[18:04] Bill Friis: Hooray.
[18:05] Unklar Klaar: prophetic
[18:05] Sidewinder Linden: you're not kidding :)
[18:07] Sidewinder Linden: you're welcome
[18:07] Sidewinder Linden: i'll try to type fast ;)
[18:07] Sidewinder Linden: btw which of these is a safe poseball? :)
[18:07] KJ Hax: :)
[18:07] Sidewinder Linden: oh well not that one - i'll stand
[18:08] Sidewinder Linden: thanks for that intro kj... the last couple of weeks have been a bit busy since the rollout, so i'm planning
[18:08] Sidewinder Linden: for this to be an informal "fireside chat" format...
[18:08] Sidewinder Linden: in other words - sidewinder didn't have time to create a multimedia "presentation" - but hopefully that will allow for more of a discussion and allow you to get more of what you want out of this
[18:09] Sidewinder Linden: i guess i'd like to start by getting a sense of how many of the people here are educators in middle or high school level schools?
[18:09] Sidewinder Linden: (if any?)
[18:09] Unklar Klaar: I'm high school
[18:09] Sidewinder Linden: oool
[18:09] Second Howlett: k-12
[18:09] Sidewinder Linden: ok
[18:09] KJ Hax: Middle School TSL club here
[18:09] Roshana Rives: teach teachers on line
[18:09] Bill Friis: University - teacher ed
[18:09] Unklar Klaar: physics teacher (ooooooooooooo)
[18:09] Sidewinder Linden: hehe
[18:10] Sidewinder Linden: so are the balance of people here in collegiate level environments?
[18:10] Moonshade Pastorelli: Yes
[18:10] Ansa Sautereau: I'm in K-12
[18:10] Bill Smirnov: K-12...district office
[18:10] Silverwing Benoir: business education, post college
[18:10] Fleep Tuque: university here
[18:10] Sidewinder Linden: ok
[18:10] Sidewinder Linden: cool
[18:10] Silverwing Benoir: statistics
[18:10] Bill Friis: A mix
[18:10] Moonshade Pastorelli: community college
[18:11] Sidewinder Linden: i guess as a starting point... i'm curious if the use of simulators of any sort have woven their way into your world yet, or is
[18:11] Sidewinder Linden: second life potentially the first "physics simulation" that you had thought about using for education?
[18:11] Silverwing Benoir: I teach simulation but not 3d
[18:11] Unklar Klaar: yes for me; also video game development
[18:11] Fleep Tuque: SL has been my first experience with it, I facilitate technology use for professors
[18:11] Silverwing Benoir: I've built some simulated trebuchets for experiments via the web
[18:12] Silverwing Benoir: would like to do the same here. hoping this new engine will make it more realistic
[18:12] Sidewinder Linden: cool
[18:12] Bill Friis: Havoc 1 physics were not too useful. I tried some experiments but gave up.
[18:12] Fleep Tuque: We have several engineering professors who want to explore the simulation aspects
[18:12] Sidewinder Linden: well i thought that perhaps i'd describe some of the things i've stumbled upon here as part o the havok4 conversion testing
[18:12] Sidewinder Linden: i have to say the things people have built are pretty amazing
[18:12] Fleep Tuque: In fact, one of our testers asked me to relay a question, whenever you're ready to take them. :)
[18:13] Sidewinder Linden: at the same time, if you' re looking for something that provides detailed modeling, down to the "i want to specify the material and density and collision characteristics of a particular object" level - we're not there yet
[18:13] Sidewinder Linden: ok fleep - i'd like to hold that fora few minutes, if that's ok...
[18:13] Fleep Tuque nods.
[18:13] Fleep Tuque: Sure of course, thanks. :)
[18:14] KJ Hax: If anyone wants to IM me questions, I'll hold them to the end and relay them....
[18:14] Sidewinder Linden: so here are some interesting things i've found in second life in the last few months
[18:14] Sidewinder Linden: some of them you may already know about - some maybe not, but i thought this might create some sense of the possibilities
[18:14] Sidewinder Linden: better than some abstract feature list :)
[18:14] Sidewinder Linden: probably one of the most surprising to me, when i first heard about it, was second life sailing
[18:14] Sidewinder Linden: has anyone here sailed in sl?
[18:15] KJ Hax: Oh yes!
[18:15] Sidewinder Linden: heh
[18:15] KJ Hax: So relaxing!
[18:15] Silverwing Benoir: yhes
[18:15] Sidewinder Linden: well - i have to admit
[18:15] Professora Bailey: I have a friend who really enjoys that
[18:15] Sidewinder Linden: that when i first heard about sailing in second life i was
[18:15] Sidewinder Linden: umm
[18:15] Sidewinder Linden: "dubious"
[18:15] KJ Hax: Me also!
[18:15] Fleep Tuque: Hehe
[18:15] Silverwing Benoir: take real sailing. delete all the boring parts. You have sl sailing
[18:15] Sidewinder Linden: because what i imagined was "ok - push left to go left, right to go right - that's not sailing"
[18:15] Sidewinder Linden: rofl silverwing
[18:15] Sidewinder Linden: for those who haven't sailed, there is a concept called "relative wind"
[18:16] Sidewinder Linden: basiclly it's that there is a wind direction, and the speed of the boat alters the apparent wind vector
[18:16] Sidewinder Linden: well...
[18:16] Sidewinder Linden: it turns out that the sailing community has implemented a quasi weather system and physics model in scripts within the boats!
[18:16] Sidewinder Linden: the boats are networked on a basic level
[18:16] Professora Bailey: wow!
[18:16] Sidewinder Linden: there's a weather system in the form of winds, and they trade information so that they are all sailing in the same "wind"
[18:17] Sidewinder Linden: the boat scripts calculate relative wind
[18:17] Sidewinder Linden: when the sails are pulled in
[18:17] Sidewinder Linden: oh heh
[18:17] Sidewinder Linden: yes main sail,foresails, even spinnakers
[18:17] Sidewinder Linden: when you pull a sail in, the boat heels over
[18:17] Sidewinder Linden: when it tacks it bobs a bit like a real boat (in the top line ones)
[18:18] Sidewinder Linden: so in short.. it's umm surprisingly like the real thing
[18:18] Sidewinder Linden: without, as silver said - the hours of waiting in between moves ;)
[18:18] Silverwing Benoir: and if you don't do it right on my catamaran, it tips over and stops
[18:18] Sidewinder Linden: ahh right !
[18:18] Sidewinder Linden: oh and scripted start/finish lines with timers that talk to the boats to figure out sequnce and winners in races
[18:19] Sidewinder Linden: when we had an early beta ready of the havok4 simulator, i made sure to have a few boat builders take a look at it, and the first hint that they were "really working this" was that the feedback i got started with
[18:19] Sidewinder Linden: ok - that's interesting but not usable
[18:19] Sidewinder Linden: (ok - a beta - no surprise)
[18:20] Sidewinder Linden: followed by three spreadsheets of dynamics data to illustrate the behavioral differences
[18:20] Sidewinder Linden: they had instrumented the boats to be data loggers and show the behavioral differences
[18:20] Sidewinder Linden: so this pretty much told me - people are doing things with physics taht most folks hadn't imagined possible iwth the "basic" simulation that we have
[18:21] Sidewinder Linden: before i describe some other things we've seen - are there ways that these approaches could be reworked to use for the soft of educational goals that you see in your classrooms?
[18:21] Sidewinder Linden: uh oh - this means i either went way to fast or put you to sleep, eh? :)
[18:22] KJ Hax: naaaaah
[18:22] Silverwing Benoir: I'
[18:22] Sidewinder Linden: so as a specific question... let's assume you have a system that has gravity, and phsyical shapes and you can script behaviors like this - what would you build to illustrate physics or scientific principles?
[18:22] Silverwing Benoir: I'd like to be able to not have lag affect the apparent performance of physical objects
[18:22] Sidewinder Linden: ahh yes
[18:23] Sidewinder Linden: "the lag thing"
[18:23] Sidewinder Linden: lag has many faces - which one are you talking about?
[18:23] Silverwing Benoir: and be able to introduce managed variability instead of unpredictable variability
[18:23] Sidewinder Linden: the frame rate - how many updates per second?
[18:23] Silverwing Benoir: not sure yet, still fairly new here
[18:23] Bill Friis: A simple machine should behave as it would in RL.
[18:23] Sidewinder Linden: silverwing... i'm curious about your lag comment - there area few possible solutions...
[18:24] Silverwing Benoir: ok..
[18:24] Silverwing Benoir: if I shoot a cannonball in sl
[18:24] Silverwing Benoir: I want the ball to land in the right spot and take the time it should
[18:24] Silverwing Benoir: regardless of lag conditions
[18:24] Sidewinder Linden: ok
[18:24] Bill Friis: The measurement must be made by a device in world, not on the client.
[18:24] Fumon Kubo: It does. I have a ballistic launcher that works perfectly.
[18:24] Silverwing Benoir: so I can time it with a rl stopwatch
[18:24] Sidewinder Linden: this is an architecture issue - that i'd like to talk about
[18:24] Sidewinder Linden: yup
[18:25] Fleep Tuque: The examples I've seen our profs build include moving parts, like pinball machines, rotating motor gears, that sort of thing
[18:25] Sidewinder Linden: second life, at the moment, does server side physics simulation
[18:25] Sidewinder Linden: that means that all phyisics are calculated on the server
[18:25] Sidewinder Linden: the good news is that this makes things very consistent from viewer to viewer
[18:25] Sidewinder Linden: because the calculations are centralized
[18:25] Sidewinder Linden: the bad new
[18:25] Sidewinder Linden: news
[18:26] Fumon Kubo: You of course, however, do viewer-side prediction based on inertia and dead reckoning.
[18:26] Sidewinder Linden: is that the calculations are centralized, so on updates you have to wait for the server
[18:26] Sidewinder Linden: fumon's my front man
[18:26] Sidewinder Linden: yes - taht's what was coming next :)
[18:26] Fleep Tuque laughs.
[18:26] Sidewinder Linden: in the "shooter games" like halo or other of teh action games
[18:26] Sidewinder Linden: there is a client (viewer side) physics engine
[18:26] Sidewinder Linden: that "assumes what other players are doing" and calculates predictions based on some assumptions
[18:27] Sidewinder Linden: and then once in a while gets corrected by the server
[18:27] Sidewinder Linden: this results in much smooth action
[18:27] Sidewinder Linden: smoother
[18:27] Sidewinder Linden: because the viewer doesn't have to wait for the server to be sure of the next step
[18:27] Bill Friis: Measurement on the client will always be effected by some kind of lag. So move the measuring device in world, yes?
[18:27] Sidewinder Linden: right - if you measure in scripts
[18:28] Sidewinder Linden: in world you'd see pretty consistent results
[18:28] Fumon Kubo: Relatively speaking.
[18:28] Sidewinder Linden: because you're querying the server/physics engine instead of the viewer
[18:28] Sidewinder Linden: yes
[18:28] Silverwing Benoir: yes measuring in world helps that but looks unrealistic to observers
[18:28] slammed Aabye: --how come initial use of new client (on MacBookPro) was laggy, now super smoother - mac or server memory leak?
[18:28] Fumon Kubo: Considering the closest thing we have to actual direct querying only recently became available in llGetObjectDetails
[18:28] Sidewinder Linden: ok im going to ask for a minute that i get a chance to catch people up before we go to advanced questions ok?
[18:29] Sidewinder Linden: one question people ask repeatedly is "well then why don't you do client-side prediction"
[18:29] Sidewinder Linden: the answer is pretty simple - and in two parts...
[18:29] Sidewinder Linden: until fairly recently (couple of years) there weren't a lot of people pushing hard for "shooter game like responsiveness"
[18:29] Sidewinder Linden: other things were driving the interest of folks, and frankly
[18:29] Sidewinder Linden: at the moment still the bulk of folks in world are not focused on that style of activity
[18:30] Sidewinder Linden: however one of the big challenges is the physics engine itself
[18:30] Silverwing Benoir: you don't need client side physics for virtual sex
[18:30] Sidewinder Linden: this project has shown me that the idea of having a "compaitible" viewer
[18:30] Bill Friis: heh
[18:30] Sidewinder Linden: heh tahnks silver
[18:30] Sidewinder Linden: i suppose it all depends on the level of realism you want, but that's for another day
[18:30] Sidewinder Linden: in ANY case
[18:31] Sidewinder Linden: if we were to put in a client side predictive engine it would need to mirror the behavior of the server very closely, because otherwise the "corrections" by the server would be constant and very noticeable
[18:31] Sidewinder Linden: our viewer is open source, and at the moment, the open source physics engines are not particularly clean compared to havok
[18:31] Sidewinder Linden: so we runinto a mismatch problem
[18:32] Fumon Kubo: Yes but if you actually provide accurate data and also build a COMPETENT physics system into the client, the only thing that would have to be corrected for would be script based impulses which could also be cached.
[18:32] Sidewinder Linden: not to mention the fact that adding a physics engine to the client will make it more memory hungry, which is perhaps not a good thing for many people
[18:32] Scalar Tardis: it appears prediction is already happening, when walking forward uncontrollably across multiple sims though I'm not pressing any keys. if that's not prediction of movement..
[18:32] Sidewinder Linden: yes - well it's more like "not stopping the movement" but in a sense there's a tweak of it there
[18:33] Fumon Kubo: It is prediction, dead reaconing. However, that's not a "physics engine" because it dosn't take into effect the actual presence of solid objects.
[18:33] Sidewinder Linden: where it starts getting pretty tricky is to
[18:33] Sidewinder Linden: right
[18:33] Sidewinder Linden: which is the issue i was about to bring up - for instance in this release we've gotten much cleaner about how we repreesent the shape of some complex objects
[18:33] Sidewinder Linden: getting that all to match would be tricky - not impossible but pretty hard
[18:34] Sidewinder Linden: so the bottom line is - yup the best case would be to have both... but there are many tradeoffs
[18:34] Sidewinder Linden: for instance
[18:34] Sidewinder Linden: btw jazmemo - your mic is on
[18:34] KJ Hax: WE're on it....
[18:35] Sidewinder Linden: for instance, we'd end up having to send phsyics information downstream to the viewer and vice versa in addition to the current stream, which would up the network traffic level...
[18:35] Fumon Kubo: You're saying that BASIC bouding box collision engine with the approximations that you're using is difficult? I find that hard to beileve.
[18:35] Scalar Tardis: implementing a local physics engine would probably mean another whole project for LL. You cannot just give us all a havok license, I don't think. it'd have to be some free engine
[18:35] Sidewinder Linden: none of these are insurmountable - but i just wanted to explain that there are some real engineering hurdles, in a fully dynamic world like this, are there
[18:35] Fumon Kubo: Sidewinder, why are you not just leveraging the existing scene graph for that information?
[18:35] Sidewinder Linden: scalar - right... there are those business issues too
[18:35] Sidewinder Linden: well fumon, the immediate answer is actually simple
[18:36] Sidewinder Linden: the havok4 project was not a features project - it was a project to do two things
[18:36] Fumon Kubo: View frustrum?
[18:36] Sidewinder Linden: 1) drop the simulator crash rate
[18:36] Sidewinder Linden: 2) try not to break the world while doing it
[18:36] Scalar Tardis: i dunno what LL is paying for Havok but I'd assume it's in the hundreds of thousands, just for this. supplying for every client.. I doubt havok would allow that
[18:36] Bill Friis: Good goals.
[18:36] Sidewinder Linden: that's not to say we can't do much more now that we have h4 in place - just an explanation that we had those two as the primes
[18:37] Sidewinder Linden: if you missed the results post - here's the blog post with the numerics behind the results of the work: http://blog.secondlife.com/2008/04/10/simulator-stability-improvement-and-crash-rate-reduction-results/
[18:37] Scalar Tardis: I'd like to see better control of the view frustrum. a fixed triangle is not useful for a racing game where stuff far to the side is not seen much anyway.
[18:37] Sidewinder Linden: hmm interesting...
[18:37] Fumon Kubo: In answer to your earlier question about what would be beneficial to classroom physics demonstrations, lower the collision distance.
[18:38] Sidewinder Linden: collision distance - the 0.1m collision tolerance?
[18:38] Scalar Tardis: or completely separate the frustrum from the user, so that they can wander around inside a rezzed up object space and not have to worry about things appearing and disappearing as they move
[18:38] Fumon Kubo: At least add the hack for the avatars to make it seem like their feet were actually touching the ground back in.
[18:38] Fleep Tuque: Yes, that was my question from earlier, the collision distance.
[18:38] Sidewinder Linden: ahh
[18:38] Fumon Kubo: I was one of the original sim testers but I don't enjoy the JIRA system.
[18:39] Sidewinder Linden: i'd have to do some checking to be sure, but i believe that that is actually part of havok4
[18:39] Fumon Kubo: If you're standing on a prim, you still hover.
[18:39] Sidewinder Linden: if i recall correctly we could work around it by tinkering with the visual representation but that's fraught with complexity
[18:39] Fleep Tuque: (what the heck is a view frustrum? if I may ask)
[18:39] Scalar Tardis: I have been searching the web for discussion of the collision tolerance but that info is probably only available to engine/game developers
[18:39] Fumon Kubo: COMPLEXITY? It's a two line edit in the viewer code.
[18:39] Sidewinder Linden: i'll try to put some info together on that scalar
[18:40] Scalar Tardis: the frustrum is the light colored triangle you see in the mini-map
[18:40] Fleep Tuque: Oh.
[18:40] Fleep Tuque: :)
[18:40] Sidewinder Linden: umm fumon - to accurately handle collision tolerance adjustments for the arbitrary shapes in world is not so simple
[18:40] Fumon Kubo: I was just talking the avatar hack.
[18:40] Sidewinder Linden: it also means changing volume calculations, mass calculations and dealing with building tools
[18:40] Fumon Kubo: Not the collision shapes.
[18:40] Sidewinder Linden: oh just foot positions?
[18:40] Sidewinder Linden: well - heh
[18:40] Fumon Kubo: Yes.
[18:41] Sidewinder Linden: so here we get into the reality of having 130 TB of user data
[18:41] Fumon Kubo: Well it was correct before H4.
[18:41] Sidewinder Linden: it turns out that any time we tinker with avatar geometry - physically or otherwise things are affected
[18:41] Sidewinder Linden: ok - this is one we should probably handle offline then... there are some differences with h4 that are built-in, and many considerations of compatibility with existing products
[18:41] Scalar Tardis: the collision tolerance is a region that havok monitors for activity. it's like a fuzzy edge as best I can tell. by making it external to a shape it makes sure that objects don't seem to "sink in" due to the fuzzy detection.. AFAIK. YMMV.
[18:42] Sidewinder Linden: we did adjust foot positions several times to try to reach a reasonable compromise
[18:42] Sidewinder Linden: yup scalar - a lot of what they did with havok4 were things to allow increased calculation sophistication while trying to lower the calculation cost (how hard it hits the server or computer)
[18:43] Sidewinder Linden: by having a collision tolerance, it can "ignore" things outside the tolerance in calculations
[18:43] Scalar Tardis: the avatar floats above the ground because they too are physical objects and also have a 0.100 m collision tolerance
[18:43] Sidewinder Linden: yes
[18:44] Sidewinder Linden: it sounds like part of what we are talking about here is non-technical - there is a philosophy point
[18:44] Sidewinder Linden: that is pretty important
[18:45] Scalar Tardis: probably the solution is the invisible substitute collision volume, where an invisible shape is bound to a phantom object. That would allow cubes stacking tight by making the invisible prim 0.100 m smaller than the phantom cube.
[18:45] Sidewinder Linden: as educators - you may be thinking about precision of simulation
[18:45] Bill Friis: I have heard Havoc 4 is just as inaccurate, but will let a larger number of objects misbehave.
[18:45] Sidewinder Linden: yes - scalar - and we decided that that was too complex to do within the scope we were working under, and thought it more important to reduce simulator crash rates
[18:46] Sidewinder Linden: i'd like to table more collision tolerance discussion if that's ok, as i suspect it's not a general interest topic
[18:46] Fumon Kubo: Havok 4 has the same problems with precision as all computer simulations in real time, the differences involve how quickly it calculates and how well it handles being de-prioritized and crashing/errors.
[18:46] Sidewinder Linden: right
[18:46] Bill Friis: For educators, a small number of objects behaving accurately is what is important.
[18:46] Scalar Tardis: heh, phjantom objects cannot be connected to physical objects. EXCEPT when flex enabled. now if only I can make a flexy 100% stiff and solid..
[18:46] Joss Charman: i like the idea of sending questions to KJ Hax first
[18:46] Sidewinder Linden: heh
[18:47] Sidewinder Linden: we could do this that way - i'd actually like to flip to a question that i'd like to ask...
[18:47] KJ Hax: Yes folks, IM me Q's please. :)
[18:47] Flight Band: All Go
[18:47] Sidewinder Linden: one of the realities of a world such as second life is that it is a general purpose simulation
[18:47] Sidewinder Linden: it is not optimized for detailed, high accuracy (very high frame rate) calculation of object dynamics
[18:48] Sidewinder Linden: it's instead optimized to handle things like "not falling over when people dump 1000 random shapes in a region from 100m in the air"
[18:48] Sidewinder Linden: and having reasonable dynamics with a scripting language that allows you to build interesting things
[18:48] Sidewinder Linden: now that's not to say you can't do real simulation
[18:49] Sidewinder Linden: but it is not really a substitute for high precision simulation
[18:49] Sidewinder Linden: with that all said... you can build things taht are visually compelling and illustrate points, with low effort and 'reasonable" realism
[18:49] Sidewinder Linden: an analogy i'd draw is to the build tools themselves
[18:49] Sidewinder Linden: <- comes from a CAD background...
[18:50] Koni Lanzius: hi Eren!
[18:50] Sidewinder Linden: when i look at the build tools in second life, i'm constantly amazed at what people build, because they are pretty basic compared to "real computer aided design" systems
[18:50] Eren Padar: Hi dere Koni!
[18:50] Sidewinder Linden: but they're simple and effective
[18:50] Eren Padar: Hi dere Sidewinder!
[18:50] Eren Padar: Nicetameecha!
[18:50] Sidewinder Linden: hi eren
[18:50] Knowclue Kidd: lol
[18:50] Sidewinder Linden: if you think about the physics modeling in the same way, you might find that you can build many interesting things that are visually compelling and satisfy the teaching objectives of understanding
[18:51] Sidewinder Linden: without necessarily having the "smooth like butter" or "high precision" of a "heavy physics simulator" (which by the way you'll spend many moons programming in c++ to do even the basics :)
[18:52] Sidewinder Linden: this is the kind of thing i hear from folks building physics-based projects in second life - does this general philosophy work for the type of education that you do, or is precision simulation "the only and right way" to get the points across in your environments?
[18:52] Silverwing Benoir: I don't need precision simulation at all
[18:53] Sidewinder Linden: i suspect there are some folks from both sides here... but it is a philosophy question first... so i wanted to lay it out that way ;)
[18:53] Fumon Kubo: I don't believe that we need "high performance, precision" simulation. I simply argue that there are better ways of making the cosmetic appeal of the current system better.
[18:53] Silverwing Benoir: frankly, I'm more concerned with system reliability. Can I schedule a class and be confident that the system will be running well enough to hold it!
[18:53] Scalar Tardis: well, its difficult to do many things due to the high imprecision of the simulation. the framerate can vary wildly on a per-frame basis and seems to be the cause of many objects "escaping" from enclosures
[18:53] Sidewinder Linden: ahh ok
[18:53] Eren Padar pretty satisfied with "works moderately well". : )
[18:53] Sidewinder Linden: good points all
[18:53] Sidewinder Linden: there are a few things here
[18:54] Scalar Tardis: suddenly for a moment the server drops to 1 FPS and an object can penetrate ten times its previous speed, and so hops over what was previously an acceptable barrier
[18:54] Fleep Tuque: I think some of our profs would say the more precise the better, but as a more general user, I'd say better functionality, things actually _working_ like marbles not falling out of containers, would be good.
[18:54] Sidewinder Linden: btw scalar - the "escapes" problem is a curent bug that we're going to look at... there is definitely something going on there that wasn't in earlier versions
[18:54] Eren Padar: good
[18:55] Bill Friis: A wheel should stay on its axel.
[18:55] Sidewinder Linden: right - the falling through, getting through walls problem is a current bug - i've seen it
[18:55] Sidewinder Linden: we do want to work on that but have had to handle some other things first
[18:55] Eren Padar: Yeah, I've even flown through walls lately. Pretty much lots. :D
[18:55] Fumon Kubo: Another bug is the flight inertia buffer is done... completely wrong.
[18:55] Sidewinder Linden: one thing is that with any system like this, the simulation is not continuous, it is discrete
[18:55] Sidewinder Linden: ok wait a minute....
[18:55] Fleep Tuque: I guess those are the things that I see as most problematic, marbles out of containers, rotating parts that for some reason get jammed..
[18:55] Sidewinder Linden: throwing a bug list at me isn't probably what everone wants right?
[18:55] Sidewinder Linden: be happy to handle that in a bug forum or offline ;)
[18:55] Eren Padar: yups.:D right
[18:55] Fumon Kubo: I will speak with you later.
[18:56] Sidewinder Linden: kk tahnks
[18:56] Sidewinder Linden: one question (not sure whose it was) was about general system stability
[18:56] Sidewinder Linden: agreed - that's something that we all need
[18:56] Scalar Tardis: well some of these I don't know if it is a bug or just injerent to the havok engine
[18:56] Sidewinder Linden: while i can't go through the whole project plan for that, there are several teams working on those issues
[18:57] Sidewinder Linden: it is quite clear that we need to be better at it - however... the part of system stability that we (the team i'm working with) have done something tangible about is simulator stability
[18:57] Sidewinder Linden: in the face of physics operations, which used to be the weak point of the simulator
[18:58] Sidewinder Linden: we used to have on the order of 14-1800 region crashes a day, and now it's down in the handfuls based on phsics issues
[18:58] Fumon Kubo: It is indeed impressive that you've been able to decouple your server code from the physical simulation... something that most systems are designed to do anyways.
[18:58] Sidewinder Linden: scalar are you umm actually typing (speaking of bugs ;)
[18:59] Scalar Tardis: as I understand it the physics engine is really only meant to keep people from walking through walls and sinking through floors. Stuff like marble toys and gears are just icing but not the core engine function
[18:59] Sidewinder Linden: ahh ok
[18:59] Scalar Tardis: yes :)
[18:59] Bill Friis: So your recommendation would be that we take what the physics engine does and try to work within it.
[18:59] Scalar Tardis: I'm already doing that. :)
[18:59] Sidewinder Linden: bill - yes - and that's not to say we can't do more with it - we will
[18:59] Sidewinder Linden: but i think you'll find faster success and easier paths to the goal if you think from the perspective of
[19:00] Sidewinder Linden: "how can i take advantage of what is
[19:00] Sidewinder Linden: than to build at the limits and hope that it will be better than it is ;)
[19:00] Scalar Tardis: I was wondering if a permission system could be built into the collision processing.
[19:00] Sidewinder Linden: for instance the "things falling off rails and such" issues - because the simulation is discrete - samples of motion over time slices
[19:00] Sidewinder Linden: if something is moving fast and toward a thin surface
[19:01] Sidewinder Linden: the next "time slice" may mean that the object is already past the surface, and no collision is generated
[19:01] Sidewinder Linden: this is something that all "game systems" have - even the client-side shooter games
[19:01] Fumon Kubo: I don't think that the question should be "Does the community want more presicion". I think that history has shown that a system that has far more flexibilty will allow the community to create new and unanticipated features. I think the focus should be to give us the best possible balance and then to continue to increase performance and features in multiple areas then to apply "magic patches" to problems.
[19:01] Sidewinder Linden: however.... they have the advantage of having a "world" that is all built by the developers
[19:01] Sidewinder Linden: so if they have a particular wall that people "shoot through" they make it thicker
[19:02] Sidewinder Linden: here anyone can build anything, and accidentally hit edges like that that are not obvious until you dig in pretty far
[19:02] KJ Hax: Sidewinder, if I may ...
[19:02] Sidewinder Linden: agreed fumon
[19:02] Sidewinder Linden: sure kj
[19:02] KJ Hax: This hour went REALLY fast!
[19:02] KJ Hax: We can keep going but
[19:02] KJ Hax: I wanted to respect everyone's
[19:02] KJ Hax: commiuttment to the end time.
[19:02] KJ Hax: What do you all think?
[19:02] KJ Hax: Wanna keep going?
[19:03] KJ Hax: I have a queston from Ansa....
[19:03] Tara5 Oh: Very interesting I missed the first part so I can keep going!
[19:03] Sidewinder Linden: hehe it's up to you guys
[19:03] Sidewinder Linden: voice is fine with me
[19:03] Tara5 Oh: Can anyone send me the chat!
[19:03] Sidewinder Linden: i've been typing all day ;)
[19:03] Scalar Tardis: well, if I could have the ear of a Linden willing to stay and talk...
[19:03] Tara5 Oh: me please!
[19:04] Sidewinder Linden: sure...
[19:04] Sidewinder Linden: would you like that in text?
[19:04] Roshana Rives: great idea
[19:04] Sidewinder Linden: or voice?
[19:04] KJ Hax: text
[19:04] KJ Hax: for the record!
[19:04] KJ Hax: :)
[19:04] Fleep Tuque: I can send a transcript, will need a sec to edit out IMs and whatnot. :)
[19:04] KJ Hax: ty
[19:04] Sidewinder Linden: heh ok
[19:04] Eren Padar: And in case someone is deaf and cannot hear voice. : )
[19:04] Sidewinder Linden: well...
[19:04] Sidewinder Linden: hmm if i rez something - willthere be complaints of favoritism?
[19:04] Joss Charman: text would get more kudos than voice :)
[19:05] Sidewinder Linden: it's just an example of a sophisticated vehicle
[19:05] Bill Friis: Feel free, if you can.
[19:05] KJ Hax: go ahead and rez!
[19:05] Sidewinder Linden: that i used as a "compatibility challenge" for the team
[19:05] Scalar Tardis: mm, wonder woman's invisible plane
[19:05] Sidewinder Linden: takes a while to rez sometimes
[19:06] Sidewinder Linden: hmm
[19:06] Sidewinder Linden: looks at simulator stats and wonders why? :)
[19:06] KJ Hax: scripts are disabled
[19:06] Bill Friis: Eren is in.
[19:06] KJ Hax: just fyi
[19:06] Sidewinder Linden: well jk... that umm "shouldn't matter" ;)
[19:06] Sidewinder Linden: but could be with the way this one is built
[19:06] Eren Padar: I just see some kind of white grid in form of a plane?
[19:06] Sidewinder Linden: ok i'll make it go away ratehr than taking time troubleshooting
[19:07] Sidewinder Linden: so that is an f-16 fighter jet
[19:07] Sidewinder Linden: that happens to be built with not just some nice visuals
[19:07] Sidewinder Linden: it's got something like 8000 lines of script code in it iirc
[19:07] Sidewinder Linden: it has a full turbine start up cycle - so that when you start it you hear the whole start cycle of a jet
[19:07] Sidewinder Linden: it has a combat system
[19:08] Sidewinder Linden: so tath when you fly - you can do a "shooter style" combat environemnt
[19:08] Sidewinder Linden: some basic aerodynamics
[19:08] Sidewinder Linden: now these are not "precision" in the sense that we talked about
[19:08] Sidewinder Linden: but it does illustrate the idea of a pretty complex piece of work
[19:08] Sidewinder Linden: and again using the idea of networking objects together to be able to deliver a behavior that's pretty complex
[19:09] Sidewinder Linden: another i saw a while back was a simulation of gravitation of planets
[19:09] Sidewinder Linden: this one actually bypassed the second life physics engine ecxept
[19:09] Sidewinder Linden: that the planets were set up as "floaters" like hover ships
[19:09] Sidewinder Linden: have you all seen hover shiops?
[19:09] Sidewinder Linden: ships?
[19:09] Eren Padar: yups
[19:10] Ansa Sautereau: Yes
[19:10] Sidewinder Linden: ok so basically each planet was a hover vehicle
[19:10] Scalar Tardis: nope, but continue. ;)
[19:10] Sidewinder Linden: with scripting and networking between them to communicate where they were
[19:10] Sidewinder Linden: with respect to each other
[19:10] Sidewinder Linden: ahh just a min
[19:10] Koni Lanzius: nice!
[19:10] Eren Padar: oooh
[19:10] Sidewinder Linden: a hover ship doesn't fly with aerodynamics
[19:10] Ansa Sautereau: Beautiful!
[19:11] Sidewinder Linden: it just "hovers"
[19:11] Sidewinder Linden: and has controls for forward back etc
[19:11] KJ Hax: WHOA NICE!!!!
[19:11] Sidewinder Linden: it "flies" prety much the way an avatar flies in terms of controls
[19:11] Sidewinder Linden: so in any case... each planet was a hover ship
[19:11] KJ Hax: please leave it up
[19:11] Sidewinder Linden: with bouyancy set to neutral
[19:11] Sidewinder Linden: sure
[19:11] KJ Hax: I need some more pics!
[19:11] Sidewinder Linden: hehe
[19:12] Sidewinder Linden: i should turn it around for you
[19:12] Sidewinder Linden: but in any case
[19:12] Sidewinder Linden: this one's a freebie btw ;)
[19:12] Sidewinder Linden: the planets were networked to relay position information to each other
[19:12] Sidewinder Linden: and would apply impulses to create "gravitational forces" between them
[19:13] Sidewinder Linden: it was pretty simplistic, but you could start one around and watch it be affected by other planets on the way around
[19:13] Sidewinder Linden: that's one example of using these tools in a way that leverages what they are
[19:13] Sidewinder Linden: for instance
[19:13] Sidewinder Linden: you could change the size of a planet
[19:13] Sidewinder Linden: and it's gravitation would change
[19:13] Sidewinder Linden: kinda neat eh?
[19:13] Kumangaro Hayes: mmhmm
[19:13] Eren Padar: yups
[19:14] Clare Lane: very cool
[19:14] Sidewinder Linden: and actually that's really using scripting to simulate physics, right? because the "physics" was really just used to keep things lofted at the right height
[19:14] Sidewinder Linden: let's see - other things
[19:15] Fumon Kubo: I think that has less to do with circumventing and more to do with the fac that we have no control over the server-side simulation varriables.
[19:15] Sidewinder Linden: well... you're probably seen cars and motorcycles in second life?
[19:15] Bill Friis: many
[19:15] Fleep Tuque: Yes, and I'm a terrible driver. :)
[19:15] Eren Padar: Lots of which we have trouble controlling now, but that's expectd with any new physics system
[19:15] Eren Padar figures all that will smooth out with time. : )
[19:15] Sidewinder Linden: ahh eren - that is one side effect of change... the builders who were invovled with the beta
[19:16] Koni Lanzius: whoa!
[19:16] Sidewinder Linden: asked us to up the motor speed
[19:16] Fumon Kubo: My "Back to the future" cloned hoverboard still works perfectly.
[19:16] Eren Padar: Ah! Well, it's upped. LOL
[19:16] Sidewinder Linden: knowing that that would make some things too fast - if the script asks for an excessive speed
[19:16] Sidewinder Linden: in h1
[19:16] Sidewinder Linden: it's clamped to 64 m /S
[19:16] Sidewinder Linden: in h4 the max is 256m/S
[19:16] Bill Friis: snappy
[19:16] Sidewinder Linden: many builders on the old simulator would combine the "motor" with "tapping the back o the vehicle" with impulses to make it go faster
[19:16] Eren Padar: So does that feed down on a ratio basis then?
[19:16] Sidewinder Linden: now if they do that it's umm "much faster"
[19:17] Sidewinder Linden: not sure what you mean eren?
[19:17] Eren Padar: Don't mean to interrupt
[19:17] Fumon Kubo: It's a clamp.
[19:17] Sidewinder Linden: stranger i think your mic is on
[19:17] Scalar Tardis: at that speed isn't the real limiter the sim-crossing handoff speed?
[19:17] Fumon Kubo: All over that limit would be considered the limit.
[19:17] shamblesguru Voom: Is it possible to take a copy of the C-Tech F2050 anti grav ship to investigate later .... seems some interesting stuff happening in the content
[19:17] Eren Padar: What I've been mainly noticing is vehicles turning much faster... 90 degrees at the bare touch of an arrow
[19:17] Sidewinder Linden: sim crossing handofs are a whole differnet thing
[19:17] Sidewinder Linden: shambles
[19:17] Sidewinder Linden: i think if you search for c-tech they have a store and you can pick one up with docs and all
[19:18] Sidewinder Linden: they have a freebie and then paid fancier versions
[19:18] shamblesguru Voom: thx
[19:18] Sidewinder Linden: hmm eren - yes the thing is that people have built vehicles in many ways
[19:18] Sidewinder Linden: some builders did some "very havok1-specific" things and others did not
[19:18] Sidewinder Linden: the current behaviors are somewhat of a tradeoff
[19:19] Sidewinder Linden: and the other part is that builders are coming out with new products that take advantage of the new behaviors
[19:19] Fumon Kubo: Fact of life, unintentional functionality built out of getting around limitations.
[19:19] Eren Padar: Yeah
[19:19] Sidewinder Linden: one race car builder said that their scripts got much simpler with h4, and the simulator load is lower
[19:19] Sidewinder Linden: yup
[19:19] Sidewinder Linden: so interms of physics
[19:19] Sidewinder Linden: you could build something like a motorcycle as a simple machine demo - to illustrate geearing as one example
[19:20] Sidewinder Linden: but on a purely "other" level
[19:20] shamblesguru Voom: so with the new physics engine .... if you fly upsidedown does your avi fall out if no seat belt ;-)
[19:20] Sidewinder Linden: and you may sneer at this from a purist perspective
[19:20] Eren Padar: LOL
[19:20] Sidewinder Linden: but - if you're presented with something that looks like this as an example
[19:20] Bill Friis: Seatbelt simulations would be really valuable. :-)
[19:20] Sidewinder Linden: some students may be much more interested than a "pure cylinder rotating on a shaft" as you'd get in a traditional simulation
[19:21] Sidewinder Linden: so i think it really brings a few things together
[19:21] Sidewinder Linden: you can build interesing content
[19:21] Sidewinder Linden: and make it visually interesting
[19:21] Sidewinder Linden: with less work than most other environments
[19:21] Sidewinder Linden: and allow students to tinker with it - and play with the results "in front of each other"
[19:21] Eren Padar: Question: in your opinion Sidewinder, will vehicles in general,programmed correctly, run better on Havok 4 than Havok1 or about the same?
[19:22] Sidewinder Linden: the feedback that i'm geting is that people who are tuning for h4 are finding that things work better, smoother
[19:22] Sidewinder Linden: motor speed is one thing
[19:22] Eren Padar: Kewl. : )
[19:22] Sidewinder Linden: rather than having to add impulses to get speed
[19:22] Sidewinder Linden: you can just set a high speed and let the motor code regulate it
[19:22] Fumon Kubo: Speed interpolation between velocities is a lot better ramped then it was and the higher cap makes it easier to tune the behavior.
[19:22] Sidewinder Linden: on the region crossing issue
[19:22] Kumangaro Hayes: i like the sound of that
[19:22] Eren Padar: Yeah
[19:23] Sidewinder Linden: i should probably hit region crossing between the eyes, because everyone wants that "fixed" :)
[19:23] Sidewinder Linden: right?
[19:23] Bill Friis: Right!
[19:23] Eren Padar: Yup yup yup!
[19:23] Sidewinder Linden: in our current architecture
[19:23] Kumangaro Hayes squeels
[19:23] Sidewinder Linden: region crossings are gated by a couple of things
[19:23] Fumon Kubo: It's a problem inherent in the server infastructure in trasfering the information to other simulators because you don't do dynamic processing of region based on actual need...
[19:23] Sidewinder Linden: 1) script complexity of the "assemblage" crossing the boundary
[19:23] Sidewinder Linden: in a non-vehicle example...
[19:24] Sidewinder Linden: many perforamnce areas these days are set up as a stage on one region, and the audience on a region adjacent
[19:24] Sidewinder Linden: to improve performance
[19:24] Sidewinder Linden: at one theater the mc was complaining bitterly of how bad region crossings to walk into the audience were
[19:24] Scalar Tardis: yes... I've thought about that.. a stadium surrounding a playfield in another sim
[19:24] Sidewinder Linden: we did a quick experiment
[19:24] Eren Padar: Kewl idea
[19:24] Sidewinder Linden: i had him remove all of his huds
[19:24] Sidewinder Linden: he was wearing five
[19:24] Eren Padar: LOL
[19:24] Sidewinder Linden: and walk across the boundary
[19:25] Sidewinder Linden: he practically fell over with surprise
[19:25] Sidewinder Linden: because "he just walked across"
[19:25] Bill Friis: Like to have seen that.
[19:25] Kumangaro Hayes: i love you sidewinder
[19:25] Sidewinder Linden: so - why is this?
[19:25] Sidewinder Linden: it's because today, to cross
[19:25] Sidewinder Linden: a region boundary we have to state-snapshot the scripts on the avatar - all of them
[19:25] Fumon Kubo: You're using a destructive method of transfering scripts and not using server time on the original server while the script is being unrolled on the other server.
[19:25] Sidewinder Linden: move them to a new server
[19:25] Fleep Tuque is usually using 6 HUDs at least.
[19:25] Sidewinder Linden: and then instantiate them and "hot start them" on the new server
[19:26] Scalar Tardis: I seem to recall that region crossing involves actually destroying the old UUID from one sim and making an all-new UUID version in the next one.
[19:26] Sidewinder Linden: hmm not sure scalar, but not much surprsies me these days
[19:26] Sidewinder Linden: in any case
[19:26] Fumon Kubo: It packs the script up as it's currently running, sends it to the new sim and then unpacks it again.
[19:26] Eren Padar: LOL
[19:26] Sidewinder Linden: you can see that the process of serializing the script and avatar environment
[19:26] Sidewinder Linden: moving it
[19:26] Sidewinder Linden: deserializing it and restarting on the new simulator is a big deal
[19:27] Sidewinder Linden: so...
[19:27] Fleep Tuque: Get rid of all these HUDs. :)
[19:27] Sidewinder Linden: i guess the moral is that with this architecture, reducing script weight is important
[19:27] Sidewinder Linden: yup
[19:27] Eren Padar: Yeah, its cumbersome. Plans in the works for redoing that?
[19:27] Fumon Kubo: The data transfer takes time, the roll up takes time and the unrolling takes time. It shouldn't take that much time if it was done correctly but it was engineered improperly in my opinion.
[19:27] Fleep Tuque cries.
[19:27] Fleep Tuque: ;)
[19:27] Clare Lane: HA! Just like a star Trek transporter heehee
[19:27] Sidewinder Linden: or (nudge nudge) convince hud vendors to offer lightweight huds or region boundary sensing huds that "go light" on crossings" ;)
[19:27] Eren Padar has noticed that the fewer prims and scripts worn, the less teleport refusals and crashes happen.
[19:27] Bill Friis: Oh, good idea.
[19:27] Fumon Kubo: Or provide client-side scripting so that we don't have to USE the crummy hud system.
[19:28] Sidewinder Linden: there are some architectural things that we can and want to do to make this much more efficient
[19:28] Sidewinder Linden: fumon - that's on epossible idea...
[19:28] Sidewinder Linden: but the current reality is that this is a driver for region crossing delays
[19:28] Fumon Kubo: I have been playing with my own implementation of LUA and XML based interfaces to server based LSL scripts.
[19:28] Sidewinder Linden: 2) complexity of the assemblage going across the boundary
[19:28] Sidewinder Linden: heh cool :)
[19:29] Scalar Tardis: is there some reason the sims have to be 256x256? A huge flight sim of 25600 by 25600 would eliminate region-cross problems
[19:29] Sidewinder Linden: fumon have you tracked the mono project at all - might be worth hunting down periapse and company to talk about what they're doing performance-wise
[19:29] Fumon Kubo: Because of backwards compatability and because they don't want to upset conciarge.
[19:29] Fleep Tuque: (I think they had some discussion about region size, shape, etc. in the Architectural Working Group
[19:29] Fleep Tuque: )
[19:29] Fumon Kubo: Mono is a "magic patch" fix.
[19:29] Sidewinder Linden: scalar - yes huge sims would be great
[19:29] Scalar Tardis: ah, ok, I'll research the chat logs
[19:30] Fleep Tuque: ((Think it was in SL-DEV archives, Scalar)
[19:30] Sidewinder Linden: the reality is we have some assumptions in the code that would need to be engineered out - in many places outside the simulator - for that to be realistic
[19:30] Eren Padar would love to see a "quad sim"
[19:30] Sidewinder Linden: it's been discussed
[19:30] Eren Padar: That's good news. : )
[19:30] Eren Padar: Hopefully someday implemented.
[19:30] Fumon Kubo: It's not that difficult... most of the problems would be fixing the existing problems with using region co-ordinate systems the way they are tiered now.
[19:31] Kumangaro Hayes: what about prim counts tho woudl they stay the same or be quadrupled
[19:31] Sidewinder Linden: ok
[19:31] Sidewinder Linden: lets slow down a bit ok - can't answer that fast
[19:31] Sidewinder Linden: ;)
[19:31] Kumangaro Hayes: hahaha
[19:31] Eren Padar: : )
[19:31] Kumangaro Hayes: k
[19:31] Scalar Tardis: well afaik the entire grid code might need to be redesigned based on how sims are "located" on the grid in general.
[19:31] Sidewinder Linden: prim counts
[19:31] Bill Friis as to go.
[19:31] Eren Padar: (It's all very exciting). :D
[19:31] Sidewinder Linden: prim counts
[19:31] Sidewinder Linden: remember our core goals for the first havok4 project?
[19:32] Scalar Tardis: stop the crashing.
[19:32] Sidewinder Linden: the hidden "anti-goal" was "attempt to do NO NEW FEATURES" :)
[19:32] Scalar Tardis: that was the main one. and you succeeded. no more crashing
[19:32] Sidewinder Linden: because every new feature is a place to find new bugs and create new crash vectors
[19:32] Eren Padar: Yeah agreed there.
[19:32] Sidewinder Linden: so - a lot of these are really cool possibilities, and we hope to be able to head there, but killing off crash modes was the most important thing
[19:32] Fumon Kubo: Please start using unit-tests.
[19:33] Sidewinder Linden: and doing it while having even "most" things working was hard enough! :)
[19:33] Sidewinder Linden: fumon - we really need to have a side discussion about these issues ok?
[19:33] Sidewinder Linden: and fwiw we do have unit tests, but creating 100% coverage for this system is "daunting"
[19:33] Eren Padar: If not impossible.:D
[19:33] Fleep Tuque: I'm afraid I must head out too. Thanks for visiting with us Sidewinder, it's always good to hear from you about what's happening. Appreciate your time and all the furious typing. :) Have a good evening all.
[19:34] Eren Padar: Waay to many variables to think of everything.
[19:34] Eren Padar: Nitey nite!
[19:34] Sidewinder Linden: thanks fleep
[19:34] KJ Hax: Nite Fleep!
[19:34] Sidewinder Linden: btw kj can i take the hover ship down?
[19:34] Tara5 Oh: Nite Fleep!
[19:34] KJ Hax: Yes
[19:34] Eren Padar: Sidewinder, can I ask a "when" question?
[19:34] Sidewinder Linden: sure
[19:35] Eren Padar: Any eta on when the new client will allow above 768 building?
[19:35] Sidewinder Linden: hmm i wonder if this will rezz here
[19:35] Sidewinder Linden: ahh
[19:35] Kumangaro Hayes: oh you better be handing these out!
[19:35] Eren Padar: oh nice car
[19:35] Koni Lanzius: wow
[19:35] Sidewinder Linden: eren - soon... not kidding
[19:35] KJ Hax: omg
[19:35] Sidewinder Linden: i think there's a bug on that
[19:35] KJ Hax: I want one
[19:35] KJ Hax: please!
[19:35] Sidewinder Linden: i think the next build should but don't quote me ;)
[19:35] Eren Padar: KK thanks Sidwinder. "Soon" is good enoug. :D
[19:35] Sidewinder Linden: this is another example of what people have built
[19:35] Sidewinder Linden: these are used in races
[19:35] KJ Hax: where?>
[19:35] Sidewinder Linden: the builder says he does 100-200 laps to tune each new model
[19:36] Eren Padar habs best seat in whole house. :D
[19:36] Sidewinder Linden: they have controlled slip
[19:36] Sidewinder Linden: and managed handling for turns
[19:36] Sidewinder Linden: so although some of these things are "lots of flash"
[19:36] Sidewinder Linden: i think it's interested to see "what's possible" as more than "can i collide some pool balls and show collisionson a pool table" (which many folks are selling in world as well)
[19:37] KJ Hax: yes, quite!
[19:37] Sidewinder Linden: here's another example
[19:38] KJ Hax: Folks I have to say, this is the first Speaker Series event EVER to run 1h30m!
[19:38] Sidewinder Linden: so another thing that i'd i guess like to "pitch"
[19:38] Sidewinder Linden: heh
[19:38] Eren Padar: Sidewinder rocks
[19:38] Sidewinder Linden: is the idea of using "Mouselook" for education
[19:38] Sidewinder Linden: have you all tried mouselook steered vehicles?
[19:38] Sidewinder Linden: or used mouselook to "steer" your avatar?
[19:38] Eren Padar: Hmmm only fliers
[19:38] Sidewinder Linden: not a lot of folks do
[19:38] Sidewinder Linden: ok so imagine this, if you will
[19:38] Eren Padar: Have used mouselook for avatar and space craft.. nto for cars
[19:38] Sidewinder Linden: some collision based demo
[19:39] Sidewinder Linden: maybe pool balls or some such
[19:39] Sidewinder Linden: one of them is a vehicle that the student "rides"
[19:39] Sidewinder Linden: and they then expeirence the collision first-hand
[19:39] Sidewinder Linden: maybe provide a more interactive experience? :)
[19:39] Eren Padar: ?? mouselooking the poolball?
[19:39] Koni Lanzius: yes!
[19:39] Sidewinder Linden: sure - maybe make it a vehicle so they stay upright
[19:40] Sidewinder Linden: or mayb enot
[19:40] Eren Padar: :D
[19:40] Eren Padar: Sounds fun. LOL
[19:40] Sidewinder Linden: at a party a month or so ago someone did an "avatar rolling ramp"
[19:40] Sidewinder Linden: where people sat on a big sphere and rollied down the hill
[19:40] Sidewinder Linden: it was pretty funny
[19:40] Scalar Tardis: there was someone asking about making a vehicle like that.
[19:40] Sidewinder Linden: but that's i guess an aside
[19:40] Ansa Sautereau: Fantastic idea! Thanks!
[19:40] Sidewinder Linden: heh
[19:40] Eren Padar: Oh yeah. We hve one of those at Dwagonville. It rolls down a huge spiral tube about 100m
[19:41] Fumon Kubo: I've made a "bounce chamber" by sitting on a prim and using a large (megaprim) hollowed sphere with a script that pushes you towards the center again when you hit the edges. Loads of fun.
[19:41] Eren Padar: Yeah. :D
[19:41] Sidewinder Linden: that sounds cool
[19:41] Scalar Tardis: that gets to issues of hollow spheres and maintainging a level seating platform inside a rolling object. That requires a point-to-point JOINT. uh, did I say that? :)
[19:41] Sidewinder Linden: yup scalar you said that
[19:41] Sidewinder Linden: and i BET you know what my answer is by now! :)
[19:41] Eren Padar: Yeah, that's a question... is the hollowed sphere thing getting on the mend?
[19:41] Sidewinder Linden: eren - what in particular?
[19:42] Eren Padar: If you make a hollow spehere, cut it in half...
[19:42] Eren Padar: Instead of being a true hollow, you wind up standing on the surface, as if it were filled with solid plastic
[19:42] Sidewinder Linden: hmm well there are two issues here
[19:42] Scalar Tardis: so, when will joints be returning in their new form? :)
[19:42] Sidewinder Linden: one is that that may be bugged...
[19:42] Sidewinder Linden: but
[19:42] Sidewinder Linden: also know what in h4 we have a system that tries to ease server overload
[19:42] Sidewinder Linden: it's called "RCCS"
[19:43] Eren Padar: Yeah, think it's a bug. Thought you may have knoweldge of a potential fix.
[19:43] Sidewinder Linden: when the physics frames per second goes below about 20
[19:43] Eren Padar: RCCS is new to me
[19:43] Sidewinder Linden: hollowed complex objects get simplified
[19:43] Eren Padar: Hmmmm
[19:43] Fumon Kubo: LOD level basically.
[19:43] Eren Padar: I'll watch for that.
[19:43] Sidewinder Linden: so for instance a hollow cube becomes handled (physics only) as a solid cube
[19:43] Sidewinder Linden: even though it looks hollow visually
[19:43] Eren Padar: Yeah
[19:43] Sidewinder Linden: yes - it's a basic LOD reducer under load
[19:43] Scalar Tardis: I'm working on an LOD-degrader detector
[19:43] Sidewinder Linden: heh
[19:43] Eren Padar: I'll check i ton the physics load and get back with you on it.
[19:43] Sidewinder Linden: kk
[19:44] Sidewinder Linden: if the sphere is "much bigger than an avatar" it should not get LOD reucd
[19:44] Sidewinder Linden: there was a funny reason for that btw
[19:44] Sidewinder Linden: at first we'd LOD reduce anything, even a room sized hollow cube
[19:44] Eren Padar: Happens even on megaspheres. Anything from 2m to 100m
[19:44] Fumon Kubo: Low prim building.
[19:44] Sidewinder Linden: we discovered umm much to our suprise
[19:44] Sidewinder Linden: that we'd created a funny griefer mode
[19:44] Eren Padar: OH? LOL
[19:45] Sidewinder Linden: where you could drop a pile of large cubes anywhere on a region
[19:45] Sidewinder Linden: and people would be shot up out of their houses
[19:45] Eren Padar: 8O
[19:45] Sidewinder Linden: because the hollowed cube low prim rooms, would suddenly go solid!
[19:45] Sidewinder Linden: so now we don't LOD large things
[19:45] Eren Padar: LOL
[19:45] Sidewinder Linden: umm someon - scalar - asked about an LOD detector
[19:45] Scalar Tardis: yes well a huge megaprim building could end up ejecting everyone
[19:45] Sidewinder Linden: we're actually adding some stats to the new viewer builds, including
[19:46] Scalar Tardis: yes it looks feasible
[19:46] Sidewinder Linden: the count of LOD reduced objects on the region
[19:46] Sidewinder Linden: it's in the release candidate (dazzle) i think...
[19:46] Fumon Kubo: I think he means more in the area of compensation script side.
[19:46] Kumangaro Hayes: Thanks so much for the class sidewinder! i gotta get some work done.
[19:46] Sidewinder Linden: kk yw kuma
[19:46] Sidewinder Linden: actually kj... i probably should hop as well - i hope this is along the lines that you were looking for, although completely off the cuff ;)
[19:47] KJ Hax: OMG!
[19:47] KJ Hax: Yes!
[19:47] KJ Hax: Unreal!
[19:47] Scalar Tardis: basically it is a physical sphere inside small cubes surrounding it. if it gets cubed it will start colliding where a sphere would not, then it shouts notification to all physics enabled toys in the sim
[19:47] Veeyawn Spoonhammer: Woot!
[19:47] shamblesguru Voom: am still soaking this all up ... and still 3 hours before lunch
[19:47] Tara5 Oh: Thank you Sidewinder for a very interesting talk.
[19:47] Sidewinder Linden: cool scalar
[19:47] KJ Hax: Absolutely a highlight of the series!
[19:47] Ansa Sautereau: Thanks a million! Very exciting!
[19:47] Sidewinder Linden: wow thanks guys
[19:47] KJ Hax: Thank you and thank you audience for coming!
[19:47] Koni Lanzius: thank you, sidewinder!
[19:47] KJ Hax: I will grab a script log
[19:47] Geoffrey Mayo: thx Sidewinder ;)
[19:47] Eren Padar: Sidewinder, thanks for all the time. Sorry I didn't get in on the first of it. : )
[19:47] KJ Hax: and post it somewhere!
[19:47] Eren Padar: Applause!!
[19:47] KJ Hax: I have the whole thing!
[19:47] KJ Hax: All text!
[19:47] Sidewinder Linden: so if i can encourage one thing - play with it - be creative - do "cool stuff" with what it can do, and i suspect you'll build some really interesting education pieces :)
[19:47] Eren Padar: Kewl KJ! I'd appreciate a copy. : )
[19:48] Koni Lanzius: Applause!
[19:48] KJ Hax: Watch the wiki - http://secondlife.iste.wikispaces.net/events
[19:48] Sidewinder Linden: kk thanks guys - have a good night
[19:48] KJ Hax: That is where I will post it!
[19:48] Veeyawn Spoonhammer: Woot!
[19:48] KJ Hax: Night Sidewinder!
[19:48] KJ Hax: Night all!
[19:48] Koni Lanzius: goodnight!
[19:48] Eren Padar: Ty Ty !