Content Creation Improvement Informal User Group/Archive/July 10th, 2012 Meeting

From Second Life Wiki
Jump to: navigation, search
15:01  Geenz Spad  welcome everyone to today's meeting!
  Keli Kyrie  Hello everyone. :)
  Asha  Hi Keli
  woozl Selkit T. Woozl  Geenz, your chair seems hungry
  Geenz Spad  we're going to be taking a slightly different approach to today's meeting, wherein the first 15 minutes will be dedicated to JIRA discussion
  Elie Spot  wow that stage is like 8 miles away from the commoners
15:02  Geenz Spad  I blame Selkit :p
  woozl Selkit T. Woozl   :3
  Asha  lol
  woozl Selkit T. Woozl  Always blame the mustelid. :c
  Braydon Randt  i can still hit um with my shoe
  Geenz Spad  we'll move the chairs up on the next one
  woozl Selkit T. Woozl  Hey! I'm not Dubya or Saddam. No shoes! D:
  Geenz Spad  so anyways, anyone have any JIRAs related to content creation up?
  Yuzuru  Hello Nalates.
  Nal  Hey Yuz
  Asha  Heya Nal
  Nal  Asha :)
  Martin RJ  hi all
15:03  Siddean Munro  I have a really really really old one that will never get fixed! :)
  tauber Paola von Tauber  well i have one but it's not on the agenda: https://jira.secondlife.com/browse/VWR-27189
  woozl Selkit T. Woozl  I have a JIRA from 2007 still alive, Siddean. :C
    Completely not content-related, however.
  Geenz Spad  it doesn't have to be on the agenda, so long as it's relating to content creation
  Keli Kyrie  There is also this Jira https://jira.secondlife.com/browse/VWR-29210
  Geenz Spad  cloth simulation was actually briefly touched on at one of the open-dev meetings a while back
  tauber Paola von Tauber  it's about cloth simulation :D
15:05  woozl Selkit T. Woozl  I think sometime before next meeting, I need to re-engineer this build. Was made for SL9B. :P
15:06  Geenz Spad  I can bring up Keli's JIRA at the next open-dev; not sure if Oz or one of the open source developers know about that one
    I'll keep that one in mind
  Keli Kyrie  Oz does I have talked to him about that one
  Geenz Spad  ah
  Martin RJ  Geenz we have talked with Oz about it at the last meeting
15:07    he said it was the bug that blocked the dev-branch from going to beta
  Geenz Spad  it's a shame I overslept then :P
    right
  Martin RJ  but since the beta is out, they probably fixed it.
  Geenz Spad  can anyone confirm this one in the beta?
15:08  Keli Kyrie  There was a new Beta released on July 8... I don't know if it is in that one but before July it was
  Nal  We need a JIRA helper here...
  Martin RJ  I assume it's fixed in Second Life 3.3.4 (261355)
  Geenz Spad  I'm going to ask Oz about that
15:09  Siddean Munro  https://jira.secondlife.com/browse/SH-2103
  Koli Melune  Strange, doesn't repro in 3.3.5.
  Martin RJ  but I didn't have a chance to test it myself
  Geenz Spad  it may have been fixed, but if not, bring it up at the next open dev
    or even if it is, bring it up anyways :)
15:10  Keli Kyrie  Cool :)
  Geenz Spad  SH-2103 -- looks like a problem with the avatar mesh's UVs
  woozl Selkit T. Woozl  Inclined to agree with Geenz, but I'd have to poke at the models themselves.
    UV overflow, maybe.
15:11  Geenz Spad  keep the JIRA's coming everyone
  woozl Selkit T. Woozl  Nasty one to try and fix, without breaking a heap of legacy content. Hate to say it, but the base SL avatar mesh is showing its age a bit.
  Maxwell Graf  The devil you say!
  Siddean Munro  I don't believe it's anything more than just changing some of the alphas
  Geenz Spad  hm
  woozl Selkit T. Woozl  Not quite that simple.
  Geenz Spad  will need to look into that one a bit further
15:12  Vivi  It should be possible to update the avatar whilst retaining legacy UV mapping for legacy content? Or offer a switch to allow "use legacy model"
  Geenz Spad  from what I'm told, touching the existing avatar mesh is a bit of a no-no for now
  woozl Selkit T. Woozl  That's one possible solution; The usual Linden policy of "Does it break legacy" probably applies in all cases.
  Vivi  and that gap is the alpha layers anyway, it's not the avatar model, the gap isn't there on the jacket layer
15:13  Geenz Spad  I believe they're working on a few things behind the scenes to make it possible in the future htough
    and if it's the alpha layers, then we can probably put together a fix
  Elie Spot  Improvement can be made without breaking content, its most likely
  woozl Selkit T. Woozl  Yup.
  Elie Spot  ^ what vivienne said
15:14  Siddean Munro  There are a bunch of textures under characters in the SL viewer, Shirt_bottom_alpha is the one with the issue
  Geenz Spad  that helps
  woozl Selkit T. Woozl  Insufficient seam allowance made, likely.
  Geenz Spad  have you tried modifying it?
  Vivi  I've tried modifiying that to resolve the issue myself but replacing the texture causes a client crash, would have to be in source and compiled with the fix as the target file, size mismatch or something
  Siddean Munro  I can't remember, it was 2007 :p
15:15  Geenz Spad  ...huh
  woozl Selkit T. Woozl  The years do blur a bit, don't they. ;)
  Geenz Spad  well that's interesting
  April Looming  Is there any plan to add more bones to support the new deformer?
  Allegory Malaprop  there is something special in the way they've cmopressed the image
  Geenz Spad  I don't have any information on that, sorry April
  woozl Selkit T. Woozl  To my knowledge, April, adding new bones to the existing avatar mesh is easier said than done.
15:16  Geenz Spad  shall we get in another JIRA before moving on?
15:17  woozl Selkit T. Woozl  Future changes may alleviate that, but going with the mesh as it is, altering the skeleton will break legacy content in several areas.
  dahlia1337 resident  adding bones shouldnt break existing content that dont reference them
  Elie Spot  adding bones wouldnt break anything
    ^
  woozl Selkit T. Woozl  That one was actually explored back in 2006 or so for us silly furries to have actual tails. :P
15:18  Geenz Spad  alright, if no one else has anything to bring up in terms of JIRAs, we'll continue discussing morph targets from last week
15:19  Siddean Munro  Looks like that's it for Jiras
15:20  Geenz Spad  for those of you who weren't here, a morph target is a special shape that defines how a mesh should "deform" when a certain parameter is increased or decreased
  Siddean Munro  Blender calls them Shape Keys
  Geenz Spad  our discussions from last week mostly looked at how this could work with regards to content creators and its usage in-world
15:21  Maxwell Graf  mesh genitals, finally.
  Asha  lol
  woozl Selkit T. Woozl  ...*facepalm*
    Trust me, I think SL's already well on that task.
  Braydon Randt  O.o :P
15:22  Geenz Spad  now, to some degree this could be seen as yet another mesh deformer alternative; although it's one that would in theory be more flexible than cBones and Qarl's deformer, and would put content creators in complete control over how a mesh "deforms"
  Asha  Yes, that's old news Max. :p
  Maxwell Graf  with morph targets though, think of the advancements!
  woozl Selkit T. Woozl  Anyone have two cents on the matter? Operating mechanics? Influence on land impact?
  Siddean Munro  It's hard to speculate on that really without some kind of testing opportunity
15:23  woozl Selkit T. Woozl  Hard to get a testing opportunity together without a basic framework.
  Cyclic Gearz-Bellamy  And really, Land Impact is only really relevant for items to be rezzed, if the morph is to be used for clothing/avatarshapes it won't matter about LI so much
  Geenz Spad  at first at least
  Asha  I just remember that during beta, the Linden at charge at that time said that it would be bad to use morphs. He never went into detail other than it would increase the downloads.
    *in charge
  Nal  Do we have any idea whether it will require changes to the viewer, server, or data to implement?
  woozl Selkit T. Woozl  I'm of the firm opinion that avatars need to have a quota of some type, but that's another matter entirely.
  Geenz Spad  I've gotten a bit of feedback on that Nal
15:24  Braydon Randt  meaning .. it would increase the work that hadnt been planned for
  Asha  Well...I think he just prefered custom skeletons and didn't understand how morphs are more robust for certain types of deformations,
15:25  Geenz Spad  it would require some changes on the format side of things, and the server would need to be able to recognize the format
  woozl Selkit T. Woozl  Asha; Morphs do tend to add a bit more to filesize.
  Esprite Xavier  morph targets are kinda intense compared to skeleton animations though isn't it?
  woozl Selkit T. Woozl  And bandwidth is a concern factored into land impact.
  Nal  ...so that is viewer, server, and data, yes?
  Geenz Spad  work wise?
  Esprite Xavier  work wise, render wise
  Geenz Spad  and yes Nal, though the server work would probably be a bit minimal depending on how this is handled
15:26  Esprite Xavier  more data too
  Asha  Oh, I'm sure it adds to file size.
  Geenz Spad  and work-wise, it shouldn't any more difficult to work with than rigging; infact in theory it should be easier
  woozl Selkit T. Woozl  As mentioned, yes, more data. Bandwidth on both CPU throughput and network traffic.
  Geenz Spad  render-wise, it'd actually be faster than Qarl's deformer
15:27  Esprite Xavier  compared to bones though it's not really super efficient, just wondering what kinda limitations there will be for the number of targets, doing an animation with them would be pretty intense right?
  Asha  But, people are already using huge file sizes for things that they don't need to. At least, increased file size for added morphs make sense. Perhaps, they could make a file size limit.
  Nal  Is there anyway to test what the render time differecne would be?
  Geenz Spad  not really
15:28    let me elaborate about how these would work
  Braydon Randt  that goes back to what is acceptable use on AV meshes really doesnt it
15:29  Geenz Spad  work-wise, a morph target is typically a special kind of vertex map (or whatever the equivalent is in your modeling software of choice); you literally are just moving around vertices that makes up the morph target
    there's no need to change your rigging, it should "just work" with the new morph
    it'd effectively allow one to "sculpt" how they think a piece of clothing should deform
15:30  Nal  So, we would build a Morph Target like we do an LoD layer?
  Braydon Randt  or would be have set morph targets .. like we done bones
  Asha  The dae file schematic supports morphs.
  Geenz Spad  rendering-wise, it'd be faster than Qarl's solution, simply because it only adds a single step: interpolating the new vertex positions
15:31  Siddean Munro  I think it'd be a setting on the actual mesh rather than be a separate mesh like LOD meshes are
  Geenz Spad  and sort of Nal, really it varies from modeling package to modeling package
  Allegory Malaprop  on most programs it is, yes. it's more akin to setting an extra animation keyframe on the existing mesh
15:32  Elie Spot  ^^ cool!
  Braydon Randt  i have a question .... can we have " zones" that we can apply to vertex groups ..... with that in mind .... any value can be applied to the vertex group for that " zone"
  Nal  It sounds like one would have a mesh that is sculpted into various shapes on various key frames. SO, rather than weight paint, we model, yes/no?
  Geenz Spad  I imagine some kind of "morph group" functionality would be required from the get-go for the sake of making it work with the avatar mesh properly, Braydon
15:33  Siddean Munro  Braydon the way they work is you set a shape key for Belly and then change the model to what it would look like with the belly slider at max
  Geenz Spad  Nal: sort of, have you ever worked with a program called ZBrush?
  Siddean Munro  or morph target, rather than shape key
  Nal  Basically no. I'm mostly Blender
  Geenz Spad  ah, then have you ever used Blender's sculpting functionality?
  Braydon Randt  <blender & zbrush
15:34  Nal  I have... Scult Mode.
  Asha  In Blender, the shape keys (morphs)are seperate from the weights Nal.
  Siddean Munro  And I think weighting is still necessary
  Asha  *separate
  Elie Spot  it makes sense that weighting would still be necessary
  Asha  Yes, you'd still need to weight the item.
  Elie Spot  as morph targets wouldn't affect the movement of your garment just the sizing
  Geenz Spad  well, creating a morph target would be creating a new shape key, and literally "sculpting" how you want it to look when someone sets a parameter to their max, minimum, or what have you
  Asha  But, you only need to weight it once for the most part.
  Esprite Xavier  I guess in some ways morph targets would be easier to understand the alternatives are kinda complicated to create for and or not supported easily
15:35  Asha  Yes, animating a face with bones is like the tenth level of hell to the average person when compared to doing it with morph targets.
  Braydon Randt  shape keys aka morph targets would give a better flexability and the ability to better decide how the garment would "morph" at different values
  Elie Spot  ^^
15:36    so lets make it happen xD
  Siddean Munro  So what do we need to get some kind of testing or prototype?
  Allegory Malaprop  on the difficult technical side: your custom morph targets would have to match up with the avatar's morph targets. anywhere you don't have an alpha, that gets just as fussy as weighting for joint movement.
  woozl Selkit T. Woozl  As Asha said. Fiddle with fussy muscle and bone groups to get a smile, or sculpt the smile and tween.
  Asha  Yes, but matching the morphs in your 3D program is not difficult.
  Braydon Randt  hense why i said about " zones" ... so when you increase your ass .... your face done gett bigger *lol*
  Geenz Spad  for a prototype, we first need to write a proposal to give to the lab
  Elie Spot  ^ allegory - it would be good to have the deformer as an alternative in that case, because it would be a lot of work, but also the flexibility would be amazing
15:37  Siddean Munro  Braydon that's up to the creator of the item :)
  Allegory Malaprop  the deformed matches up to the current morph targets automatically- the avatar mesh is not morph targeted (especially for extreme shapes) especially well, which is why the deformer has the issues it does- matching up to that with cutom morph targets will run into similar issues.
15:38  Siddean Munro  I'm hoping we can have the option of both Alle. :)
  Geenz Spad  now, this proposal needs to include a clear vision as to how it should work with regards to user-facing content, how it should work with regards to the content creators, and how it should work on the lab's side of things
15:39  Asha  It would be nice if you could have access to the morphs through LSL.
  Geenz Spad  the user-facing side I feel is fairly obvious; the user adjusts their shape, and their clothing deforms
  woozl Selkit T. Woozl  I would imagine any viable proposal at all, Asha, would have to have a plan for how it ties into LSL.
  Elie Spot  is there any way we can incorporate physics into this? i know that was something LL placed a higher priority on
  Geenz Spad  well, I think that depends really
  Elie Spot  *avatar physics
  Geenz Spad  on both Elie and Selkit's comments
  Braydon Randt  so .... clothes cant fit ... but hell .. your boobs will bounce well
15:40  Anouk Spot  hahah

[15:40] AshaAsha snorts

  Geenz Spad  we should think about what would be best for a "version one" release
  Toast Bard  haha
  woozl Selkit T. Woozl  Build a frame to hang the shinies on.
  Braydon Randt  no wonder .... LL priority ... tits and ass * lol*
  Elie Spot  if LL accepts a versoin 1 without physics im fine w/ it haha
    just wondering if it can be incorporated
    at some poitn
    typos!
  Geenz Spad  I'm sure LL would want that in version 1 :P
  Asha  I'd assume Qarl's deformer would have to address avatar physics.
15:41  Elie Spot  i have a feeling they'll want that too yah
  Geenz Spad  Qarl's deformer actually does it automatically
  Asha  Yes right
  Allegory Malaprop  physics is using a similar system for its movement. it would not be impossible to build in at the same time
  Elie Spot  so in defining our own morph targets would we also have to define phyrics movement?
  woozl Selkit T. Woozl  Physics is something that more or less has to enter in the first step. Anyone else remember prim joints and Havok 4? ;)
  Geenz Spad  somewhere in the proposal I'm sure Elie
15:42  Elie Spot  or could this be automatic
    i cant stop making typos my apologies!!!
  woozl Selkit T. Woozl  Quite alright!
  Asha  The problem with physics is that we'd need a reference in our modelers too make the morphs around.
  Geenz Spad  but I can speak with some people at LL to find out if they'd really want that in version 1; they do have an influence on what will go into the proposal from a requirements point of view
  Asha  *to even
  Siddean Munro  Geenz is there anyone that you would suggest work on this proposal? I can provide some input but I think we need to get moving on it soon.
15:43  Elie Spot  good idea Geenz, best to find out how they'll respond to a version 1 with or without physics support
    maybe before we do the proposal? idk
  Geenz Spad  right, I can't suggest anyone in particular off the top of my head, but I do think we need to keep this one open ended; how about for the next meeting, we open the floor to everyone's ideas?
  Siddean Munro  Ok
15:44  Geenz Spad  from there Siddean and I can begin working on a draft
  Asha  Sounds good Geenz
  Nal  The proposal is good. But, how do we get Qarl moving on the Mesh Deformer? He is waiting for us to make a consensus, which sucks.
  Siddean Munro  Cool with me
  Elie Spot  sounds good!
    well in this case, morph targets wouldn't replace the deformer right?
  Geenz Spad  Nal, that actually gets into #5 on the agenda, but if Paola wouldn't mind us going ahead with that, we can discuss that now
  Braydon Randt  at the moment ... we have nothing ...... so we cant go wrong .. we just need to be consistant
15:45  Asha  Well, I think the mophs are a bit of a different deal from cbones versus deformer (albeit some overlap).
  Nal  Or do it in oder NP with that
  Elie Spot  ah right
  Asha  *morphs
  woozl Selkit T. Woozl  Elie, Asha; Features in SL are rarely mutually exclusive. There's likely room for both.
  Geenz Spad  right, well, on the subject of the deformer
  Elie Spot  agreed selkit
  Geenz Spad  during our last meeting, people seemed to want both
  Asha  It would be great to have both Selkit.
  Elie Spot  i def want both
  woozl Selkit T. Woozl  Applications for both, in fact. Prims are not obsolete simply because we have mesh and sculpts.
  Geenz Spad  now, the current hold up on the deformer is, do people want Qarl to continue with it
15:46  Elie Spot  custom morph targets are really advanced, i don't think everyone should have to do that
    heck yes
  Siddean Munro  I do because at the moment we have nothing at all and it's really holding back content development
  Geenz Spad  what Qarl currently wants is a consensus on whether people want his solution, or would have another solution, or want both
  Elie Spot  yeah, and even with morph targets i feel like it would still be used
  woozl Selkit T. Woozl  My opinion leans towards both.
  Elie Spot  for people who dont want to do all that work
  Maxwell Graf  I think it would be crazy to abandon all the work he has done so far, and not simply because a lot of us paid for it.
  Vivi  Given that the deformer work Qarl is doing works with physics and all avatar sliders, vs the alternate method that uses the bones of the collision skeleton that doesn't work with physics and doesn't work with all sliders
  Asha  To my mind, Qarl's deformer and morphs would work pretty well and cover many issues from custom AVs, fixing size issues and opening up mesh animation.
15:47  Elie Spot  it will be extremely valuable to have both, for a lot of reasons imho
    and for content creators to have a choice
  Vivi  and complimentary solutions are not a bad thing really
  Siddean Munro  Absolutely
  Elie Spot  ^^ agree
  Geenz Spad  if you have an opinion regarding this, I implore you to speak up! I'll most definitely talk to Qarl about it
  Siddean Munro  Can we comment on his blog about it?
15:48  Geenz Spad  that you can!
  Elie Spot  who wants both? do we all?
  woozl Selkit T. Woozl  I don't see why not.
  Cyclic Gearz-Bellamy  Yeah, I'm in favour of both
  Elie Spot  i do!
  Anouk Spot  i would want both
  Siddean Munro  I wasn't aware that he was waiting for anyone to give him the go ahead
  Nal  I would like all that I can get...
  tauber Paola von Tauber  from a mere consumer point of view, yes please! :D
  Asha  Both would be cool with me.
  Maxwell Graf  As I stated on Metareality last week, Id prefer, at this point, ANY solution that is going to 1) work the best and 2) do so with a minimum of additional work required by the designer during the creation process.
  woozl Selkit T. Woozl  Both seems to be fairly unanimous as a view.
  Vivi  and from what I understand the deformer is much closer to being reality, morph targets would require longer development time, the deformer project would give content creators (and users!) an enhanced toolset much sooner
  Maxwell Graf  if thats both, then im for it.
15:49  Braydon Randt  as am I
  Elie Spot  ^^ agreed vivi
  sachi Vixen  I blame Max
     :P
  Maxwell Graf  lo
  Asha  lol
  dahlia1337 resident  what are the performance implications of having both, and do they conflict?
  woozl Selkit T. Woozl  Maxwell; I'm afraid that no one individual solution will do what you've mentioned, but when there's more than one tool, picking the right one for the right task works wonders.
  Elie Spot  deformer will be a shorter term release hopefully! is max waiting on our consensus about morph targets or was he waiting on more info about the possibility of cbones to pursue a different deformer entirely
    one that might work better altogether than his current?
    woops
    not max
    qarl
    lol
15:50  sachi Vixen  I think need whatever we can get asap
  Esprite Xavier  Waiting for features or toolsets to be completed or started on kinda sucks when you can have some solutions now :D
  Geenz Spad  there's no negative performance implications of having both specifically, just the performance implications that Qarl's produces its self; but he's done his best to mitigate that, and it only occurs when someone uploads a mesh to use his method specifically
  Nal  ASAP is good
  Geenz Spad  right then, sounds like we have a consensus
    I'll speak to Qarl this week about going for both
  Siddean Munro  Ok
  dahlia1337 resident  did the initial delay problem of the deformer get resolved?
15:51  woozl Selkit T. Woozl  To be fair, Geenz, artist practices have a huge impact on performance anywhere.
  Maxwell Graf  My biggest concern at this point is just how much work is going to be involved on the design end in order to get a system that works. If it requires X amount of additional weight painting and adjustment, various sizes, alpha maps, etc etc etc. then this becomes less and less of a viable solution as the workload increases, regardless of how well the end result works.
  Vivi  collision skeleton is workable but it's really kind of an ugly hack and it doesn't work as well as Qarl's method, so...as an interim solution while people wait on the deformer to ship in the viewer, it's useful but I don't see it as more than a stopgap
  Geenz Spad  http://www.qarl.com/qLab/?p=86 here's the most recent blog post I believe of the deformer
  dahlia1337 resident  qarl was mentioning adding an octree or something
  Geenz Spad  so if any of you want to go post something up there, feel free
  Nal  Dahlia, the 0.2 Mesh deformer had a lag. In the 0.3 version Qarl had it pretty well handled.
  woozl Selkit T. Woozl  Max; That's one of the topics that will need to be covered; a general consensus on just how things are implemented, the depth and complexity...
15:52  Elie Spot  the only issue with the latest deformer were the distortions on extreme shapes, everything else looked great
    I think when Qarl saw redpoly's demo, he was wondering if whatever they did could be used as part of the deformer
  woozl Selkit T. Woozl  Elie; That can be mitigated with tweaks to mesh design.
  Elie Spot  it cant, its because the base av is wonky in certain places
  Siddean Munro  The base avatar morphs are pretty ugly
  Elie Spot  so smoothing out the default avatar is one of the possible solutions
  woozl Selkit T. Woozl  Right, base av can't be helped, unfortunately.
  Elie Spot  which we're working on
  Maxwell Graf  When the pipeline for development becomes 10%% design and 90%% production methods, I think we run into a serious viability issue, not just for the individual designer, but for the entire population of people who may want to or are designing fashion.
  dahlia1337 resident  I would think the deformer has more potential for better deformations and additional features like facial expressions and lip sync
15:53  Elie Spot  it can selkit
  Vivi  the base avatar is not perfectly symmetrical among other things
  woozl Selkit T. Woozl  Topology on that poor old beastie is a nightmare in spots.
  Siddean Munro  Yes it really is
  Asha  That's why Qarl stated he would add the feature to have any base that you want. So, you could use a base closer to the extreme shape for those.
15:54  Siddean Munro  Dahlia, in order for your custom mesh face not to turn out really ugly, you have to build it exactly to Ruth.
  Nal  Having a selectable base avatar would be great.
  Asha  right
  dahlia1337 resident  exactly? lol well maybe with the vertices near ruth's vertices
  Braydon Randt  so start ugly ..... is the key
  Geenz Spad   :P
  woozl Selkit T. Woozl  I doubt we're ever going to quite outrun Ruth. :c
15:55  Siddean Munro  And then it looks just like the SL face, which kinda defeats the point of making a custom face :)
  Esprite Xavier  not unless someone comes up with some kinda transitional solution
  Braydon Randt  ruth a good old bird ..... bad hair day EVERY day
  woozl Selkit T. Woozl  Makes me a bit nostalgic.
  Geenz Spad  so anyways, shall we get on with Paola's topic?
  Siddean Munro  Yes
  Geenz Spad  Find a way to allow custom imported mesh avatars to wear existing content (textures/mesh). We need better, more detailed avatars, whilst retaining legacy content and immersive uniqueness. (Paola Tauber)
  Vivi  Not necessarily, it can look similar BUT increasing the number of tris and verts will improve the overall model considerably
  Maxwell Graf  I guess what im saying is that it is very easy in this situation to get completely caught up in the technical aspects of implementation, while very easy to forget or push aside the human factors of actually using this type of a system on a daily basis. It should be considered as importantly as the technical aspects.
  Martin RJ  I totally agree with Maxwell
15:56  woozl Selkit T. Woozl  Vivi; It's not just a matter of polycount, unfortunately.
  Esprite Xavier  Be nice to see a new avatar mesh that had 2 uv sets, a new one and one that could make use of the old content by matching the uvs in some fashion, though that's not really a perfect solution either
  Geenz Spad  Maxwell: the workflow will be discussed some more at the next meeting (that or we can arrange a meeting specific to morph targets; it really is a large topic to discuss)
  woozl Selkit T. Woozl  Geenz; I may suggest we dedicate a meeting.
  dahlia1337 resident  well I would agree that the "bind" position (if Im not over-extending the term) of Ruth's lips are a bit too close for good deformation using the deformer :/
  Siddean Munro  Can we move onto the next topic? We're going to run out of time :)
  Geenz Spad  not a bad idea Selkit
15:57    anyways, so Paola's topic
  tauber Paola von Tauber  when you all stop giggling at that idea, my request it's coming from a consumer point of view. we can have custom mesh avatars, but we can't use legacy content. there must be a smart way to combine both
  dahlia1337 resident  legacy content like skins?
  Siddean Munro  Paola, the only way to use existing textures on new meshes is for the UV's to be exactly the same. I can't see how we can do that
  Geenz Spad  in the context of mesh avatar replacements, it wouldn't be a bad idea to extend some of the luxuries we have on the avatar mesh to them as well
  Nal  I don't think so Paola...
15:58  tauber Paola von Tauber  like cloth simulation, i understand it's blue sky thinking more than anything else
  woozl Selkit T. Woozl  Paola; Morph targets would alleviate at least some of that, allowing you to define your own clothing folds and wrinkles.
  sachi Vixen  I don't think you are ever going to be able to use legacy content with custom avatars. Things just don't physically work that way.
  dahlia1337 resident  the seams on the UV maps for Ruth are horrible, would be nearly impossible to match
  Geenz Spad  hm
  Siddean Munro  She's talking about being able to use existing skins on new custom built avatars, for example
  Geenz Spad  well, let's think of this from another angle shall we?
  Siddean Munro  Or your texture layer shirt
15:59  woozl Selkit T. Woozl  Existing skins would have to be applied to a matching UV.
  Nal  Cloth sim is CPU intense. In blender it drives my machine to its knees. I doubt we will see in SL anytime soon.
  Allegory Malaprop  people are matching up the seams already. but there's the entire multi layering issue, the avatar mesh functions differently.
  sachi Vixen  If a mesh avatar is made that is better than ruth, even if you could bind it to the deformer you are never going to be able to wear Ruth's skins on it
  Braydon Randt  truth of the matter is .. legacy content . has been replaced, and there slittle we can do to make old stuff work on your frames
    new*
  sachi Vixen  Or Ruth's texture clothes
  Tya Fallingbridge  ruth needs to be cleaned up
  Maxwell Graf  i dont mean to add a secondary agenda, just responding to the fact that I see a ton of discussion about this or that technical element, and very little consideration for how it works when you have to use it every day, how it fits into a pipeline and how much of that pipeline it makes up. The entire issue should be prefaced with the original intent, which is to make this result easier to accomplish.
  woozl Selkit T. Woozl  Texture clothes require a matching UV.
  tauber Paola von Tauber  what do you mean Geenz?
  dahlia1337 resident  have you seen how jagged the seams on Ruth's arms are?
  Geenz Spad  how about we think about how we can extend some of the cool things we have on the base avatar that we don't have on mesh ones?
16:00  Allegory Malaprop  a large percentage of mesh avatars i've seen ARE using the matching UV with a retpop'd model
  woozl Selkit T. Woozl  I have. As I mentioned earlier, dahlia, Ruth is a mess.
  Braydon Randt  i agree with Max ...... making it simple ... is more important .. than anything else .. as over complicating a system will kill the desire to use it
  Allegory Malaprop  but you can only wear one extra "clothing layer" above the skin because layering doesn't work the same
  Elie Spot  allegory, like people took ruth and modded it a little?
  Vivi  Petites for instance are the most popular mesh avatars and they're based on the existing base shape and UV mapping
  Tya Fallingbridge  yes
  Allegory Malaprop  complete retopology, but matching UVs
  Siddean Munro  The petites are modified SL avatars
  Allegory Malaprop  all the petite avatars, for instance
16:01  tauber Paola von Tauber  how about we think about how we can extend some of the cool things we have on the base avatar that we don't have on mesh ones?
    what do you mean by that Geenz?
  Geenz Spad  well, baked textures for example
  Allegory Malaprop  they use a very different topology that has less distortion (higher poly as well as different choices made in overall topology, smoother weight painting, etc.)
  Siddean Munro  If I were to build a custom avatar there is no way it'd be using SL texture layers
16:02  Braydon Randt  its like wrapping a diamond in brown paper
  Asha  lol
    No kidding
  Siddean Munro  Baked layers would be a good idea actually
  Allegory Malaprop  meshes do not support texture "layering" in the manner the avatar does. decals, undershirt/shirt/jacket
  Geenz Spad  sure they're not "amazing", but it'd allow content creators to make a base avatar that others can easily create clothing for
  woozl Selkit T. Woozl  Baked mesh layers would make me drool in unseemly ways.

[16:02] AshaAsha chuckles

  Braydon Randt  why not ...... as you can alpha your mesh layers ?
16:03  Geenz Spad  Allegory: however, the existing mechanisms for baked texture layers would apply to meshes as well, it wouldn't be all that complex
  Braydon Randt  well ... i duppodsr face issues
  Esprite Xavier  meshes do support multiple uvs as well
  woozl Selkit T. Woozl  We're currently limited to a single material per-face right now, Braydon.
  Braydon Randt  , wtfv did i just type *lol*
  Asha  no idea Bray ^^
  Allegory Malaprop  layering alphas brings up alpha issues
  Finlay MacFanatic  duppodsr
  Allegory Malaprop  i.e. alphas do not play nicely.
  Geenz Spad  it's not explicitly tied to the avatar mesh, having actually looked at how that works, it could be extended to meshes as well
  woozl Selkit T. Woozl  Allegory; Hence, baking them in layers much like the existing avatar.
  Braydon Randt  well thats a whole new issues to play with isnt it
16:04  Martin RJ  some news: SL has 30 million registered accounts now. And: did you all read about that: SL doesn't have an age verification anymore http://bit.ly/slageverific
  woozl Selkit T. Woozl  I'm afraid that's not very relevant to our current topic, Martin.
  Allegory Malaprop  yes selkit, which is what i meant by decals, and layering in the same manner as the existing avatar mesh, not using alpha layers as a solution, because it isn't really.
  Siddean Munro  Ok here's an example. I make a mesh avatar - I want to include cleavage, freckles, makeups wihtout having to change the head mesh or complete texture
  Martin RJ  no not really, but it's past 4pm :)
  woozl Selkit T. Woozl  I didn't suggest using alpha layers, Allegory.
16:05  Geenz Spad  Martin: the meeting lasts till 4:15 :P
  Martin RJ  oooh
  woozl Selkit T. Woozl  Siddean; That's precisely what layering would permit.
  Allegory Malaprop  no, but someone else in the discussion did :)
  Siddean Munro  So I would want to do it in layers like we do now
    Exactly
  woozl Selkit T. Woozl  Hell, we need that for meshes in general, not just avatars.
  Vivi  1-bit instead of 8-bit alphas layer properly BUT 1-bit alphas also have issues with image resolution and edge aliasing
  woozl Selkit T. Woozl  Build a car, want to add decals or numbering or weathering? Bake a combined material. <3
16:06  Allegory Malaprop  you can get around that with baking the layers, vivi
  Geenz Spad  so, I guess, it turns into, who thinks clothing layers like we have for the existing avatar mesh would be a good idea for rigged meshes as well? specifically those that are intended to replace the avatar mesh?
  tauber Paola von Tauber  me! :D
    (i think!)
  woozl Selkit T. Woozl  I'd actually think extending that beyond avatars and into general materials is a better, more broad-purpose idea.
  Siddean Munro  I think it's a good idea,it's work for close fitting clothing layers like underwear that can be worn underneath rigged mesh outer clothing
  Braydon Randt  so what we want .... is the SL avatar .. but custom .... with all the bells and whistles it has ... without it looking like Bicentenial man
16:07  Geenz Spad  basically, yes Braydon, and I'd see clothing layers as a good starting point to that
  Esprite Xavier  so just a way to use all the existing texture slots but use a custom avatar mesh?
  Vivi  I think it's a good idea, yes, if possible
  tauber Paola von Tauber  i'd be happy with that
  Nal  Since the Lab is building new compositing servers... this could be a complex issue just now.
  Esprite Xavier  well I just "just" but it's probably more complicated
  woozl Selkit T. Woozl  Complexity is a cost of progress, unfortunately.
  Geenz Spad  Esprite: yes, although I think we'd also want the ability to have your own UVs for it as well
  Braydon Randt  what we dont want ...... is to rush a half assed solution for anything .. or we get the " we dont want to break existing content" in 3 years
16:08  woozl Selkit T. Woozl  Geenz, I think we have another one-topic meeting we need to schedule. ;)
  Geenz Spad  we still have some time to discuss it
  Vivi  I'm not sure how it would be a complex issue as I assume the new compositing servers won't change the compositing apart from taking the load of doing it off the simulator and ending the client-side compositing currently used, other than that it shouldn't generate any specific issues?
  Allegory Malaprop  it's a solution that could be extended to buildings, which could lower texture size as well.
  Geenz Spad  though I will want to speak with Toast shortly afterwards about their issue
16:09    and that's correct Vivi
  Toast Bard   :D
  woozl Selkit T. Woozl  Allegory; Composite textures for non-avatar meshes would make me melt in glee.
  Esprite Xavier  yeah, it'd be interesting to see where it goes but ya know, kinda got to the point of expecting nothing until it's there hehe
  Geenz Spad  I know the person who was working on the shining project in that regard, I can talk to him how this would effect that, and if they can factor this into their design
  Nal  Vivienne, it is going to be politically complex. Getting the Lab to change mid process is not easy.
16:10  Geenz Spad  you never know Nal, how the existing compositing system actually works is very simple
    it's literally layering textures on top of each other in a specific order
  Nal  I'll give you that Geenz... see what they say.
  woozl Selkit T. Woozl  Merging down by layer priority.
16:11  Esprite Xavier  flatten image :)
  Geenz Spad  so it won't be too much trouble to speak to the lab about it
  woozl Selkit T. Woozl  It'd be nice if we could specify blend mode per layer, but that, may be significantly more troublesome to ask for.
  Geenz Spad  right, let's keep it simple for now
  Braydon Randt  might be something that can merged into the materials system we want real bad
  Siddean Munro  So are you going to talk to someone at the Lab about that Geenz?
  Tya Fallingbridge  k.i.s.s.
  Geenz Spad  I will be
16:12  Siddean Munro  Ok cool :)
  tauber Paola von Tauber  yes. for the dumb girl over here (me): keep it simple and give me a better body on which i can plaster my beloved skin :D
  woozl Selkit T. Woozl  If asking fails, please break out the nuclear puppy eyes. We *need* that.
  Geenz Spad  we only have a few minutes left, so Toast? Do you have a means to reproduce the alpha problem?
  Maxwell Graf  RL calls. Have a good day all.
  Esprite Xavier  Wish ya the best in that reguard :D
  Asha  Bye Max
  Maxwell Graf  ...HOLA!!...
  Braydon Randt  laters max
16:13  Esprite Xavier  I wonder if that seam issue is somehow related to mip mapping or something
  sachi Vixen  I don't think that will ever be feasible Paola. To wear your existing skin with a custom avatar.
  Toast Bard  sure lemme just grap a screenshot
    one sec
  woozl Selkit T. Woozl  Paola, wearing your skin with a custom avatar would entail precisely matching the UVs and closely matching surface topology to avoid deformation.
  Toast Bard  http://gyazo.com/3b2286ef8d19a2a636190f751258c736
  sachi Vixen  Unless the creator remade it for the custom specifically
  Anouk Spot  i can see it right now on Lamb's hair
  woozl Selkit T. Woozl  Essentially it would be a slightly upgraded version of Ruth.
  Anouk Spot  it conflicts with the background tree
  Geenz Spad  I haven't seen your issue personally, but I know of others who have been seeing it recently; if I have a solid reproduction case, I can look into it, or forward it to the correct person
  woozl Selkit T. Woozl  And it would still have to share Ruth's faults.
  Toast Bard  you can see in that picture that the alpha on lambs hair is cutting out the things behind it, and if there were alphas on her hair behidn it, the prim would cut them out too
16:14    as if it were an invisiprom
  Lamb Bellic  allegory's hair as well
  Toast Bard  its incredibly difficult as a designer to work with this
  Siddean Munro  isn't this the long standing opengl alpha sorting issue?
  Koli Melune  I *think* runitai has fixed that.
  Esprite Xavier  alpha sorting :(
  Toast Bard  because she couldnt wear like, say, a sheer shirt with that
  Allegory Malaprop  rigged mesh renders alpha like this- it's not the sorting thing really
  Toast Bard  yeah
  Koli Melune  I seem to remember some MAINT JIRA about something with that.
  Geenz Spad  hm, it *looks* like fast alpha, I'll ask Runitai about it
  Toast Bard  its def not a flickering issue
  Allegory Malaprop  rigged completely fails to render any alphas behind it all- no flicker, nothing. and yes, there's a jira, though i don't recall the number
  Anouk Spot  doesn't look fixed to me
16:15  Koli Melune  The fix isn't out of QA. :p
  Geenz Spad  and yes, I see it Allegory's hair
  Toast Bard  you can see it in every mesh hair with an alpha
  Allegory Malaprop  rigged mesh

[16:15] Koli MeluneKoli Melune wonders if he can find the JIRA.

  Toast Bard  or any mesh item
  Allegory Malaprop  it's exclusively a rigged mesh issue: hair's most obvious, as people are just avoiding alphas elsewhere because of it
  Geenz Spad  if there's a specific JIRA for that, give it to me please
16:16  Siddean Munro  Is this to do with mesh rigged hair that makes eyelashes disappear?
  Toast Bard  yeah i'd bet
  Siddean Munro  I think I have a jira in about that
  Geenz Spad  looks like we're going over out time slot again :) not a problem though
  Anouk Spot  i've seen this happen with non rigged mesh as well
    like with glasses with alpha, it's not flickering that happens, it just blanks out that area
  Geenz Spad  right, if there isn't a JIRA for it, please file one; either I'll look into a fix, or I'll speak with someone who works in that area
16:17  dahlia1337 resident  doesn anyone know when are mesh attachments drawn wrt other parts of an avatar?
  Koli Melune  Oh, the JIRA I was thinking of is different, but somewhat the same: https://jira.secondlife.com/browse/MAINT-622
  Toast Bard  so theres no insider information on how close a fix is to completion?
  sachi Vixen  Isn't the blanking out the same reaction as the flickering but because the prims/mesh are different the reaction is slightly different?
  Geenz Spad  I haven't been told anything, sorry
  Siddean Munro  https://jira.secondlife.com/browse/SH-2836
  Anouk Spot  oh i'm not sure
  Toast Bard  ah alright
16:18  Siddean Munro  that's the one I reported
  Allegory Malaprop  i've noticed an increase in alpha rendering being a little
    off, in general
  Geenz Spad  I think that sums up Toast's issue, can you look over it and make sure Toast?
  Toast Bard  well yeah my issue was just to find out if there was actually a fix being worked on and how far along it was
  Allegory Malaprop  things flickering out and completely turning invisible- i'm wondering if that might be more your issue, Anouk, this particular thing seems to be exclusively rigged, consistently
16:19  Toast Bard  but since theres no insider information i guess all i can do is wait
  Geenz Spad  I'll ask around about it
  Toast Bard  thank you :D
  Anouk Spot  it could be, alle!
  Geenz Spad  anyways, thank you all for coming to today's meeting!
16:20  Toast Bard  will you be making another meeting just for deformer talk?
  woozl Selkit T. Woozl  Thank you for not throwing rotten fruit. :D
  tauber Paola von Tauber  bye and thank you
  Geenz Spad  I'll be discussing that with the relevant people, Toast
  Toast Bard  noice!
  woozl Selkit T. Woozl  T'would be shiny to have a dedicated chat about that.
  Geenz Spad  and as always, we do have an open agenda, so feel free to get some items up for next week's meeting
16:21  Asha  Thanks for the meeting guys.
  Braydon Randt  aye ... thank you
  Geenz Spad  for those of you who missed that URL, https://wiki.secondlife.com/wiki/Content_Creation_Improvement_Informal_User_Group
  woozl Selkit T. Woozl  You're very welcome, Asha, Braydon and everyone.
  Esprite Xavier  Yep thanks, fun to talk about this stuff :)
  Geenz Spad  we will have the logs posted up in an hour or two