Chat log from the meeting on 2012-07-03
From OpenSimulator
(Difference between revisions)
(Created page with "<pre> [11:43] Dahlia Trimble: hi [11:43] Justin Clark-Casey: well, knowing if there are high physics times would help [11:43] Nebadon Izumi: ya its actually ok if we have less th...") |
|||
Line 1: | Line 1: | ||
<pre> | <pre> | ||
− | [11:43] Dahlia Trimble: hi | + | [2012/07/03 11:07] Justin Clark-Casey: hi folks |
− | [11:43] Justin Clark-Casey: well, knowing if there are high physics times would help | + | [2012/07/03 11:08] Mike Kayaker: The holiday is tomorow! |
− | [11:43] Nebadon Izumi: ya its actually ok if we have less than 15 avatars | + | [2012/07/03 11:08] BlueWall Slade: Hi Justin |
− | [11:43] Justin Clark-Casey: if not, then you could proabbly rule out physics | + | [2012/07/03 11:08] BlueWall Slade: yeah |
− | [11:43] Justin Clark-Casey: hi dahlia | + | [2012/07/03 11:08] BlueWall Slade: a lot of people take a week |
− | [11:44] Nebadon Izumi: once we cross a certain threshold of avatars | + | [2012/07/03 11:08] Justin Clark-Casey: holiday? |
− | [11:44] Nebadon Izumi: it just explodes | + | [2012/07/03 11:08] Justin Clark-Casey: ah, 4th july? |
− | [11:44] BlueWall Slade: Hello Dahlia | + | [2012/07/03 11:08] BlueWall Slade: Independ-- yeah |
− | [11:44] Nebadon Izumi: everyone complains prims are not loading | + | [2012/07/03 11:09] BlueWall Slade: I looked into that inventory issue last week, but hit dead ends - and my debugger was out of sorts |
− | [11:44] Nebadon Izumi: cant see any avatars | + | [2012/07/03 11:10] Nebadon Izumi: hello |
− | [11:44] BlueWall Slade: those stats should give some clues too. | + | [2012/07/03 11:10] BlueWall Slade: got that straight now, maybe I can look a little further into it |
− | [11:44] Justin Clark-Casey: It wouldn't enormously surprise me if some other process is soaking up all the sim time | + | [2012/07/03 11:10] Justin Clark-Casey: hi neb |
− | [11:44] BlueWall Slade: and you can use the in-world debugginh things | + | [2012/07/03 11:10] BlueWall Slade: Hi Nebadon |
− | [11:44] Justin Clark-Casey: like lots of invalid group requests or something | + | [2012/07/03 11:10] Nebadon Izumi: heh totally wasnt paying attention to the time |
− | [11:44] Nebadon Izumi: but why doesnt Lbsa do it | + | [2012/07/03 11:10] Richardus Raymaker: hi |
− | [11:44] Nebadon Izumi: we can have 40 avatars on lbsa | + | [2012/07/03 11:11] Nebadon Izumi: Guess maybe a lot of people may not show up, maybe holiday stuff |
− | [11:44] Nebadon Izumi: same box | + | [2012/07/03 11:11] Nebadon Izumi: hard to say, surprised more people arent here |
− | [11:44] Justin Clark-Casey: I do not know | + | [2012/07/03 11:12] Justin Clark-Casey: yes, seems to fluctuate quite randomly wek to eek |
− | [11:44] Nebadon Izumi: never ever once has done it on Lbsa | + | [2012/07/03 11:12] BlueWall Slade: so what's up? |
− | [11:45] Nebadon Izumi: but everytime we use Bade it happens | + | [2012/07/03 11:12] Justin Clark-Casey: not a lot :) |
− | [11:45] Bea Witched: flying isn't allowed on lbsa so height is restricted | + | [2012/07/03 11:13] Justin Clark-Casey: took a holiday myself yesterday - not sure it helped much |
− | [11:45] Nebadon Izumi: 4-5 meetings in a row it happened | + | [2012/07/03 11:13] BlueWall Slade: haa |
− | [11:45] Nebadon Izumi: finally we just moved it here for first time this month | + | [2012/07/03 11:13] BlueWall Slade: it's always good to do |
− | [11:45] Nebadon Izumi: we had 25 avatars | + | [2012/07/03 11:13] Nebadon Izumi: not much going on here |
− | [11:45] Nebadon Izumi: stats were perfect | + | [2012/07/03 11:13] Justin Clark-Casey: also meens I have 3 days of stuff to catch up rather than 2 :) |
− | [11:45] Nebadon Izumi: literally no lag | + | [2012/07/03 11:13] BlueWall Slade: :S |
− | [11:45] Nebadon Izumi: you would think if any region was going to have issues | + | [2012/07/03 11:14] Justin Clark-Casey: nebadon: did you try the latest opensim? Did it make any change to memory consumption? |
− | [11:45] Nebadon Izumi: it would be this one | + | [2012/07/03 11:14] Nebadon Izumi: I did, and i didnt see much if any change |
− | [11:46] Nebadon Izumi: lol | + | [2012/07/03 11:14] Justin Clark-Casey: ok |
− | [11:46] Nebadon Izumi: bade has no neighbors | + | [2012/07/03 11:14] Nebadon Izumi: memory is still pretty high at times |
− | [11:46] Nebadon Izumi: its completely isolated | + | [2012/07/03 11:14] Nebadon Izumi: maybe a little better, but nothing drastic |
− | [11:46] Nebadon Izumi: low prim, low scripts | + | [2012/07/03 11:14] Justin Clark-Casey: yeah, I thikn the real reasons are elsewhere - this bug has only been around for 6 weeks or so |
− | [11:46] Nebadon Izumi: so bizzare | + | [2012/07/03 11:15] Nebadon Izumi: I actually had to back pedal this region |
− | [11:46] Richardus Raymaker: thats same as with my physical train test (i stopped with that) 1 train no timeings. when i rezz aseconds train phyics time jump up to 180,ms. like am used toin the past with 1 train | + | [2012/07/03 11:15] Nebadon Izumi: was really wierd |
− | [11:46] Justin Clark-Casey: well, not seeing objects is often a synptom of bad network comms | + | [2012/07/03 11:15] Nebadon Izumi: I think the Leap second actually whacked this box out |
− | [11:46] Justin Clark-Casey: object update packets being completely dropped and the resend system failing | + | [2012/07/03 11:16] Nebadon Izumi: it was on Sunday |
− | [11:46] Nebadon Izumi: ya this felt more like a hung thread or something | + | [2012/07/03 11:16] Nebadon Izumi: right before the Monthly town hall meeting here |
− | [11:47] Nebadon Izumi: and the only thing i can really pinpoint and say its completely different than the other plazas | + | [2012/07/03 11:16] Nebadon Izumi: all the regions on this box were using super high CPU |
− | [11:47] Nebadon Izumi: is that we meet above 100m above ground | + | [2012/07/03 11:16] Nebadon Izumi: was sooo bizzare |
− | [11:47] Richardus Raymaker: i wish group IM did not be handles by region. | + | [2012/07/03 11:16] Nebadon Izumi: even the empty void place holder regions were using 200% cpu |
− | [11:47] Nebadon Izumi: otherwise nothing is different | + | [2012/07/03 11:16] Justin Clark-Casey: so did you actually go back to an earlier version of jurst restart the simulator? |
− | [11:47] Nebadon Izumi: anyway i wasnt looking for a solution | + | [2012/07/03 11:16] Nebadon Izumi: ya I did that 1st |
− | [11:47] Nebadon Izumi: i just wanted to present the idea | + | [2012/07/03 11:16] Nebadon Izumi: nothing changed |
− | [11:47] Richardus Raymaker: hi dahlia | + | [2012/07/03 11:17] Nebadon Izumi: i restarted everything and still super high CPU |
− | [11:47] Nebadon Izumi: i'll try to gather more info | + | [2012/07/03 11:17] Nebadon Izumi: so i rebooted the box |
− | [11:47] Justin Clark-Casey: yeah | + | [2012/07/03 11:17] Nebadon Izumi: and then everything was fine |
− | [11:48] Dahlia Trimble: object updates dropping? is this a prim/texture heave region? | + | [2012/07/03 11:17] Justin Clark-Casey: I know there were some underlying linux issues with the leap second and java |
− | [11:48] Dahlia Trimble: *heavy | + | [2012/07/03 11:17] Nebadon Izumi: but i was reading that a huge portion of the internet servers went crazy after the leap second thing |
− | [11:48] Nebadon Izumi: no Dahlia, its quite low prims low textures | + | [2012/07/03 11:17] Nebadon Izumi: now they are saying it was the Linux Kernel was the issue |
− | [11:48] Nebadon Izumi: its maybe 7000 prims | + | [2012/07/03 11:17] Justin Clark-Casey: maybe the same was true of other things as well |
− | [11:48] Nebadon Izumi: few 100 textures maybe | + | [2012/07/03 11:17] Justin Clark-Casey: yes |
− | [11:48] Dahlia Trimble: so much for that idea :/ | + | [2012/07/03 11:17] Nebadon Izumi: saw a bunch of articles today saying it was the linux kernel |
− | [11:48] Nebadon Izumi: ya ive tried everyting i could think of | + | [2012/07/03 11:17] Nebadon Izumi: a lot of servers went crazy |
− | [11:49] Nebadon Izumi: i literally inspected every script on the sim | + | [2012/07/03 11:18] Nebadon Izumi: i have no other explination from what i was reading |
− | [11:49] Nebadon Izumi: i went through the database | + | [2012/07/03 11:18] Nebadon Izumi: for most servers you had to reboot |
− | [11:49] Nebadon Izumi: and located every script | + | [2012/07/03 11:18] Nebadon Izumi: and then it was fine |
− | [11:49] Nebadon Izumi: to make sure it wasnt some kind of visitor scanner or something | + | [2012/07/03 11:18] Nebadon Izumi: which is exactly what happend on this box |
− | [11:49] Nebadon Izumi: some kind of radar | + | [2012/07/03 11:18] Nebadon Izumi: was really bizzare |
− | [11:49] Nebadon Izumi: i couldnt find any heavy scripts | + | [2012/07/03 11:18] Nebadon Izumi: it just seems odd to me this was the day after the leapsecond thing |
− | [11:49] Nebadon Izumi: its all like sit scripts and hover text | + | [2012/07/03 11:18] Nebadon Izumi: and i really had not looked at the box for probably 24+ hours prior to that |
− | [11:49] Richardus Raymaker: also checked for lost AO's and other objects people lose ? | + | [2012/07/03 11:19] Justin Clark-Casey: it quite possibly was leap second related |
− | [11:49] Nebadon Izumi: ya, i couldnt find anything unusual | + | [2012/07/03 11:20] Nebadon Izumi: ya its been fine since the reboot |
− | [11:50] Nebadon Izumi: thats not on all the other plazas | + | [2012/07/03 11:20] Nebadon Izumi: and never seen it do anything like that before |
− | [11:50] Justin Clark-Casey: anyway, any other opensim topics anybody wants to bring up today? | + | [2012/07/03 11:20] Nebadon Izumi: i have no other explantaion |
− | [11:50] Mike Kayaker: I have a question about timers... | + | [2012/07/03 11:20] Richardus Raymaker: sorry, tried small font. |
− | [11:50] Nebadon Izumi: hrmm I dont think I have anything else myself | + | [2012/07/03 11:21] Richardus Raymaker: 40" (temporarily) is a bit to big as screen |
− | [11:51] Nebadon Izumi: shoot Mike | + | [2012/07/03 11:21] Richardus Raymaker: lol |
− | [11:51] Mike Kayaker: I seem to be running into a script resource I don't understand | + | [2012/07/03 11:21] Richardus Raymaker: sorry 37" |
− | [11:51] Richardus Raymaker: well justin like i sayed in mailinmglist. new version seems to work betetr . sofar i have doen tests on the sim no crashing anymore | + | [2012/07/03 11:22] Richardus Raymaker: where on phgone to. sorry to ask neb. did you found what where wrong ? |
− | [11:51] Dahlia Trimble: I need advice on adding properties to SOG | + | [2012/07/03 11:22] Nebadon Izumi: no clue, when i rebooted the server it was fine |
− | [11:51] Mike Kayaker: If I have a bunch of scripts running that all have a short (1 second timer) | + | [2012/07/03 11:22] Richardus Raymaker: still 06-30 ? |
− | [11:51] Dahlia Trimble: like for pathfincing, whether an object is a walkable | + | [2012/07/03 11:23] Nebadon Izumi: ya i never reverted it again |
− | [11:51] Mike Kayaker: the FPS starts to drop | + | [2012/07/03 11:23] Nebadon Izumi: i forgot to do that |
− | [11:52] Dahlia Trimble: how it affects database and oars | + | [2012/07/03 11:23] Nebadon Izumi: i'll do that in a bit |
− | [11:52] Mike Kayaker: I want to have two kinds of timers I guess, | + | [2012/07/03 11:25] Nebadon Izumi: we did have 25 people here though |
− | [11:52] Mike Kayaker: those that I care if the timer is on time, | + | [2012/07/03 11:25] Nebadon Izumi: at the Town Hall |
− | [11:52] Richardus Raymaker: strange mike. i have used 0.5s timers (no llSLeep) and never seen frame drops | + | [2012/07/03 11:25] Nebadon Izumi: it held up really good |
− | [11:52] Mike Kayaker: and others that I don't care | + | [2012/07/03 11:25] Richardus Raymaker: after downgrade it worked perfect yes |
− | [11:52] Justin Clark-Casey: I would find it very surprising if fps drops due to script timers | + | [2012/07/03 11:25] Nebadon Izumi: actually at 1st it didnt |
− | [11:53] Justin Clark-Casey: unless you're using hundereds and hundreads of them and never removing them | + | [2012/07/03 11:25] Nebadon Izumi: it wasnt until i rebooted the server it ran normal |
− | [11:53] Mike Kayaker: Is there some way I can allow the don't care timers to volunteer to be time sliced? | + | [2012/07/03 11:26] Justin Clark-Casey: yeah, I would be pretty sure this is the leap second issue rather than opensim per se |
− | [11:53] Mike Kayaker: like calling llSleep(.0001) or something? | + | [2012/07/03 11:26] Nebadon Izumi: Dan Banner had a very good theory we need to test Justin |
− | [11:53] Nebadon Izumi: i dont think so | + | [2012/07/03 11:26] Nebadon Izumi: normally we have our Town Hall at Bade Plaza |
− | [11:53] Richardus Raymaker: llSLeep.00001) never works | + | [2012/07/03 11:26] Nebadon Izumi: but lately for last 6+ months |
− | [11:53] Nebadon Izumi: there is a limit how low sleep can go | + | [2012/07/03 11:26] Nebadon Izumi: it runs horrible there |
− | [11:53] Dahlia Trimble: use not_at_target() for a per-frame event | + | [2012/07/03 11:26] Nebadon Izumi: we get like 15 people and it starts lagging |
− | [11:53] Nebadon Izumi: its 0.01? | + | [2012/07/03 11:26] Nebadon Izumi: and running horrible |
− | [11:53] Nebadon Izumi: or 0.05? | + | [2012/07/03 11:26] Nebadon Izumi: ive spent hours trying to figure it out |
− | [11:54] Nebadon Izumi: i cant remember | + | [2012/07/03 11:26] Nebadon Izumi: erased all kinds of stuff |
− | [11:54] Mike Kayaker: Well, of cource I do a vector calculation and a resize or something on every tick | + | [2012/07/03 11:27] Nebadon Izumi: basically the sim is almost empty |
− | [11:54] Nebadon Izumi: maybe its even less | + | [2012/07/03 11:27] Nebadon Izumi: BUT |
− | [11:54] Richardus Raymaker: fastest i have seen is around 0.01 maby but i resits to use llSLeep at all. only in rare cases | + | [2012/07/03 11:27] Nebadon Izumi: Dan had the idea, that maybe it has something to do with the fact we have our meeting up at like 100+ meters in the sky |
− | [11:54] Justin Clark-Casey: smallest sleep would be 1ms (0.001) | + | [2012/07/03 11:27] Nebadon Izumi: thats the only variable I have not tested yet |
− | [11:54] Nebadon Izumi: hrmm i ok didnt think it went that far | + | [2012/07/03 11:27] Justin Clark-Casey: do you observe frame time numbers during these tests? |
− | [11:54] Nebadon Izumi: i thought it was 0.01 | + | [2012/07/03 11:27] Richardus Raymaker: only thing todo is test it 1 time at lower level before things change |
− | [11:54] Mike Kayaker: but not-at-target would slow down as the FPS goes down... | + | [2012/07/03 11:28] Nebadon Izumi: here at Wright Plaza? |
− | [11:54] Richardus Raymaker: i agree with nebadon. i have not seen any difference betwen 0.01 and lower | + | [2012/07/03 11:28] Nebadon Izumi: or at Bade? |
− | [11:55] Dahlia Trimble: why would you want to update anything faster than frame rate? | + | [2012/07/03 11:28] Richardus Raymaker: no bade |
− | [11:55] Richardus Raymaker: at_target stuff. thats still so unreliable for me | + | [2012/07/03 11:28] Justin Clark-Casey: bade |
− | [11:55] Mike Kayaker: Actully, I don't care how long the slep sleeps, just want timers on the other side of the sim to stay accurate when have a lot of ddon't care timers running | + | [2012/07/03 11:28] Nebadon Izumi: yes its the same server that runs Lbsa Plaza |
− | [11:55] Justin Clark-Casey: yes. In fact, anything less than 0.001 will get trunacated to 0, no sleep at all | + | [2012/07/03 11:28] Nebadon Izumi: and we never have those issues there |
− | [11:56] Justin Clark-Casey: mike: I wouldn't be surprised if timers are not very accurate atm | + | [2012/07/03 11:28] Justin Clark-Casey: and? what are the frame time numbers under load? |
− | [11:56] Mike Kayaker: thing is, I build one item, timers work good enough, | + | [2012/07/03 11:28] Nebadon Izumi: but we move 15-20 people to bade up at 100m and it runs like pure crap |
− | [11:56] Justin Clark-Casey: I think the way they operate is somewhat inaccurate | + | [2012/07/03 11:28] Nebadon Izumi: I don't know honestly |
− | [11:56] Richardus Raymaker: Justin, timers, sensors, at_target are never 100% save for me | + | [2012/07/03 11:28] Nebadon Izumi: all I know is that after so many people log in |
− | [11:56] Dahlia Trimble: its not a real-time simulation running on a real-time os | + | [2012/07/03 11:28] Nebadon Izumi: prims stop loading |
− | [11:56] BlueWall Slade: working on some stuff a couple of months back - looks like the events priority is burried below other things | + | [2012/07/03 11:29] Nebadon Izumi: you cant see avatars |
− | [11:56] Mike Kayaker: then I rez a bunch of the item, and things get very strange | + | [2012/07/03 11:29] Justin Clark-Casey: well, it's easy to see - just open the stats pane |
− | [11:57] BlueWall Slade: at_target was delayed for a long time | + | [2012/07/03 11:29] Justin Clark-Casey: if the physics time is high then that would indicate an ode issue |
− | [11:57] Dahlia Trimble: not at_target(), use not_at_target() *note underscores* | + | [2012/07/03 11:29] Nebadon Izumi: we took the region down for now |
− | [11:57] BlueWall Slade: when it finally showed up (way past the target point) it spammed the output pretty bad with many lines | + | [2012/07/03 11:29] Nebadon Izumi: oh you mean when no avatars are there? |
− | [11:58] Justin Clark-Casey: the problem is timers are run on the same loop as a lot of other lsl stuff | + | [2012/07/03 11:29] Justin Clark-Casey: if not and there is still lots of spare time per frame then it's something else |
− | [11:58] Justin Clark-Casey: if there's a lot going on then I suspect they could be delayed | + | [2012/07/03 11:29] Justin Clark-Casey: well no, whenever you get these issues |
− | [11:58] Mike Kayaker: What resoure am I using up that causes the timers to slow down? | + | [2012/07/03 11:29] Nebadon Izumi: ya i would need to setup a test |
− | [11:58] Justin Clark-Casey: and minimum timer resolution on this loop is 100ms | + | [2012/07/03 11:29] Nebadon Izumi: we took the sim down |
− | [11:58] Richardus Raymaker: lsl is to bad not really multitasking | + | [2012/07/03 11:29] Nebadon Izumi: because its completely useless |
− | [11:59] Dahlia Trimble: any event has to be queued anyway | + | [2012/07/03 11:29] Richardus Raymaker: are the timeing numbers now more reliable ? |
− | [11:59] Justin Clark-Casey: url listeners, sensors, dataserver requets | + | [2012/07/03 11:29] Justin Clark-Casey: yes |
− | [11:59] Justin Clark-Casey: and yes, then there's a delay/inaccuracy with timer event queueing | + | [2012/07/03 11:30] Nebadon Izumi: what i really need to do is setup a new sim |
− | [11:59] Simulator Version v0.5 shouts: OpenSim 0.7.4 Dev 1f34c82: 2012-06-13 04:05:02 +0100 (Unix/Mono) | + | [2012/07/03 11:30] Nebadon Izumi: and test |
− | [11:59] Richardus Raymaker: i use sensor(s0 to, but the maby get called once every second or longer | + | [2012/07/03 11:30] Justin Clark-Casey: on here, for instance, physics time is only about 0.5 - 1.5 ms |
− | [12:00] Justin Clark-Casey: I suspect this area could be improved | + | [2012/07/03 11:30] Nebadon Izumi: and also test in bade too |
− | [12:01] Mike Kayaker whispers: So is there a way to do a lag-friendly timer? Like get a timer evey second, but not slow down other scripts if the sim is buisy? | + | [2012/07/03 11:30] Justin Clark-Casey: with spare time of 17ms or so |
− | [12:01] Dahlia Trimble: probably not if the sim is busy | + | [2012/07/03 11:30] Justin Clark-Casey: every 18 ms frame |
− | [12:01] Mike Kayaker: I mean, a timer that does allow itself to get slowed down, so other timers will not get delayes. | + | [2012/07/03 11:30] Justin Clark-Casey: so we aren't stressing physics unduly |
− | [12:02] Justin Clark-Casey: I'm really not sure where the slow down is happening | + | [2012/07/03 11:30] Nebadon Izumi: its just so wierd |
− | [12:02] Richardus Raymaker: Mike TimerEvent is possible most save, but not reliable if you ned to have strict timeings | + | [2012/07/03 11:30] Nebadon Izumi: Bade plaza is like empty |
− | [12:02] Justin Clark-Casey: this would require rearchitecting the way that timers are handled | + | [2012/07/03 11:30] Nebadon Izumi: there are almost no scripts |
− | [12:02] Justin Clark-Casey: possibly so they sit on a separate thread rather than with every other event | + | [2012/07/03 11:31] Nebadon Izumi: not many prims |
− | [12:02] Justin Clark-Casey: or even setup actual system resource timers | + | [2012/07/03 11:31] Nebadon Izumi: and it runs horrible |
− | [12:02] Mike Kayaker: Every timer event in my sim starts getting delays. | + | [2012/07/03 11:31] Nebadon Izumi: so strange |
− | [12:02] BlueWall Slade: when we can build on C# 5.0 we should do that anyway | + | [2012/07/03 11:31] Nebadon Izumi: i'll try to come up with some numbers |
− | [12:02] Richardus Raymaker: Mike , do you have physics on the sim ? | + | [2012/07/03 11:31] Justin Clark-Casey: what's cpu load like? |
− | [12:03] Dahlia Trimble: given that all events are queued I'd say its near impossible | + | [2012/07/03 11:31] Richardus Raymaker: it feels alone... |
− | [12:03] Justin Clark-Casey: well this doesn't need c# 5.0, bluewall | + | [2012/07/03 11:31] Nebadon Izumi: see what i can fine |
− | [12:03] Mike Kayaker: Yes, but nothing physical rezzed | + | [2012/07/03 11:31] Justin Clark-Casey: yes, can't tell anything without data |
− | [12:03] Richardus Raymaker: ehh ? | + | [2012/07/03 11:31] Nebadon Izumi: its been over a month since we used Bade |
− | [12:03] Richardus Raymaker: phsiscal but not physical. | + | [2012/07/03 11:31] Nebadon Izumi: i cant recall any of the numbers now |
− | [12:03] Mike Kayaker: The sim has ODE, but I don't have any physical objects in the sim ATM | + | [2012/07/03 11:32] Justin Clark-Casey: well, I don't know why jhere would work just fine but bade would not |
− | [12:03] Richardus Raymaker: ok, thats what i mean | + | [2012/07/03 11:32] Nebadon Izumi: its the only region we have trouble with, but its also the only region that we do not gather near the ground |
− | [12:04] Richardus Raymaker: its a dedicated server ? not VPS ? | + | [2012/07/03 11:32] Justin Clark-Casey: well, 100m should hardly make any difference |
− | [12:04] Mike Kayaker: Im not sure, VPS??? | + | [2012/07/03 11:32] Nebadon Izumi: ya you would think it wouldnt |
− | [12:04] Justin Clark-Casey: anyway, I'd like to answer dahlia quickly before she disappearas | + | [2012/07/03 11:32] Justin Clark-Casey: but... who knows |
− | [12:05] Richardus Raymaker: go justin | + | [2012/07/03 11:32] Justin Clark-Casey: it would be worth the comparison |
− | [12:05] Justin Clark-Casey: dahlia: at the moment, extended sog is difficult. In fact, my favourite solution would be to serialize a separate piece of data to the region, a bit like the moap xml | + | [2012/07/03 11:33] BlueWall Slade: also, it has lots of megaprims |
− | [12:05] Justin Clark-Casey: the alternative is to add more hard coded fields to prims or primshapes | + | [2012/07/03 11:33] Nebadon Izumi: yea |
− | [12:05] Dahlia Trimble: ewwww | + | [2012/07/03 11:33] Nebadon Izumi: ya there are a bunch of mega prims too |
− | [12:06] Dahlia Trimble: its a SOG property | + | [2012/07/03 11:33] Justin Clark-Casey: megaprims would certainly be an interesting factor |
− | [12:06] Justin Clark-Casey: which has to be persisted? | + | [2012/07/03 11:33] BlueWall Slade: and where the meeting is, there may be some interactions with the bounding boxes |
− | [12:06] Dahlia Trimble: yes | + | [2012/07/03 11:34] Richardus Raymaker: lots of boxes there |
− | [12:06] Justin Clark-Casey: I know, unofrutnately there is no separate table for sogs | + | [2012/07/03 11:34] Justin Clark-Casey: well, as I said, need more data |
− | [12:06] Dahlia Trimble: ok how about to a prim with parent id of 0? | + | [2012/07/03 11:34] Nebadon Izumi: but Lbsa Plaza also has lots of mega prims too |
− | [12:06] Justin Clark-Casey: if it was a serialization one could conceivably add it to the root prim only | + | [2012/07/03 11:34] Justin Clark-Casey: if somehow physics was soaking up lots of time then that would show up in the stats |
− | [12:07] Dahlia Trimble: so SOP then? | + | [2012/07/03 11:34] Nebadon Izumi: that whole pyramid is mega prims |
− | [12:07] Justin Clark-Casey: when you stat serializating properties you get into the expense of that, thuogh I personally don't think it's a huge issue | + | [2012/07/03 11:35] Nebadon Izumi: all the big boxes that surround the sim |
− | [12:07] Justin Clark-Casey: the problem is putting this stuff on the region itself | + | [2012/07/03 11:35] Nebadon Izumi: the floor |
− | [12:07] Justin Clark-Casey: you have to have a field in the regions db tables | + | [2012/07/03 11:35] Nebadon Izumi: and its the same server |
− | [12:08] Dahlia Trimble: if I add one how do I handle migrations and oars? | + | [2012/07/03 11:35] Nebadon Izumi: running same version as bad |
− | [12:08] Justin Clark-Casey: you can add a migration like the existing ones to add an extra field. Serialization covers oars, inventory, attachment teleports, etc. | + | [2012/07/03 11:35] BlueWall Slade: the framework that suspends it |
− | [12:09] Dahlia Trimble: kk | + | [2012/07/03 11:35] BlueWall Slade: those probably go outside the sim |
− | [12:09] Justin Clark-Casey: that needs an explicit change to the serialization/deser code in OpenSim/Region/Frametwork/Scene/Serialization/SceneObjectSerializer | + | [2012/07/03 11:35] Nebadon Izumi: hrmm |
− | [12:10] Justin Clark-Casey: I would ask that this isn't put directly into master in case there are any issues - something in a branch could be more easily checkable | + | [2012/07/03 11:36] Nebadon Izumi: ya hard to say |
− | [12:10] Dahlia Trimble: I'll ask you in IRC before adding anything | + | [2012/07/03 11:36] BlueWall Slade: I think I remember something about them being > 256m |
− | [12:11] Dahlia Trimble: anyway it would be fields that LL is adding but I dont think they use the normal protocols to share them with viewers | + | [2012/07/03 11:36] Nebadon Izumi: ya they are tall |
− | [12:11] Justin Clark-Casey: ok, I would certainly like the opportunity to think about the future data format | + | [2012/07/03 11:36] Nebadon Izumi: i think they are like 300m tall |
− | [12:12] Justin Clark-Casey: dahlia: using caps instead of udp? | + | [2012/07/03 11:36] Nebadon Izumi: but i do not recall how wide |
− | [12:12] BlueWall Slade: does OAR get the V3 environment settings ATM ? | + | [2012/07/03 11:36] Nebadon Izumi: i would be surprised though if they were not phantom |
− | [12:12] Justin Clark-Casey: bluewall: I don;'t think so | + | [2012/07/03 11:37] Nebadon Izumi: unfortunately it will be a bit before I can bring it back up |
− | [12:12] Dahlia Trimble: looks more like some proprietary havok encoding | + | [2012/07/03 11:37] Nebadon Izumi: we actually replaced it with another region for OSG5B |
− | [12:12] Dahlia Trimble: over caps | + | [2012/07/03 11:37] BlueWall Slade: ha |
− | [12:12] Justin Clark-Casey: uhhhhhhh :) | + | [2012/07/03 11:38] Richardus Raymaker: that building is higher then 256m ? |
− | [12:12] Dahlia Trimble: ya | + | [2012/07/03 11:38] Nebadon Izumi: yes |
− | [12:13] Richardus Raymaker: brrr, havok :) | + | [2012/07/03 11:38] Nebadon Izumi: well |
− | [12:13] Dahlia Trimble: but still the properties need to exist | + | [2012/07/03 11:38] Nebadon Izumi: actually not sure |
− | [12:13] Justin Clark-Casey: yes | + | [2012/07/03 11:38] Nebadon Izumi: the arch supports are 1/2 a cylinder |
− | [12:13] Justin Clark-Casey: well, maybe they could just be new fields - that would just be historical practice after all | + | [2012/07/03 11:38] Nebadon Izumi: so 300m 1/2 cylinder |
− | [12:13] Justin Clark-Casey: just a bit of a pita | + | [2012/07/03 11:38] Nebadon Izumi: is only 150m |
− | [12:14] Dahlia Trimble: need a process for this kind of thing | + | [2012/07/03 11:38] Nebadon Izumi: so it is probably less than 256 |
− | [12:14] Justin Clark-Casey: yes | + | [2012/07/03 11:39] Richardus Raymaker: but nebadon. on your vergicle sim you use >512m prims |
− | [12:14] Justin Clark-Casey: ok, I need to go and get some stuff done | + | [2012/07/03 11:39] Nebadon Izumi: vergicle? |
− | [12:14] Dahlia Trimble: bye :) | + | [2012/07/03 11:40] Richardus Raymaker: verhicle |
− | [12:14] BlueWall Slade: bye Justin | + | [2012/07/03 11:41] Nebadon Izumi: ahh |
− | [12:14] Bea Witched: be well Justin | + | [2012/07/03 11:41] BlueWall Slade: verticle? |
− | [12:14] Dahlia Trimble: me too lol... bye all :) | + | [2012/07/03 11:41] Nebadon Izumi: ya but I think Bluewalls point is that in bade |
− | [12:15] Mike Kayaker: thnks all | + | [2012/07/03 11:41] Nebadon Izumi: the prims might go over the edge of the sim |
− | [12:15] BlueWall Slade: bye Dahlia | + | [2012/07/03 11:41] Justin Clark-Casey: vestigal? |
− | [12:15] Justin Clark-Casey: mike: do you ever compile opensim or do you stick wityh binaries? | + | [2012/07/03 11:41] Nebadon Izumi: lol |
− | [12:15] Mike Kayaker: binaries | + | [2012/07/03 11:41] BlueWall Slade: and maybe the way they intersect |
− | [12:15] Nebadon Izumi: see you guys | + | [2012/07/03 11:42] Justin Clark-Casey: I dunno - I honestly have very little experience with megaprim sims |
− | [12:15] Nebadon Izumi: thanks for coming | + | [2012/07/03 11:43] Nebadon Izumi: ya there is something unique going on at bade |
− | [12:15] Richardus Raymaker: bye justin | + | [2012/07/03 11:43] Nebadon Izumi: that i have never experienced in any other region |
− | [12:15] Bea Witched: ty have a good day all | + | [2012/07/03 11:43] Nebadon Izumi: its very very odd |
− | [12:15] Justin Clark-Casey: ok, I was going to suggest a tweakable paramter but I'm not sure how much it would help | + | [2012/07/03 11:43] Justin Clark-Casey: I haven't heard of any issues with many users at considerable height |
− | [12:15] BlueWall Slade: I need to take off too - take care all | + | [2012/07/03 11:43] Nebadon Izumi: ive gone over the configurations and every prim in the entire sim |
− | [12:15] Justin Clark-Casey: maybe I should expose it as a config thing | + | [2012/07/03 11:43] Justin Clark-Casey: higher than 100m in fact |
− | [12:15] Mike Kayaker: next time | + | [2012/07/03 11:43] Nebadon Izumi: looking for scripts or something |
− | [12:16] Richardus Raymaker: bye blue | + | [2012/07/03 11:43] Dahlia Trimble: hi |
− | [12:16] Justin Clark-Casey: l | + | [2012/07/03 11:43] Justin Clark-Casey: well, knowing if there are high physics times would help |
− | [12:16] Justin Clark-Casey: k | + | [2012/07/03 11:43] Nebadon Izumi: ya its actually ok if we have less than 15 avatars |
− | [12:16] Justin Clark- | + | [2012/07/03 11:43] Justin Clark-Casey: if not, then you could proabbly rule out physics |
+ | [2012/07/03 11:43] Justin Clark-Casey: hi dahlia | ||
+ | [2012/07/03 11:44] Nebadon Izumi: once we cross a certain threshold of avatars | ||
+ | [2012/07/03 11:44] Nebadon Izumi: it just explodes | ||
+ | [2012/07/03 11:44] BlueWall Slade: Hello Dahlia | ||
+ | [2012/07/03 11:44] Nebadon Izumi: everyone complains prims are not loading | ||
+ | [2012/07/03 11:44] Nebadon Izumi: cant see any avatars | ||
+ | [2012/07/03 11:44] BlueWall Slade: those stats should give some clues too. | ||
+ | [2012/07/03 11:44] Justin Clark-Casey: It wouldn't enormously surprise me if some other process is soaking up all the sim time | ||
+ | [2012/07/03 11:44] BlueWall Slade: and you can use the in-world debugginh things | ||
+ | [2012/07/03 11:44] Justin Clark-Casey: like lots of invalid group requests or something | ||
+ | [2012/07/03 11:44] Nebadon Izumi: but why doesnt Lbsa do it | ||
+ | [2012/07/03 11:44] Nebadon Izumi: we can have 40 avatars on lbsa | ||
+ | [2012/07/03 11:44] Nebadon Izumi: same box | ||
+ | [2012/07/03 11:44] Justin Clark-Casey: I do not know | ||
+ | [2012/07/03 11:44] Nebadon Izumi: never ever once has done it on Lbsa | ||
+ | [2012/07/03 11:45] Nebadon Izumi: but everytime we use Bade it happens | ||
+ | [2012/07/03 11:45] Bea Witched: flying isn't allowed on lbsa so height is restricted | ||
+ | [2012/07/03 11:45] Nebadon Izumi: 4-5 meetings in a row it happened | ||
+ | [2012/07/03 11:45] Nebadon Izumi: finally we just moved it here for first time this month | ||
+ | [2012/07/03 11:45] Nebadon Izumi: we had 25 avatars | ||
+ | [2012/07/03 11:45] Nebadon Izumi: stats were perfect | ||
+ | [2012/07/03 11:45] Nebadon Izumi: literally no lag | ||
+ | [2012/07/03 11:45] Nebadon Izumi: you would think if any region was going to have issues | ||
+ | [2012/07/03 11:45] Nebadon Izumi: it would be this one | ||
+ | [2012/07/03 11:46] Nebadon Izumi: lol | ||
+ | [2012/07/03 11:46] Nebadon Izumi: bade has no neighbors | ||
+ | [2012/07/03 11:46] Nebadon Izumi: its completely isolated | ||
+ | [2012/07/03 11:46] Nebadon Izumi: low prim, low scripts | ||
+ | [2012/07/03 11:46] Nebadon Izumi: so bizzare | ||
+ | [2012/07/03 11:46] Richardus Raymaker: thats same as with my physical train test (i stopped with that) 1 train no timeings. when i rezz aseconds train phyics time jump up to 180,ms. like am used toin the past with 1 train | ||
+ | [2012/07/03 11:46] Justin Clark-Casey: well, not seeing objects is often a synptom of bad network comms | ||
+ | [2012/07/03 11:46] Justin Clark-Casey: object update packets being completely dropped and the resend system failing | ||
+ | [2012/07/03 11:46] Nebadon Izumi: ya this felt more like a hung thread or something | ||
+ | [2012/07/03 11:47] Nebadon Izumi: and the only thing i can really pinpoint and say its completely different than the other plazas | ||
+ | [2012/07/03 11:47] Nebadon Izumi: is that we meet above 100m above ground | ||
+ | [2012/07/03 11:47] Richardus Raymaker: i wish group IM did not be handles by region. | ||
+ | [2012/07/03 11:47] Nebadon Izumi: otherwise nothing is different | ||
+ | [2012/07/03 11:47] Nebadon Izumi: anyway i wasnt looking for a solution | ||
+ | [2012/07/03 11:47] Nebadon Izumi: i just wanted to present the idea | ||
+ | [2012/07/03 11:47] Richardus Raymaker: hi dahlia | ||
+ | [2012/07/03 11:47] Nebadon Izumi: i'll try to gather more info | ||
+ | [2012/07/03 11:47] Justin Clark-Casey: yeah | ||
+ | [2012/07/03 11:48] Dahlia Trimble: object updates dropping? is this a prim/texture heave region? | ||
+ | [2012/07/03 11:48] Dahlia Trimble: *heavy | ||
+ | [2012/07/03 11:48] Nebadon Izumi: no Dahlia, its quite low prims low textures | ||
+ | [2012/07/03 11:48] Nebadon Izumi: its maybe 7000 prims | ||
+ | [2012/07/03 11:48] Nebadon Izumi: few 100 textures maybe | ||
+ | [2012/07/03 11:48] Dahlia Trimble: so much for that idea :/ | ||
+ | [2012/07/03 11:48] Nebadon Izumi: ya ive tried everyting i could think of | ||
+ | [2012/07/03 11:49] Nebadon Izumi: i literally inspected every script on the sim | ||
+ | [2012/07/03 11:49] Nebadon Izumi: i went through the database | ||
+ | [2012/07/03 11:49] Nebadon Izumi: and located every script | ||
+ | [2012/07/03 11:49] Nebadon Izumi: to make sure it wasnt some kind of visitor scanner or something | ||
+ | [2012/07/03 11:49] Nebadon Izumi: some kind of radar | ||
+ | [2012/07/03 11:49] Nebadon Izumi: i couldnt find any heavy scripts | ||
+ | [2012/07/03 11:49] Nebadon Izumi: its all like sit scripts and hover text | ||
+ | [2012/07/03 11:49] Richardus Raymaker: also checked for lost AO's and other objects people lose ? | ||
+ | [2012/07/03 11:49] Nebadon Izumi: ya, i couldnt find anything unusual | ||
+ | [2012/07/03 11:50] Nebadon Izumi: thats not on all the other plazas | ||
+ | [2012/07/03 11:50] Justin Clark-Casey: anyway, any other opensim topics anybody wants to bring up today? | ||
+ | [2012/07/03 11:50] Mike Kayaker: I have a question about timers... | ||
+ | [2012/07/03 11:50] Nebadon Izumi: hrmm I dont think I have anything else myself | ||
+ | [2012/07/03 11:51] Nebadon Izumi: shoot Mike | ||
+ | [2012/07/03 11:51] Mike Kayaker: I seem to be running into a script resource I don't understand | ||
+ | [2012/07/03 11:51] Richardus Raymaker: well justin like i sayed in mailinmglist. new version seems to work betetr . sofar i have doen tests on the sim no crashing anymore | ||
+ | [2012/07/03 11:51] Dahlia Trimble: I need advice on adding properties to SOG | ||
+ | [2012/07/03 11:51] Mike Kayaker: If I have a bunch of scripts running that all have a short (1 second timer) | ||
+ | [2012/07/03 11:51] Dahlia Trimble: like for pathfincing, whether an object is a walkable | ||
+ | [2012/07/03 11:51] Mike Kayaker: the FPS starts to drop | ||
+ | [2012/07/03 11:52] Dahlia Trimble: how it affects database and oars | ||
+ | [2012/07/03 11:52] Mike Kayaker: I want to have two kinds of timers I guess, | ||
+ | [2012/07/03 11:52] Mike Kayaker: those that I care if the timer is on time, | ||
+ | [2012/07/03 11:52] Richardus Raymaker: strange mike. i have used 0.5s timers (no llSLeep) and never seen frame drops | ||
+ | [2012/07/03 11:52] Mike Kayaker: and others that I don't care | ||
+ | [2012/07/03 11:52] Justin Clark-Casey: I would find it very surprising if fps drops due to script timers | ||
+ | [2012/07/03 11:53] Justin Clark-Casey: unless you're using hundereds and hundreads of them and never removing them | ||
+ | [2012/07/03 11:53] Mike Kayaker: Is there some way I can allow the don't care timers to volunteer to be time sliced? | ||
+ | [2012/07/03 11:53] Mike Kayaker: like calling llSleep(.0001) or something? | ||
+ | [2012/07/03 11:53] Nebadon Izumi: i dont think so | ||
+ | [2012/07/03 11:53] Richardus Raymaker: llSLeep.00001) never works | ||
+ | [2012/07/03 11:53] Nebadon Izumi: there is a limit how low sleep can go | ||
+ | [2012/07/03 11:53] Dahlia Trimble: use not_at_target() for a per-frame event | ||
+ | [2012/07/03 11:53] Nebadon Izumi: its 0.01? | ||
+ | [2012/07/03 11:53] Nebadon Izumi: or 0.05? | ||
+ | [2012/07/03 11:54] Nebadon Izumi: i cant remember | ||
+ | [2012/07/03 11:54] Mike Kayaker: Well, of cource I do a vector calculation and a resize or something on every tick | ||
+ | [2012/07/03 11:54] Nebadon Izumi: maybe its even less | ||
+ | [2012/07/03 11:54] Richardus Raymaker: fastest i have seen is around 0.01 maby but i resits to use llSLeep at all. only in rare cases | ||
+ | [2012/07/03 11:54] Justin Clark-Casey: smallest sleep would be 1ms (0.001) | ||
+ | [2012/07/03 11:54] Nebadon Izumi: hrmm i ok didnt think it went that far | ||
+ | [2012/07/03 11:54] Nebadon Izumi: i thought it was 0.01 | ||
+ | [2012/07/03 11:54] Mike Kayaker: but not-at-target would slow down as the FPS goes down... | ||
+ | [2012/07/03 11:54] Richardus Raymaker: i agree with nebadon. i have not seen any difference betwen 0.01 and lower | ||
+ | [2012/07/03 11:55] Dahlia Trimble: why would you want to update anything faster than frame rate? | ||
+ | [2012/07/03 11:55] Richardus Raymaker: at_target stuff. thats still so unreliable for me | ||
+ | [2012/07/03 11:55] Mike Kayaker: Actully, I don't care how long the slep sleeps, just want timers on the other side of the sim to stay accurate when have a lot of ddon't care timers running | ||
+ | [2012/07/03 11:55] Justin Clark-Casey: yes. In fact, anything less than 0.001 will get trunacated to 0, no sleep at all | ||
+ | [2012/07/03 11:56] Justin Clark-Casey: mike: I wouldn't be surprised if timers are not very accurate atm | ||
+ | [2012/07/03 11:56] Mike Kayaker: thing is, I build one item, timers work good enough, | ||
+ | [2012/07/03 11:56] Justin Clark-Casey: I think the way they operate is somewhat inaccurate | ||
+ | [2012/07/03 11:56] Richardus Raymaker: Justin, timers, sensors, at_target are never 100% save for me | ||
+ | [2012/07/03 11:56] Dahlia Trimble: its not a real-time simulation running on a real-time os | ||
+ | [2012/07/03 11:56] BlueWall Slade: working on some stuff a couple of months back - looks like the events priority is burried below other things | ||
+ | [2012/07/03 11:56] Mike Kayaker: then I rez a bunch of the item, and things get very strange | ||
+ | [2012/07/03 11:57] BlueWall Slade: at_target was delayed for a long time | ||
+ | [2012/07/03 11:57] Dahlia Trimble: not at_target(), use not_at_target() *note underscores* | ||
+ | [2012/07/03 11:57] BlueWall Slade: when it finally showed up (way past the target point) it spammed the output pretty bad with many lines | ||
+ | [2012/07/03 11:58] Justin Clark-Casey: the problem is timers are run on the same loop as a lot of other lsl stuff | ||
+ | [2012/07/03 11:58] Justin Clark-Casey: if there's a lot going on then I suspect they could be delayed | ||
+ | [2012/07/03 11:58] Mike Kayaker: What resoure am I using up that causes the timers to slow down? | ||
+ | [2012/07/03 11:58] Justin Clark-Casey: and minimum timer resolution on this loop is 100ms | ||
+ | [2012/07/03 11:58] Richardus Raymaker: lsl is to bad not really multitasking | ||
+ | [2012/07/03 11:59] Dahlia Trimble: any event has to be queued anyway | ||
+ | [2012/07/03 11:59] Justin Clark-Casey: url listeners, sensors, dataserver requets | ||
+ | [2012/07/03 11:59] Justin Clark-Casey: and yes, then there's a delay/inaccuracy with timer event queueing | ||
+ | [2012/07/03 11:59] Simulator Version v0.5: shouts: OpenSim 0.7.4 Dev 1f34c82: 2012-06-13 04:05:02 +0100 (Unix/Mono) | ||
+ | [2012/07/03 11:59] Richardus Raymaker: i use sensor(s0 to, but the maby get called once every second or longer | ||
+ | [2012/07/03 12:00] Justin Clark-Casey: I suspect this area could be improved | ||
+ | [2012/07/03 12:01] Mike Kayaker: whispers: So is there a way to do a lag-friendly timer? Like get a timer evey second, but not slow down other scripts if the sim is buisy? | ||
+ | [2012/07/03 12:01] Dahlia Trimble: probably not if the sim is busy | ||
+ | [2012/07/03 12:01] Mike Kayaker: I mean, a timer that does allow itself to get slowed down, so other timers will not get delayes. | ||
+ | [2012/07/03 12:02] Justin Clark-Casey: I'm really not sure where the slow down is happening | ||
+ | [2012/07/03 12:02] Richardus Raymaker: Mike TimerEvent is possible most save, but not reliable if you ned to have strict timeings | ||
+ | [2012/07/03 12:02] Justin Clark-Casey: this would require rearchitecting the way that timers are handled | ||
+ | [2012/07/03 12:02] Justin Clark-Casey: possibly so they sit on a separate thread rather than with every other event | ||
+ | [2012/07/03 12:02] Justin Clark-Casey: or even setup actual system resource timers | ||
+ | [2012/07/03 12:02] Mike Kayaker: Every timer event in my sim starts getting delays. | ||
+ | [2012/07/03 12:02] BlueWall Slade: when we can build on C# 5.0 we should do that anyway | ||
+ | [2012/07/03 12:02] Richardus Raymaker: Mike , do you have physics on the sim ? | ||
+ | [2012/07/03 12:03] Dahlia Trimble: given that all events are queued I'd say its near impossible | ||
+ | [2012/07/03 12:03] Justin Clark-Casey: well this doesn't need c# 5.0, bluewall | ||
+ | [2012/07/03 12:03] Mike Kayaker: Yes, but nothing physical rezzed | ||
+ | [2012/07/03 12:03] Richardus Raymaker: ehh ? | ||
+ | [2012/07/03 12:03] Richardus Raymaker: phsiscal but not physical. | ||
+ | [2012/07/03 12:03] Mike Kayaker: The sim has ODE, but I don't have any physical objects in the sim ATM | ||
+ | [2012/07/03 12:03] Richardus Raymaker: ok, thats what i mean | ||
+ | [2012/07/03 12:04] Richardus Raymaker: its a dedicated server ? not VPS ? | ||
+ | [2012/07/03 12:04] Mike Kayaker: Im not sure, VPS??? | ||
+ | [2012/07/03 12:04] Justin Clark-Casey: anyway, I'd like to answer dahlia quickly before she disappearas | ||
+ | [2012/07/03 12:05] Richardus Raymaker: go justin | ||
+ | [2012/07/03 12:05] Justin Clark-Casey: dahlia: at the moment, extended sog is difficult. In fact, my favourite solution would be to serialize a separate piece of data to the region, a bit like the moap xml | ||
+ | [2012/07/03 12:05] Justin Clark-Casey: the alternative is to add more hard coded fields to prims or primshapes | ||
+ | [2012/07/03 12:05] Dahlia Trimble: ewwww | ||
+ | [2012/07/03 12:06] Dahlia Trimble: its a SOG property | ||
+ | [2012/07/03 12:06] Justin Clark-Casey: which has to be persisted? | ||
+ | [2012/07/03 12:06] Dahlia Trimble: yes | ||
+ | [2012/07/03 12:06] Justin Clark-Casey: I know, unofrutnately there is no separate table for sogs | ||
+ | [2012/07/03 12:06] Dahlia Trimble: ok how about to a prim with parent id of 0? | ||
+ | [2012/07/03 12:06] Justin Clark-Casey: if it was a serialization one could conceivably add it to the root prim only | ||
+ | [2012/07/03 12:07] Dahlia Trimble: so SOP then? | ||
+ | [2012/07/03 12:07] Justin Clark-Casey: when you stat serializating properties you get into the expense of that, thuogh I personally don't think it's a huge issue | ||
+ | [2012/07/03 12:07] Justin Clark-Casey: the problem is putting this stuff on the region itself | ||
+ | [2012/07/03 12:07] Justin Clark-Casey: you have to have a field in the regions db tables | ||
+ | [2012/07/03 12:08] Dahlia Trimble: if I add one how do I handle migrations and oars? | ||
+ | [2012/07/03 12:08] Justin Clark-Casey: you can add a migration like the existing ones to add an extra field. Serialization covers oars, inventory, attachment teleports, etc. | ||
+ | [2012/07/03 12:09] Dahlia Trimble: kk | ||
+ | [2012/07/03 12:09] Justin Clark-Casey: that needs an explicit change to the serialization/deser code in OpenSim/Region/Frametwork/Scene/Serialization/SceneObjectSerializer | ||
+ | [2012/07/03 12:10] Justin Clark-Casey: I would ask that this isn't put directly into master in case there are any issues - something in a branch could be more easily checkable | ||
+ | [2012/07/03 12:10] Dahlia Trimble: I'll ask you in IRC before adding anything | ||
+ | [2012/07/03 12:11] Dahlia Trimble: anyway it would be fields that LL is adding but I dont think they use the normal protocols to share them with viewers | ||
+ | [2012/07/03 12:11] Justin Clark-Casey: ok, I would certainly like the opportunity to think about the future data format | ||
+ | [2012/07/03 12:12] Justin Clark-Casey: dahlia: using caps instead of udp? | ||
+ | [2012/07/03 12:12] BlueWall Slade: does OAR get the V3 environment settings ATM ? | ||
+ | [2012/07/03 12:12] Justin Clark-Casey: bluewall: I don;'t think so | ||
+ | [2012/07/03 12:12] Dahlia Trimble: looks more like some proprietary havok encoding | ||
+ | [2012/07/03 12:12] Dahlia Trimble: over caps | ||
+ | [2012/07/03 12:12] Justin Clark-Casey: uhhhhhhh :) | ||
+ | [2012/07/03 12:12] Dahlia Trimble: ya | ||
+ | [2012/07/03 12:13] Richardus Raymaker: brrr, havok :) | ||
+ | [2012/07/03 12:13] Dahlia Trimble: but still the properties need to exist | ||
+ | [2012/07/03 12:13] Justin Clark-Casey: yes | ||
+ | [2012/07/03 12:13] Justin Clark-Casey: well, maybe they could just be new fields - that would just be historical practice after all | ||
+ | [2012/07/03 12:13] Justin Clark-Casey: just a bit of a pita | ||
+ | [2012/07/03 12:14] Dahlia Trimble: need a process for this kind of thing | ||
+ | [2012/07/03 12:14] Justin Clark-Casey: yes | ||
+ | [2012/07/03 12:14] Justin Clark-Casey: ok, I need to go and get some stuff done | ||
+ | [2012/07/03 12:14] Dahlia Trimble: bye :) | ||
+ | [2012/07/03 12:14] BlueWall Slade: bye Justin | ||
+ | [2012/07/03 12:14] Bea Witched: be well Justin | ||
+ | [2012/07/03 12:14] Dahlia Trimble: me too lol... bye all :) | ||
+ | [2012/07/03 12:15] Mike Kayaker: thnks all | ||
+ | [2012/07/03 12:15] BlueWall Slade: bye Dahlia | ||
+ | [2012/07/03 12:15] Justin Clark-Casey: mike: do you ever compile opensim or do you stick wityh binaries? | ||
+ | [2012/07/03 12:15] Mike Kayaker: binaries | ||
+ | [2012/07/03 12:15] Nebadon Izumi: see you guys | ||
+ | [2012/07/03 12:15] Nebadon Izumi: thanks for coming | ||
+ | [2012/07/03 12:15] Richardus Raymaker: bye justin | ||
+ | [2012/07/03 12:15] Bea Witched: ty have a good day all | ||
+ | [2012/07/03 12:15] Justin Clark-Casey: ok, I was going to suggest a tweakable paramter but I'm not sure how much it would help | ||
+ | [2012/07/03 12:15] BlueWall Slade: I need to take off too - take care all | ||
+ | [2012/07/03 12:15] Justin Clark-Casey: maybe I should expose it as a config thing | ||
+ | [2012/07/03 12:15] Mike Kayaker: next time | ||
+ | [2012/07/03 12:16] Richardus Raymaker: bye blue | ||
+ | [2012/07/03 12:16] Justin Clark-Casey: l | ||
+ | [2012/07/03 12:16] Justin Clark-Casey: k | ||
+ | [2012/07/03 12:16] Justin Clark-Casey: Justin Clark-Casey waves | ||
</pre> | </pre> |
Revision as of 12:21, 3 July 2012
[2012/07/03 11:07] Justin Clark-Casey: hi folks [2012/07/03 11:08] Mike Kayaker: The holiday is tomorow! [2012/07/03 11:08] BlueWall Slade: Hi Justin [2012/07/03 11:08] BlueWall Slade: yeah [2012/07/03 11:08] BlueWall Slade: a lot of people take a week [2012/07/03 11:08] Justin Clark-Casey: holiday? [2012/07/03 11:08] Justin Clark-Casey: ah, 4th july? [2012/07/03 11:08] BlueWall Slade: Independ-- yeah [2012/07/03 11:09] BlueWall Slade: I looked into that inventory issue last week, but hit dead ends - and my debugger was out of sorts [2012/07/03 11:10] Nebadon Izumi: hello [2012/07/03 11:10] BlueWall Slade: got that straight now, maybe I can look a little further into it [2012/07/03 11:10] Justin Clark-Casey: hi neb [2012/07/03 11:10] BlueWall Slade: Hi Nebadon [2012/07/03 11:10] Nebadon Izumi: heh totally wasnt paying attention to the time [2012/07/03 11:10] Richardus Raymaker: hi [2012/07/03 11:11] Nebadon Izumi: Guess maybe a lot of people may not show up, maybe holiday stuff [2012/07/03 11:11] Nebadon Izumi: hard to say, surprised more people arent here [2012/07/03 11:12] Justin Clark-Casey: yes, seems to fluctuate quite randomly wek to eek [2012/07/03 11:12] BlueWall Slade: so what's up? [2012/07/03 11:12] Justin Clark-Casey: not a lot :) [2012/07/03 11:13] Justin Clark-Casey: took a holiday myself yesterday - not sure it helped much [2012/07/03 11:13] BlueWall Slade: haa [2012/07/03 11:13] BlueWall Slade: it's always good to do [2012/07/03 11:13] Nebadon Izumi: not much going on here [2012/07/03 11:13] Justin Clark-Casey: also meens I have 3 days of stuff to catch up rather than 2 :) [2012/07/03 11:13] BlueWall Slade: :S [2012/07/03 11:14] Justin Clark-Casey: nebadon: did you try the latest opensim? Did it make any change to memory consumption? [2012/07/03 11:14] Nebadon Izumi: I did, and i didnt see much if any change [2012/07/03 11:14] Justin Clark-Casey: ok [2012/07/03 11:14] Nebadon Izumi: memory is still pretty high at times [2012/07/03 11:14] Nebadon Izumi: maybe a little better, but nothing drastic [2012/07/03 11:14] Justin Clark-Casey: yeah, I thikn the real reasons are elsewhere - this bug has only been around for 6 weeks or so [2012/07/03 11:15] Nebadon Izumi: I actually had to back pedal this region [2012/07/03 11:15] Nebadon Izumi: was really wierd [2012/07/03 11:15] Nebadon Izumi: I think the Leap second actually whacked this box out [2012/07/03 11:16] Nebadon Izumi: it was on Sunday [2012/07/03 11:16] Nebadon Izumi: right before the Monthly town hall meeting here [2012/07/03 11:16] Nebadon Izumi: all the regions on this box were using super high CPU [2012/07/03 11:16] Nebadon Izumi: was sooo bizzare [2012/07/03 11:16] Nebadon Izumi: even the empty void place holder regions were using 200% cpu [2012/07/03 11:16] Justin Clark-Casey: so did you actually go back to an earlier version of jurst restart the simulator? [2012/07/03 11:16] Nebadon Izumi: ya I did that 1st [2012/07/03 11:16] Nebadon Izumi: nothing changed [2012/07/03 11:17] Nebadon Izumi: i restarted everything and still super high CPU [2012/07/03 11:17] Nebadon Izumi: so i rebooted the box [2012/07/03 11:17] Nebadon Izumi: and then everything was fine [2012/07/03 11:17] Justin Clark-Casey: I know there were some underlying linux issues with the leap second and java [2012/07/03 11:17] Nebadon Izumi: but i was reading that a huge portion of the internet servers went crazy after the leap second thing [2012/07/03 11:17] Nebadon Izumi: now they are saying it was the Linux Kernel was the issue [2012/07/03 11:17] Justin Clark-Casey: maybe the same was true of other things as well [2012/07/03 11:17] Justin Clark-Casey: yes [2012/07/03 11:17] Nebadon Izumi: saw a bunch of articles today saying it was the linux kernel [2012/07/03 11:17] Nebadon Izumi: a lot of servers went crazy [2012/07/03 11:18] Nebadon Izumi: i have no other explination from what i was reading [2012/07/03 11:18] Nebadon Izumi: for most servers you had to reboot [2012/07/03 11:18] Nebadon Izumi: and then it was fine [2012/07/03 11:18] Nebadon Izumi: which is exactly what happend on this box [2012/07/03 11:18] Nebadon Izumi: was really bizzare [2012/07/03 11:18] Nebadon Izumi: it just seems odd to me this was the day after the leapsecond thing [2012/07/03 11:18] Nebadon Izumi: and i really had not looked at the box for probably 24+ hours prior to that [2012/07/03 11:19] Justin Clark-Casey: it quite possibly was leap second related [2012/07/03 11:20] Nebadon Izumi: ya its been fine since the reboot [2012/07/03 11:20] Nebadon Izumi: and never seen it do anything like that before [2012/07/03 11:20] Nebadon Izumi: i have no other explantaion [2012/07/03 11:20] Richardus Raymaker: sorry, tried small font. [2012/07/03 11:21] Richardus Raymaker: 40" (temporarily) is a bit to big as screen [2012/07/03 11:21] Richardus Raymaker: lol [2012/07/03 11:21] Richardus Raymaker: sorry 37" [2012/07/03 11:22] Richardus Raymaker: where on phgone to. sorry to ask neb. did you found what where wrong ? [2012/07/03 11:22] Nebadon Izumi: no clue, when i rebooted the server it was fine [2012/07/03 11:22] Richardus Raymaker: still 06-30 ? [2012/07/03 11:23] Nebadon Izumi: ya i never reverted it again [2012/07/03 11:23] Nebadon Izumi: i forgot to do that [2012/07/03 11:23] Nebadon Izumi: i'll do that in a bit [2012/07/03 11:25] Nebadon Izumi: we did have 25 people here though [2012/07/03 11:25] Nebadon Izumi: at the Town Hall [2012/07/03 11:25] Nebadon Izumi: it held up really good [2012/07/03 11:25] Richardus Raymaker: after downgrade it worked perfect yes [2012/07/03 11:25] Nebadon Izumi: actually at 1st it didnt [2012/07/03 11:25] Nebadon Izumi: it wasnt until i rebooted the server it ran normal [2012/07/03 11:26] Justin Clark-Casey: yeah, I would be pretty sure this is the leap second issue rather than opensim per se [2012/07/03 11:26] Nebadon Izumi: Dan Banner had a very good theory we need to test Justin [2012/07/03 11:26] Nebadon Izumi: normally we have our Town Hall at Bade Plaza [2012/07/03 11:26] Nebadon Izumi: but lately for last 6+ months [2012/07/03 11:26] Nebadon Izumi: it runs horrible there [2012/07/03 11:26] Nebadon Izumi: we get like 15 people and it starts lagging [2012/07/03 11:26] Nebadon Izumi: and running horrible [2012/07/03 11:26] Nebadon Izumi: ive spent hours trying to figure it out [2012/07/03 11:26] Nebadon Izumi: erased all kinds of stuff [2012/07/03 11:27] Nebadon Izumi: basically the sim is almost empty [2012/07/03 11:27] Nebadon Izumi: BUT [2012/07/03 11:27] Nebadon Izumi: Dan had the idea, that maybe it has something to do with the fact we have our meeting up at like 100+ meters in the sky [2012/07/03 11:27] Nebadon Izumi: thats the only variable I have not tested yet [2012/07/03 11:27] Justin Clark-Casey: do you observe frame time numbers during these tests? [2012/07/03 11:27] Richardus Raymaker: only thing todo is test it 1 time at lower level before things change [2012/07/03 11:28] Nebadon Izumi: here at Wright Plaza? [2012/07/03 11:28] Nebadon Izumi: or at Bade? [2012/07/03 11:28] Richardus Raymaker: no bade [2012/07/03 11:28] Justin Clark-Casey: bade [2012/07/03 11:28] Nebadon Izumi: yes its the same server that runs Lbsa Plaza [2012/07/03 11:28] Nebadon Izumi: and we never have those issues there [2012/07/03 11:28] Justin Clark-Casey: and? what are the frame time numbers under load? [2012/07/03 11:28] Nebadon Izumi: but we move 15-20 people to bade up at 100m and it runs like pure crap [2012/07/03 11:28] Nebadon Izumi: I don't know honestly [2012/07/03 11:28] Nebadon Izumi: all I know is that after so many people log in [2012/07/03 11:28] Nebadon Izumi: prims stop loading [2012/07/03 11:29] Nebadon Izumi: you cant see avatars [2012/07/03 11:29] Justin Clark-Casey: well, it's easy to see - just open the stats pane [2012/07/03 11:29] Justin Clark-Casey: if the physics time is high then that would indicate an ode issue [2012/07/03 11:29] Nebadon Izumi: we took the region down for now [2012/07/03 11:29] Nebadon Izumi: oh you mean when no avatars are there? [2012/07/03 11:29] Justin Clark-Casey: if not and there is still lots of spare time per frame then it's something else [2012/07/03 11:29] Justin Clark-Casey: well no, whenever you get these issues [2012/07/03 11:29] Nebadon Izumi: ya i would need to setup a test [2012/07/03 11:29] Nebadon Izumi: we took the sim down [2012/07/03 11:29] Nebadon Izumi: because its completely useless [2012/07/03 11:29] Richardus Raymaker: are the timeing numbers now more reliable ? [2012/07/03 11:29] Justin Clark-Casey: yes [2012/07/03 11:30] Nebadon Izumi: what i really need to do is setup a new sim [2012/07/03 11:30] Nebadon Izumi: and test [2012/07/03 11:30] Justin Clark-Casey: on here, for instance, physics time is only about 0.5 - 1.5 ms [2012/07/03 11:30] Nebadon Izumi: and also test in bade too [2012/07/03 11:30] Justin Clark-Casey: with spare time of 17ms or so [2012/07/03 11:30] Justin Clark-Casey: every 18 ms frame [2012/07/03 11:30] Justin Clark-Casey: so we aren't stressing physics unduly [2012/07/03 11:30] Nebadon Izumi: its just so wierd [2012/07/03 11:30] Nebadon Izumi: Bade plaza is like empty [2012/07/03 11:30] Nebadon Izumi: there are almost no scripts [2012/07/03 11:31] Nebadon Izumi: not many prims [2012/07/03 11:31] Nebadon Izumi: and it runs horrible [2012/07/03 11:31] Nebadon Izumi: so strange [2012/07/03 11:31] Nebadon Izumi: i'll try to come up with some numbers [2012/07/03 11:31] Justin Clark-Casey: what's cpu load like? [2012/07/03 11:31] Richardus Raymaker: it feels alone... [2012/07/03 11:31] Nebadon Izumi: see what i can fine [2012/07/03 11:31] Justin Clark-Casey: yes, can't tell anything without data [2012/07/03 11:31] Nebadon Izumi: its been over a month since we used Bade [2012/07/03 11:31] Nebadon Izumi: i cant recall any of the numbers now [2012/07/03 11:32] Justin Clark-Casey: well, I don't know why jhere would work just fine but bade would not [2012/07/03 11:32] Nebadon Izumi: its the only region we have trouble with, but its also the only region that we do not gather near the ground [2012/07/03 11:32] Justin Clark-Casey: well, 100m should hardly make any difference [2012/07/03 11:32] Nebadon Izumi: ya you would think it wouldnt [2012/07/03 11:32] Justin Clark-Casey: but... who knows [2012/07/03 11:32] Justin Clark-Casey: it would be worth the comparison [2012/07/03 11:33] BlueWall Slade: also, it has lots of megaprims [2012/07/03 11:33] Nebadon Izumi: yea [2012/07/03 11:33] Nebadon Izumi: ya there are a bunch of mega prims too [2012/07/03 11:33] Justin Clark-Casey: megaprims would certainly be an interesting factor [2012/07/03 11:33] BlueWall Slade: and where the meeting is, there may be some interactions with the bounding boxes [2012/07/03 11:34] Richardus Raymaker: lots of boxes there [2012/07/03 11:34] Justin Clark-Casey: well, as I said, need more data [2012/07/03 11:34] Nebadon Izumi: but Lbsa Plaza also has lots of mega prims too [2012/07/03 11:34] Justin Clark-Casey: if somehow physics was soaking up lots of time then that would show up in the stats [2012/07/03 11:34] Nebadon Izumi: that whole pyramid is mega prims [2012/07/03 11:35] Nebadon Izumi: all the big boxes that surround the sim [2012/07/03 11:35] Nebadon Izumi: the floor [2012/07/03 11:35] Nebadon Izumi: and its the same server [2012/07/03 11:35] Nebadon Izumi: running same version as bad [2012/07/03 11:35] BlueWall Slade: the framework that suspends it [2012/07/03 11:35] BlueWall Slade: those probably go outside the sim [2012/07/03 11:35] Nebadon Izumi: hrmm [2012/07/03 11:36] Nebadon Izumi: ya hard to say [2012/07/03 11:36] BlueWall Slade: I think I remember something about them being > 256m [2012/07/03 11:36] Nebadon Izumi: ya they are tall [2012/07/03 11:36] Nebadon Izumi: i think they are like 300m tall [2012/07/03 11:36] Nebadon Izumi: but i do not recall how wide [2012/07/03 11:36] Nebadon Izumi: i would be surprised though if they were not phantom [2012/07/03 11:37] Nebadon Izumi: unfortunately it will be a bit before I can bring it back up [2012/07/03 11:37] Nebadon Izumi: we actually replaced it with another region for OSG5B [2012/07/03 11:37] BlueWall Slade: ha [2012/07/03 11:38] Richardus Raymaker: that building is higher then 256m ? [2012/07/03 11:38] Nebadon Izumi: yes [2012/07/03 11:38] Nebadon Izumi: well [2012/07/03 11:38] Nebadon Izumi: actually not sure [2012/07/03 11:38] Nebadon Izumi: the arch supports are 1/2 a cylinder [2012/07/03 11:38] Nebadon Izumi: so 300m 1/2 cylinder [2012/07/03 11:38] Nebadon Izumi: is only 150m [2012/07/03 11:38] Nebadon Izumi: so it is probably less than 256 [2012/07/03 11:39] Richardus Raymaker: but nebadon. on your vergicle sim you use >512m prims [2012/07/03 11:39] Nebadon Izumi: vergicle? [2012/07/03 11:40] Richardus Raymaker: verhicle [2012/07/03 11:41] Nebadon Izumi: ahh [2012/07/03 11:41] BlueWall Slade: verticle? [2012/07/03 11:41] Nebadon Izumi: ya but I think Bluewalls point is that in bade [2012/07/03 11:41] Nebadon Izumi: the prims might go over the edge of the sim [2012/07/03 11:41] Justin Clark-Casey: vestigal? [2012/07/03 11:41] Nebadon Izumi: lol [2012/07/03 11:41] BlueWall Slade: and maybe the way they intersect [2012/07/03 11:42] Justin Clark-Casey: I dunno - I honestly have very little experience with megaprim sims [2012/07/03 11:43] Nebadon Izumi: ya there is something unique going on at bade [2012/07/03 11:43] Nebadon Izumi: that i have never experienced in any other region [2012/07/03 11:43] Nebadon Izumi: its very very odd [2012/07/03 11:43] Justin Clark-Casey: I haven't heard of any issues with many users at considerable height [2012/07/03 11:43] Nebadon Izumi: ive gone over the configurations and every prim in the entire sim [2012/07/03 11:43] Justin Clark-Casey: higher than 100m in fact [2012/07/03 11:43] Nebadon Izumi: looking for scripts or something [2012/07/03 11:43] Dahlia Trimble: hi [2012/07/03 11:43] Justin Clark-Casey: well, knowing if there are high physics times would help [2012/07/03 11:43] Nebadon Izumi: ya its actually ok if we have less than 15 avatars [2012/07/03 11:43] Justin Clark-Casey: if not, then you could proabbly rule out physics [2012/07/03 11:43] Justin Clark-Casey: hi dahlia [2012/07/03 11:44] Nebadon Izumi: once we cross a certain threshold of avatars [2012/07/03 11:44] Nebadon Izumi: it just explodes [2012/07/03 11:44] BlueWall Slade: Hello Dahlia [2012/07/03 11:44] Nebadon Izumi: everyone complains prims are not loading [2012/07/03 11:44] Nebadon Izumi: cant see any avatars [2012/07/03 11:44] BlueWall Slade: those stats should give some clues too. [2012/07/03 11:44] Justin Clark-Casey: It wouldn't enormously surprise me if some other process is soaking up all the sim time [2012/07/03 11:44] BlueWall Slade: and you can use the in-world debugginh things [2012/07/03 11:44] Justin Clark-Casey: like lots of invalid group requests or something [2012/07/03 11:44] Nebadon Izumi: but why doesnt Lbsa do it [2012/07/03 11:44] Nebadon Izumi: we can have 40 avatars on lbsa [2012/07/03 11:44] Nebadon Izumi: same box [2012/07/03 11:44] Justin Clark-Casey: I do not know [2012/07/03 11:44] Nebadon Izumi: never ever once has done it on Lbsa [2012/07/03 11:45] Nebadon Izumi: but everytime we use Bade it happens [2012/07/03 11:45] Bea Witched: flying isn't allowed on lbsa so height is restricted [2012/07/03 11:45] Nebadon Izumi: 4-5 meetings in a row it happened [2012/07/03 11:45] Nebadon Izumi: finally we just moved it here for first time this month [2012/07/03 11:45] Nebadon Izumi: we had 25 avatars [2012/07/03 11:45] Nebadon Izumi: stats were perfect [2012/07/03 11:45] Nebadon Izumi: literally no lag [2012/07/03 11:45] Nebadon Izumi: you would think if any region was going to have issues [2012/07/03 11:45] Nebadon Izumi: it would be this one [2012/07/03 11:46] Nebadon Izumi: lol [2012/07/03 11:46] Nebadon Izumi: bade has no neighbors [2012/07/03 11:46] Nebadon Izumi: its completely isolated [2012/07/03 11:46] Nebadon Izumi: low prim, low scripts [2012/07/03 11:46] Nebadon Izumi: so bizzare [2012/07/03 11:46] Richardus Raymaker: thats same as with my physical train test (i stopped with that) 1 train no timeings. when i rezz aseconds train phyics time jump up to 180,ms. like am used toin the past with 1 train [2012/07/03 11:46] Justin Clark-Casey: well, not seeing objects is often a synptom of bad network comms [2012/07/03 11:46] Justin Clark-Casey: object update packets being completely dropped and the resend system failing [2012/07/03 11:46] Nebadon Izumi: ya this felt more like a hung thread or something [2012/07/03 11:47] Nebadon Izumi: and the only thing i can really pinpoint and say its completely different than the other plazas [2012/07/03 11:47] Nebadon Izumi: is that we meet above 100m above ground [2012/07/03 11:47] Richardus Raymaker: i wish group IM did not be handles by region. [2012/07/03 11:47] Nebadon Izumi: otherwise nothing is different [2012/07/03 11:47] Nebadon Izumi: anyway i wasnt looking for a solution [2012/07/03 11:47] Nebadon Izumi: i just wanted to present the idea [2012/07/03 11:47] Richardus Raymaker: hi dahlia [2012/07/03 11:47] Nebadon Izumi: i'll try to gather more info [2012/07/03 11:47] Justin Clark-Casey: yeah [2012/07/03 11:48] Dahlia Trimble: object updates dropping? is this a prim/texture heave region? [2012/07/03 11:48] Dahlia Trimble: *heavy [2012/07/03 11:48] Nebadon Izumi: no Dahlia, its quite low prims low textures [2012/07/03 11:48] Nebadon Izumi: its maybe 7000 prims [2012/07/03 11:48] Nebadon Izumi: few 100 textures maybe [2012/07/03 11:48] Dahlia Trimble: so much for that idea :/ [2012/07/03 11:48] Nebadon Izumi: ya ive tried everyting i could think of [2012/07/03 11:49] Nebadon Izumi: i literally inspected every script on the sim [2012/07/03 11:49] Nebadon Izumi: i went through the database [2012/07/03 11:49] Nebadon Izumi: and located every script [2012/07/03 11:49] Nebadon Izumi: to make sure it wasnt some kind of visitor scanner or something [2012/07/03 11:49] Nebadon Izumi: some kind of radar [2012/07/03 11:49] Nebadon Izumi: i couldnt find any heavy scripts [2012/07/03 11:49] Nebadon Izumi: its all like sit scripts and hover text [2012/07/03 11:49] Richardus Raymaker: also checked for lost AO's and other objects people lose ? [2012/07/03 11:49] Nebadon Izumi: ya, i couldnt find anything unusual [2012/07/03 11:50] Nebadon Izumi: thats not on all the other plazas [2012/07/03 11:50] Justin Clark-Casey: anyway, any other opensim topics anybody wants to bring up today? [2012/07/03 11:50] Mike Kayaker: I have a question about timers... [2012/07/03 11:50] Nebadon Izumi: hrmm I dont think I have anything else myself [2012/07/03 11:51] Nebadon Izumi: shoot Mike [2012/07/03 11:51] Mike Kayaker: I seem to be running into a script resource I don't understand [2012/07/03 11:51] Richardus Raymaker: well justin like i sayed in mailinmglist. new version seems to work betetr . sofar i have doen tests on the sim no crashing anymore [2012/07/03 11:51] Dahlia Trimble: I need advice on adding properties to SOG [2012/07/03 11:51] Mike Kayaker: If I have a bunch of scripts running that all have a short (1 second timer) [2012/07/03 11:51] Dahlia Trimble: like for pathfincing, whether an object is a walkable [2012/07/03 11:51] Mike Kayaker: the FPS starts to drop [2012/07/03 11:52] Dahlia Trimble: how it affects database and oars [2012/07/03 11:52] Mike Kayaker: I want to have two kinds of timers I guess, [2012/07/03 11:52] Mike Kayaker: those that I care if the timer is on time, [2012/07/03 11:52] Richardus Raymaker: strange mike. i have used 0.5s timers (no llSLeep) and never seen frame drops [2012/07/03 11:52] Mike Kayaker: and others that I don't care [2012/07/03 11:52] Justin Clark-Casey: I would find it very surprising if fps drops due to script timers [2012/07/03 11:53] Justin Clark-Casey: unless you're using hundereds and hundreads of them and never removing them [2012/07/03 11:53] Mike Kayaker: Is there some way I can allow the don't care timers to volunteer to be time sliced? [2012/07/03 11:53] Mike Kayaker: like calling llSleep(.0001) or something? [2012/07/03 11:53] Nebadon Izumi: i dont think so [2012/07/03 11:53] Richardus Raymaker: llSLeep.00001) never works [2012/07/03 11:53] Nebadon Izumi: there is a limit how low sleep can go [2012/07/03 11:53] Dahlia Trimble: use not_at_target() for a per-frame event [2012/07/03 11:53] Nebadon Izumi: its 0.01? [2012/07/03 11:53] Nebadon Izumi: or 0.05? [2012/07/03 11:54] Nebadon Izumi: i cant remember [2012/07/03 11:54] Mike Kayaker: Well, of cource I do a vector calculation and a resize or something on every tick [2012/07/03 11:54] Nebadon Izumi: maybe its even less [2012/07/03 11:54] Richardus Raymaker: fastest i have seen is around 0.01 maby but i resits to use llSLeep at all. only in rare cases [2012/07/03 11:54] Justin Clark-Casey: smallest sleep would be 1ms (0.001) [2012/07/03 11:54] Nebadon Izumi: hrmm i ok didnt think it went that far [2012/07/03 11:54] Nebadon Izumi: i thought it was 0.01 [2012/07/03 11:54] Mike Kayaker: but not-at-target would slow down as the FPS goes down... [2012/07/03 11:54] Richardus Raymaker: i agree with nebadon. i have not seen any difference betwen 0.01 and lower [2012/07/03 11:55] Dahlia Trimble: why would you want to update anything faster than frame rate? [2012/07/03 11:55] Richardus Raymaker: at_target stuff. thats still so unreliable for me [2012/07/03 11:55] Mike Kayaker: Actully, I don't care how long the slep sleeps, just want timers on the other side of the sim to stay accurate when have a lot of ddon't care timers running [2012/07/03 11:55] Justin Clark-Casey: yes. In fact, anything less than 0.001 will get trunacated to 0, no sleep at all [2012/07/03 11:56] Justin Clark-Casey: mike: I wouldn't be surprised if timers are not very accurate atm [2012/07/03 11:56] Mike Kayaker: thing is, I build one item, timers work good enough, [2012/07/03 11:56] Justin Clark-Casey: I think the way they operate is somewhat inaccurate [2012/07/03 11:56] Richardus Raymaker: Justin, timers, sensors, at_target are never 100% save for me [2012/07/03 11:56] Dahlia Trimble: its not a real-time simulation running on a real-time os [2012/07/03 11:56] BlueWall Slade: working on some stuff a couple of months back - looks like the events priority is burried below other things [2012/07/03 11:56] Mike Kayaker: then I rez a bunch of the item, and things get very strange [2012/07/03 11:57] BlueWall Slade: at_target was delayed for a long time [2012/07/03 11:57] Dahlia Trimble: not at_target(), use not_at_target() *note underscores* [2012/07/03 11:57] BlueWall Slade: when it finally showed up (way past the target point) it spammed the output pretty bad with many lines [2012/07/03 11:58] Justin Clark-Casey: the problem is timers are run on the same loop as a lot of other lsl stuff [2012/07/03 11:58] Justin Clark-Casey: if there's a lot going on then I suspect they could be delayed [2012/07/03 11:58] Mike Kayaker: What resoure am I using up that causes the timers to slow down? [2012/07/03 11:58] Justin Clark-Casey: and minimum timer resolution on this loop is 100ms [2012/07/03 11:58] Richardus Raymaker: lsl is to bad not really multitasking [2012/07/03 11:59] Dahlia Trimble: any event has to be queued anyway [2012/07/03 11:59] Justin Clark-Casey: url listeners, sensors, dataserver requets [2012/07/03 11:59] Justin Clark-Casey: and yes, then there's a delay/inaccuracy with timer event queueing [2012/07/03 11:59] Simulator Version v0.5: shouts: OpenSim 0.7.4 Dev 1f34c82: 2012-06-13 04:05:02 +0100 (Unix/Mono) [2012/07/03 11:59] Richardus Raymaker: i use sensor(s0 to, but the maby get called once every second or longer [2012/07/03 12:00] Justin Clark-Casey: I suspect this area could be improved [2012/07/03 12:01] Mike Kayaker: whispers: So is there a way to do a lag-friendly timer? Like get a timer evey second, but not slow down other scripts if the sim is buisy? [2012/07/03 12:01] Dahlia Trimble: probably not if the sim is busy [2012/07/03 12:01] Mike Kayaker: I mean, a timer that does allow itself to get slowed down, so other timers will not get delayes. [2012/07/03 12:02] Justin Clark-Casey: I'm really not sure where the slow down is happening [2012/07/03 12:02] Richardus Raymaker: Mike TimerEvent is possible most save, but not reliable if you ned to have strict timeings [2012/07/03 12:02] Justin Clark-Casey: this would require rearchitecting the way that timers are handled [2012/07/03 12:02] Justin Clark-Casey: possibly so they sit on a separate thread rather than with every other event [2012/07/03 12:02] Justin Clark-Casey: or even setup actual system resource timers [2012/07/03 12:02] Mike Kayaker: Every timer event in my sim starts getting delays. [2012/07/03 12:02] BlueWall Slade: when we can build on C# 5.0 we should do that anyway [2012/07/03 12:02] Richardus Raymaker: Mike , do you have physics on the sim ? [2012/07/03 12:03] Dahlia Trimble: given that all events are queued I'd say its near impossible [2012/07/03 12:03] Justin Clark-Casey: well this doesn't need c# 5.0, bluewall [2012/07/03 12:03] Mike Kayaker: Yes, but nothing physical rezzed [2012/07/03 12:03] Richardus Raymaker: ehh ? [2012/07/03 12:03] Richardus Raymaker: phsiscal but not physical. [2012/07/03 12:03] Mike Kayaker: The sim has ODE, but I don't have any physical objects in the sim ATM [2012/07/03 12:03] Richardus Raymaker: ok, thats what i mean [2012/07/03 12:04] Richardus Raymaker: its a dedicated server ? not VPS ? [2012/07/03 12:04] Mike Kayaker: Im not sure, VPS??? [2012/07/03 12:04] Justin Clark-Casey: anyway, I'd like to answer dahlia quickly before she disappearas [2012/07/03 12:05] Richardus Raymaker: go justin [2012/07/03 12:05] Justin Clark-Casey: dahlia: at the moment, extended sog is difficult. In fact, my favourite solution would be to serialize a separate piece of data to the region, a bit like the moap xml [2012/07/03 12:05] Justin Clark-Casey: the alternative is to add more hard coded fields to prims or primshapes [2012/07/03 12:05] Dahlia Trimble: ewwww [2012/07/03 12:06] Dahlia Trimble: its a SOG property [2012/07/03 12:06] Justin Clark-Casey: which has to be persisted? [2012/07/03 12:06] Dahlia Trimble: yes [2012/07/03 12:06] Justin Clark-Casey: I know, unofrutnately there is no separate table for sogs [2012/07/03 12:06] Dahlia Trimble: ok how about to a prim with parent id of 0? [2012/07/03 12:06] Justin Clark-Casey: if it was a serialization one could conceivably add it to the root prim only [2012/07/03 12:07] Dahlia Trimble: so SOP then? [2012/07/03 12:07] Justin Clark-Casey: when you stat serializating properties you get into the expense of that, thuogh I personally don't think it's a huge issue [2012/07/03 12:07] Justin Clark-Casey: the problem is putting this stuff on the region itself [2012/07/03 12:07] Justin Clark-Casey: you have to have a field in the regions db tables [2012/07/03 12:08] Dahlia Trimble: if I add one how do I handle migrations and oars? [2012/07/03 12:08] Justin Clark-Casey: you can add a migration like the existing ones to add an extra field. Serialization covers oars, inventory, attachment teleports, etc. [2012/07/03 12:09] Dahlia Trimble: kk [2012/07/03 12:09] Justin Clark-Casey: that needs an explicit change to the serialization/deser code in OpenSim/Region/Frametwork/Scene/Serialization/SceneObjectSerializer [2012/07/03 12:10] Justin Clark-Casey: I would ask that this isn't put directly into master in case there are any issues - something in a branch could be more easily checkable [2012/07/03 12:10] Dahlia Trimble: I'll ask you in IRC before adding anything [2012/07/03 12:11] Dahlia Trimble: anyway it would be fields that LL is adding but I dont think they use the normal protocols to share them with viewers [2012/07/03 12:11] Justin Clark-Casey: ok, I would certainly like the opportunity to think about the future data format [2012/07/03 12:12] Justin Clark-Casey: dahlia: using caps instead of udp? [2012/07/03 12:12] BlueWall Slade: does OAR get the V3 environment settings ATM ? [2012/07/03 12:12] Justin Clark-Casey: bluewall: I don;'t think so [2012/07/03 12:12] Dahlia Trimble: looks more like some proprietary havok encoding [2012/07/03 12:12] Dahlia Trimble: over caps [2012/07/03 12:12] Justin Clark-Casey: uhhhhhhh :) [2012/07/03 12:12] Dahlia Trimble: ya [2012/07/03 12:13] Richardus Raymaker: brrr, havok :) [2012/07/03 12:13] Dahlia Trimble: but still the properties need to exist [2012/07/03 12:13] Justin Clark-Casey: yes [2012/07/03 12:13] Justin Clark-Casey: well, maybe they could just be new fields - that would just be historical practice after all [2012/07/03 12:13] Justin Clark-Casey: just a bit of a pita [2012/07/03 12:14] Dahlia Trimble: need a process for this kind of thing [2012/07/03 12:14] Justin Clark-Casey: yes [2012/07/03 12:14] Justin Clark-Casey: ok, I need to go and get some stuff done [2012/07/03 12:14] Dahlia Trimble: bye :) [2012/07/03 12:14] BlueWall Slade: bye Justin [2012/07/03 12:14] Bea Witched: be well Justin [2012/07/03 12:14] Dahlia Trimble: me too lol... bye all :) [2012/07/03 12:15] Mike Kayaker: thnks all [2012/07/03 12:15] BlueWall Slade: bye Dahlia [2012/07/03 12:15] Justin Clark-Casey: mike: do you ever compile opensim or do you stick wityh binaries? [2012/07/03 12:15] Mike Kayaker: binaries [2012/07/03 12:15] Nebadon Izumi: see you guys [2012/07/03 12:15] Nebadon Izumi: thanks for coming [2012/07/03 12:15] Richardus Raymaker: bye justin [2012/07/03 12:15] Bea Witched: ty have a good day all [2012/07/03 12:15] Justin Clark-Casey: ok, I was going to suggest a tweakable paramter but I'm not sure how much it would help [2012/07/03 12:15] BlueWall Slade: I need to take off too - take care all [2012/07/03 12:15] Justin Clark-Casey: maybe I should expose it as a config thing [2012/07/03 12:15] Mike Kayaker: next time [2012/07/03 12:16] Richardus Raymaker: bye blue [2012/07/03 12:16] Justin Clark-Casey: l [2012/07/03 12:16] Justin Clark-Casey: k [2012/07/03 12:16] Justin Clark-Casey: Justin Clark-Casey waves