Chat log from the meeting on 2012-12-04

From OpenSimulator

Revision as of 14:50, 24 October 2015 by Fritigern (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
[11:01]  Hiro Protagonist is Online
[11:01]  Connecting to in-world Voice Chat...
[11:02]  Connected
[11:02]  Nebadon Izumi: ya I think a bunch of people kind of all logged in at once
[11:02]  Nebadon Izumi: seems ok now though
[11:02]  Justin Clark-Casey: hello folks
[11:02]  Hiro Protagonist: Hi Justin
[11:02]  BlueWall Slade: Hi JCC
[11:02]  Nebadon Izumi: hello
[11:02]  Vivian Klees: hello
[11:02]  Sarah Kline: hi all
[11:02]  Arielle Popstar: Hi Justin
[11:03]  Arielle Popstar: Hi Viv
[11:03]  Sarah Kline: justin
[11:03]  Joe.Radik @joeradik.dyndns.org: Hello everybody
[11:03]  Vivian Klees: hi Arielle
[11:03]  Arielle Popstar: Hi Joe
[11:04]  Nebadon Izumi: Key and I just updated all the plazas a few minutes ago
[11:04]  Nebadon Izumi: got the offline IM issue fixed
[11:04]  Nebadon Izumi: diva fixed it
[11:04]  Sarah Kline: oh good
[11:04]  Justin Clark-Casey: what issue was that?
[11:04]  Nebadon Izumi: i'll post up a new release in a bit
[11:04]  Nebadon Izumi: remember when we changed the groupofflinemessage setting?
[11:05]  Nebadon Izumi: apparantly doing that was completely disabling offline ims completely
[11:05]  Richardus Raymaker is Online
[11:05]  Justin Clark-Casey: oh, not good
[11:05]  Nebadon Izumi: she seemed to know instantly what was wrong and put in a patch
[11:05]  Vivian Klees: okay I get what you said in irc earlier
[11:06]  Richardus Raymaker is Offline
[11:07]  Sarah Kline: i got theses messages earlier today but it did correct itself later
[11:07]  Sarah Kline: [23:56] Second Life: Unable to rez: problem accessing inventory or locating assets
[00:28] Second Life: Notecard is missing from the database.
[11:07]  Sarah Kline: its been quite common
[11:07]  Nebadon Izumi: the new OpenSimulator version of Firestorm seems to be working pretty good
[11:07]  Joe.Radik @joeradik.dyndns.org: Why do I see someone...Nebadon, I think...as an Unknown User in chat.
[11:07]  Joe.Radik @joeradik.dyndns.org: ?
[11:07]  Richardus Raymaker is Online
[11:07]  Nebadon Izumi: might need to clear your cache Joe.Radik
[11:07]  Nebadon Izumi: stuff like that can occur because of hypergrid stuff
[11:07]  Nebadon Izumi: cache gets wierd
[11:07]  Joe.Radik @joeradik.dyndns.org: Thanks. Will try that.
[11:08]  Justin Clark-Casey: mmm, interesting
[11:08]  Richardus Raymaker: hmm strange. V3 viewers seems to place some huds wrong on the screen. offset changed
[11:08]  Sarah Kline: the search is a welcome return
[11:08]  Arielle Popstar: clearing cache is only resolving it for a couple logins then back to unknowns again
[11:08]  Nebadon Izumi: one wierd thing i noticed about Firestorm last night i completely purged any trace of firestorm from my computer and re-installed fresh the opensim version
[11:08]  Justin Clark-Casey: I wonder why cache would get wierd - UUIDs should be unique
[11:08]  Nebadon Izumi: and for some reason its not storing my user credentials
[11:08]  Nebadon Izumi: everytime i load viewer i have to type my name and password
[11:09]  Nebadon Izumi: other than that its working well, i reported it to Tank Master
[11:09]  Richardus Raymaker: hmm, justin. is there not a change with HG that you get double UUID's ?
[11:09]  Joe.Radik @joeradik.dyndns.org: Justin, I use Firestorm for several grids. May have something to do with it.
[11:09]  Sarah Kline: its saving mine ok
[11:09]  Nebadon Izumi: ya I am not sure why it happens, I think it might have something to do with how it looks up the name of the uuid
[11:09]  Richardus Raymaker: well, V3 still bad. UI scale bug still present
[11:09]  Justin Clark-Casey: richardus: I'm not sure what you mean, but the answer is probably no
[11:09]  Nebadon Izumi: and it checks its own grid maybe and caches that information as unknown user because it doesnt exist on the other grid
[11:09]  Nebadon Izumi: usually clearing the viewer cache helps
[11:10]  Justin Clark-Casey: nebadon: cool, nice to see that
[11:10]  Nebadon Izumi: one thing I miss from Imprudence viewer
[11:10]  Nebadon Izumi: is the ability to only clear the name cache
[11:10]  Richardus Raymaker: brb, to a normal viewer
[11:10]  Nebadon Izumi: and not have to dump the entire thing everytiem
[11:10]  Richardus Raymaker is Offline
[11:10]  Nebadon Izumi: need to ask Firestorm to consider instituting that, i think its helpful
[11:10]  Richardus Raymaker is Online
[11:10]  Justin Clark-Casey: The Unknown User actually does come from opensim though
[11:10]  Arielle Popstar: cant creeate a seperate script for that?
[11:10]  Nebadon Izumi: ya
[11:11]  Nebadon Izumi: though ive seen it happen just by logging directly into multiple grids too
[11:11]  Nebadon Izumi: like SL and Here
[11:11]  Nebadon Izumi: ive seen it happen that way too
[11:11]  Justin Clark-Casey: or at least, it's used as a string in the codebase if creator identification has an hg form but doesn't parse properly
[11:11]  Richardus Raymaker: the problem with viewers. many share still things between grids.
[11:11]  Nebadon Izumi: though its been a while since i really noticed it, i don't go to many other grids
[11:12]  Justin Clark-Casey: yeah, I know that for the vanilla sl viewer, caching is done per name rather than per uuid
[11:12]  Richardus Raymaker: afraid firestorm share cache between windows and opensim version.
[11:12]  Justin Clark-Casey: so I'm not surprised things can get confused if you use the same name between grids
[11:12]  Justin Clark-Casey: it really should be cached per uuid
[11:12]  Nebadon Izumi: ya
[11:12]  Nebadon Izumi: or per grid
[11:12]  Hiro Protagonist: no kidding
[11:12]  Justin Clark-Casey: yeah
[11:12]  Nebadon Izumi: each grid gets its own cache
[11:12]  Richardus Raymaker: also yopu need to be carefull with installing firestorm. its default installing in the windows directory.
[11:12]  Hiro Protagonist: nothing like an architecture that uses a different primary key in one place
[11:13]  BlueWall Slade: I think it was a mistake to use the name as an identifier anyway
[11:13]  BlueWall Slade: should just use uuid or an email address
[11:13]  Justin Clark-Casey: unfortunately linden are coding for a single grid :)
[11:13]  Nebadon Izumi: heh ya, i guess its one of those things that LL just never thought their viewer would be connecting to some other grid
[11:13]  Richardus Raymaker: nebadon, if i look in my windows user dir. i dont get the diea the have a seperate chace
[11:13]  BlueWall Slade: yep, lol
[11:13]  Justin Clark-Casey: or at the most, grids which share idnetical backend information
[11:13]  Joe.Radik @joeradik.dyndns.org: I use the same name on several grids. I've seen messages from one grid appear when logging into another.
[11:14]  Richardus Raymaker: well, firestorm could at least choice default a different installation directory for opensim
[11:14]  Nebadon Izumi: ya i always used the same name as well
[11:14]  Richardus Raymaker: maby also user dir
[11:15]  Sarah Kline: They said they will be automating something rich they just didnt have time this time
[11:15]  Master Dubrovna: Yeah Opensim Firestorm and SL Firestorm should be separate
[11:15]  Richardus Raymaker: Well for now people need to be carefull if the have windows version
[11:15]  Nebadon Izumi: Hopefully we'll see another Firestorm for opensim update soon
[11:15]  Richardus Raymaker: i hope LL fix that ui scale bug soon, so i can wave V1 bye bye
[11:16]  Justin Clark-Casey: having separate codebases is a recipe for much more work
[11:16]  Nebadon Izumi: and chasing lindens isnt?
[11:16]  Nebadon Izumi: lol
[11:16]  Mike Kayaker: I installed bot the LL and OS Mac versions, they land in the same application directory, but I just renamed one before installing the other.
[11:16]  Justin Clark-Casey: yes, but linden still have the whip hand when it comes to ppl working on the viewer
[11:16]  Joe.Radik @joeradik.dyndns.org: Nebadon, does mesh physics really work in the OS version? Their website says it's not included.
[11:16]  Nebadon Izumi: ya i tested it
[11:16]  Nebadon Izumi: it worked
[11:16]  Richardus Raymaker: i changed the installation dir before installatio
[11:16]  Vivian Klees: haha another game for cuteulala to create whack a linden
[11:17]  Sarah Kline: yes they still kept the old uploader in ours
[11:17]  Richardus Raymaker: mesh physics i hear soemthing weird with opensim
[11:17]  Richardus Raymaker: lol vivian
[11:17]  Sarah Kline: hehe i did this yesterday in sandbox
[11:17]  Richardus Raymaker: is there a topic about makeing this with beldner already ?
[11:18]  Nebadon Izumi: see Robert
[11:18]  Nebadon Izumi: this was uploaded with Firestorm
[11:18]  Claudia Sanz: neat
[11:18]  Robert Adams: is there a place in OSGrid to get some meshes to experiment with?
[11:18]  Nebadon Izumi: Reticulation
[11:19]  Sarah Kline: they simply have the LL havoc code in the SL version now
[11:19]  Nebadon Izumi: is my mesh sandbox
[11:19]  Nebadon Izumi: theres a landmark
[11:19]  Robert Adams: thanks much
[11:19]  Vivian Klees: was anything ever mentioned about mesh uploading getting around some of the time delays?
[11:19]  Richardus Raymaker: for now i kep useing zen for mesh
[11:19]  Justin Clark-Casey: Time delays?
[11:20]  Vivian Klees: some of the checks the viewer does
[11:20]  Dahlia Trimble is Online
[11:20]  Sarah Kline: i often have to relog
[11:20]  Sarah Kline: and then it works fine
[11:20]  Justin Clark-Casey: What checks are those
[11:20]  Justin Clark-Casey: ?
[11:21]  Sarah Kline: the delay is the uploading surely
[11:21]  Vivian Klees: really not sure but sometimes a say a 600kb file will take 3-4 minutes per check
[11:21]  Sarah Kline: heavier the mesh the longer takes to look at
[11:21]  Justin Clark-Casey: never experienced a check delay myself
[11:21]  Nebadon Izumi: ya I just think thats the opensource stuff they are using to handle the mesh
[11:21]  Sarah Kline: depends how clean your mesh is
[11:21]  Nebadon Izumi: Havok is highly optimized
[11:21]  Justin Clark-Casey: admittedly I've only used the ll viewer to upload
[11:21]  Nebadon Izumi: the opensource stuff the TPVs use is just not nearly as optimized
[11:22]  Nebadon Izumi: im sure in time it will improve
[11:22]  Sarah Kline: it was reverse engineered
[11:22]  BlueWall Slade is Offline
[11:22]  Vivian Klees: something then we need to mention then
[11:22]  Sarah Kline: quite fortunate they managed it
[11:22]  BlueWall Slade is Online
[11:22]  Richardus Raymaker: is that a mesh xmas hat nebadon ?
[11:23]  Nebadon Izumi: no dahlia made this many years ago
[11:23]  Nebadon Izumi: its sculpty
[11:23]  Richardus Raymaker: ok. that one i have
[11:23]  Robert Adams: I seriously need a Christmas hat
[11:23]  Arielle Popstar: Marcus made a mesh one i think
[11:23]  Nebadon Izumi: i'll send you one
[11:23]  Richardus Raymaker: robert, you maby know why the trains looked wrong on your screen ?
[11:23]  Dahlia Trimble: hi
[11:23]  Hiro Protagonist: I seriously need some kind of beverage heh\
[11:24]  Dahlia Trimble searches inventory for christmas hats...
[11:24]  Hiro Protagonist: coffee I think :)
[11:24]  Hiro Protagonist: brb
[11:24]  Arielle Popstar: Hi Dahlia
[11:24]  Robert Adams: I do not... Richardus
[11:25]  Nebadon Izumi: did you get the hat I sent Robert?
[11:25]  Robert Adams: what Richardus is talking about, yesterday he took me to his sandbox to show me his wonderfuly working trains...
[11:25]  Robert Adams: the physical trains appeared to wander off the track in my viewer even though they were actually still on the tracks
[11:26]  Nebadon Izumi: interesting
[11:26]  Robert Adams: I' musing the latest OS-Firestorm... I don't know why that would happen
[11:26]  Richardus Raymaker: i need to check it with firestorm myself.
[11:26]  Arielle Popstar: in ode or bullet?
[11:26]  Robert Adams: ode
[11:26]  Richardus Raymaker: ode
[11:26]  Dahlia Trimble: anyone want a santa hat?
[11:27]  Justin Clark-Casey: they wandered off the tracks for all viewers of the scene?
[11:27]  Richardus Raymaker: in singularity it looks fine, zen i dont see anything wrong. trying right now firestorm..
[11:28]  Robert Adams: all right! looking good
[11:28]  Nebadon Izumi: lol
[11:29]  Richardus Raymaker: robert, i dont see any differeny between singularity and latest firestorm here.
[11:30]  Robert Adams: Dahlia, is there any reason why the meshmerizer can't just use the base display mesh if there is no physical mesh?
[11:30]  Hiro Protagonist: /back
[11:30]  Robert Adams: well, it might just be me :)
[11:30]  Dahlia Trimble: well I guess it could
[11:30]  Nebadon Izumi: speaking of, last night i did a git clone of mono master 3.0.2
[11:31]  Nebadon Izumi: and opensimulator will compile and start now using mono 3
[11:31]  Richardus Raymaker: hehe
[11:31]  Richardus Raymaker: i only see a missing guide robert :O
[11:31]  Justin Clark-Casey: great!
[11:31]  Nebadon Izumi: but there are physics issues
[11:31]  Dahlia Trimble: I didnt add it before because didnt know which LOD to use
[11:31]  Justin Clark-Casey: not so great!
[11:31]  Nebadon Izumi: ya not so great
[11:31]  Nebadon Izumi: it seems that cut and hollow prims are not recognized
[11:31]  Nebadon Izumi: and physically they are the full prim
[11:31]  Justin Clark-Casey: that's rather wierd
[11:31]  Justin Clark-Casey: very wierd in fact
[11:31]  Nebadon Izumi: I think i had similar issue with Bulletsim
[11:31]  Nebadon Izumi: bulletsim testing didnt go as well
[11:32]  Simulator Version v0.5 shouts: OpenSim 0.7.5 Dev          583e441: 2012-12-04 08:36:54 -0800 (Unix/Mono)
[11:32]  Nebadon Izumi: i think maybe because the entire sim was 1 solid peice
[11:32]  Nebadon Izumi: from all the cut and hollow prims
[11:32]  Robert Adams: I dont' like to hear that :-(
[11:32]  Justin Clark-Casey: is this actually a 'proper' mono release or is this just the version signifier they current have on master?
[11:32]  Nebadon Izumi: i did a git clone from github
[11:33]  Nebadon Izumi: it was master as of like 14 hours ago
[11:33]  Richardus Raymaker: my sandbox use a git from 2-3 days ago
[11:33]  Justin Clark-Casey: interesting. When did you know it last wasn't working?
[11:33]  Nebadon Izumi: I did a git clone of libgdiplus as well
[11:33]  Nebadon Izumi: 2.10.9
[11:33]  Taarna Welles is Online
[11:34]  Nebadon Izumi: was last known good working version of mono
[11:34]  Nebadon Izumi: we are running 2.10.8 here
[11:34]  Nebadon Izumi: ya its very odd
[11:34]  Hiro Protagonist: Question for you Justin, diff topic: Has any progress been made with the versioning architecture of the add-ins?
[11:34]  Nebadon Izumi: even mono-sgen ran great
[11:34]  Nebadon Izumi: its just physics, particularly meshmerizer seemed to fail
[11:35]  Nebadon Izumi: physics worked well otherwise
[11:35]  Nebadon Izumi: just no hollows and cuts
[11:35]  Justin Clark-Casey: hiro: afaik that would be working now with all the version information updating
[11:35]  Dahlia Trimble: on regular prims?
[11:35]  Justin Clark-Casey: though I can't say for sure as I haven't been following that super closely
[11:35]  Hiro Protagonist: Excellent Justin, thanks
[11:35]  Nebadon Izumi: yes on cubes, cylinders, spheres
[11:35]  Nebadon Izumi: if you cut a prim 50% and hollow 90%
[11:36]  Dahlia Trimble: on mono?
[11:36]  Nebadon Izumi: when you collided with it, it was like the full prim still
[11:36]  Nebadon Izumi: ya this was on mono 3.0.2
[11:36]  Robert Adams: that's really odd.... the code is just if's
[11:36]  Dahlia Trimble: that is odd
[11:36]  Nebadon Izumi: i only tested on 64 bit
[11:36]  Nebadon Izumi: i can try again on 32 bit later
[11:37]  Hiro Protagonist: FYI I'll be testing percona server on my region server after the meeting; percona server (http://www.percona.com/software/percona-server) is a high performance drop-in replacement for mysql. I'll start up a forum thread when I have something to report, so keep an eye open if you're interested.
[11:37]  Robert Adams: you are a seeker of pain, Nebadon
[11:37]  Hiro Protagonist: LOL\
[11:37]  Justin Clark-Casey: I wonder if the mono fix was https://github.com/mono/mono/commit/b392a01a851570a8cdc60c340be098147da74421
[11:37]  Nebadon Izumi: lol I am excited to get on mono 3
[11:37]  Justin Clark-Casey: anyway, no matter - just glad that kind of resolved itself :)
[11:37]  Nebadon Izumi: be interesting if it was
[11:37]  Nebadon Izumi: ya, we are real close to being able to use mono 3
[11:38]  Nebadon Izumi: i hope the physics thing can resolve itself soon as well
[11:38]  Nebadon Izumi: the sim ran good with sgen
[11:38]  Justin Clark-Casey: that's really odd though - because it suggests some issue in the c# section
[11:38]  Dahlia Trimble: why do we want to rush to mono 3?
[11:38]  Justin Clark-Casey: mono-sgen is a much better gc
[11:38]  Nebadon Izumi: performance and memory usage should be considerably better on mono 3
[11:39]  Dahlia Trimble: well sgen wasnt much of an improvement
[11:39]  Nebadon Izumi: im certainly not suggesting we force everyone to use mono 3
[11:39]  Nebadon Izumi: i just think it would be good if we could
[11:39]  Richardus Raymaker: where still on .net 3.5 right ?
[11:40]  Nebadon Izumi: ya pretty much
[11:40]  Richardus Raymaker: .net 4 some mprovements to use ?
[11:40]  Nebadon Izumi: i imagine it is
[11:40]  Richardus Raymaker: wants .net4 :)
[11:41]  BlueWall Slade: the new mono has C# 5 support
[11:41]  BlueWall Slade: and supports asynch methods
[11:41]  Dahlia Trimble: so I noticed some conversations on IRC about latest firestorm not being able to upload meshes, but I didnt catch the whole conversation. Is it working?
[11:41]  BlueWall Slade: so a sleep in the scrips could wait in the background while other things run
[11:41]  Hiro Protagonist: yes, bad language in a warning about havoc not being available
[11:42]  Nebadon Izumi: ya it works, I mentioned my concerns about the confusion to Tank Master
[11:42]  Richardus Raymaker: Bluewall .net 4 or 5 ?
[11:42]  Nebadon Izumi: hopefully they can tone that down a bit in next version
[11:42]  Nebadon Izumi: explain better what they meant
[11:42]  Richardus Raymaker: because the sleep like you talk about sound sa bit microthreading for me.
[11:43]  Nebadon Izumi: when you go to upload a mesh for the first time it prompts you with an ok button that kind of stops you dead in your tracks like its flat out not going to work at all
[11:43]  Nebadon Izumi: but it still works as well as it previously did
[11:43]  Sarah Kline: lol it invites you to go to SL instead
[11:43]  Taarna Welles: @Sarah: Yep
[11:44]  Nebadon Izumi: I think they still have quite a bit to do on the OpenSim version
[11:44]  Sarah Kline: but all in all its working ok considering
[11:44]  Nebadon Izumi: this is the 1st one, we shouldnt expect anything less than a few problems
[11:44]  Nebadon Izumi: hehe
[11:44]  Justin Clark-Casey: so they have officially split now into havok and non-havok versions?
[11:44]  Nebadon Izumi: yes
[11:45]  Justin Clark-Casey: ok
[11:45]  BlueWall Slade: afk...
[11:45]  Robert Adams: There are a bunch of Firestore mods for Aurora also
[11:45]  Nebadon Izumi: http://wiki.phoenixviewer.com/fs_downloads_opensim
[11:45]  Nebadon Izumi: ya I think those are nothing real new
[11:45]  Nebadon Izumi: a lot of the older TPVs had that as well
[11:46]  Nebadon Izumi: some of those go back as far as Imprudence even
[11:48]  Robert Adams: IF the viewer is going to support it, varregions would be nice in OpenSim
[11:48]  Robert Adams: but it's not like I don't have anything else to do :)
[11:48]  Justin Clark-Casey: yeah, I think there are still many conceptual issues
[11:48]  Sarah Kline: be nice to have those
[11:48]  Nebadon Izumi: getting them working in grid mode is kind of complicated
[11:48]  Richardus Raymaker: to disrupt things a bit, anything new about the slow errors ? it seems for me fine on few region if i turn caps off
[11:48]  Justin Clark-Casey: e.g. how do they show up on the map, how do you stop collisions with other var regions, etc.
[11:48]  Nebadon Izumi: ya thats what I was wondering myself
[11:49]  Nebadon Izumi: I assume having neighbors is not really an option but i dunno
[11:49]  OtakuMegane Desu: Var regions? Mulltiples of 256 for dimensions.
[11:49]  OtakuMegane Desu: Or some other standard size
[11:49]  Justin Clark-Casey: otaku: no, any size, no standard
[11:49]  Hiro Protagonist is Offline
[11:49]  Justin Clark-Casey: nearest meter I suppose
[11:49]  Robert Adams: it's a virtual world... any defintion will do... 6 dimensions maybe ;-)
[11:49]  Richardus Raymaker: only thing you can do is check if the 256x256 sqm the region want to use is empty.
[11:49]  Sarah Kline: Maybe Rev figured out how to do it
[11:49]  OtakuMegane Desu: Can't realistically be done in grid mode. It's a nonoption outside of standalone or a remote island, simple as that.
[11:50]  Richardus Raymaker: i think vars can only registrate on the grid in 256x256 blocks for now
[11:50]  OtakuMegane Desu: It might be possible but I can't imagine it being anything but convoluted and error-prone
[11:50]  Sarah Kline: each tile on grid doesnt represent 256m
[11:50]  Sarah Kline: just a pointer to a var region maybe
[11:51]  Justin Clark-Casey: there are a lot of issues to overcome - so I'd like to see it really working in many situations
[11:51]  OtakuMegane Desu: If you eliminated the hard coordinate system then yes, though it'd become insanely hard to make a map if you do that
[11:51]  Justin Clark-Casey: and you also end up having to do things differently server side - perhaps you cant' specify region co-ords such as 1000,1000 anymore
[11:51]  Richardus Raymaker: i think it would be just fine to have mega's better working.. thats for most enopugh
[11:51]  Dahlia Trimble: probably a lot easier to make it multiples of 256
[11:51]  Justin Clark-Casey: unless you have a restriction that a region can still only 'start' at a 256m co-ord
[11:51]  Justin Clark-Casey: yes, a lot easier to do 256 multiplers
[11:51]  Justin Clark-Casey: multiples
[11:52]  Justin Clark-Casey: which is kind of what megaregions are, but without explicit viewer support
[11:52]  OtakuMegane Desu: I don't know too many people who honestly would have an issue with 256 multiple.
[11:52]  Hiro Protagonist is Online
[11:52]  Richardus Raymaker: i would say forgot vars, i dont see need for it and just use the 256 mega's
[11:52]  Richardus Raymaker: thats most used to
[11:52]  OtakuMegane Desu: Megaregions have their annoyances too. Like the base region you have to go to thena tp to where you actually aimed
[11:52]  Justin Clark-Casey: yeah
[11:53]  Justin Clark-Casey: to fix that probably requires both viewer and simulator changes
[11:53]  Arielle Popstar: bet it would be resolvable if someone gave megas some love
[11:53]  OtakuMegane Desu: Megaregions are a good solution for the current viewer limits but if the coding for variable regions can get into viewers, why use the kludge?
[11:53]  Richardus Raymaker: Well otaku, i think the solution for that is something like telehub
[11:53]  Richardus Raymaker: not sure if it works ....
[11:54]  Justin Clark-Casey: because doing full variable regions requires a lot of design change, not just coding change
[11:54]  Richardus Raymaker: var are a kludge compared to 256x256. because that rster is already used by the viewer.
[11:54]  Justin Clark-Casey: that's not to say it wouldnt' be good - but a lot of work
[11:54]  Justin Clark-Casey: and then you get incompatibility with other viewers, etc.
[11:55]  OtakuMegane Desu: But the viewers only recognize a single 256x256 space. Nthing larger. Thus the kludge megaregions use to fool it
[11:55]  Arielle Popstar: singularity dropped their support because it created other issues
[11:55]  Dahlia Trimble: dropped support for variable?
[11:55]  Arielle Popstar: according to one of their devs
[11:55]  Robert Adams: the other solution is to make region crossings smooth as butter
[11:56]  Dahlia Trimble: lol
[11:56]  Hiro Protagonist: do it Robert xD
[11:56]  Vivian Klees: region crossing isn't that good in SL
[11:56]  Hiro Protagonist: it used to be not that bad
[11:56]  Richardus Raymaker: Robert, that sounds impossible task
[11:56]  Justin Clark-Casey: robert: ha - I suspect that's... ambitious
[11:56]  Richardus Raymaker: unless you use mega's
[11:56]  Dahlia Trimble: its fine if you dont have any attachmehts
[11:57]  Robert Adams: what, you a vegan or something?
[11:57]  Justin Clark-Casey: dahlia: oh really?
[11:57]  Dahlia Trimble: well it is on some regions I have
[11:57]  Justin Clark-Casey: I haven't done a lot of testing in that area recently
[11:57]  Justin Clark-Casey: because that would suggest a problem in a specific area, which in principle one should be able to overcome
[11:57]  Richardus Raymaker: Dahlia. and now a region thats a bit further away from you.
[11:57]  Justin Clark-Casey: if attachment-less crossings are much better
[11:57]  Richardus Raymaker: simborders never got smooth like mega's
[11:57]  OtakuMegane Desu: I've never seen a 100% smooth region crossing anywhere under any circumstance. There's always something that has to be transferred
[11:58]  Dahlia Trimble: maybe its more scripted attachments
[11:58]  OtakuMegane Desu: It may be a fraction of a second but it's still there
[11:58]  Dahlia Trimble doesnt wear scripted attachments
[11:58]  Robert Adams was thinking about special casing region crossings on the same simulator... just pass structures rather than total serialization, etc
[11:58]  Richardus Raymaker: it most cost me i think 3-10 seconds before i crossed region to new simulator
[11:58]  Arielle Popstar: my understandin of avn resolution is some sort of prediction thing so the next region prepares itself for the handover
[11:58]  Hiro Protagonist: Robert Adams funny you should mention that
[11:59]  Justin Clark-Casey: robert: yeah
[11:59]  Hiro Protagonist: I was just trying to articulate a similar notion to myself
[11:59]  Dahlia Trimble: I have 9 regions on osgrid all on the same simulator, its not bad there at all
[11:59]  Justin Clark-Casey: prediction will start getting very complex
[11:59]  Justin Clark-Casey: since then you have to handle issues such as if the avatar doens't actually cross, etc.
[12:00]  Nebadon Izumi: border crossing is usually not bad for 1 avatar
[12:00]  Nebadon Izumi: try crossing 10 avatars at once
[12:00]  Nebadon Izumi: heh
[12:00]  Richardus Raymaker: Arielle, Secondlife still have no smooth simborders..
[12:00]  OtakuMegane Desu: lol
[12:00]  Dahlia Trimble: lol
[12:00]  Justin Clark-Casey: heh, never tried that
[12:00]  Nebadon Izumi: its not good usually
[12:00]  Hiro Protagonist: I was thinking that there might be certain variables involved that arent bound to the avatar that might be pre-calculated (for lack of a better description) and held on a sort of scoreboard
[12:00]  OtakuMegane Desu: Well, we're not really trying to fix SL either, so :P
[12:00]  Richardus Raymaker: aha i see mistake. you dont run more sims on 1 simulator..
[12:00]  Justin Clark-Casey: crossing between separate simulators will always be much worse
[12:01]  Justin Clark-Casey: might be practically impossible to get those to be 'smooth'
[12:01]  Richardus Raymaker: well, so the only solution for smooth big regions are mega's
[12:01]  Hiro Protagonist: another thing you could do
[12:01]  OtakuMegane Desu: Different instances on one server are usually very close to borders in a singlle instance but both are still imperfect
[12:01]  Dahlia Trimble: I vaguely remember Teravus telling me there was some zone near the edges of a region where it starts to share things
[12:01]  Hiro Protagonist: preload attachments
[12:01]  Justin Clark-Casey: yes, as I think with a mega everything is running on a single scene loop
[12:01]  Hiro Protagonist: you've got the child in every adjacent region
[12:01]  Justin Clark-Casey: or at least a single physics loop
[12:01]  Robert Adams: would require prediction, pre-staging and complex handshakes
[12:01]  Hiro Protagonist: so why serialize attachments at all
[12:02]  Robert Adams: like they do for virtual machine migration
[12:02]  Richardus Raymaker: only problem with mega's is maby the load you geet if it start to fill
[12:02]  Dahlia Trimble: script state needs to be serialized
[12:02]  Hiro Protagonist: right on, radams
[12:02]  Robert Adams: I've seen that done across machines almost instantaniously
[12:02]  Basil Sosides: hello, good Evening
[12:02]  Richardus Raymaker: robert, and useless bandwidth useage ?
[12:02]  Justin Clark-Casey: hello basil
[12:02]  OtakuMegane Desu: If megaregions could allow a single direct tp to the chosen coords and not have every region named I'd be content with them.
[12:02]  Hiro Protagonist: greeetings
[12:02]  Dahlia Trimble: hi
[12:03]  BlueWall Slade: \o
[12:03]  Richardus Raymaker: ANd parcels, sound and a few ore things. but the are still more easy to fix i think then make var's
[12:03]  Hiro Protagonist: whats the deal with vars? what are they?
[12:04]  Richardus Raymaker: lol hiro, that can give 2 answhere's :)
[12:04]  BlueWall Slade: primshapes, primitems
[12:04]  Sarah Kline: you just enter what size you want your region to be
[12:04]  Arielle Popstar: aurora's answer tyo megas
[12:04]  Hiro Protagonist: ahhhh variable size regions, gotcha
[12:04]  Justin Clark-Casey: alright, I need to bounce out,
[12:04]  BlueWall Slade: see ya JCC
[12:05]  Hiro Protagonist: me too, much work to be done
[12:05]  Sarah Kline: Bye Justin
[12:05]  Vivian Klees: tc Justin
[12:05]  Richardus Raymaker: bye justin
[12:05]  Justin Clark-Casey: thanks for the conversation, folks
[12:05]  Arielle Popstar: waves
[12:05]  Taarna Welles: Bye Justin
[12:05]  Hiro Protagonist: me waves
[12:05]  Hiro Protagonist is Offline
[12:05]  Justin Clark-Casey waves
Personal tools
General
About This Wiki