Chat log from the meeting on 2023-06-06

From OpenSimulator

Jump to: navigation, search
[11:04 AM PT]  Andrew Hellershanks: Hello, everyone.
[11:04 AM PT]  Arielle Popstar: - 
[11:04 AM PT]  Arielle Popstar: v66- 
[11:04 AM PT]  Gavin.Hird Hi Andrew
[11:05 AM PT]  Cuga.Rajal is that god color work on standalones or just grids configs?
[11:05 AM PT]  Andrew Hellershanks: I had a service call to deal with and also trying to have my lunch.
[11:05 AM PT]  Vincent.Sylvester Region configs
[11:06 AM PT]  Andrew Hellershanks: Arielle is speaking in tongues. ;)
[11:06 AM PT]  Arielle Popstar: ^^
[11:06 AM PT]  Arielle Popstar: dropped my keyboard
[11:06 AM PT]  Cuga.Rajal So should work on standalone HG regions?
[11:06 AM PT]  Ubit Umarov: so as i told, fs did release beta 6.6.13 that does have the working fix for the non power of 2 regions sizes
[11:06 AM PT]  Ubit Umarov: we can even see the textures on minimpa
[11:06 AM PT]  Arielle Popstar: works for me
[11:06 AM PT]  Ubit Umarov: and minimap
[11:07 AM PT]  Arielle Popstar: why did it work before though?
[11:07 AM PT]  Ubit Umarov: thind dayturn as same simple fix also
[11:07 AM PT]  Vincent.Sylvester So long as it is in the OpenSim.ini and not overwritten it should work, you can check the running config with show config on the console that might give you a hint if you set it correctly
[11:07 AM PT]  Ubit Umarov: well on 6.6.12 there was a simple typo that just made it crash
[11:07 AM PT]  Vincent.Sylvester The module that has that along with the display names etc. retrieves that data either directly from robust or through user level I am not sure what it did, been some years
[11:07 AM PT]  Arielle Popstar: yes but it has always worked in past before those bad beta releases
[11:08 AM PT]  Arielle Popstar: what changed?
[11:08 AM PT]  Cuga.Rajal "show config" Invalid command
[11:08 AM PT]  Vincent.Sylvester Try config show
[11:08 AM PT]  Ubit Umarov: well older versions did ignre the warning, somewhere in code.. recent ones lost that ignore.. never found where..
[11:09 AM PT]  Cuga.Rajal that worked
[11:09 AM PT]  Ubit Umarov: instead made a fix, that Beq and Gavin did apply
[11:09 AM PT]  Arielle Popstar: so its fixed but we dont know why?
[11:09 AM PT]  Andrew Hellershanks: Hello, Kayaker.
[11:09 AM PT]  Ubit Umarov: i di dfix the real issue
[11:09 AM PT]  Andrew Hellershanks: Hello, Motoko.
[11:09 AM PT]  Kayaker Magic: Hi all, did I miss much?
[11:09 AM PT]  Motoko.Karu Hihi.. sorry for the lateness
[11:09 AM PT]  Motoko.Karu Hello Andrew
[11:10 AM PT]  Cuga.Rajal console needs a pipe and grep option lol
[11:10 AM PT]  Ubit Umarov: tthe issue is on textures the viewer draws.. so we can just tell it to use a power of 2 size
[11:10 AM PT]  Andrew Hellershanks: Kayaker, no you haven't missed much.
[11:10 AM PT]  Ubit Umarov: simple fix..
[11:10 AM PT]  Vincent.Sylvester I think it can write to file if you pipe it, well at least the yengine commands can I know that
[11:11 AM PT]  Cuga.Rajal its ok, I coopy/paste it into a text editor
[11:11 AM PT]  Ubit Umarov: isnta that god color a viewer side thing?
[11:11 AM PT]  Cuga.Rajal so GodNames is active and lists my name in the config, but it never changes color
[11:11 AM PT]  Gavin.Hird there are some god colors  in color.xml for the viewer
[11:11 AM PT]  Arielle Popstar: can change it viewer side
[11:12 AM PT]  Cuga.Rajal Is there somethign I'm missing? Heres the config:
[11:12 AM PT]  Cuga.Rajal [GodNames]
  Enabled = true
  FullNames = Cuga Rajal
  Surnames = Rajal
[11:12 AM PT]  Ubit Umarov: well something totally outside my priorities :)
[11:12 AM PT]  Arielle Popstar: i had to in s/l for the Lindens
[11:12 AM PT]  Ubit Umarov: ( god text color..   i mean )
[11:12 AM PT]  Arielle Popstar: yes
[11:12 AM PT]  Vincent.Sylvester You might need an underscore in there for the full name or wrapped in "", tried that?
[11:13 AM PT]  Cuga.Rajal havent tried
[11:13 AM PT]  Cuga.Rajal I'll try that
[11:13 AM PT]  Ubit Umarov: so.. abotu code changes..
[11:13 AM PT]  Gavin.Hird MenuBarGodBgColir, SyntaxLSLGodMode, MenuNonProductionGodBgColor
[11:14 AM PT]  Ubit Umarov: again not many.. well some.. let me see
[11:14 AM PT]  Cuga.Rajal I have my name in quotes in OpenSim.ini, output of config show doesnt have quotes
[11:14 AM PT]  Cuga.Rajal I'll try the underscore
[11:15 AM PT]  Ubit Umarov: by request i of the person that did original patch, i changes osDie to only work on prims rezzed by a script on that linkset
[11:15 AM PT]  Ubit Umarov: code was allowing osdie to kill that linkset..
[11:16 AM PT]  Ubit Umarov: that was on  mantis 9080
[11:17 AM PT]  Ubit Umarov: There was a somewhat big code contribution to add linksetdata,  - as sl
[11:17 AM PT]  Ubit Umarov: sadly i did refuse it, bc the problems sl idea has, and a few on that particular implementation
[11:18 AM PT]  Ubit Umarov: i did howver commit some constants, etc comaptible with that code, so to reserve for future
[11:18 AM PT]  Ubit Umarov: that was on mantis 9081, btw
[11:19 AM PT]  Cuga.Rajal I'll be right back
[11:19 AM PT]  Andrew Hellershanks: hb
[11:20 AM PT]  Ubit Umarov: well and a few more "cosmetic changes"  here and there
[11:20 AM PT]  Andrew Hellershanks: Hello, Jagga
[11:20 AM PT]  Ubit Umarov: like on sensor repeat code
[11:20 AM PT]  Gavin.Hird I have all the changes up and running
[11:20 AM PT]  Ubit Umarov: :)
[11:20 AM PT]  Gavin.Hird no problems, but problably have not tested specifics
[11:21 AM PT]  Vincent.Sylvester Also the lllinkplaysound thing and some typos
[11:21 AM PT]  Ubit Umarov: i actually did not update this region
[11:21 AM PT]  Ubit Umarov: ahg yes
[11:21 AM PT]  Ubit Umarov: lllinkplaysound was not as spec
[11:22 AM PT]  Gavin.Hird is pretty much running the latest. There is a rez zone there
[11:22 AM PT]  Ubit Umarov: sl as a extra parameter that makes it also do loop sound etc
[11:22 AM PT]  Ubit Umarov: so i added that
[11:23 AM PT]  Ubit Umarov: they are also suicidal
[11:23 AM PT]  Ubit Umarov: allowing LINK_SET , to a sound can play on every prim of a linkset
[11:23 AM PT]  Ubit Umarov: well i allowed that also..
[11:24 AM PT]  Ubit Umarov: not sure that is a good idea.. but well it is spec
[11:24 AM PT]  Ubit Umarov:
[11:25 AM PT]  Ubit Umarov: i left llLinkPlaySound(LSL_Integer linknumber, string sound, double volume) as a alias to
[11:25 AM PT]  Ubit Umarov: llLinkPlaySound(LSL_Integer linknumber, string sound, double volume, 0)
[11:25 AM PT]  Andrew Hellershanks: brb
[11:26 AM PT]  Ubit Umarov: ( double there means lsl float )
[11:26 AM PT]  Vincent.Sylvester I did not see a caveat there saying it only worked for specific links and link_set was listed, suppose when it was implemented someone did think maybe not to go crazy with it
[11:26 AM PT]  Cuga.Rajal you can get some cool sound phasing effects when you have the same sound from multiple prims
[11:27 AM PT]  Ubit Umarov: well i changed to work closer to that spec ( if not exactly as it )
[11:27 AM PT]  Vincent.Sylvester Viewers with fmod should modulate sound accordingly as well yeah
[11:27 AM PT]  Cuga.Rajal is  llLinkPlaySound something relatively new?
[11:28 AM PT]  Cuga.Rajal just curious, for some reason I never heard of it till recently
[11:28 AM PT]  Ubit Umarov: think i did add it a few months ago..  but a simpler one
[11:28 AM PT]  Ubit Umarov: only did play, and only one prim
[11:28 AM PT]  Cuga.Rajal ok, so maybe in the last year
[11:29 AM PT]  Cuga.Rajal thx
[11:29 AM PT]  Andrew Hellershanks: I'm back
[11:29 AM PT]  Vincent.Sylvester phil poked me about link sounds stuff on irc like two three years ago about some missing parts on that
[11:29 AM PT]  Vincent.Sylvester Don't remember exactly what that was, something about stopping sounds I think
[11:30 AM PT]  Ubit Umarov: HA-1: 44b4fe523b2b2f0d1a84bbe1ddc44b2f89d1291e

* add llLinkAdjustSoundVolume, llLinkPlaySound and llLinkStopSound
[11:30 AM PT]  Ubit Umarov: 6 feb 2023
[11:30 AM PT]  Cuga.Rajal ah so it is relatively new.. thanks
[11:31 AM PT]  Cuga.Rajal I have content that could use that
[11:31 AM PT]  Ubit Umarov: i should limit it again to a simple prim...
[11:31 AM PT]  Ubit Umarov: well.. lets see :)
[11:31 AM PT]  Tess JL: llLinkPlaySound() seems to be an attempt mto consolidate nearly all sound related lsl functions. It has a flag parameter that covers msot everything
[11:32 AM PT]  Ubit Umarov: so does ours now
[11:32 AM PT]  Ubit Umarov: ( the one in feb did not had last parameter )
[11:32 AM PT]  Cuga.Rajal Its just  llLinkPlaySound() with different parameters for trigger, play, loop and stop
[11:32 AM PT]  Tess JL: Does that mean it's recommended to use it instead of the old ones?
[11:33 AM PT]  Ubit Umarov: no
[11:33 AM PT]  Cuga.Rajal only when you need to trigger sound from a different prim than the script is in
[11:33 AM PT]  Ubit Umarov: main thing about it is the *link* part
[11:33 AM PT]  Cuga.Rajal can reduce script count
[11:33 AM PT]  Ubit Umarov: to issue a sound on other prim
[11:34 AM PT]  Ubit Umarov: to do more things than play, is SL beaking naming loging
[11:34 AM PT]  Ubit Umarov: logic
[11:34 AM PT]  Ubit Umarov: just lazy code
[11:34 AM PT]  Vincent.Sylvester Now I have a small question about prim updates. I have a script that updates 25 textures on an object, which works fine for the most part, but when I put multiple objects out eventually they only up 24 or even less. Are there differences in how OpenSim and SL send prim updates for textures and color? I know we don't send certain terse or partial updates, could this be the cause? I added small random sleeps to the update function in attempt to break the updates up into their own packets, which seems to help a bit.
[11:35 AM PT]  Ubit Umarov: some texture changes can be terse updates now
[11:35 AM PT]  Ubit Umarov: other changes are a full update
[11:36 AM PT]  Ubit Umarov: and no.. we send far more than 24
[11:36 AM PT]  Ubit Umarov: you udp may lose it
[11:36 AM PT]  Ubit Umarov: or you viewer get overloaded and lose them
[11:37 AM PT]  Vincent.Sylvester It starts to break at like 200  when I have 8 or more objects out they sometimes miss one or two, but only when they run at the same time after a region restart
[11:37 AM PT]  Vincent.Sylvester I added sleep with frand in state entry and that helped
[11:37 AM PT]  Ubit Umarov: i tested that 256 prims transparence change and got no fails
[11:37 AM PT]  Ubit Umarov:
[11:38 AM PT]  Vincent.Sylvester That's the weird part as well, I couldn't reproduce on local machine either, granted local is windows and region is linux
[11:38 AM PT]  Lyr Lobo: Thank you for the link
[11:38 AM PT]  Ubit Umarov: udp is per user..
[11:39 AM PT]  Ubit Umarov: a issue similar to holes on terrain
[11:39 AM PT]  Ubit Umarov: viewer just gets overloaded and loses things, but flagging it did them
[11:40 AM PT]  Ubit Umarov: i had to slow down lludp more to avoid that terrain things
[11:40 AM PT]  Cuga.Rajal Have you tried using  llSetLinkPrimitiveParamsFast (add Fast)? there is less of an internal delay
[11:40 AM PT]  Tess JL: Holes on terrain?
[11:40 AM PT]  Ubit Umarov: les internaly delay makes it worse lol
[11:40 AM PT]  Andrew Hellershanks: How about we table the rest of the talk about the sound issue for the moment. Cuga has some information for us and last week Tess said they had some topics for the meeting. I don't want to have Tess wait another week.
[11:40 AM PT]  Ubit Umarov: less
[11:41 AM PT]  Andrew Hellershanks: Wait a moment, Tess. I'll let Cuga give a quick update then you can go.
[11:41 AM PT]  Kayaker Magic: What holes on terrain issue are you talking about Ubit? Is there a mantis on that?
[11:41 AM PT]  Vincent.Sylvester My main concern is we are operating to spec on that and not disobey max packet sizes?
[11:41 AM PT]  Tess JL: It'll have to wait another week, Andrew. I haven't really got around to it yet. It's not urgent though
[11:41 AM PT]  Ubit Umarov: yes with high bandwith set on viewer udp, we would get holes
[11:41 AM PT]  Cuga.Rajal I use llSetLinkPrimitiveParamsFast a lot and never had issue
[11:41 AM PT]  Ubit Umarov: bc opensim just did sent terrain faster than viewers could handle
[11:41 AM PT]  Andrew Hellershanks: Tess, ok. Just wanted to make sure you had an opportunity to ask your questions.
[11:42 AM PT]  Tess JL: :-)
[11:42 AM PT]  Ubit Umarov: so i added a extra check
[11:42 AM PT]  Cuga.Rajal I have just a short update on Bullet
[11:42 AM PT]  Andrew Hellershanks: Go ahead, Cuga.
[11:42 AM PT]  Cuga.Rajal theres neen some work in the background on a new version, an update
[11:42 AM PT]  Cuga.Rajal been*
[11:42 AM PT]  Ubit Umarov: region stops sendind udp if it has X entries waiting for AC K
[11:42 AM PT]  Kayaker Magic: I have a problem with holes in terrain, but only on a few terrains, trying to make it repeatable to submit as a mantis.
[11:42 AM PT]  Ubit Umarov: that way it can wait for viewer to be ready
[11:43 AM PT]  Ubit Umarov: reduce udp bandwith
[11:43 AM PT]  Ubit Umarov: or get better net :p
[11:43 AM PT]  Ubit Umarov: well or gpu
[11:43 AM PT]  Andrew Hellershanks: Let's allow Cuga to finish providing the update on physics.
[11:43 AM PT]  Ubit Umarov: or cpy
[11:43 AM PT]  Ubit Umarov: cpu even
[11:43 AM PT]  Cuga.Rajal MrBlue has effectively squashed the biggest Bullet bug by adding a Extended Physics function to disable it
[11:43 AM PT]  Orbert.Tatham I had something like the same idea - outgoing queue for the viewer requests
[11:44 AM PT]  Cuga.Rajal so with that addition we think the new version is about ready
[11:44 AM PT]  Gavin.Hird what is that bug?
[11:44 AM PT]  Ubit Umarov: to disable what ?
[11:44 AM PT]  Cuga.Rajal Bullet 3.25 doesn;t appear any different from 2.86, but there are some bug fixes
[11:45 AM PT]  Cuga.Rajal I mean, the differences are visually obvious
[11:45 AM PT]  Cuga.Rajal but there are
[11:45 AM PT]  Cuga.Rajal sorry, meant to sat differences are not visually obvious
[11:45 AM PT]  Cuga.Rajal has some bug fixes
[11:45 AM PT]  Gavin.Hird what is the biggest bug squashed?
[11:45 AM PT]  Cuga.Rajal MrBlue adapted my build process for other OS's
[11:46 AM PT]  Cuga.Rajal and there is a github actions now for building the new Bullet for multi-platforms
[11:46 AM PT]  Cuga.Rajal the new Bullet will show new Bullet and wrapper version
[11:46 AM PT]  Cuga.Rajal when queried with os scripting command
[11:47 AM PT]  Cuga.Rajal and you can now turn off physics sleep on invividual objects
[11:47 AM PT]  Cuga.Rajal that was the biggest issue
[11:47 AM PT]  Ubit Umarov: well that is not a fix
[11:47 AM PT]  Cuga.Rajal correct
[11:47 AM PT]  Ubit Umarov: manual turn off is not that nice
[11:47 AM PT]  Gavin.Hird turn off sleep - what does that actually do?
[11:47 AM PT]  Cuga.Rajal but the sleep code was so deeply embedded in their code, disabling it wouldf have been a major project
[11:48 AM PT]  Ubit Umarov: i have a mess of code on ubode to do proper fix :p
[11:48 AM PT]  Gavin.Hird why would I want to do that?
[11:48 AM PT]  Cuga.Rajal and would have induced other instabilities
[11:48 AM PT]  Vincent.Sylvester Oh great more rogue vibrating prims killing cpu
[11:48 AM PT]  Cuga.Rajal so being able to disable it, doesnt fix problem, but gives option to disable that feature
[11:48 AM PT]  Cuga.Rajal on a per-object basis
[11:48 AM PT]  Ubit Umarov: bc you you remove a prim with a physical one or a avatar on top ot it, those will stay iin air
[11:48 AM PT]  Ubit Umarov: for example
[11:49 AM PT]  Ubit Umarov: sl also had that problem for ages
[11:49 AM PT]  Cuga.Rajal you only need to disable the physics sleep on a one or few use cases it's an issue
[11:49 AM PT]  Cuga.Rajal like my labyrinth example we rezzed here a few weeks ago
[11:49 AM PT]  Ubit Umarov: well and that is not related to bullet version.. just our code use of it possible
[11:49 AM PT]  Cuga.Rajal correct
[11:49 AM PT]  Ubit Umarov: well old issue
[11:49 AM PT]  Cuga.Rajal new solutioin
[11:50 AM PT]  Ubit Umarov: ubode has that fixed since.. ever
[11:50 AM PT]  Ubit Umarov: well ever with name ubode
[11:50 AM PT]  Cuga.Rajal bullet sleep issue is worse than ubide
[11:50 AM PT]  Ubit Umarov: don't remember if i did that still at aurora
[11:50 AM PT]  Cuga.Rajal but with this option it effectively fixes the few use cases that have the issue, but requires some scripting
[11:50 AM PT]  Cuga.Rajal thats as far as MrBlue is willing to go
[11:51 AM PT]  Ubit Umarov: yes but manual disable of sleep is hardly a fix
[11:51 AM PT]  Cuga.Rajal fixing the sleep in Bullet code the right way is too big of effort
[11:51 AM PT]  Gavin.Hird could you describe the use cases, cause I still don't get why it is needed
[11:51 AM PT]  Ubit Umarov: wel it is some code
[11:51 AM PT]  Cuga.Rajal remember Bullet is not ours, it is a big project that we extract some code from
[11:51 AM PT]  Vincent.Sylvester Bandaid on a compound fracture
[11:52 AM PT]  Andrew Hellershanks: Hello, Joe.
[11:52 AM PT]  Cuga.Rajal just relaying what MrBlue told me
[11:52 AM PT]  Gavin.Hird Hi Joe
[11:52 AM PT]  Ubit Umarov: see my 2 boxes gavin=
[11:52 AM PT]  Joe Magarac: Hello. Late today, sorry.
[11:52 AM PT]  Gavin.Hird yup
[11:52 AM PT]  Ubit Umarov: opt one is physical
[11:52 AM PT]  Cuga.Rajal I did test hid physics-turn-off option and it did solve the issue for me
[11:52 AM PT]  Cuga.Rajal no noticeable CPU change
[11:52 AM PT]  Ubit Umarov: this is ubode
[11:52 AM PT]  Ubit Umarov: after a little delay it seen that i did remove the bottom prim
[11:52 AM PT]  Gavin.Hird so the physical one falls as it should
[11:53 AM PT]  Ubit Umarov: bullet will not do that
[11:53 AM PT]  Cuga.Rajal so anyway, the new version of Bullet is not far off, in its final testing
[11:53 AM PT]  Ubit Umarov: until something colides with the box
[11:53 AM PT]  Gavin.Hird right
[11:53 AM PT]  Gavin.Hird I was about to say till you kick it
[11:53 AM PT]  Ubit Umarov: old issue.. as i said sl that it for ages also
[11:53 AM PT]  Cuga.Rajal originally we thought there would be a need for sandboxing and thorough testoing, but MrBlue and I dont see much change between versions
[11:54 AM PT]  Gavin.Hird where does the sleep come in
[11:54 AM PT]  Ubit Umarov: .. sl had ..
[11:54 AM PT]  Cuga.Rajal so the concensus is to just generate the new Bullet for all arcvhes
[11:54 AM PT]  Gavin.Hird that you turn off
[11:54 AM PT]  Ubit Umarov: but sleep is a big thing to keep cpu under acceptable load
[11:54 AM PT]  Cuga.Rajal Not from my experiencer
[11:54 AM PT]  Cuga.Rajal that was the intention
[11:54 AM PT]  Cuga.Rajal but it was written in a time when CPUs were much slower
[11:55 AM PT]  Cuga.Rajal I hacked the Bullet code and turned sleep off for everything
[11:55 AM PT]  Ubit Umarov: does not matter if cpus are faste
[11:55 AM PT]  Cuga.Rajal ran that for a week, and did not see any change
[11:55 AM PT]  Cuga.Rajal that was just a test, not for public :)
[11:55 AM PT]  Ubit Umarov: it still impacts your electricity bill
[11:56 AM PT]  Ubit Umarov: and number of active prims before you notice the lag
[11:56 AM PT]  Cuga.Rajal so turning off for only the prims/cases that need it should not be signififcant issue
[11:56 AM PT]  Ubit Umarov: but you power bill will go up :p
[11:56 AM PT]  Cuga.Rajal and MrBlue is not going to re-do the whole physics code of Bullet
[11:56 AM PT]  Cuga.Rajal I mean sleep code
[11:56 AM PT]  Gavin.Hird so you turn off the sleep for a prim you want to fall like Ubit demonstrated?
[11:56 AM PT]  Cuga.Rajal and I dont see a need
[11:57 AM PT]  Ubit Umarov: well those changes are already on master
[11:57 AM PT]  Ubit Umarov: i think
[11:57 AM PT]  Cuga.Rajal if you set a prom to have physics sleep turned off, it will fall without delay
[11:57 AM PT]  Cuga.Rajal yes, already omplemented
[11:57 AM PT]  Gavin.Hird ok, thanks
[11:57 AM PT]  Gavin.Hird you spent 100 lines explaining what could be explained in 2 lines
[11:57 AM PT]  Cuga.Rajal So this sleep change is not actually part of the new Bullet
[11:57 AM PT]  Gavin.Hird ;-)
[11:57 AM PT]  Cuga.Rajal but was a bug that he was looking for solution
[11:57 AM PT]  Ubit Umarov: no it is not
[11:58 AM PT]  Cuga.Rajal before finishing Bullet
[11:58 AM PT]  Vincent.Sylvester Users gonna set that for everything as standard once they hear its a fix to some problem and then we have regions filled with vibrating prims eating cpu by the byteload
[11:58 AM PT]  Cuga.Rajal lol
[11:58 AM PT]  Vincent.Sylvester Probably should limit for how many prims per scene you can do that or something
[11:58 AM PT]  Cuga.Rajal The new Bullet does not disable sleep!
[11:58 AM PT]  Ubit Umarov: actually bullet module is very agressive on sleep
[11:58 AM PT]  Cuga.Rajal and the option to disable is on a per-object basis
[11:59 AM PT]  Ubit Umarov: it even stops prims in mid air on a slow fall :)
[11:59 AM PT]  Cuga.Rajal so that was the main  hurdle before finishing up Bullet 3.25
[11:59 AM PT]  Ubit Umarov: that is a per object flag on engines
[11:59 AM PT]  Cuga.Rajal now that is being worked on
[11:59 AM PT]  Ubit Umarov: ( wel there maybe a global one and global settings)
[11:59 AM PT]  Cuga.Rajal so final Bullet update coming soon
[11:59 AM PT]  Cuga.Rajal version 3.25
[12:00 PM PT]  Cuga.Rajal it's not really much different than 2.86 but likely has some bug fixes
[12:00 PM PT]  Gavin.Hird how do you turn off per object sleep? In a script?
[12:00 PM PT]  Cuga.Rajal yes
[12:00 PM PT]  Andrew Hellershanks: Big jump in version number so one would expect some bug fixes. :)
[12:00 PM PT]  Cuga.Rajal actually a good thing that it isnt much different
[12:00 PM PT]  Gavin.Hird is there new syntax to something to do that?
[12:01 PM PT]  Ubit Umarov: cuga is talking abotu unmanaged code version
[12:01 PM PT]  Cuga.Rajal MrBlue updated the wiki with instructions
[12:01 PM PT]  Cuga.Rajal recently
[12:01 PM PT]  Cuga.Rajal the sleep fix uses an API to Bullet that already exusted in 2.86
[12:01 PM PT]  Gavin.Hird ok
[12:01 PM PT]  Cuga.Rajal so you dont have to wait for the new Bullet to use the new sleep disable
[12:02 PM PT]  Cuga.Rajal although the 3.25 effort is what motivated that addition
[12:02 PM PT]  Ubit Umarov: unrelated changes :)
[12:02 PM PT]  Cuga.Rajal unrelated in code, related in process
[12:02 PM PT]  Cuga.Rajal so thats all I wanted to announce about Bullet
[12:02 PM PT]  Cuga.Rajal unless tyhere are quesations
[12:03 PM PT]  Ubit Umarov: you know that also happens to avatars right?
[12:03 PM PT]  Ubit Umarov: :)
[12:03 PM PT]  Cuga.Rajal Does ubODE have a sleep disable option?
[12:04 PM PT]  Ubit Umarov: why should it have such a band thing?=
[12:04 PM PT]  Ubit Umarov: bad
[12:04 PM PT]  Ubit Umarov: ;)
[12:04 PM PT]  Cuga.Rajal difference of opinion
[12:04 PM PT]  Cuga.Rajal then no delay when you pull the block out from under :)
[12:05 PM PT]  Ubit Umarov: we do like to not burn cpus for no reaso
[12:05 PM PT]  Cuga.Rajal we should all be Ruths then
[12:05 PM PT]  Ubit Umarov: ytou add that and all things will use it
[12:05 PM PT]  Cuga.Rajal ?
[12:05 PM PT]  Arielle Popstar: dont give him ideas
[12:05 PM PT]  Andrew Hellershanks: hehe
[12:06 PM PT]  Cuga.Rajal There are a few use cases where physucs sleep just doesn;t work
[12:06 PM PT]  Gavin.Hird we are all ruths
[12:06 PM PT]  Cuga.Rajal my labyrinth was one example
[12:06 PM PT]  Cuga.Rajal broke on ubODE here too
[12:06 PM PT]  Gavin.Hird and that is primarily a viewer side CPU/GPU issue
[12:06 PM PT]  Cuga.Rajal disabling physics sleep on just the ball fixes it
[12:07 PM PT]  Vincent.Sylvester People still mixing mesher and physics, give them an option and you can advise all you want, detrimental or catastrophic if it adds even the perception of making something better it will be used relentlessly
[12:07 PM PT]  Cuga.Rajal anyway, details on the Bullet physucs sleep is here:
[12:08 PM PT]  Cuga.Rajal scroll down to physDisableDeactivation(TRUE);
[12:08 PM PT]  Cuga.Rajal you have to enable Extended Physics
[12:08 PM PT]  Cuga.Rajal works like a charm
[12:08 PM PT]  Cuga.Rajal the fans on my computer never spun up
[12:08 PM PT]  Arielle Popstar: learning curve is steep enough, lets not make it worse
[12:08 PM PT]  Cuga.Rajal :)
[12:09 PM PT]  Cuga.Rajal so thats all I had unless there are questions
[12:09 PM PT]  Gavin.Hird so physDisableDeactivation must be added to viewer script syntax
[12:09 PM PT]  Selby.Evans Bye all
[12:10 PM PT]  Tess JL: Bye Selbye
[12:10 PM PT]  Cuga.Rajal I donlt think it's a viewer thing
[12:10 PM PT]  Jamie.Jordan have a great week yall
[12:10 PM PT]  Gavin.Hird if you use it in a script, maybe the viewer should highlight it?
[12:10 PM PT]  Cuga.Rajal when you enable extended physics, the script compiles with that comand
[12:10 PM PT]  Gavin.Hird or is it a setting
[12:11 PM PT]  Cuga.Rajal if you donlt enable extended physics the script will not compile
[12:11 PM PT]  Ubit Umarov: extended physics does not show up on script syntaxe
[12:11 PM PT]  Cuga.Rajal that adds some scripting functions like phys*
[12:12 PM PT]  Cuga.Rajal that start with phys
[12:12 PM PT]  Vincent.Sylvester He is talking about the ScriptSyntax.xml which defines OpenSim specific functions and constants so the viewer can properly syntax highlight them in the script text box
[12:12 PM PT]  Gavin.Hird exactly
[12:12 PM PT]  Vincent.Sylvester When you add anything you want to add it there as well to make sure it is highlighted, don't think you can have a module intercept and add to that though
[12:12 PM PT]  Ubit Umarov: yes.. " extended physics" are not included
[12:12 PM PT]  Cuga.Rajal I dont think the viewers support coloring/help/etc
[12:12 PM PT]  Gavin.Hird physDisableDeactivation(TRUE); seems like something the viewer script editor should highlight
[12:12 PM PT]  Cuga.Rajal for the phys* scripting functions
[12:12 PM PT]  Cuga.Rajal but they still work
[12:13 PM PT]  Gavin.Hird whaddoyomean does not support
[12:13 PM PT]  Ubit Umarov: they are even on a module disabled by default
[12:13 PM PT]  Cuga.Rajal Extended physics functions have been around for many years
[12:13 PM PT]  Cuga.Rajal physDisableDeactivation is just a new addition
[12:13 PM PT]  Gavin.Hird it is a fuction of the viewer script editor, where it does a lookup on known syntax and highlights it accordingly
[12:13 PM PT]  Ubit Umarov: and that is a bullet only module also
[12:14 PM PT]  Cuga.Rajal correct
[12:14 PM PT]  Ubit Umarov: i did disabled it if bullet was not present..
[12:14 PM PT]  Ubit Umarov: but did allow it again, bc no big issues on being there
[12:14 PM PT]  Cuga.Rajal right
[12:14 PM PT]  Gavin.Hird so if there exist such scripting additions related to bullet the viewer does not highlight, let's get the into the that file ?
[12:15 PM PT]  Vincent.Sylvester ExtendedPhysics would need to re-implement the function that sends the XML file and add it's parts to it when enabled, which would require making that function overridable in the first place
[12:15 PM PT]  Vincent.Sylvester Adding highlighting for modules that may not be enabled will just confuse users Ubit is right
[12:15 PM PT]  Ubit Umarov: ok need go rl
[12:15 PM PT]  Gavin.Hird not any more than non-impletelmted LSL
[12:15 PM PT]  Ubit Umarov: cya later
[12:16 PM PT]  Cuga.Rajal If people are using these functions, they are probably advanced enough they dont need viewer support
[12:16 PM PT]  Gavin.Hird cheers Ubit
[12:16 PM PT]  Cuga.Rajal tc Ubit
[12:16 PM PT]  Vincent.Sylvester hands Cuga a blast shield
[12:16 PM PT]  Vincent.Sylvester You'll need that when the users find it
[12:16 PM PT]  Andrew Hellershanks: ok, Ubit. See you next week.
[12:16 PM PT]  Cuga.Rajal MrBlue added it before I knew!
[12:17 PM PT]  Cuga.Rajal sends blast shield to MrBlue
[12:17 PM PT]  Andrew Hellershanks: :)
[12:17 PM PT]  Andrew Hellershanks: On that note, I'll wrap up todays meeting.
[12:17 PM PT]  Andrew Hellershanks: Thank you all for coming. See you again next week.
Personal tools
About This Wiki