User:Andrew Linden/Office Hours/2008 02 07

From Second Life Wiki
< User:Andrew Linden‎ | Office Hours
Revision as of 08:28, 8 February 2008 by Andrew Linden (talk | contribs) (formatted with Zero's perl script)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Transcript of Andrew Linden's office hours:

[17:21] Andrew Linden: hey rex
[17:21] Rex Cronon: hi andrew
[17:21] Rex Cronon: hi sai
[17:21] Andrew Linden: sorry I'm late... had trouble logging in
[17:21] Rex Cronon: i just got here
[17:22] Andrew Linden: kitto is with me in the Beta
[17:22] Andrew Linden: I think he's still having trouble
[17:22] Saijanai Kuhn: oddness in SL. The mono sim I was in kept on rebooting or thats what the message said
[17:22] Rex Cronon: your scripts were to advanced for it:)
[17:23] Gaius Goodliffe: Something seems missing...
[17:23] Saijanai Kuhn: riight. for (i = 0 blah blah
[17:23] Andrew Linden: well, what should we do? Hold the hours here? Or put up a sign and move it to Beta?
[17:23] Rex Cronon: are kitos trains missbehaving?
[17:23] Saijanai Kuhn: hoping to create a genuine robot in mono and filmm it in monmo and LSL2 on the main grid
[17:24] Saijanai Kuhn: I heard someone had a LSL-based BVH file reader at one point. It might actually be workable with mono
[17:24] Rex Cronon: i think u should have a poster here that says on what gird are u holding the office hours
[17:25] Gaius Goodliffe: I guess it depends on whether people are starting to be able to filter in now, or if we just got lucky...
[17:25] Andrew Linden: hang on, lemme see what's up with kitto
[17:25] Rex Cronon: kito is here
[17:25] Indigo Lucerne: waves... only managed to log in a few min ago myself
[17:25] Andrew Linden: ah ok
[17:26] Andrew Linden: let's hold it here there
[17:26] Andrew Linden: then
[17:26] Kitto Flora: Hello
[17:26] Kitto Flora: Made it
[17:26] Rex Cronon: hi
[17:26] Rex Cronon: how r your trains doing?
[17:26] Gaius Goodliffe: Gravity is for wimps. :)
[17:26] Kitto Flora: Slowly they are doing it
[17:27] Andrew Linden: ok lemme start with a quick update
[17:27] Andrew Linden: I've got some items I want to talk about...
[17:28] Andrew Linden: first of all... I was planning on holding an impromptu vehicle session tonight if anyone here wanted to give me some broken vehicles
[17:28] Andrew Linden: I'd hold it in the beta grid, and I'd update one region to be at the tip of the code branch... beyond the current deploy to the beta grid
[17:28] Andrew Linden: I've fixed a few vehicle bugs
[17:29] Andrew Linden: and want to have them in effect when testing the latest broken vehicles
[17:29] Rex Cronon: i have never scripted any vehicles
[17:29] Andrew Linden: ok, well I'll announce that again at the end, and if anyone wants to we can go over, otherwise we'll skip it
[17:29] Kaiser Bogomil: :)
[17:29] Andrew Linden: next... some bugs fixed in the codebase but not yet released:
[17:30] Andrew Linden: SVC-1333 (gap between rezzed boxes)
[17:30] Gaius Goodliffe: My last set of buts don't appear to be fixed yet on the current beta, but if you've got a new version to deploy I'll be happy to check
[17:30] Andrew Linden: SVC-1293
[17:30] Andrew Linden: SVC-1253 (no damage to avs sitting on phantom)
[17:30] Andrew Linden: SVC-1363
[17:30] Andrew Linden: SVC-1239 (broken mouselook vehicles)
[17:31] Andrew Linden: the STATUS_ROTATE_* FALSE bug is fixed
[17:31] Rex Cronon: so, no sitting won't make u invulnerable?
[17:31] Andrew Linden: and also SVC-54 (target omega objects don't update sometimes)
[17:31] Andrew Linden: Yeah, the bug was that sitting avs could be invulnerable
[17:32] Andrew Linden: Ok... now on to the things I wanted to talk about...
[17:32] Kaiser Bogomil: hmmm... kind of a feature though too
[17:32] Andrew Linden: first, Simon noted that the max vehicle linear motor is 40 m/sec
[17:32] Andrew Linden: which he thought was kinda too slow
[17:32] Andrew Linden: the way it works right now is if you set it to 100 it will be clamped to 40 internally
[17:33] Andrew Linden: He was wondering if we could raise it, and why it was so low
[17:33] Creem Pye: yeah, it is pretty slow if your linear motor timescale is fairly high
[17:33] Andrew Linden: the main reason it was so low was that in Havok1 it was pretty easy to drive a complex vehicle into a state of penetration when going at high speeds
[17:33] Andrew Linden: and this would lead to a deep think
[17:33] Andrew Linden: and would eventually crash the simulator
[17:33] Andrew Linden: the deep think crash is gone in Havok4
[17:34] Kaiser Bogomil: non-anticipatory collision logic does that
[17:34] Creem Pye: plenty of vehicles go faster; they just rely on llApplyImpulse for a boost =)
[17:34] Gaius Goodliffe: I've never tried over 25, but then I mostly build airships -- they aren't known for their speed. :) I'm all for less restrictive limits in any case.
[17:34] Andrew Linden: so we could raise the speed, however... some current vehicles would "break" in the sense that they might suddenly start moving faster then they had
[17:34] Andrew Linden: if you set to 100 and had it clamped to 40, for example
[17:35] Andrew Linden: The question is... should we just raise it?
[17:35] Gaius Goodliffe: Maybe make it possible to adjust the limit? Default to 40, but you can make a call to raise it to 100 or whatever?
[17:35] Creem Pye: I think the people who set it past the limit really want to go fast anyway =)
[17:35] Saijanai Kuhn: llSetDefaultVehicleParasm([...]);
[17:35] Andrew Linden: Alternatively we could add a flag that is... DISCARD_LEGACY_VEHICLE_LIMITS or something, which would be FALSE by default
[17:36] Kaiser Bogomil: if you go too fast - you should crash
[17:36] Gaius Goodliffe nds.
[17:36] Gaius Goodliffe: *nods
[17:36] Andrew Linden: and you would have to set it TRUE to get access to the higher speeds and whatever else gets lifted
[17:36] Creem Pye: llSetVehicleFloatParam(MAX_LINEAR_MOTOR, )
[17:36] Kaiser Bogomil: sounds resonable to me
[17:36] Kaiser Bogomil: don't want to break the old code
[17:36] Gaius Goodliffe: I like Creem's idea better, but I don't know if that'd be more difficult to implement.
[17:37] Saijanai Kuhn: the question is, will there be more than one place where you might want to make changes to defult behavior?
[17:37] Kaiser Bogomil: leave it to the scripter - he'll have the options
[17:37] Andrew Linden: At the moment there is only one for vehicles that I know about
[17:37] Indigo Lucerne: Yeah, I was thinking too that it'd be useful to have a few flags that set things to "old style" Havok1 behavior (where you have constraints or current limits)... slow typing, heh
[17:37] Saijanai Kuhn: sure, which is why you use teh set params model with a list parameter
[17:37] Gaius Goodliffe: llSetVehicleFloatParam(MAX_ANGULAR_MOTOR, ...) would be nice too.
[17:37] Andrew Linden: I already lifted the global hover height limitation
[17:37] Creem Pye: Gaius, what's the maximum now?
[17:38] Creem Pye: 2 pi?
[17:38] Gaius Goodliffe: Four pi I think.
[17:38] Kaiser Bogomil: e
[17:38] Andrew Linden: 4 pi / sec? = 2 revolutions per second
[17:38] Andrew Linden: pretty fast
[17:38] Gaius Goodliffe: It has, at times, been insufficient for me.
[17:38] Kaiser Bogomil: yes - me 2
[17:39] Gaius Goodliffe: Especially when combined with the vertical attractor.
[17:39] Andrew Linden: really? for what? twitchy steering?
[17:39] Creem Pye: well, it'll never quite reach that value due to angular friction and deflection, I suppose...
[17:39] Kaiser Bogomil: noooo - for blue angel stuff
[17:39] Andrew Linden: oh, gets attenuated sometimes?
[17:39] Gaius Goodliffe nods.
[17:39] Kaiser Bogomil: yes
[17:39] Kaiser Bogomil: soooo...
[17:40] Kaiser Bogomil: can I ask some questions?
[17:40] Andrew Linden: sure
[17:40] Kaiser Bogomil: k - I've always wondered why objects that are static with respect to each other - one resting on the other for example - don't touch
[17:41] Andrew Linden: oh, the collision tolerance. The gap between stacked dynamic objects?
[17:41] Kaiser Bogomil: yes
[17:41] Andrew Linden: The havok physics engine uses the collision tolerance to limit the number of penetrations
[17:42] Andrew Linden: so the objects are colliding and push apart if they are < the collision_tolerance away from each other
[17:42] Andrew Linden: it is an artifact of how the physics engine works
[17:42] Kaiser Bogomil: can that be .001 M?
[17:42] Kaiser Bogomil: & that's not going to change?
[17:42] Saijanai Kuhn: the moral is, don't use physics to get perfect behavior.
[17:42] Kaiser Bogomil: sigh
[17:42] Andrew Linden: no... you get problems if you make it too small... snagging and grippy collisions
[17:43] Andrew Linden: as well as explosive results from the penetration resolution code
[17:43] Kaiser Bogomil: yes - I've experienced that :(
[17:43] Gaius Goodliffe hates it when code explodes.
[17:43] Kitto Flora: The gap in H4 appears to be about the sameas H1, but tripping on edges is significantly worse
[17:43] Saijanai Kuhn: Maya and other professional level animation software STILL has issues. You use even the best physics simulators to get close as possible, then hand-tweak for the movies
[17:43] Andrew Linden: my hope is that eventually we could reduce it or eliminate it by moving to a different physics engine
[17:43] Kaiser Bogomil: I don't even use the physics for stuff like that
[17:43] Creem Pye: an in-house system? =)
[17:44] Kaiser Bogomil: I have to do my own physics
[17:44] Andrew Linden: Havok is made for games, which have a lot more leeway when it comes to tuning the content so you don't notice the collision_toleerance
[17:44] Gaius Goodliffe: You *can* change the laws of physics? ;)
[17:44] Kaiser Bogomil: what about Aegis?
[17:44] Kaiser Bogomil: their engine anticipates
[17:45] Andrew Linden: all other possible physics engine candidates must wait for Havok4 to complete
[17:45] Kitto Flora: Is it possible in SLED code to adjust what constitute a collision?
[17:45] Andrew Linden: there is a lot of cleanup involved in the Havok4 project
[17:45] Andrew Linden: and it will be much easier to try other engines after Havok4 is done
[17:45] Kaiser Bogomil: ahhhh good!
[17:45] Saijanai Kuhn: well that lead eventually to the open source of the server? I thought you said physics was a huge issue there because you can't provide a plugin API right now
[17:46] Sidewinder Linden: /hi guys sorry - wrangling traffic then deploy...
[17:46] Rex Cronon: hi sidewinder
[17:46] Hiro Pendragon: Man, it took forever to get here
[17:46] Andrew Linden: yes, Havok is a blocker for opening the simulator source ATM
[17:46] Kitto Flora: Hi Sidewinder
[17:46] Andrew Linden: so... we'd have to finish Havok4, then implement our abstracted API in a different (more free) physics engine beore we could effectively open the simulator code.
[17:47] Kaiser Bogomil: yay!
[17:47] Andrew Linden: Here is another issue I wanted to bring up...
[17:47] Saijanai Kuhn: I'm sure the OPenSim folks would be happy to help at that point
[17:47] Andrew Linden: Currently the speed of the Havok4 avatar is limited
[17:47] Saijanai Kuhn: or maybe not. FreeBSD license vs GPL. Sigh
[17:47] Andrew Linden: I've had trouble balancing the forces to get jetpacks and other gadgets moving at the same speeds
[17:48] Gaius Goodliffe: The licenses tend to be compatible. It's the people that frequently aren't. :p
[17:48] Andrew Linden: at the moment, in the tip of the branch, I've got slow things pushing too fast, and fast things pushing too slow
[17:48] Hiro Pendragon: How about a local tp function instead? :)
[17:48] Clarknova Helvetic: Isn't the limit on avatar speed just a cap? It seems to stop physics for the avatar momentarily if its velocity is too great.
[17:48] Andrew Linden: and I'm having trouble balancing them all
[17:49] Andrew Linden: one option is to just loosen the limits on the avatar speeds
[17:49] Andrew Linden: which would cause most slower objects to push too fast, and most fast objects to push fast enough
[17:49] Andrew Linden: The question is... should we release some of the limits on the avatar motion? If so, how should we go about it?
[17:50] Kaiser Bogomil: ok...
[17:50] Kaiser Bogomil: I'll jump in here...
[17:50] Andrew Linden: That is, how should we make the decision? How much feedback to ask for?
[17:50] Kaiser Bogomil: if the physics is correct that shouldn't be an issue
[17:50] Saijanai Kuhn: can avatar motion be added to LSL as well as be set by client options?
[17:50] Creem Pye: you're talking about non-sitting avatars, correct?
[17:50] Rex Cronon: things should be ok, as long as blitzes don't start working again
[17:50] Saijanai Kuhn: avatar motion modding*
[17:50] Kaiser Bogomil: you should open up the speeds/velocities
[17:50] Clarknova Helvetic: You'll get plenty of feedback when hundreds of gadget owners start complaining in the jira.
[17:51] Gaius Goodliffe: hehe
[17:51] Andrew Linden: I'm talking about moving avatars... flying or walking.
[17:51] Sidewinder Linden: some of this has to do with things that folks did to work around some havok1 oddities, that aren't buggy in h4, but were...
[17:51] Kaiser Bogomil: & they are solid moving objects
[17:51] Hiro Pendragon: Most of the avatar push stuff is all hacks anyway. Anything standard that we could use would be better than the old way.
[17:51] Clarknova Helvetic: One thing you can start by doing is taking off the speed limits for obejcts that the avatar is linked to
[17:51] Kaiser Bogomil: yes
[17:51] Clarknova Helvetic: either by attachment or by sitting
[17:52] Sidewinder Linden: clarknova... what if that caused some existing well known jetpack-type products to operate "dizzyingly fast"? :)
[17:52] Kaiser Bogomil: use a LEGACY_FLAG
[17:52] Andrew Linden: right... existing content that has been tuned to be just right will be way off... instead of somewhat off
[17:52] Saijanai Kuhn: seems like it might be the same issue as before: default vs modded parameters
[17:52] Clarknova Helvetic: Isn't that the point of jetpacks? Obviously we can't respond to your questions about technical issues in code we can't see. But if you want a suggestion, there's mine.
[17:52] Kaiser Bogomil: yes
[17:53] Andrew Linden: I don't think I can match everything just the way it was
[17:53] Kaiser Bogomil: ok - I'm ok with that :)
[17:53] Gaius Goodliffe: Too fast is probably preferable to too slow more than half the time, I'm guessing.
[17:53] Saijanai Kuhn: depends on how laggy things are
[17:54] Clarknova Helvetic: I would always prefer too fast. I don;t even mind being orbited you know.
[17:54] Kaiser Bogomil: laggy is always gonna be a problem - the world is a big place
[17:54] Clarknova Helvetic: The only people who argue against blitz-style devices are the people that aren't affected by them.
[17:54] Saijanai Kuhn: can these be made simiulator-side options?
[17:54] Kaiser Bogomil: hehe
[17:54] Kaiser Bogomil: there!
[17:54] Kaiser Bogomil: yes!
[17:54] Kaiser Bogomil: client side or server side only!
[17:55] Clarknova Helvetic: physics is always server side.
[17:55] Hiro Pendragon: Honestly... Andrew... why isn't there SL 2.0?
[17:55] Kaiser Bogomil: I'd so love to not worry about jerky motion & NOT the physics
[17:55] Hiro Pendragon: Just freeze the code for SL 1.0, and then let people decide to upgrade.
[17:55] Rex Cronon: clarknove, i have been affected by them. it would be ok with me, if scripts weren't broken by it, and if u don't become invisible when u get down
[17:55] Saijanai Kuhn: I proposed that a while back. The arguments are on teh wiki ;-)
[17:56] Hiro Pendragon: Just like anything else in software or the internet, yeah, some old stuff may break.
[17:56] Andrew Linden: the problem with making them server side options... is that it won't be clear where your jetpack will work right, and where it will malfunction
[17:56] Sidewinder Linden: and then you'd end up with some really bizarre behavior crossing region boundaries with varying settings ;)
[17:56] Kaiser Bogomil: well - the other side is that a compeitor will come along & do it right
[17:56] Clarknova Helvetic: so the way to fix that is to fix the client render error that stops you from seeing your own avatar. changing physics is a very roundabout way of solving that problem
[17:56] Saijanai Kuhn looks sad.
[17:56] Sidewinder Linden: and probably a slew of phantom bug reports and resident problem reports
[17:56] Andrew Linden: Hiro, we don't develop SL in a way that would allow us to do a 2.0... upgrades are more incremental
[17:57] Hiro Pendragon: Yeah, that's my point.\
[17:57] Clarknova Helvetic: yes, and you don't want to split the network into two parts.
[17:57] Saijanai Kuhn: well, you HAVE a SL 2.0 already. The het grid. But no-one noticed (which was the point)
[17:57] Kaiser Bogomil: so Stargate worlds might win - ya know?
[17:57] Gaius Goodliffe: lol
[17:57] Sidewinder Linden: ok ... so i would like to get back to one of the key questions... which is what to do about max avatar and vehicle speeds...
[17:57] Hiro Pendragon: SL is now 7 years old. That's a long lifespan for any software verison.
[17:57] Hiro Pendragon: version
[17:57] Sidewinder Linden: we *could* just do something and say "here it is"
[17:57] Sidewinder Linden: this is a chance to make a group decision... :)
[17:58] Kitto Flora: How about find the 6 most popular flight assist devices - and get them working as close to old way as possible. The rest will have to fix their code.
[17:58] Gaius Goodliffe: Expand the limits. Too fast is generally better than not fast enough...
[17:58] Creem Pye votes for a linear motor speed limit of 100m/s :D
[17:58] Gaius Goodliffe: (Exceptions abound, but we'll deal...)
[17:58] Kaiser Bogomil: I agree - I like fast
[17:58] Sidewinder Linden: i've heard some reasons, even in these office hours, that are pretty persuasive, for relaxing upper speed limits
[17:58] Sidewinder Linden: for instance
[17:58] Gearsawe Stonecutter: ha my is missing from database so it is dead
[17:59] Sidewinder Linden: if we have a 4096 ceiling... having a faster ascent would be nice
[17:59] Hiro Pendragon: Sidewinder, well, regardless of the result, you're going to have some unhappy people.
[17:59] Sidewinder Linden: for push assist "quasi-transporters"
[17:59] Clarknova Helvetic: I say try to make physics as much like Havok1 as possible in behavior: especially when it comes to the limits on avatar speed. I don't know why you can't balance the fources in H4. I'd be interested in knowing. But from a user's persepctive, whcih is all I have, I know a lot of things will be broken by the current limits on AV speed, and people are already very upset about it
[17:59] Sidewinder Linden: what would you say to a proposal that looks something like this (for avatar speeds)
[17:59] Sidewinder Linden: clarknova
[18:00] MartinRJ Fayray: For people with a slower computer vehicles that are too fast wouldnt have any sense - most sims wont even allow very fast vehicles to go more than a few seconds.
[18:00] Creem Pye: well nothing's stopping them from fixing the behavior of scripts which relied on a lower limit in the past =)
[18:00] Sidewinder Linden: it's partly that folks relied on some buggy behavior to get the results they got... and including those "bugs" is partly very dificult and then makes new products not able to work any better than the h1 versions
[18:00] Clarknova Helvetic: Bugs have a way of evolving into features :)
[18:00] Sidewinder Linden: that's one of the questions, creeme.... for proucts where the builder is still around
[18:00] Sidewinder Linden: the could be updated
[18:00] Sidewinder Linden: for old ones and freebies, theypretty much "are what they are"
[18:01] Andrew Linden: well, not really Sidewinder... not quite bugs in Havok1.... that was how it worked
[18:01] Sidewinder Linden: ok well some and some...
[18:01] Andrew Linden: the vehicle was limited to 40 m/sec via the linear motor
[18:01] Andrew Linden: some vehicle makers add an llApplyimpulse() to get higher speeds
[18:01] Sidewinder Linden: right
[18:02] Andrew Linden: the 40 m/sec isn't a bug as much as a design limitation
[18:02] Sidewinder Linden: and there's a question as wto whether that should override the motor speed or not... which is "right" behavior?
[18:02] Creem Pye: I still think that somebody setting it to 60m/s really wants to go fast in any case =)
[18:02] Saijanai Kuhn: Just wanted to point out that re: height limnits, some people might implement a 3D grid using opensim. So keep that in mind...
[18:02] Creem Pye: and there's always the hard limit of 255m/s, right?
[18:02] Sidewinder Linden: sai - not following your point re that?
[18:03] Clarknova Helvetic: "right" is determined by what people want to buy. if people want fast cars and planes that take advandage of additional impulse, and they're willing to pay for them, and HAVE paid for them, why should an arbitrary judgement of what is correct and incorrect behavior break thier content?
[18:03] Saijanai Kuhn: well, every time you raise the hight limit, you push back where the next sim would start... might not be relevant but just an intuition
[18:03] Rex Cronon: u mean, stacking sims one on top of the other?
[18:03] Saijanai Kuhn: right. Not an option with Zero's vision, but some grid operators might want to change that
[18:03] Sidewinder Linden: ok saijani - that's a whole different issue...
[18:04] Saijanai Kuhn: I realize. Just a thought of how design decisions might change options for others...
[18:04] Sidewinder Linden: we're talking about the specific issue of vehicle and avatar speeds - that's a long term design and philosophy question, and "way outside" the havok4 project
[18:04] Creem Pye: impulses have hte disadvantage that you can't set things like the force's decay timescale
[18:04] Clarknova Helvetic: but it directly applies to the Havok4 project in a concrete way.
[18:04] Clarknova Helvetic: Decsions about it are being made now.
[18:05] Sidewinder Linden: ok... let's back up a few steps...
[18:05] Saijanai Kuhn hideds under the table
[18:05] Gaius Goodliffe: Right. It's better to discard the limit. Better still to discard the limit IF the vehicle script sets the appropriate flag.
[18:05] Saijanai Kuhn: hides*
[18:05] Kitto Flora: Seems to me, if a builder scripted a linear motor at 60Ms/, and in H1 it only went 40M/s, if in H4 it suddenly does go at 60M/s either he will be happy, or he cant really complain about getting what he asked for, if that now gives him problems.
[18:05] Sidewinder Linden: what do you see as the advantages and disadvantages of raising speed limits on avatars and vehicles?
[18:05] Sidewinder Linden: that's what i was thinking, kitto, and builders are saying theywant faster in some cases
[18:06] Kitto Flora: So - if it will work, give it to them.
[18:06] Sidewinder Linden: and using impulse, if what you really want is a controlled set higher speed, is not a very direct path there...
[18:06] Kaiser Bogomil: course you could add a new API for the "new" stuff
[18:06] MartinRJ Fayray: I want faster short-distance teleporters not faster vehicles
[18:06] Sidewinder Linden: martinrj - yes that's the avatar side of this discussion
[18:06] Andrew Linden: hrm... maybe we could boost the vehiclt top speed about 5m/sec every update until we were up to 80 or 100
[18:06] Clarknova Helvetic: The advantage is that it's crazy fun to shoot to the moon with my flight assist or Lumiere Noir spaceship. The disadvantage is that it is extremely unamusing to find it anchored to my avatar like a bird on a chain.
[18:07] Saijanai Kuhn: Is that a floor safe on your back, or are you just happy to see me, Dah Oh?
[18:07] Sidewinder Linden: hehe andrew :)
[18:07] Andrew Linden: if we do it slow enough maybe few will notice ;-)
[18:07] Gaius Goodliffe: Advantage: more possible behaviors accurately simulated. Disadvantage: Old thigns *might* break. I suspect the problem isn't that widespread, though. How often are people really asking for 60m/s when they really want 40?
[18:07] Day Oh: :D <3
[18:07] stimpy Tripp: (hello)
[18:07] Rex Cronon: hi
[18:07] Clarknova Helvetic: On the contrary, the problem is absolutely widespread.
[18:08] Kaiser Bogomil: deprecate the older api & introduce the faster or unbounded one
[18:08] Clarknova Helvetic: Nearly every user owns something that likes to go fast, and they like using them.
[18:08] Sidewinder Linden: ok.. so.. there is an experimental way to figure this out
[18:08] Gaius Goodliffe nods.
[18:08] Sidewinder Linden: we could do a release - maybe to a few hosts on the beta preview, with high speed limits, and see what you all think about it...?
[18:08] Kitto Flora: All I have is Carbon Rod :)
[18:08] Saijanai Kuhn: can I wear the ant?
[18:08] Sidewinder Linden: heh
[18:08] Creem Pye: sounds good to me, sidewinder
[18:08] Gaius Goodliffe: yes.
[18:08] Gaius Goodliffe: Very good idea.
[18:09] Clarknova Helvetic: That would be splendid. Will there be a way for users to tell you pseicifcally that they like one version vs. another?
[18:09] Kitto Flora: YEs... try it out for a while
[18:09] Clarknova Helvetic: *specifically
[18:09] Sidewinder Linden: andrew... is tinkering with max speed for both vehicles and avatar attachment pushes pretty straightforward?
[18:09] Kitto Flora: See how many splatter their bits at sim boundary :)
[18:09] Sidewinder Linden: i.e. wouldn't take much time to do?
[18:09] Saijanai Kuhn: That about dialog box might mention max speed and other changed parameters...
[18:09] Saijanai Kuhn: likewise with the sim change warning
[18:10] Andrew Linden: deja vu
[18:10] Sidewinder Linden: the new 1.19 rc1 finally has a less intrusive announcmenet about that
[18:10] Sidewinder Linden: ? andrew ?
[18:10] Gaius Goodliffe: w00t!
[18:10] Kitto Flora: The hetgrid sim change notice was a major pain
[18:10] Clarknova Helvetic: you're talking about the different-server-version popup yes?
[18:10] Andrew Linden: I was having a moment of deja vu.
[18:10] Sidewinder Linden: ahh
[18:11] Sidewinder Linden: yes clarknova - for now it's been modified to a yellow popup in the lower right - same placement as the contact login logout
[18:11] Sidewinder Linden: not a dialog with an ok button
[18:11] Saijanai Kuhn: sounds reasonable (what I would have tried too)
[18:11] Gaius Goodliffe: Good, that old notice was a menace while flying fast...
[18:12] Clarknova Helvetic nods
[18:12] Saijanai Kuhn: with the new speed: you are in a new sim <click> you are in a NEW new sim <click>
[18:12] Gaius Goodliffe: (More likely to crash due to the notice than the changed behavior lol)
[18:12] Sidewinder Linden: andrew - is it fairly straightforward to do a build with high avatar and vehicle max speeds?
[18:12] Andrew Linden: yes Sidewinder
[18:13] Kaiser Bogomil: change a constant?
[18:13] Sidewinder Linden: how fast should the new settings be?
[18:13] Andrew Linden: I was just now bouncing Balance on the Beta grid for some vehicle session testing.
[18:13] Hiro Pendragon: Sidewinder, will we still have the sit-tp hack?
[18:14] Kaiser Bogomil: hehe
[18:14] Sidewinder Linden: which one hiro?
[18:14] Hiro Pendragon: Using the offset to move an av instantly
[18:14] Kaiser Bogomil: go through a wall with a site location
[18:14] Hiro Pendragon: then unsitting.
[18:14] Kaiser Bogomil: sit
[18:14] Andrew Linden: I could re-update Balance in about 10 min if we wanted to play with things
[18:14] Creem Pye: 100m/s for vehicles sounded fast enough to me... others might want more though ;)
[18:14] Hiro Pendragon: If we still have that... I'd think max speed should be ... as fast as a rocket.
[18:14] Kaiser Bogomil: or a very fast impulse to drive you through a wall
[18:14] Kaiser Bogomil: yes
[18:14] Hiro Pendragon: If we dont, then we need tp speed. Entire sim within 0.1s
[18:14] Andrew Linden: however, I was having a little change of heart...
[18:14] Sidewinder Linden: does offset respect max avatar speed andrew?
[18:14] Sidewinder Linden: ?
[18:14] Saijanai Kuhn: could you put w or 3 sims side by side with different speeds at the same time?
[18:15] Gaius Goodliffe: I like Creem's idea for an llSetVehicleFloatParam for adjusting that maximum.
[18:15] Saijanai Kuhn: 2* or 3 sims
[18:15] Andrew Linden: I was thinking maybe we should just leave vehicles hobbled, and I could try one last time to get the avatar back to normal
[18:15] Sidewinder Linden: huh?
[18:15] Saijanai Kuhn: thats a client side change. More than 10 inutes...
[18:15] Andrew Linden: just waffling here...
[18:16] Andrew Linden: the hassle of trying to push a real change here is large
[18:16] Kaiser Bogomil: I get that feeling
[18:16] Saijanai Kuhn: any LSL chnage takes mon ths, I think
[18:16] Kaiser Bogomil: oh well
[18:16] Saijanai Kuhn: especially in the middle of mono beta testing
[18:16] Kaiser Bogomil: open source could really help things along
[18:17] Hiro Pendragon: Andrew, let's face it - all vehicle makers will want to retune their vehicels anyway
[18:17] stimpy Tripp: opensource havoc? (:
[18:17] Clarknova Helvetic: Please don't waffle too much. The feedback will be very positive if you undo the speed limit cap. If you just leave things as they are, or nearly as they are when you roll out Havok 4 on the main, you're going to think the outcry againsgt breaking the warPos was a gentel sumemr breeze.
[18:17] Kaiser Bogomil: NOT HAVOC!
[18:17] Hiro Pendragon: so, as long as the limits of movement are reasonably large, we should be ok
[18:17] Saijanai Kuhn: client is open source, but mono is server side, and so is LSL script executiton
[18:17] Andrew Linden: warPos?
[18:17] Rex Cronon: warp
[18:17] Gaius Goodliffe: warpPos... the llSetPrimitiveParams trick for bypassing 10m movement limit.
[18:18] Saijanai Kuhn: move 10m move 10m all in the same list
[18:18] Andrew Linden: oh
[18:18] Kaiser Bogomil: or the multiple tasks to get smooth motion
[18:18] Gaius Goodliffe: That was a fun breakage...
[18:18] Gaius Goodliffe: :p
[18:18] Sidewinder Linden: ok... soo....
[18:18] Hiro Pendragon: Will we get llSetPos for phys objects with Havok 4?
[18:18] Clarknova Helvetic: Yes. Linden labs had to roll back the "fix" because when it was broken for reasons of philosophical correctness, 20k users sent angry emails to Linden Labs.
[18:18] Kaiser Bogomil: NOooooooo
[18:18] Sidewinder Linden: do you think we can make some sort of decision here, or should we just go back and "do something" and se what you think about it?
[18:18] Kaiser Bogomil: you don't want physics for setpos!!!!!!!!!!!1
[18:19] Sidewinder Linden: can we please stop the off-toipic rambling...?
[18:19] Kitto Flora: llSetPos is already there in H4
[18:19] Kitto Flora: also llSetRot
[18:19] Clarknova Helvetic: You'll see the same thing when you roll out Havok4 as is. There was recently a Jira filed by a very upset vehicles sim owner who was desperate to opt out of the early adopters program
[18:19] Andrew Linden: well, I'm going to have to leave the office hour soon
[18:19] Clarknova Helvetic: of course it was the wrong venue for the request, but that's how deesperate he was.
[18:19] Kaiser Bogomil: well... I'm gonna go... what I want is 2.0 stuff so... good luck folks :)
[18:20] Sidewinder Linden: clarknova, we are not rolling out havok4 "as-is"
[18:20] Gaius Goodliffe nods.
[18:20] Andrew Linden: however, I've got Balance on the Beta grid coming up with the bleeding edge simulator code
[18:20] Sidewinder Linden: there are significant vehicles fixes not even in the current deploy that just got out
[18:20] Clarknova Helvetic: I ean as is vis a vis the avatar movement speed limit.
[18:20] Andrew Linden: and I'm willing to spend the next hour or so if anyone here wants to test their buggy vehicles there
[18:20] Gaius Goodliffe: Okay! To Balance! :)
[18:20] Andrew Linden: the bleeding edge tip has not been QA'ed
[18:20] Clarknova Helvetic: Oh yes. I'd like that very much, andrew. there's a root prim issue i think you should see.
[18:21] Sidewinder Linden: the build that andrew is talking about has some newer work that is not in the deploy that just went to the early adopters to try out
[18:21] Andrew Linden: so if you find any permissions bugs please let me know
[18:21] Kitto Flora: I be there in a while, look at whats there, maybe test some stuff, see if it still works ok
[18:22] Andrew Linden: ok, so I've fixed mouselook vehicles, and also some cars that were tending to stand up on their nose or tails
[18:22] Saijanai Kuhn: Andew, I think the iddea of setting a new limit and jsut letting people know about it on the beta grid would be fine. Or here if thats workable
[18:22] Andrew Linden: if you have any vehicles you know are broken, you can test them in Balance, and if they are still broken you can give me a copy and I'll use them to troubleshoot the bugs
[18:24] Andrew Linden: Otherwise... see you all later.
[18:24] MartinRJ Fayray: Goodnight.
[18:24] Kitto Flora: Byebye
[18:24] Rex Cronon: bye andrew
[18:24] Creem Pye: ok i'll check it out - thanks for your help =)
[18:24] Gaius Goodliffe: See you on the other side. ;)
[18:24] Indigo Lucerne waves
[18:24] Hiro Pendragon: thanks Andrew
[18:24] Rex Cronon: bye everybody
[18:25] Rex Cronon: and have fun crashing vehicles;)