Chat log from the meeting on 2020-11-17

From OpenSimulator

Jump to: navigation, search

[11:04] Gavin.Hird @grid.xmir.org:8002: Hi Andrew
[11:04] Andrew Hellershanks: Hello, everyone.
[11:04] Arielle Popstar: you see us sitting aqbove the chairs Kay?
[11:04] Gavin.Hird @grid.xmir.org:8002: only you Arielle
[11:04] Kayaker Magic: Just you Arielle
[11:05] Gavin.Hird @grid.xmir.org:8002: steevithak too
[11:05] Kayaker Magic: I'm running on a pre-release of FS 6.4.12, it may have broken voice again...
[11:06] Andrew Hellershanks: Kayaker, that is a more recent version than I have.
[11:06] Gavin.Hird @grid.xmir.org:8002: try turn off your AO sit override if on
[11:06] Kayaker Magic: Why would that make her see me sitting high?
[11:06] Arielle Popstar: trying it both ways
[11:06] Andrew Hellershanks: Hello, Ubit.
[11:06] Ubit Umarov: hi
[11:06] Kayaker Magic: Interesting tatoos, Ubit
[11:06] Gavin.Hird @grid.xmir.org:8002: Hi Ubit
[11:07] Ubit Umarov: :)
[11:07] Arielle Popstar: trying it both ways
[11:07] Arielle Popstar: hi
[11:07] Arielle Popstar: different then last week whatever it is
[11:07] Arielle Popstar: sits been played with Ubit?
[11:07] Ubit Umarov: sits?
[11:08] Gavin.Hird @grid.xmir.org:8002: Arielle is positioned quite off when she sits here
[11:08] Ubit Umarov: no idea
[11:08] Ubit Umarov: no code changes on those
[11:08] Gavin.Hird @grid.xmir.org:8002: did you update your viewer since last week Arielle?
[11:08] Arielle Popstar: yes
[11:09] Arielle Popstar: new beta
[11:09] Gavin.Hird @grid.xmir.org:8002: then you've got a beta sit
[11:09] Ubit Umarov: maybe a hover issue again?
[11:09] Arielle Popstar: but if you see me sitting off center then your viewer too
[11:10] Ubit Umarov: well siting back is not a hover issue
[11:10] Ubit Umarov: well sit works for me
[11:10] Kayaker Magic: I have some good news: Seth Nygard saw me whining about Vivox voice in Linux FS. He says it is because the vivox module was compiled in 32 bit mode. It just needs some missing dependancies to run, instructions to do this at https://wiki.firestormviewer.org/firestorm_in_64-bit_ubuntu_1804
[11:10] Arielle Popstar: ahh was an ao under my other one
[11:11] Gavin.Hird @grid.xmir.org:8002: lol
[11:11] Gavin.Hird @grid.xmir.org:8002: nested AOs
[11:11] Kayaker Magic: This has voice working for me everywhere but here.... might be because I upgraded to a newer FS last night.
[11:11] Ubit Umarov: dunno
[11:11] Ubit Umarov: the prob i see with vivox is private ims
[11:11] Ubit Umarov: local works
[11:12] Kayaker Magic: Now I see you sitting properly Arielle! What did you do?
[11:12] Ubit Umarov: and private ims is viewers things
[11:12] Gavin.Hird @grid.xmir.org:8002: she removed the double AO
[11:12] Ubit Umarov: they talk directly with vivox and do peer to peer
[11:13] Ada.Radius @grid.kitely.com:8002: Hi everyone sorry i'm late
[11:13] Ubit Umarov: kay it was a keyboard to chair interface issue
[11:13] Ubit Umarov: :p
[11:13] Gavin.Hird @grid.xmir.org:8002: Hi Ada
[11:13] Ubit Umarov: hi
[11:13] Ubit Umarov: so what news do you have about opensim?
[11:14] Andrew Hellershanks: There was an assortment of changes made this past week.
[11:14] Andrew Hellershanks: Kayaker, yay!
[11:14] Ubit Umarov: oh? tell us about them andrew
[11:14] Ubit Umarov: (hihihi)
[11:14] Arielle Popstar: @ Kay- an ao under my regular one which was still turned on
[11:15] Andrew Hellershanks: Kayaker, I thought you should have been able to get voice working.
[11:15] Kayaker Magic: Yeah, every time I re-boot Ubuntu, it decides to use the mic on the WEBcam that doesn't work, even though I keep setting the analog mic as the default.
[11:15] Andrew Hellershanks: Ubit, I often can't summarize the changes easily due to lack of info in the commit messages.
[11:15] Arielle Popstar: Hi Ada
[11:15] Ubit Umarov: well code changes..
[11:16] Kayaker Magic: I especially like the commit messages that just say "oops"
[11:16] Ubit Umarov: a fix on language preferences.. code was seening that as private always
[11:16] Arielle Popstar: yes informative those
[11:17] Andrew Hellershanks: I see one change that looks like it meant to improve the avoidance of having usernames appear as Unknown...
[11:17] Kayaker Magic: Object Oriented Programming System?
[11:17] Ubit Umarov: and changes adding more information about region current grid
[11:17] Andrew Hellershanks: Kayaker, one of the recent commits says "small changes". :)
[11:17] Ubit Umarov: to try help viewers to do hg tps
[11:17] Arielle Popstar: same with "small changes"
[11:17] Arielle Popstar: very helpful for troubleshooting
[11:18] Gavin.Hird @grid.xmir.org:8002: the small change is something the CPU hardly notice
[11:18] Ubit Umarov: that may had mean some changes on ini files
[11:18] Ubit Umarov: also added more checks on the values typed on those
[11:18] Andrew Hellershanks: Some changes appear to be related to getting some settings from the ini files. Particularly the grid info.
[11:18] Arielle Popstar: why not just say so then so people will look?
[11:18] Ubit Umarov: old code could send totally broken urls to viewers
[11:19] Ubit Umarov: ie waht you typed on inis was sent directly to viewers
[11:19] Ubit Umarov: bad karma
[11:19] Ubit Umarov: but still like that on a ton of palces
[11:20] Ubit Umarov: now does more checks like if the host is a valid host spec, in same cases if does have dns resolve
[11:20] Andrew Hellershanks: Ubit, in that "small change" commit the last name used to end with a ) to match the ( at the start of the first name. The commit now loses the closing ) as the last name is just set to the empty string.
[11:20] Arielle Popstar: is this for hypergrid jumps?
[11:20] Ubit Umarov: hmm i made a typo on that andrew ?
[11:20] Arielle Popstar: oops
[11:21] Andrew Hellershanks: :)
[11:21] Ubit Umarov: im asking :p
[11:21] Andrew Hellershanks: Yes, You left the opening ( so the first name should have been updated to end in ).
[11:21] Arielle Popstar: is that why my name in Hgluv account is now Arielle Arielle?
[11:22] Ubit Umarov: yeah seems i forgot to remove the (
[11:22] Kayaker Magic: Isn't Hgluv running a different branch of OpenSim that Must Not Be Named here?
[11:22] Arielle Popstar: well hgluv people dont see the prob, only the main branch/trunk people do
[11:23] Ubit Umarov: on home grids name should be the normal first last
[11:23] Ubit Umarov: but after HG things may get confused and show hg name, even on home grid
[11:23] Arielle Popstar: thats not my hg name though
[11:24] Arielle Popstar: i am arielle popstar in all grids
[11:24] Ubit Umarov: ahh sorry
[11:24] Ubit Umarov: that was a bug
[11:24] Ubit Umarov: fixed weeks ago
[11:24] Gavin.Hird @grid.xmir.org:8002: I've seen the double name a couple times here too, but after a while, it switched back to single name, so maybe some delayed fetch/update of HG names?
[11:24] Ubit Umarov: actually was on for a few days only
[11:25] Ubit Umarov: typical c&p issue
[11:25] Arielle Popstar: ok
[11:25] Ubit Umarov: did it and forgot to change first to last
[11:25] Ubit Umarov: :p
[11:25] Arielle Popstar: i have a question about Gloebits as explained in this blog: https://blog.fireandicegrid.net/upgrade-to-globit-and-opensim-version/
[11:26] Ubit Umarov: yeah been at fireandice helping on that
[11:26] Ubit Umarov: with Bill help also
[11:26] Arielle Popstar: will crashing be an issue for others?
[11:26] Ubit Umarov: we have it on mantis also
[11:26] Arielle Popstar: is that a problem from the gloebit devs?
[11:26] Ubit Umarov: Sara and bill did report 6.3.9 crashes
[11:27] Ubit Umarov: but problem is the module
[11:27] Ubit Umarov: it must be compiled with opensim
[11:27] Ubit Umarov: the dll just smokes
[11:28] Arielle Popstar: do it will be a required update to keep it all compatible?
[11:28] Ubit Umarov: well that post tells almost all
[11:28] Ubit Umarov: on code that is changing just using a dll compiled for other code may fail
[11:29] Ubit Umarov: the dll they provide for 0.9.2 now crashes
[11:29] Ubit Umarov: needs to be compiled with opensim
[11:29] Ubit Umarov: in fact the miracle was dlls working on dif code base
[11:30] Ubit Umarov: that will work on a stable release
[11:30] Ubit Umarov: modules are hard dlls
[11:31] Ubit Umarov: they both add code for opensim to use and they also use opensim code
[11:31] Ubit Umarov: a dll that only provides code, is more resilent
[11:31] Ubit Umarov: like our mysql connector etc
[11:33] Ubit Umarov: btw seems fs 6.3.9 no longer crashs on proper gloebits module
[11:34] Ubit Umarov: did i answer? :)
[11:34] Andrew Hellershanks: You certainly did.
[11:34] Kayaker Magic: A couple of questions for Gavin: In the viewer preferences on FS, in the Rendering page, there is an option to restrict maximum texture resolution to 512px. How does that work? Does the server downsample 1024 textures before sending them?
[11:35] Ubit Umarov: server does nothing like that
[11:35] Gavin.Hird @grid.xmir.org:8002: To ask me about FS settings is really not the right address but...
[11:35] Ubit Umarov: but jpeg2k has resoution things
[11:35] Kayaker Magic: So the server sends the 1024px image, wasting bandwidth, then the viewer dosnsamples?
[11:36] Ubit Umarov: No
[11:36] Gavin.Hird @grid.xmir.org:8002: SL now supports both 512 and 1024 avatar textures, so it is possibly a setting to not get the 1024 version on machhines with lesser capabilities
[11:36] Gavin.Hird @grid.xmir.org:8002: as they us 4x the memory per texture
[11:36] Ubit Umarov: bc j2k format, viewers do not ask for the full asset file
[11:36] Gavin.Hird @grid.xmir.org:8002: use*
[11:37] Kayaker Magic: Or it could be an old setting to save on video texture memory.....
[11:37] Ubit Umarov: they ask for start of file..
[11:37] Ubit Umarov: decode its details
[11:37] Gavin.Hird @grid.xmir.org:8002: what is the settings name?
[11:37] Ubit Umarov: the ask for the parts that do have the resolution they need
[11:37] Arielle Popstar: wouldnt that make everyones experience better if the texture size was limited?
[11:38] Ubit Umarov: never will ask for 1024px anyways
[11:38] Ubit Umarov: unless you zoom very close on a large object
[11:38] Ada.Radius @grid.kitely.com:8002: 512x512 is too limited, compared with other games. But 1024 used on BoM loads too slow.
[11:38] Arielle Popstar: i thought we had 1024 tattoo bakes?
[11:38] Kayaker Magic: This is a checkbox in the Rendering page of Preferences "Restrict maximum texture resolution to 512px (64bit only, requires Restart)
[11:38] Ubit Umarov: yeah rebake is now a pain on my net
[11:38] Gavin.Hird @grid.xmir.org:8002: ok
[11:38] Ada.Radius @grid.kitely.com:8002: we do have tattoo bakes that can be 1024. if you're willing to wait
[11:39] Ubit Umarov: 10 1024px textures take a while to upload on 800kbps
[11:39] Arielle Popstar: can one limitit viewer side?
[11:39] Ubit Umarov: no option
[11:39] Arielle Popstar: why would you have to upload?
[11:39] Ubit Umarov: all are 1024px except eyes still 128px
[11:40] Ubit Umarov: bc that is how rebakes work duhhh
[11:40] Arielle Popstar: or from server you mean
[11:40] Ubit Umarov: TO the server
[11:40] Ubit Umarov: gezz
[11:40] Arielle Popstar: isnt the texture already on the server?
[11:40] Ubit Umarov: download is per dischard level
[11:40] Ada.Radius @grid.kitely.com:8002: There's also a question about how, exactly Bakes on Mesh is identifying the uuid of the textures in the Universal or Skin or whatever layers. I'm working with Oren at Kitely on it - Kitely's going to have to change the way they handle deliveries from the marketplace
[11:40] Gavin.Hird @grid.xmir.org:8002: It was commited to the repository 2 years ago
[11:41] Ubit Umarov: No it is now
[11:41] Gavin.Hird @grid.xmir.org:8002: so not an entirely new setting
[11:41] Ubit Umarov: bakes are made per user viewer
[11:41] Ubit Umarov: hmm ?
[11:41] Gavin.Hird @grid.xmir.org:8002: commit message is: Add option to restrict display of inworld textures to 512px for 64bit viewers (default on 32bit already)
[11:41] Ubit Umarov: what you mean Ada?
[11:42] Arielle Popstar ish confused
[11:42] Ubit Umarov: don't see why changes
[11:42] Ubit Umarov: except on the meshes
[11:42] Arielle Popstar: kitely is a way different branch anyway
[11:43] Ada.Radius @grid.kitely.com:8002: Kitely marketplace, Oren said, changes the uuid of the textures on delivery, to prevent non-export assets from exporting to other grids. I don't understand why it would matter - if I set up BoM mesh, I can change the layers and the mesh still finds the right texture and it bakes on.
[11:43] Ubit Umarov: well transfer may ltry to fecth those
[11:43] Ubit Umarov: think i took them out.. will recheck
[11:43] Ubit Umarov: ahh their own issue
[11:43] Ubit Umarov: simple fix
[11:44] Ubit Umarov: they just need to compare to the magic uuids and dont change those
[11:44] Kayaker Magic: I think BoM had some magic UUIDs, Kitely has to test for those now.
[11:44] Kayaker Magic: what he said
[11:44] Ada.Radius @grid.kitely.com:8002: advantage to BoM - greatly reduced avatar complexity, and as far as I can tell, once the mesh is baked, loads very fast
[11:44] Arielle Popstar: on local grid?
[11:45] Ubit Umarov: one baked cna replace a ton of other textures
[11:45] Ada.Radius @grid.kitely.com:8002: any grid i've been to so far. Right now I'm wearing one Universal layer, that's texturing my avatar body, ears and tail
[11:45] Ubit Umarov: and onion(?) meshes
[11:45] Ada.Radius @grid.kitely.com:8002: ya
[11:45] Ada.Radius @grid.kitely.com:8002: no more onions. mwahaha
[11:45] Gavin.Hird @grid.xmir.org:8002: I have yet to find a male BOM mesh that does not have a lot of feminine curves
[11:46] Arielle Popstar: does ruth or roth v3 have a alpha hud?
[11:46] Ada.Radius @grid.kitely.com:8002: Ruth and Roth are BoM now, though apparently the universal layer, which I wanted, so we could include fingernail and toe nail textures, interfered with one of the HUDs
[11:46] Ada.Radius @grid.kitely.com:8002: but the av I have on now doesn't need a HUD
[11:46] Ubit Umarov: and BoM magic is just that
[11:46] Ubit Umarov: special magic uuids
[11:46] Arielle Popstar: there were people in s/l asking for it
[11:47] Arielle Popstar: in the ruth/roth group
[11:47] Ubit Umarov: where the mesh is attached to a avatar, viewer sees the uuid and applies respective baked
[11:47] Ubit Umarov: simple (in theory )
[11:47] Ada.Radius @grid.kitely.com:8002: it's on Ruth2v4 and Roth2v2 mesh now. For the skin textures anyway
[11:47] Kayaker Magic: Kitely says magic UUIDs are fixed on their end, waiting for the next rollout.
[11:48] Ada.Radius @grid.kitely.com:8002: Needs to be tested still
[11:48] Ada.Radius @grid.kitely.com:8002: if it works I'll stop asking questions about what ya mean by "magic"?
[11:48] Kayaker Magic: LOL
[11:49] Ada.Radius @grid.kitely.com:8002: :)
[11:49] Ubit Umarov: magic bc they do magic :p
[11:49] Ubit Umarov: in several cases also no need for alpha, another small saving
[11:50] Ada.Radius @grid.kitely.com:8002: when I started working on this av, before I started mucking with the universal layer, the av complexity was 15K ish, Just transferring three textures to BoM (eyes, ear tail) reduced it by around 3K
[11:50] Kayaker Magic: There are a bunch of "fixed" UUIDs in SL and OpenSim that must always refer to certain textures, animations, etc. Apparently BoM added some more of these.
[11:50] Ada.Radius @grid.kitely.com:8002: 11 of the, yes
[11:50] Ubit Umarov: err no
[11:50] Ubit Umarov: the 11 are more special
[11:50] Kayaker Magic: more magical?
[11:50] Ada.Radius @grid.kitely.com:8002: 11 BoM textures, one for each BoM layer? they show when you rez the out on the ground
[11:51] Ubit Umarov: No
[11:51] Ubit Umarov: each for a baked type
[11:51] Ada.Radius @grid.kitely.com:8002: right
[11:51] Ubit Umarov: layer is confusing :)
[11:51] Ada.Radius @grid.kitely.com:8002: which can come from any of a number of layers - skin, eyes, clothing etc
[11:52] Selby.Evans @grid.kitely.com:8002: Hi everyone
[11:52] Ada.Radius @grid.kitely.com:8002: Hi Selby
[11:52] Ubit Umarov: hi selby.Evans
[11:52] Gavin.Hird @grid.xmir.org:8002: Hello Selby
[11:52] Kayaker Magic: Selby, you still on Daylite Wastings Time?
[11:52] Ubit Umarov: each baked corresponds to a body part mostly
[11:52] Andrew Hellershanks: Hello, Selby.
[11:52] Ubit Umarov: except the aux ones
[11:52] Ubit Umarov: not totally sure how those work
[11:52] Ada.Radius @grid.kitely.com:8002: So I'm wearing a universal layer, with skin, eyes, tail, ear and a 4x4 alpha for hair. On top of that, half a dozen layers for the Arcadia space suit clothing
[11:53] Ada.Radius @grid.kitely.com:8002: the tail, ear, and um, neck gap prim, take the three AUX slots
[11:53] Ubit Umarov: and on each body part there are wearables types
[11:53] Ada.Radius @grid.kitely.com:8002: yes
[11:53] Ubit Umarov: that help define rendering order and rules
[11:54] Ada.Radius @grid.kitely.com:8002: can be quite a few, but adding them doesn't change av complexity, because they get baked down
[11:54] Ada.Radius @grid.kitely.com:8002: yes
[11:54] Ubit Umarov: like skin, tatoo shirt
[11:54] Ada.Radius @grid.kitely.com:8002: yes
[11:54] Arielle Popstar: do textures add to complexity anyway?
[11:54] Ubit Umarov: then some of those can have several textures
[11:54] Ubit Umarov: like several shirts
[11:55] Ada.Radius @grid.kitely.com:8002: not in BoM, because they get baked together, so what rezzes are only, at most, the 11 BoM slots. And I repeated textures in some of those
[11:55] Ubit Umarov: so baking is a bit complex process
[11:55] Ada.Radius @grid.kitely.com:8002: ya. the puzzle is how baking is identifying the uuids it needs to find
[11:55] Arielle Popstar: why does the base BoM texture have to be yellow and blue?
[11:55] Ada.Radius @grid.kitely.com:8002: that made the Kitely Market delivery system break
[11:56] Ubit Umarov: well that is part of the wearables system
[11:56] Ada.Radius @grid.kitely.com:8002: It's what the viewer assigns the colors for each of the 11 BoM slots
[11:56] Ada.Radius @grid.kitely.com:8002: eyes are pink upper is blue etc
[11:56] Gavin.Hird @grid.xmir.org:8002: Does it even matter Arielle? They are just placeholders
[11:56] Selby.Evans @grid.kitely.com:8002: OsTeleportObject works beautifully in grid -- in my sim was almost instant tp with me sitting on it (Answering question from last week)
[11:56] Ada.Radius @grid.kitely.com:8002: excellent
[11:56] Ubit Umarov: and im talking abotu baking.. that is only relevant to own avatar viewer
[11:56] Andrew Hellershanks: Thank you for the update, Selby.
[11:56] Ada.Radius @grid.kitely.com:8002: and everyone else's viewer in the sim
[11:57] Ubit Umarov: other viewers only see baked
[11:57] Arielle Popstar: it looks horrible on older viewers
[11:57] Ada.Radius @grid.kitely.com:8002: ya
[11:57] Ada.Radius @grid.kitely.com:8002: but can be fabulous in the newer ones, because I get to add normal and specular maps, plus the three kinds of alpha
[11:57] Gavin.Hird @grid.xmir.org:8002: BOM is not designed for older viewer
[11:58] Ubit Umarov: and not on 0.9.2 they don't get any information about the other textures used on baking
[11:58] Ubit Umarov: or werables details
[11:58] Arielle Popstar: yes but why not make it so the base layer is a skin texture
[11:58] Gavin.Hird @grid.xmir.org:8002: get with the program and upgrade
[11:58] Ada.Radius @grid.kitely.com:8002: it depends on the mesh UV mapping
[11:58] Ubit Umarov: ( so copybots have a little more trouble )
[11:58] Arielle Popstar: cant with Lumiya
[11:58] Ada.Radius @grid.kitely.com:8002: that too :)
[11:58] Arielle Popstar: no old copybots, every one uses the newer ones
[11:58] Ubit Umarov: fun.. one was telling me how bad BoM is
[11:58] Ubit Umarov: deep reason.. copybot fail :p
[11:59] Gavin.Hird @grid.xmir.org:8002: hehe
[11:59] Ada.Radius @grid.kitely.com:8002: I can distribute a universal layer, a different cat texture, for example, for the same mesh
[11:59] Arielle Popstar: that was only a fail for 5 months
[11:59] Ubit Umarov: they can't recover all data
[11:59] Kayaker Magic: poor babies!
[11:59] Ada.Radius @grid.kitely.com:8002: and I don't have to load a HUD with my textures
[11:59] Andrew Hellershanks: hehe
[11:59] Ubit Umarov: my get a static look only
[11:59] Arielle Popstar: the lab people used that justtification too but if they been paying attention they would know it was a very short term fail
[11:59] Ubit Umarov: may
[12:00] Ubit Umarov: but fails.. bc those are bakes...
[12:00] Ubit Umarov: and no source files to recreate them
[12:00] Ada.Radius @grid.kitely.com:8002: yesssss
[12:00] Ubit Umarov: so yeah bom is bad :p
[12:00] Arielle Popstar: in the end the only viewerr adversely affected was the one android viewer in existance
[12:00] Ubit Umarov: ( for those )
[12:01] Arielle Popstar: half the freaking world on android and we dont have a working viewer
[12:01] Arielle Popstar: meaning the lab
[12:01] Ubit Umarov: well on 0.9.2 i did reduce some info opensim was sending
[12:01] Ubit Umarov: including the viewer version
[12:01] Ubit Umarov: that is gone
[12:01] Ada.Radius @grid.kitely.com:8002: plus I'm having fun making very small armatures for the mesh, that will use normal animations, and some morphs and Appearance sliders work
[12:02] Ubit Umarov: oh found a old blender file that seems to have and do all avatar morphs
[12:02] Ada.Radius @grid.kitely.com:8002: like this
[12:03] Kayaker Magic: A viewer question I suspect nobody knows the answer to: The binary mesh files LL used for the default avatar have both a list of normals and a list of binormals. Anyone know if those binormals are actually used? No modern 3D file format has those.
[12:03] Ada.Radius @grid.kitely.com:8002: @ubit which file?
[12:03] Ubit Umarov: they are
[12:03] Kayaker Magic: Have a link to that old Blender file Ubit?
[12:03] Ubit Umarov: and modern also use them
[12:03] Ubit Umarov: but they can be restored from other data
[12:04] Ubit Umarov: so not necessary included on transfer files
[12:04] Ubit Umarov: a simple case is just triangles normals
[12:05] Ubit Umarov: in most cases no need for them
[12:05] Ubit Umarov: bc normal is included on the triangle vertices and their order
[12:06] Ubit Umarov: they mb include for special cases
[12:06] Ubit Umarov: where the triangle normal is special
[12:06] Ubit Umarov: same for vextices normals, etc
[12:06] Ubit Umarov: and vertices
[12:06] Ada.Radius @grid.kitely.com:8002: you mean triangle face normals?
[12:07] Ubit Umarov: yes
[12:07] Ada.Radius @grid.kitely.com:8002: (I thought only vert normals were used inworld) OK
[12:07] Ubit Umarov: physics uses face :)
[12:07] Ada.Radius @grid.kitely.com:8002: ahhh
[12:07] Ada.Radius @grid.kitely.com:8002: OK
[12:07] Ubit Umarov: ubode gets them on the fly from the vertices
[12:08] Kayaker Magic: So LL calculated the binormals and wasted file space and memory storing them to save a little bit of calculation. Moore's Law would make that the wrong decision.
[12:08] Ubit Umarov: it is a balance btw math cost versus memory access cost
[12:08] Ada.Radius @grid.kitely.com:8002: I've been using Data Transfer Modifiers to average out the vert normals aka face corner normals. I don't need to change that then
[12:08] Ubit Umarov: it is more complex than that kayaker
[12:08] Ubit Umarov: those impact light etc
[12:09] Ubit Umarov: may not be possible to derive all from the other data
[12:09] Ubit Umarov: and if is possibel may be very slow
[12:10] Ubit Umarov: guess stiching to other parts not present on same file also
[12:10] Kayaker Magic: Well, DAE files have no binormals and if NEEDED they must be calculate on load or on use. So it must be possible.
[12:10] Ubit Umarov: how neck blendes with torsi
[12:11] Ubit Umarov: in terms of light etc
[12:11] Ada.Radius @grid.kitely.com:8002: e.g. not very well
[12:12] Ubit Umarov: who told you dae does not have binormals??
[12:12] Arielle Popstar: waves
[12:12] Ada.Radius @grid.kitely.com:8002: Kayaker's been looking at the llm files and the dae files
[12:12] Kayaker Magic: I've looked at DAE files from Blender, it has normals only.
[12:13] Ubit Umarov: ie you looked to specific dae files..
[12:13] Kayaker Magic: I'm having to calculate binormals to make LLM files (Linden Lab Mesh)
[12:13] Ada.Radius @grid.kitely.com:8002: Maya might have them?? I don't know and don't want to pay to find out
[12:14] Gavin.Hird @grid.xmir.org:8002: The reference is Maya and not Blender
[12:14] Ada.Radius @grid.kitely.com:8002: Ya. needs to be converted to Blender down the road. open source
[12:15] Ubit Umarov: how are you making the llm ?
[12:15] Andrew Hellershanks: We are at quarter past. If someone has a question who also needs to leave soon, now would be a good time to ask.
[12:15] Kayaker Magic: I'm just saying: Blender DAE files have no binormals, Blender DAE mesh files load fine in OpenSim, therefore binormals are not necessary for mesh. Viewer or somebody may be calculating them for us...
[12:16] Ubit Umarov: loaded dae meshes are not llm files
[12:16] Ada.Radius @grid.kitely.com:8002: we know
[12:16] Ubit Umarov: they are a limited subset
[12:16] Ubit Umarov: so how are u making the llm ?
[12:17] Kayaker Magic: And it is a shame that LL stared over again, instead of just doing mesh one way from the start.
[12:17] Kayaker Magic: Writing code that reads DLL and writes LLM
[12:18] Kayaker Magic: I'm tempted to write all <0,0,1> vectors as the binormals and watch what goes wrong.
[12:19] Andrew Hellershanks: Any more questions/comments about (bi-)normals before we wrap up todays meeting?
[12:20] Andrew Hellershanks: Or any other last minute questions/comments?
[12:20] Ubit Umarov: think i once created them for our test viewer..
[12:20] Ubit Umarov: but long forgotten :)
[12:20] Ada.Radius @grid.kitely.com:8002: nope I gotta go check out my new boat house. later later and thank you, I got another insight today
12:21] Andrew Hellershanks: ok, Ada. See you next week.
[12:21] Ada.Radius @grid.kitely.com:8002: bye!
[12:22] Andrew Hellershanks: If there are no other last minute items for today that will conclude todays meeting.
[12:22] Andrew Hellershanks: Thank you all for coming. See you again next week.
[12:22] Ubit Umarov: and not exporting seems to be a blender issue
[12:22] Ubit Umarov: people complaing also for fbx etc

Personal tools
General
About This Wiki