Mesh/Archive/2012-02-06

From Second Life Wiki
Jump to navigation Jump to search

Agenda

  1. As mentioned in the last OH with regards to the joints, it was mentioned on the old forums at one point that they would be re-added in a thread announcing the plan to upgrade to Havok 3 (which went so well, but eventually skipped past to the next versions in short order and we're thankful for it). I know there are people who would once again like to create wind chimes and swing sets without resorting to physics trickery (see SLURL below for in world example*) at a substantial prim cost. More so though (as applies to meshes) we would like to rig up our avatars to have flexible additions to avatars (such as tails, or hair) where appropriate. I'd appreciate any word on a review of Havok's potential systems for SL's use, starting with joints. Feynt Mistral 23:49, 3 February 2012 (PST)
  2. I was speaking with someone recently about the inefficiencies of creating a physics mesh for a house versus replacing it with prims, citing a dramatic ballooning of the LI for even a simple room boxed out with planes. This seems to be a common occurrence with people in the various mesh groups, eschewing the low poly physics mesh option for prims. This makes me sad however as it seems as though a 100-200 poly physics mesh ought to be very light weight as far as LI goes. Is there a way for us to debug the decomposition into convex hulls to identify where problem spots are and also is it possible for a single plane to be a physics boundary in and of itself? Feynt Mistral 23:49, 3 February 2012 (PST)
    1. http://maps.secondlife.com/secondlife/Serenity%20Woods/118/79/25
  3. After speaking with a group of fellow fashion designers, we have been wondering a bit about the status of the current mesh deformer project, as well as concerns brought up about how the project itself doesnt quite fit what is needed to make mesh universally excepted on the grid; Or that what was presented to us in the initial discussions of what the deformer could be, is vastly different from what is. Currently with out a deformer, the "Size requirements" are alienating a lot of residents and forcing many to fit a mold that isnt quite them. On the other side of the coin, the deformer project as it is now, is not even a shell of what it could be. It's hardly sufficent, will negate a lot of work that has been done up until its possible release. What is really needed is a deformer based around multiple size options human and non to allow max compatibility to residents across the grid. [Tyr Rozenblum]
    1. don't you think you have a social responsibility to adopt it? That we are hammered by mainstream media that you should change your body to comply with a certain standard in real life is bad enough. Letting clothing designers dictate the size of prim boobs, well, just be better than the mainstream media? [Davido Chrome]
  4. Could you explain the physics costs in this image? They are of simple box prims, simple mesh box (12 triangles), and simple mesh pyramid (6 Triangles) respectively. Why are both Mesh items the same, and why are both heavier on physics than the prim version? [Davido Chrome]

Meeting notes

Transcript for Monday February 6, 2012

[12:02] Charlar The Terrible (charlar.linden) KK, let's get this train a'rollin!
[12:02] davep (runitai.linden) I saw it -- I only accept friend requests from folks I know IRL
[12:02] Koli Melune (sahkolihaa.contepomi) Ah, alright.
[12:03] Charlar The Terrible (charlar.linden) 1) As mentioned in the last OH with regards to the joints, it was mentioned on the old forums at one point that they would be re-added in a thread announcing the plan to upgrade to Havok 3 (which went so well, but eventually skipped past to the next versions in short order and we're thankful for it). I know there are people who would once again like to create wind chimes and swing sets without resorting to physics trickery (see SLURL below for in world example*) at a substantial prim cost. More so though (as applies to meshes) we would like to rig up our avatars to have flexible additions to avatars (such as tails, or hair) where appropriate. I'd appreciate any word on a review of Havok's potential systems for SL's use, starting with joints. Feynt Mistral 23:49, 3 February 2012 (PST)
[12:04] Yuzuru (yuzuru.jewell) Hello Asha.
[12:04] Asha (ashasekayi.ra) Hiya Yuzuru :)
[12:04] Elie Spot Runitai your avatar is awesome xD
[12:06] Charlar The Terrible (charlar.linden) Well, there really isn't a word to say on Havok's systems. It's not something we have on our roadmap specifically.
[12:07] Charlar The Terrible (charlar.linden) We won't review any possibilities for a number of months.
[12:08] Charlar The Terrible (charlar.linden) Next
[12:08] Charlar The Terrible (charlar.linden) 2) I was speaking with someone recently about the inefficiencies of creating a physics mesh for a house versus replacing it with prims, citing a dramatic ballooning of the LI for even a simple room boxed out with planes. This seems to be a common occurrence with people in the various mesh groups, eschewing the low poly physics mesh option for prims. This makes me sad however as it seems as though a 100-200 poly physics mesh ought to be very light weight as far as LI goes. Is there a way for us to debug the decomposition into convex hulls to identify where problem spots are and also is it possible for a single plane to be a physics boundary in and of itself? Feynt Mistral 23:49, 3 February 2012 (PST)
[12:09] Nyx Linden so from havok's perspective a box is the simplest possible collision volume
[12:09] Leeoh Fox (raven.juno) Rooms are Hard as fuck. All i have to say.
[12:09] Nyx Linden if you want to specify your own mesh, each triangle in the mesh is roughly as complex to compute as a box.
[12:09] Drongle McMahon Easy when you learn how.
[12:10] Nyx Linden though it is tough to say from screenshots alone what exactly can be done to optimize, as the exact cost is dependent on how/if you decompose, what scale you're working at, etc
[12:10] Second Life Flawzell Williamz (heyyoman) is now known as Flawzell Theodore Williamz.
[12:10] Asha (ashasekayi.ra) There's a post on Maeve's good experiments with a room on the forum.
[12:10] Drongle McMahon Triangles are cheaper than decomposed meshes if they are big enough.
[12:10] Nyx Linden there have been some great physics representations that have been done quite cheaply, there's a bit of an art to it
[12:11] Nyx Linden ultimately using prims as physics shapes shouldn't necessarily be problematic, though having a custom mesh can be better under certain circumstances
[12:12] Nyx Linden If there are specifics that we can help with regarding how we cost physics shapes let us know, but best practices tips and tricks is something we encourage people to experiment with and discuss in the builder community
[12:12] Charlar The Terrible (charlar.linden) Posting in the forums is a good way to get feedback from experts as well.
[12:13] Charlar The Terrible (charlar.linden) next item...
[12:13] ÄlveKatt (davido.chrome) Does this mean turning a cube into a yramid doesn't lower physics lag a bit?
[12:13] Charlar The Terrible (charlar.linden) 3) After speaking with a group of fellow fashion designers, we have been wondering a bit about the status of the current mesh deformer project, as well as concerns brought up about how the project itself doesnt quite fit what is needed to make mesh universally excepted on the grid; Or that what was presented to us in the initial discussions of what the deformer could be, is vastly different from what is. Currently with out a deformer, the "Size requirements" are alienating a lot of residents and forcing many to fit a mold that isnt quite them. On the other side of the coin, the deformer project as it is now, is not even a shell of what it could be. It's hardly sufficent, will negate a lot of work that has been done up until its possible release. What is really needed is a deformer based around multiple size options human and non to allow max compatibility to residents across the grid. [Tyr Rozenblum]
[12:14] Charlar The Terrible (charlar.linden) OK, there's not really much of a question in this...
[12:14] draconis.neurocam (draconis.neurocam) yeah
[12:14] Elie Spot Can I pose a question based on that topic?
[12:14] Tyr Rozenblum while not but its something that needs to be discussed
[12:14] Charlar The Terrible (charlar.linden) The deformer project is open source, so we can't really comment on it - you'd need to talk to the actual peopel working on it.
[12:14] Tyr Rozenblum it's a pretty major problem right now on the grid.
[12:15] ÄlveKatt (davido.chrome) If you could post my followup.
[12:15] Charlar The Terrible (charlar.linden) Sure...
[12:15] Charlar The Terrible (charlar.linden) 3.1) don't you think you have a social responsibility to adopt it? That we are hammered by mainstream media that you should change your body to comply with a certain standard in real life is bad enough. Letting clothing designers dictate the size of prim boobs, well, just be better than the mainstream media? [Davido Chrome]
[12:15] Charlar The Terrible (charlar.linden) Elie - go ahead and add something to the agenda, unless Davido's covers your question.
[12:15] Motor Loon As far as I know it's stuck right now on needing some serverside functions...
[12:16] Charlar The Terrible (charlar.linden) That is not entirely true - and not really related.
[12:16] Tyr Rozenblum beyond that, the hesitation to adopt mesh for clothing is based around the fact that most of it alienates many residents.
[12:16] Elie Spot If the parties involved with creating the deformer don't complete it in an acceptable way (to LL or the community), will LL be taking any steps to complete and release it?
[12:16] Charlar The Terrible (charlar.linden) Again, the status of the deformer is not something we should really talk about here.
[12:17] ÄlveKatt (davido.chrome) Is that Oz Lindens meeting then?
[12:17] Nal (nalates.urriah) The latest info from Qarl is in his recent video. You can find links to it here: http://blog.nalates.net/2012/01/24/status-update-mesh-deformer/
[12:17] Nal (nalates.urriah) As best I can tell we are witing on Qarl.
[12:18] Charlar The Terrible (charlar.linden) If they don't finish the deformer, we wouldn't have time to even look at it for some months, and even then I can't say where it would fall relative to other work at that point
[12:18] Charlar The Terrible (charlar.linden) I hope qarl can complete it, and that it's successful.
[12:18] Jus (jusden.jonstone) who within LL, if anyone, is reviewing the deformer code in terms of potential official adoption by LL and who are we to speak with about what criteria LL has in terms of adopting any deformer code?
[12:18] Charlar The Terrible (charlar.linden) Ahh, contact Oz Linden
[12:19] Charlar The Terrible (charlar.linden) Oz is in charge of Open Source, he can take your input and point you in the right direction
[12:19] Jus (jusden.jonstone) i understand its open source project but is Oz the actual one who is deciding whether it will meet LL standdards?
[12:19] Elie Spot In the meantime, a Standard Sizing system seems to be the best alternative. This is really limited though as people who are left out of the basic sizing feel discriminated against. Does LL recognize this dilemma for designers? Would anything be done in the meantime (before the deformer) to help us with this?
[12:19] Jonathan (jonathan.yap) Open Source contributions are approved by the Product Team
[12:19] Nal (nalates.urriah) When I talk to OZ on the subject he often tells me I know as much as he does. They have moved discussion to Qarl's blog.
[12:19] Nal (nalates.urriah) They=Qarl
[12:20] Charlar The Terrible (charlar.linden) That is true, specific to Qarls work.
[12:20] Tyr Rozenblum Deformers and third parties aside, the biggest problem (atleast from the fashion side of it which is a big portion of in world sales), is the lack of adoption of mesh, because of how it works with the avatar, leaving many out. Is there plans at all to look into it in the future
[12:20] Tyr Rozenblum should the deformers not come to be
[12:21] draconis.neurocam (draconis.neurocam) I think it would be up to the residents to create any standards, and up to us to adopt them, i dont understand why no one has made a standard male and female, and then clothes to fit them, and then some other designer can do the same thing, and viola diversity
[12:21] Tyr Rozenblum there is a standard but the standards still leave out quite a few residents, that is the glory of sl
[12:21] Tyr Rozenblum avatars are unique
[12:21] draconis.neurocam (draconis.neurocam) then there are not enough standards
[12:21] Charlar The Terrible (charlar.linden) you're all snowflakes
[12:21] Nal (nalates.urriah) Drac, the problem is peolpe are resisitng changing their shape. No one wants to be a standard size avatar.
[12:21] Tyr Rozenblum and unique avatars require fitted mesh..and in a lot of ways right now, people are having to adjust in a extreme ways to just enjoy the new feature.
[12:21] Boxxy Brown (danny.nolan) even with standards of sizing, you can still have a shape that's "close" to the standard but still have clipping with mesh, etc
[12:21] Elie Spot We did create a Standard Sizing system but because its a lot of work to create 5 different sizes of every clothing article, people who are left out of these ranges feel discriminated against.
[12:22] Asha (ashasekayi.ra) Draconis that has been done. Most residents don't like that conformity.
[12:22] Tyr Rozenblum and to be fair, they shouldnt have too
[12:22] Leeoh Fox (raven.juno) Most residents are sheep.
[12:22] davep (runitai.linden) IRL if you fit outside of those 5 sizes, you go to a tailor
[12:22] Elie Spot A list of stores creating for that system can be found here: http://bloggingsecondlife.wordpress.com/the-stores-creators-lists/mesh-stores-using-standard-sizing/
[12:22] draconis.neurocam (draconis.neurocam) if they don't like them they don't have to buy them, and someone else can come in and take that market
[12:22] davep (runitai.linden) if someone sent you their avatar shape, could you tailor the mesh for them for, say, L$1000
[12:23] Leeoh Fox (raven.juno) IRL there arent 5 sizes theres usually 28. Just saying. starting at 00 and moving all the way up to plus sizes 14 - 28
[12:23] Tyr Rozenblum that is..a very..very unreasonable way to look at it
[12:23] Asha (ashasekayi.ra) Unfortunately, most don't want to pay the extra cost for tailoring.
[12:23] Boxxy Brown (danny.nolan) that's a pretty big inconvenience for the consumer
[12:23] Nal (nalates.urriah) I'm surprised I have not seen more offers to tailor mesh clothes.
[12:23] Boxxy Brown (danny.nolan) the people who drive SL
[12:23] ÄlveKatt (davido.chrome) If you want to mimic RL you need cloth simulation.
[12:23] davep (runitai.linden) for pants, you get lots of sizes
[12:23] Tyr Rozenblum they shouldnt have too..if other games are capable of creating mesh to deform to an avatar shape, why cant a successful game like second life
[12:23] davep (runitai.linden) dresses and shirts, maybe 8
[12:23] Koli Melune (sahkolihaa.contepomi) Avatar shapes aren't always transferable - mine isn't, since it was part of the avatar.
[12:24] Asha (ashasekayi.ra) Nal, people will tailor for the right price. But, many wouldn't pay that price.
[12:24] davep (runitai.linden) and producing longer inseam versions of pants isn't a lot of work
[12:24] Elie Spot Doing custom sizes for individual customers would take a lot of time for the designer. My concern is that the work required to make multiple sizes, especially if 5 sizes isn't enough and we are expected to make 28, will make creating rigged mesh clothing not worthit. If we are discriminating against large portions of the market, we're spending a lot of time on clothing that only a % of people can actually wear.
[12:24] Leeoh Fox (raven.juno) Simple answer. sit the fuck down and wait for the mesh deformers to come out. ... oh my god. suddenly everything is so clear to me
[12:24] ÄlveKatt (davido.chrome) Runitai, I find your comment quite tasteless, in world where people are stressed into anorexia...
[12:24] Nyx Linden in most games, the creators of the avatars and the clothes are a small team of artists that can work within very specific constraints to get things to fit well.
[12:24] Tyr Rozenblum should we bring up blue mars
[12:25] Tyr Rozenblum just to make a comparison.
[12:25] Vincent Nacon blue mars doesn't exist....
[12:25] Vincent Nacon it's a thing of past
[12:25] Asha (ashasekayi.ra) Blue Mars does exist. LOL
[12:25] davep (runitai.linden) AlveKatt: sorry if I said something to offend
[12:25] Vincent Nacon nope... I don't remember
[12:25] Tyr Rozenblum it showed the capability of mesh deformers not just on a single layer but multiple
[12:26] Tyr Rozenblum i guess my point is a lot of money i'm guessing went into creating mesh for the grid, I know lots of testing and all that. We want it to be adopted as baddly as you guys.
[12:26] Tyr Rozenblum but we need help
[12:26] Boxxy Brown (danny.nolan) Also have to keep in mind, that it's us that perpetuate second life
[12:26] Boxxy Brown (danny.nolan) and the demand that drives us to create
[12:26] ÄlveKatt (davido.chrome) Honestly, it feels like the best thing right now would be to accept that Mesh isn't readyfor clothing design.
[12:27] Vincent Nacon it never was
[12:27] Vincent Nacon same for sculpty back then
[12:27] Asha (ashasekayi.ra) The best thing is to wait to see what Qarl does with the deformer since it is only alpha for goodness sake.
[12:27] Charlar The Terrible (charlar.linden) We deferred the mesh rigged clothing back in July
[12:27] draconis.neurocam (draconis.neurocam) you could have been there during the creation process bringing up these topics as it was being created, saying it after the fact makes it harder to change
[12:27] Tyr Rozenblum I wish I could have been
[12:28] Asha (ashasekayi.ra) Some of us did bring it up during beta.
[12:28] Tyr Rozenblum just because i wasnt there doesnt make it any less relevant
[12:28] Boxxy Brown (danny.nolan) It can still be resolved with the code being out now
[12:28] Tiberious Neruda don't mind the snapshot... just me getting a decent image of my av to illustrate something
[12:28] Charlar The Terrible (charlar.linden) Yep, and we needed to focus on other aspects of mesh.
[12:28] Tyr Rozenblum snap away <3
[12:29] Asha (ashasekayi.ra) It was brought up Tyr. They decided to focus on other parts of mesh use.
[12:29] Drongle McMahon Which ones, Charlar?
[12:29] Tyr Rozenblum like?
[12:29] Drongle McMahon Oh ,,, ?I missed the past tense.
[12:29] Charlar The Terrible (charlar.linden) In-world mesh, uploading, Land Impact, etc.
[12:30] Motor Loon yeah... lets get the basics working well before we put new shit on the table
[12:30] Drongle McMahon I thought there was something new :-(
[12:30] Nyx Linden we needed to get the basics of mesh importing correctly first
[12:30] Charlar The Terrible (charlar.linden) oh, sorry Drongle
[12:30] Charlar The Terrible (charlar.linden) this was back in July
[12:31] Vincent Nacon I've been thinking maybe we should import LOD and physic all together in one DAE file, separate them by object name like "LOD1" "LOD3" and "Physic"
[12:31] Motor Loon Yes pls!
[12:31] davep (runitai.linden) part of the collada spec covers physics
[12:31] Motor Loon and pls pls lets have pivot objects back
[12:32] Vincent Nacon there's no point in trying to import more than one objects without prepping the LOD and physic
[12:32] Vincent Nacon because that's mostly for lazy people..
[12:32] Charlar The Terrible (charlar.linden) OK, I think that's run it's course for today.
[12:33] Charlar The Terrible (charlar.linden) next item
[12:33] Charlar The Terrible (charlar.linden) 4) Could you explain the physics costs in this image? They are of simple box prims, simple mesh box (12 triangles), and simple mesh pyramid (6 Triangles) respectively. Why are both Mesh items the same, and why are both heavier on physics than the prim version? [Davido Chrome]
[12:33] Vincent Nacon oh right... sorry
[12:33] davep (runitai.linden) nice composite!
[12:33] Motor Loon http://fc08.deviantart.net/fs71/f/2012/037/b/9/physics_information_by_alvekatt-d4ov6ei.png
[12:34] Charlar The Terrible (charlar.linden) thanks!
[12:34] Charlar The Terrible (charlar.linden) for posting the link...
[12:34] Motor Loon np
[12:34] Nyx Linden as I mentioned earlier, boxes (prim boxes) are really cheap in terms of physics computations, using a mesh is often going to be more expensive for shapes that can be represented by simple boxes.
[12:35] Nyx Linden as to why the two on the right have the same cost, it is difficult to say without having the source files to look at triangle density, scale, etc
[12:36] Drongle McMahon Are the mesh ones decomposed?
[12:36] Nyx Linden also a very important question drongle :)
[12:36] ÄlveKatt (davido.chrome) All objects are three prim.
[12:36] Nyx Linden also are they using convex hull or prim settings for the linksets?
[12:37] ÄlveKatt (davido.chrome) Prim setting.
[12:37] Motor Loon yeah lots of variables there
[12:37] Nyx Linden ok, were the physics shapes decomposed?
[12:37] ÄlveKatt (davido.chrome) DDefine decomposed?
[12:37] Motor Loon That's probably a no °͜°
[12:38] Nyx Linden on the physics tab there are options to decompose the mesh into a set of convex hulls (analyze, etc)
[12:38] ÄlveKatt (davido.chrome) Those models were uploaded to test physics, o I used the same models for the visual part.
[12:38] Tiberious Neruda ...speaking of, what exactly does decomposing a physics shape do for how it acts in-world?
[12:38] Nyx Linden if you're not using those options, your physics costs are likely to be significantly higher
[12:39] Nyx Linden decomposing helps make the collisions *much* cheaper for the havok to compute
[12:39] Drongle McMahon Decomposed = "Analyse" button
[12:39] davep (runitai.linden) does blender have a way to define physical shapes like a box or a sphere and have them tagged as collision shapes?
[12:39] Nyx Linden if you're not careful, it can close holes in windows, etc, but that can be fixed with tweaking
[12:39] Vincent Nacon decomposed to "simplified" your physic mesh... when it knows it could be less. by the given setting
[12:39] Drongle McMahon Nyx ... except for perfect box.sphere/cylinder
[12:40] Tiberious Neruda but breaking that analyzed shape down further really doesn't do much?
[12:40] Drongle McMahon Those are cheaper than their own convex hulls, I think.
[12:40] Nyx Linden honestly if you want pefect boxes/spheres/cylinders I'd probably suggest using prims as collision shapes ideally.
[12:40] Vincent Nacon depends on what you're trying to do
[12:40] Asha (ashasekayi.ra) Davep: You can define a model as a collision shape within Blender's on engine for sure.
[12:40] ÄlveKatt (davido.chrome) I am trying to do low lag racetrack..
[12:40] Nyx Linden Tiberious - the settings you use for the decomposition let you control how accurate you want the result to be, or how performant you want it to be
[12:41] Tiberious Neruda mmk
[12:41] Nyx Linden you should be able to see the results of how close it gets in the preview window
[12:41] Drongle McMahon The uploader should recognise perfect physics primitives in your physics mesh and use them as such, shouldn't it?
[12:41] Vincent Nacon mesh isn't ready for high physic rendering time.
[12:42] Nyx Linden Drongle - I'll have to ask, but I wouldn't assume that to necessarily be true
[12:42] davep (runitai.linden) Asha: when you export it, does it show up in the collada file under library_physics (as a physics_model node) ?
[12:42] davep (runitai.linden) see http://www.khronos.org/files/collada_reference_card_1_4.pdf
[12:42] ÄlveKatt (davido.chrome) What is visually represented there are the simplest physics shapes i could think of to suit my purpose.
[12:42] Asha (ashasekayi.ra) Davep: I don't know. I never looked to see. But, I can try it later.
[12:42] davep (runitai.linden) kk, sounds like a good candidate for an open source project
[12:43] Nyx Linden davido - also note that your server cost is also 1.5 prims, so your total cost won't get below that.
[12:44] Nyx Linden though optimizing physics independent of that is still a worthwile discussion :)
[12:44] ÄlveKatt (davido.chrome) Yeah, but I am thinking about what the physics cost adds up to when I have put 20 of those shapes together to gorm a court. The total cost is arbitrary to what i am trying to learn here.
[12:45] Vincent Nacon the scale factor to prim weight count should be dropped but redefine server weight a bit more to balance it out a bit.
[12:45] ÄlveKatt (davido.chrome) I want to make sure that each vehicle being driven on the track uses the least possible amount of physics resources.
[12:46] Nyx Linden are these objects out somewhere permanently?
[12:46] ÄlveKatt (davido.chrome) Afraid not.
[12:47] Nyx Linden I can forward on the screenshot, but an image doesn't tell us all the information. I'd encourage you to work with people on the forums to get tips about physics optimization, there are a lot of tips and tricks to learn
[12:47] Vincent Nacon here's a tip you can use... you don't really need a physic surface on the bottom of the road if nothing going to be driving much on it
[12:48] ÄlveKatt (davido.chrome) I tried that ona simball ramp. Made people get stuck all the time...
[12:48] ÄlveKatt (davido.chrome) It was weird.
[12:48] Vincent Nacon of course, you have to be aware of the edge collision issue
[12:48] ÄlveKatt (davido.chrome) So I redesigned the ramp to have volume, and no more being stuck.
[12:49] Tiberious Neruda ....how can you have 'one-sided' collision? o.O
[12:49] Vincent Nacon Havok runs that way... as one sided collision
[12:49] ÄlveKatt (davido.chrome) Yes. I designed it with lots of overlap. That's why it's the best Mesh ramp out there. :D
[12:49] Motor Loon onesides works fine
[12:49] Tiberious Neruda huh
[12:49] Vincent Nacon when object have a "volume" it's just push you out of that area
[12:49] Vincent Nacon without volume, no push
[12:50] ÄlveKatt (davido.chrome) Motor, actually. We tested it quite extensively. Onesided doesn't work fine at all.
[12:50] Motor Loon well, probably depends on the purpose
[12:50] Vincent Nacon but of course... mesh isn't ready for physic go-time
[12:50] Asha (ashasekayi.ra) It can for walls.
[12:50] Motor Loon I was testing with avatar collisions on a sailboat... large scale... worked fine being singlesided
[12:51] Charlar The Terrible (charlar.linden) OK, well, I think we should wrap this topic up, here, anyway.
[12:51] Charlar The Terrible (charlar.linden) We've got 10 minutes left
[12:51] Vincent Nacon go ahead, next topic
[12:51] Drongle McMahon There do seem to be occasional leaks when you use a single plane wall.
[12:51] Charlar The Terrible (charlar.linden) I can get you a little metric info, if you're interested...
[12:52] Charlar The Terrible (charlar.linden) There are no more topics.
[12:52] Nal (nalates.urriah) yes
[12:52] Charlar The Terrible (charlar.linden) /me smiles at Nal
[12:52] Vincent Nacon oh
[12:52] ÄlveKatt (davido.chrome) In a ramp and just a little lag, they aren't occasional, as we learned.
[12:52] Drongle McMahon Davep ... I found the jira I made on physics shapes in collada files, ages ago. Don't know if it makes any sense. https://jira.secondlife.com/browse/CTS-176
[12:52] ÄlveKatt (davido.chrome) But then, I guess Simboards taxes the physics engine to it's maximum.
[12:53] Vincent Nacon well... think you could make LODs and physic easier, going by the name of the object instead of importing from another DAE?
[12:54] Tiberious Neruda see, I'm curious why 'faces' with alpha in a rigged mesh acting like an invisiprim is considered a bug
[12:54] davep (runitai.linden) Drongle: it makes lots of sense!
[12:54] Charlar The Terrible (charlar.linden) Only Nal wanted metrics...
[12:54] Nal (nalates.urriah) :(
[12:55] davep (runitai.linden) Drongle: that would be a good thing to throw at open source devs who are frustrated with trying to support convex decomposition
[12:55] draconis.neurocam (draconis.neurocam) i want metrics
[12:55] Asha (ashasekayi.ra) I'd like to see the metrics.
[12:55] Charlar The Terrible (charlar.linden) Ok, so latest mesh adoption:
[12:55] Koli Melune (sahkolihaa.contepomi) /me sees Leslie Linden just fixed the VBO toggle not working bug. \o/
[12:55] Charlar The Terrible (charlar.linden) These are all for Mesh enabled, or what I call "modern" viewers.
[12:55] Charlar The Terrible (charlar.linden) :-)
[12:56] Tiberious Neruda are we still at 30% holdouts?
[12:56] Charlar The Terrible (charlar.linden) and these are just the most recent day's data
[12:56] Gramma Fiddlesticks inlcudes mesh enabled viewers like Pheonix 1.6 and Singularity I assume?
[12:56] Charlar The Terrible (charlar.linden) Daily sessions are at 64%
[12:57] Charlar The Terrible (charlar.linden) Uniques residents 71%
[12:57] Charlar The Terrible (charlar.linden) overall duration was 67%
[12:57] Tiberious Neruda ....'Unique residents' means that the person has, at one time, used a mesh-enabled viewer?
[12:58] Charlar The Terrible (charlar.linden) Each day. It means a person logged in at least once with that type of viewer.
[12:58] Vincent Nacon uniques residents means one person using more than one accounts. I think
[12:58] ÄlveKatt (davido.chrome) Daily sessions? Does that mean that I count 5 times if I log out five times a day to do RL stuff?
[12:58] Charlar The Terrible (charlar.linden) if they use multiple viewers, then we could see them multiple times for unique viewers
[12:59] Charlar The Terrible (charlar.linden) Avle - yes
[12:59] Charlar The Terrible (charlar.linden) it's just all the sessions for all the peoples
[12:59] Nyx Linden yes, but you'd only count once in unique residents assuming you use a single viewer
[12:59] Tiberious Neruda I'm still concerned...
[12:59] Tiberious Neruda we've had mesh for, what, more than half a year now?
[12:59] Charlar The Terrible (charlar.linden) mesh regions are now at over 29% of the grid
[12:59] Charlar The Terrible (charlar.linden) We went live in August.
[13:00] Vincent Nacon no Tiberious, not yet
[13:00] Tiberious Neruda ah
[13:00] Nyx Linden that's why there's a bit of difference between sessions, users, and time spent
[13:00] Tiberious Neruda 5 months
[13:00] Tiberious Neruda why is it we still have 3 out of 10 people not using mesh-enabled clients?
[13:00] Gramma Fiddlesticks almost 6 months. Will be 6 in a few days
[13:00] Motor Loon yeah, its not because the rest dont WANT to see mesh... it's just that some people have issues running the mesh viewers
[13:00] Nyx Linden that's a good question Tiberious :)
[13:00] Charlar The Terrible (charlar.linden) We can guess
[13:00] ÄlveKatt (davido.chrome) I think there is one thing that might speed down adoption. People making good Mesh shit knows to charge good.
[13:00] Boxxy Brown (danny.nolan) UI, I'd wager
[13:00] Charlar The Terrible (charlar.linden) Some people don't like change
[13:00] Vincent Nacon because they're running on a old dinosaurs
[13:01] draconis.neurocam (draconis.neurocam) viewer adoption is one of the saddest things about sl
[13:01] Asha (ashasekayi.ra) Many claim they can't run the mesh viewers on their computers.
[13:01] Squirrely Wrath (squirrel.wood) If you had to handle all those smallcoins you wouldn't like change too :p
[13:01] Gramma Fiddlesticks I have decent computer and it has troubles running Viewers 2 and 3.
[13:01] Tiberious Neruda I doubt that. Unless their system's 10 or more YEARS old...
[13:01] ÄlveKatt (davido.chrome) The ss2 architecture thing?
[13:01] Koli Melune (sahkolihaa.contepomi) /me has an AMD CPU and gets great frame rates.
[13:01] Gramma Fiddlesticks mine is 2 years old.
[13:01] Sera Lok well there are lots of memory issues with the latest releases, sorry to say -shrug-
[13:01] Asha (ashasekayi.ra) Some people have some old as heck rigs.
[13:02] Boxxy Brown (danny.nolan) The engine really needs to be optimized
[13:02] Tiberious Neruda SSE2 should not be any real 'barrier'
[13:02] Squirrely Wrath (squirrel.wood) Memory: 24568 MB
[13:02] Vincent Nacon it should...
[13:02] Koli Melune (sahkolihaa.contepomi) SSE2 has been in CPUs since ... the P4.
[13:02] Vincent Nacon yeah...
[13:02] Vincent Nacon time to move on
[13:02] Squirrely Wrath (squirrel.wood) if it does not do SSE2 it needs to be upgraded. badly.
[13:02] davep (runitai.linden) non-SSE enabled systems were less than 1% of logins when we made SSE a requirement
[13:02] Koli Melune (sahkolihaa.contepomi) Heh..
[13:02] ÄlveKatt (davido.chrome) If you are correct, they should be able to find a computer that can run SL in a dumpster...
[13:03] Boxxy Brown (danny.nolan) The problem with "halting updates to rendering code, etc" is you save a few people, but hurt the larger portion
[13:03] Tiberious Neruda I mean, those people really should be required to upgrade... if they get left behind, well, tough rocks
[13:03] Koli Melune (sahkolihaa.contepomi) Main problem - rendering is stuck in the main thread. **Points to Runitai.** :p
[13:03] Boxxy Brown (danny.nolan) ^^^
[13:03] Drongle McMahon Oh yes .... memory. I have to keep relogging to avoid hyperlag from page thrashing.....
[13:03] Squirrely Wrath (squirrel.wood) any 300$ junk computer from your discounter will run SL better than those 10 year old dinosaurs
[13:03] Tiberious Neruda precisely
[13:04] Charlar The Terrible (charlar.linden) OK, this meeting is over, officially. I have to go, so the notes are going to end here...
[13:04] Koli Melune (sahkolihaa.contepomi) See you, Charlar, Nyx & Runitai.
[13:04] Nal (nalates.urriah) Thx bye
[13:04] Nyx Linden thanks for coming all!
[13:04] ÄlveKatt (davido.chrome) Tiberious, while I agree that they should be able to find a cheap p4, unless they are counted as antiques, give them your old rig instead of sayiong "tough rocks" :)
[13:04] Vincent Nacon more secrets, eh?
[13:04] Drongle McMahon ....usually when sl gets up to more than 600 mbytes
[13:04] Sera Lok have a great week , thank you
[13:04] Yuzuru (yuzuru.jewell) Thank you all
[13:04] draconis.neurocam (draconis.neurocam) another issue i think is the ui should not be anywhere near competition for the amount of cycles used by rendering, but it happens to be
[13:04] Gramma Fiddlesticks thanks for the meeting all Lindnes. Gotta go now myself. Bye evewryone.
[13:04] Squirrely Wrath (squirrel.wood) 64 bit SL client.. when? :)
[13:04] Tiberious Neruda my rig is probably due for an overhaul soon
[13:04] Boxxy Brown (danny.nolan) the renderer really needs a facelift
[13:05] Boxxy Brown (danny.nolan) without a doubt
[13:05] Nyx Linden see everyone next week!
[13:05] Charlar The Terrible (charlar.linden) BYE!!!
[13:05] davep (runitai.linden) later, folks
[13:05] Squirrely Wrath (squirrel.wood) considering 64 bit systems gained a lot of popularity as of late..
[13:05] Koli Melune (sahkolihaa.contepomi) 64-bit SL on Windows isn't possible due to CrapTime being 32-bit only.
[13:05] Sera Lok go ahead, try to tp out! bwahahaha
[13:05] Charlar The Terrible (charlar.linden) thanks very much for all the discussionss.
Retrieved from "https://wiki.secondlife.com/w/index.php?title=Mesh/Archive/2012-02-06&oldid=1162547"