark-Casey: robert: and you're going to look at mantis 7050? [11:33] Mircea Kitsune: I remember a time when ODE was better than Bullet [11:33] Mircea Kitsune: Which I think was for many years actually [11:33] Marcus Llewellyn: It worked, but I remember ODE being very crashy. It was very at your own risk to enable it. [11:33] Robert Adams: justincc: 7067... yes, I'll make a short term fix... it will take a while to get Melanie's longer term fix [11:33] Dahlia Trimble: I think first time I tried opensim I had to install physx but everything was phantom. I think it wasn't until about a year later when I tried it again [11:33] Justin Clark-Casey: robert: that's great! [11:34] Marcus Llewellyn: Which grid is using PhysX now? Is that Inworldz? [11:34] Richardus Raymaker: is 7050 really so important ? if you have var [11:34] Marcus Llewellyn: Or was that Meta7? [11:34] Justin Clark-Casey: robert: anything I can help with? I was planning to look at the issues, though i can also go on and look at other bugs too [11:34] Richardus Raymaker: marcus, sofar i know the do [11:34] Arielle.Popstar @bound2.dyndns.org:8002: inworldz [11:34] Marcus Llewellyn nods amiably. [11:34] Nebadon Izumi: https://www.youtube.com/watch?v=Y0olsWlTc0A [11:34] Richardus Raymaker: yup [11:34] Marcus Llewellyn: Thanks. :) [11:34] Nebadon Izumi: here is video of first ODE test ever [11:35] Nebadon Izumi: me and Teravus [11:35] Justin Clark-Casey: yeah, I had to identify and fix some tricky threading issues with ODE, where parts of the code that were apparantly thread-safe actually weren't [11:35] Justin Clark-Casey: if you were running more than one regino [11:35] Robert Adams: justincc: I'll do megas and floating avatars and let you handle other bugs (as if there aren't enough to go around) [11:36] Richardus Raymaker: not tested the ground sit btw. is that fixt ? http://opensimulator.org/mantis/view.php?id=6813 [11:36] Marcus Llewellyn: Besides the threading stuff, ODE is kinda moribund upstream, ain't it? [11:36] Justin Clark-Casey: robert: and the crash on region recross? I was planning on looking at that since the immediate problem hits the entity state machine (though I expect that to be a symptom rather than cause). [11:36] Justin Clark-Casey: robert: or do you think that you have a handle on it? [11:37] Mircea Kitsune: So I heard [11:37] Justin Clark-Casey: marcus: there's still a bit of dev but not much. Vastly less than bullet, I believe [11:37] Robert Adams: justincc: yes, you should take the first look at the cross-crash... I would suspect it is a state machine problem [11:37] Justin Clark-Casey: robert: ok, will do [11:37] Justin Clark-Casey: robert: thanks [11:39] Richardus Raymaker: oh kokua dont have ground sit ? [11:39] Marcus Llewellyn: Anyone know is MOSES has played with vars? Maybe with DSG too? [11:39] Marcus Llewellyn: No, Kokua doesn't. :( [11:39] Andrew Hellershanks: kokua, I've noticed a few things missing from Kokua. [11:39] Marcus Llewellyn: It's one of the few features I really miss in Kokua. [11:39] Andrew Hellershanks meant to direct that at Richardus [11:39] Richardus Raymaker: bnut vivian still did go down [11:39] Mircea Kitsune: I use Kokua, I remember seeing ground sit [11:39] Justin Clark-Casey: Ill very probably make a post about working towards 0.8 on the dev mailing list tonight - I don't want misundertandings like last time [11:39] Billy.Bradshaw @hg.viewtwo.net:8600: sit = ctl-alt-s ? [11:39] Richardus Raymaker: bot on kokua [11:39] Mircea Kitsune: I used it here actually [11:39] Vivian Klees: yes Billy [11:40] Richardus Raymaker: oh wait that stupid wasd is enabled i think. the need to disable that default [11:41] Mircea Kitsune: WASD movement is good [11:41] Dahlia Trimble: I like wasd [11:41] Richardus Raymaker: no my cam moves with ctrl-alt-s [11:41] Richardus Raymaker: why, you neve ruse it. curcor keys are for that [11:41] Dahlia Trimble: I prefer wasd :P [11:41] Marcus Llewellyn uses WASD. [11:41] Mircea Kitsune: same [11:42] Sarah Kline: spot the game players ) [11:42] Marcus Llewellyn: I also mouse steer my avatar most of the time. [11:42] Richardus Raymaker: game players use cursor keys. you cant go wrong then [11:42] Nebadon Izumi: heh ive always been a arrow keys guy [11:42] Nebadon Izumi: in SL / OS anyway [11:42] Richardus Raymaker: yeah. also its more easy [11:42] Andrew Hellershanks: Yup. arrows ftw [11:43] Sarah Kline: you cant move mouse and arrow keys easily [11:43] Marcus Llewellyn: So, for me, I'm most W and S, with the mouse. [11:43] Dahlia Trimble: arrows take 2 hands and are a PITA when you have to use number and function keys to kill thingies [11:43] Sarah Kline: lol [11:43] Marcus Llewellyn grins. [11:44] Andrew Hellershanks: dahlia, you have an odd keyboard? I have separate arrow keys from my number pad. [11:44] Nebadon Izumi: ya for me in SL in the Samurai Clans of Samurai island it was easier to do arrow keys and pg up and pg down for jumping [11:44] Andrew Hellershanks: perhaps Dahlia is using a laptop. [11:44] Dahlia Trimble has standard US key layout [11:45] Marcus Llewellyn: A full KB has both solitary arrow keys and arrow keys on the number pad. [11:45] Richardus Raymaker: laptops are not for games. only nobody see it and screaming about lag, hot , melted :) [11:45] Billy.Bradshaw @hg.viewtwo.net:8600: Gamers use right hand to kill things [11:45] Nebadon Izumi: my laptop is [11:45] Nebadon Izumi: hehe [11:45] Marcus Llewellyn won't buy a laptop without a numpad. Icky. [11:45] Nebadon Izumi: ya luckily my laptop does have full keyboard [11:46] Mircea Kitsune: mine too [11:46] Richardus Raymaker: then you have a good helaty right hand [11:46] Dahlia Trimble: I have full keyboard but I prefer WASD [11:46] Marcus Llewellyn: My numpad is usually in number mode. [11:46] Nebadon Izumi: someones voice locked open? [11:47] Nebadon Izumi: ah its person down at landing area [11:47] Billy.Bradshaw @hg.viewtwo.net:8600: Voice lock is a pain, especially in conferences, could do with some remote switch [11:48] Nebadon Izumi: I am not sure if I will be at the next meeting or not [11:48] Nebadon Izumi: I will try my best to be [11:48] Nebadon Izumi: i will be in Sweden [11:48] Nebadon Izumi: i leave on Saturday [11:48] Mircea Kitsune: ok. Good luck! [11:48] Marcus Llewellyn: Bring back meatballs. :) [11:48] Nebadon Izumi: lol [11:48] Justin Clark-Casey: sounds like an adventure! [11:48] Dahlia Trimble: ewwww 12 hour flight :( [11:48] Andrew Hellershanks: Send your avatar instead. ;) [11:48] Marcus Llewellyn: lol [11:48] Nebadon Izumi: ya should be fun though i'll be working most of the time [11:48] Nebadon Izumi: teaching students opensim [11:49] Marcus Llewellyn: I'm so sorry. [11:49] Nebadon Izumi: haha [11:49] Billy.Bradshaw @hg.viewtwo.net:8600: Watch out for small log cabins, with steam leaking around the door [11:49] Andrew Hellershanks: You get to fly to a foreign country to teach OS? Interesting. Hope you can get a chance to look around the place a bit. [11:49] Nebadon Izumi: ya I am sure I will I may end up in germany a couple days too [11:50] Mircea Kitsune: Wait... you're really going on that trip for Opensim? That's awesome!! [11:50] Nebadon Izumi: ya [11:50] Mircea Kitsune: Very nice :) [11:50] Nebadon Izumi: I am modeling 3 cities in Sweden [11:50] Mircea Kitsune: Always nice to hear this is a bigger and more implicated project than I thought [11:50] Nebadon Izumi: teaching students with Masters degrees in Urban Planning [11:50] Richardus Raymaker: watch out for photographers nebadon :O [11:50] Marcus Llewellyn: Stupid question, cuz I haven't been able to experiment much. Can var regions (and normal ones) of various sizes directly neighbor each other safely? With crossing and such? [11:51] Mircea Kitsune: I'm making a city here on OSGrid, but it will take a while [11:51] Nebadon Izumi: ive heard people say no you cant and otehrs who have said they did it and it worked [11:51] Nebadon Izumi: so I am not sure honestly I have never tested it [11:51] Marcus Llewellyn: I've kept my vars neighborless just in case. [11:51] Nebadon Izumi: nice Mircea [11:51] Mircea Kitsune: Yeah. It will be very detailed and interactive even [11:51] Dahlia Trimble: I hear they can be different sizes but I haven't tried it [11:51] Nebadon Izumi: nebadon2025.com/screenshots/sanjose2_002.png [11:51] Nebadon Izumi: nebadon2025.com/screenshots/sanjose2_003.png [11:52] Andrew Hellershanks: Getting near top of the hour. I have some questions (for justin) before he disappears. [11:52] Dahlia Trimble: ask Robert [11:52] Billy.Bradshaw @hg.viewtwo.net:8600: crossing between var regions of the same size works for us [11:52] Nebadon Izumi: those are shots of one of my models that one is in San Jose CA [11:52] Justin Clark-Casey: still here, reading philips blog post about identity in the metaverse [11:52] Mircea Kitsune: Problem is I'm going for such detail (like full apartments in each block) that I've already exceeded what would normally be the prim limit of a region on the LL grid. Thankfully tho Opensim lets me do that, even if it loads slowly [11:52] Nebadon Izumi: doing same thing with students from SJSU [11:52] Justin Clark-Casey: quite interesting, though super vague [11:52] Mircea Kitsune: And I only started the first building lol [11:53] Justin Clark-Casey: andrew? [11:53] Andrew Hellershanks: Justin, one of the things I've done this past week is to set my machine up so I can build the LSL parser/compiler tool in opensim-libs, and I also tracked down the latest copy of CSTools. [11:53] Marcus Llewellyn: Getting hi-res satellite terrain imagery is difficult. :( [11:53] Nebadon Izumi: yes it is [11:53] Nebadon Izumi: its expensive [11:54] Mata Hari: hack the CIA computers [11:54] Andrew Hellershanks: Justin, is the parser/code generation in OS and opensim-libs supposed to track each other? I noticed a patch to CSCodeGenerator.cs in OS that was not applied to opensim-libs. [11:54] Marcus Llewellyn: The free stuff is gonna be like 30m per pixel. Which is just a blur at full scale. [11:54] Mircea Kitsune: I remember there was once a plan to import real life cities into Opensim... via satellite or something. I think I looked into it sime 6 years ago? No news since [11:54] Nebadon Izumi: that is what we are doing Mircea [11:54] Justin Clark-Casey: andrew; three shoujldn't be a CSCodeGenerator in opensim-libs since that's an open side component [11:54] Mircea Kitsune: Oh, nice [11:54] Nebadon Izumi: its GIS/LIDAR data we are using [11:54] Nebadon Izumi: to scale [11:54] Andrew Hellershanks: Mircea, See if you could use the information that is part of the google earth project? [11:54] Nebadon Izumi: the San Jose model is 2 miles by 1.5 miles [11:55] Mircea Kitsune: That would be nice [11:55] Nebadon Izumi: its 12 768x768 var regions [11:55] Nebadon Izumi: in 3x4 pattern [11:55] Mircea Kitsune: My own skills don't go half that far tho [11:55] Billy.Bradshaw @hg.viewtwo.net:8600: UK gov have done this http://www.ordnancesurvey.co.uk/innovate/developers/minecraft-map-britain.html [11:55] Nebadon Izumi: its 3300 buildings hehe [11:55] Marcus Llewellyn: Google earth's imagry is .5m per pixel, which is still a bit low. Even so, it's also *strictly* licensed. Google pays for all of it. [11:55] Andrew Hellershanks: Justin, I wanted to use the lsl2cs tool in opensim-libs to test changes to the code parsing/generation. [11:55] Justin Clark-Casey: andrew: ok I see the file you mean and the fact that I made a change to it. However, I can't remember why it's there... I don't think that it's a good reason though for some reason I kept it [11:56] Nebadon Izumi: actually .5 is pretty good [11:56] Justin Clark-Casey: andrew: if it's important you could just apply the same changes [11:56] Marcus Llewellyn: MY avatar keeps typing. I'm not typing, darnit! lol [11:56] Justin Clark-Casey: andrew: btw, there's a horrible gotcha with regenerating the parse that I need to tell you about [11:56] Andrew Hellershanks: Justin, I'm not sure if it is important. [11:56] Nebadon Izumi: I generally use 1m GIS and 1 foot aerial photography [11:56] Andrew Hellershanks: Uh oh... [11:56] Mata Hari: it's talking behind your back, Marcus [11:56] Justin Clark-Casey: essentially, the penultimate number on the long list needs to be changed from 1 to 0 manually [11:56] Justin Clark-Casey: or 0 to 1 - I can't remember which way around [11:57] Justin Clark-Casey: and I can't even remember why... suffice to say that if you don't do it, the parse fails [11:57] Andrew Hellershanks: justin, the long list?? [11:57] Marcus Llewellyn: .5m ain't lousy, especially compared to whatyou'll get from something like USGS. But when you're standing on it in a var, it's still blurry. :/ [11:57] Nebadon Izumi: ya for free 0.5 is great [11:57] Nebadon Izumi: you will not find better [11:58] Marcus Llewellyn: On the other hand, free terrain elevation data works quite well for mesh terrain. [11:58] Justin Clark-Casey: andrew: in the generated lsl.lexer.cs, there is a huge list of numbers in yyLSLTokens.arr, which I believe represent tokens [11:58] Justin Clark-Casey: when the file is generated, I believe it is the third to last number which is 1 [11:59] Justin Clark-Casey: that has to be changed manually to 0 - the worst thing is that I can't remember why [11:59] Andrew Hellershanks: Justin, I have an idea now why line numbers reported in error messages are so out of whack. The generated code doesn't use #line directives and the generated code doesn't include the blank lines of the original input file so line numbers get off quickly. [11:59] Justin Clark-Casey: andrew: you can't miss the large array of numbers in the file - it is the only one [12:00] Andrew Hellershanks: Hm... that sounds like a change for the state machine that may indicate a bug(?) in lsl.lexer [12:00] Justin Clark-Casey: #line directives? [12:00] Nebadon Izumi: Andrew didnt you say the code in opensim-libs worked properly? [12:00] Nebadon Izumi: but what was in core doesnt? [12:01] Andrew Hellershanks: Justin, to help the compiler know the real line numbers in the source when it deals with the generated code. [12:01] Justin Clark-Casey: I'm not sure how it's generated atm, but I believe there is a line map which does that [12:02] Simulator Version v0.5 shouts: OpenSim 0.8.0 Dev 6557eba: 2014-03-28 13:04:19 +0300 (Unix/Mono) [12:02] Andrew Hellershanks: nebadon, Testing the output.exe to parse LSL code and checking C#, doing something like "int a=1; b=a; list c=[a,b];" in a global context is supported. OpenSim isn't accepting it. [12:03] Justin Clark-Casey: that's very curious! [12:03] Marcus Llewellyn just realized Dahlia is unarmed! Who kidnapped the real Dahlia! [12:03] Andrew Hellershanks: Justin, Melanie mentioned a line map but I'm not sure where it is or how to get the OS scripting ending to use it. [12:03] Dahlia Trimble: shouldnt it be a comma after a=1 ? [12:03] Justin Clark-Casey: are you sure that it's not generating invalid c# code? [12:04] Peter.Nagy @lpgrid.com:8002: Szió baa [12:04] baajos.Bagley @lpgrid.com:8002: szió;) [12:04] Andrew Hellershanks: Dahlia, I meant to type "int b=a" and I'm shortcuttint as it should be "integer". [12:04] Peter.Nagy @lpgrid.com:8002: Hívd riát [12:04] baajos.Bagley @lpgrid.com:8002: olyan vagyok mint Enola;)) [12:04] Peter.Nagy @lpgrid.com:8002: Vagy hogy hívják itt. [12:04] baajos.Bagley @lpgrid.com:8002: de hogy hívjam? [12:04] Peter.Nagy @lpgrid.com:8002: Nálam a neved van csak [12:04] Andrew Hellershanks: Justin, I looked at the generated C# code and it looked ok to me. IIRC, I pasted it in to a new project in monodevelop and was able to compile it. [12:04] Dahlia Trimble: I didnt think LSL allowed you to evaluate a variable outside of a block [12:04] Peter.Nagy @lpgrid.com:8002: TP gomb [12:05] Mircea Kitsune: Think I'm gonna log off now. Later [12:05] Dahlia Trimble: later [12:05] Marcus Llewellyn: Cya Mircea [12:05] Andrew Hellershanks: Dahlia, it doesn't currently. That's the bug. It breaks a lot of scripts coming from SL. C# seems to allow it. [12:05] Richardus Raymaker: what andrew wrote is just more command on one line, seperated by ; [12:05] Justin Clark-Casey: bye mircea [12:05] Dahlia Trimble: Andrew, oh, I didnt think it worked in SL either [12:06] Andrew Hellershanks: There are lots of scripts in SL that do that. Particularly for list initialization. [12:06] Dahlia Trimble: inside a block? [12:06] Justin Clark-Casey: ok, that all sounds very curious. btw, there's also bin/OpenSim.Tools.lslc.exe in OpenSimulator itself which translates lsl to c# [12:06] Justin Clark-Casey: would be interesting to know if that generates the same thing or something different [12:06] Andrew Hellershanks: You often see "integer varname=4; list somelistvar = [ varname ];" [12:06] Marcus Llewellyn: SL scripting is a lot looser about variable scope, if that's what you guys mean. [12:06] Andrew Hellershanks: Just look at almost any AO script. [12:07] Richardus Raymaker: abdrew. thats just 2 lines on one row. i dotn like it because readbilioty. but not see a problem in that. [12:07] Dahlia Trimble: I get kinda ill when I look at most AO scripts :/ [12:08] Andrew Hellershanks: Justin, OpenSim.Tools.lslc.exe? Hm... I haven't noticed that. [12:08] Nebadon Izumi: haha dahlia [12:08] Marcus Llewellyn: When they do while(true) is when I run screaming. [12:09] Dahlia Trimble: the 0.01 second timer usually makes me run to the porcelain altar [12:09] Marcus Llewellyn: Hehe [12:09] Andrew Hellershanks: Richardus, this is open chat. I don't know how to force line breaks in chat output so I'm putting the stuff on one line to show code snippets. [12:09] Andrew Hellershanks: Richardus, it isn't how the actual LSL scripts are written. [12:10] Richardus Raymaker: aha [12:10] Justin Clark-Casey: ok, I need to pop off now [12:10] Marcus Llewellyn: Take care, jcc :) [12:10] Justin Clark-Casey: thanks for the meeting, folks [12:10] Dahlia Trimble: laterz jcc [12:10] Nebadon Izumi: ok Justin, talk soon I am sure :) [12:10] Mata Hari: bye Justin [12:10] Caro Fayray: bye justin [12:10] Richardus Raymaker: bye justin [12:10] Andrew Hellershanks: Justin, btw, I captured a lot of info about the parsing and code generation. I'll eventually type it up and dump it in to the wiki. [12:11] Justin Clark-Casey: andrew: that would be fantastic - I don' t think there's much, if anything, there atm [12:11] Andrew Hellershanks: I don't know of anything. [12:11] Andrew Hellershanks: Bye, Justin. [12:11] Dahlia Trimble: bye all :) [12:11] Richardus Raymaker: bye dahlia [12:12] Mata Hari: I'm off too....bye everyone [12:12] Eliopod Beaumont: Bye all [12:12] Caro Fayray: bye dahlia [12:12] Justin Clark-Casey: there are some pages under http://opensimulator.org/wiki/Developer_Documentation#Scripting but not much that tells you how things are working internally [12:12] Nebadon Izumi: by everyone who is leaving :) [12:12] Justin Clark-Casey: hope you have a good trip, neb [12:12] Andrew Hellershanks wishes he had someone who would pay money for working on Open Source projects. :) [12:12] Nebadon Izumi: thanks Justin [12:12] Andrew Hellershanks: Yes, Have a great trip, nebadon. [12:12] Sarah Kline: bye justin [12:12] Marcus Llewellyn: I thought scripts where excecuted by leprechauns enslaved by MelT. :) [12:12] Justin Clark-Casey: it's noe easy, but hopefully the opportunities can increase in the future [12:13] Justin Clark-Casey: that's the way to grow the ecosystem, I think