User:Andrew Linden/Office Hours/2008 09 23
Jump to navigation
Jump to search
Transcript of Andrew Linden's office hours:
[11:03] | Andrew Linden: | Hello Gaius |
[11:03] | Gaius Goodliffe: | Hi there, long time no see :) |
[11:03] | Andrew Linden: | It's been a while, yes. |
[11:03] | Gaius Goodliffe: | Ooo crayons? |
[11:03] | Andrew Linden: | yeah, this is my crayon gun |
[11:03] | Gaius Goodliffe: | XD |
[11:04] | Andrew Linden: | I really like it, so I've been wearing it for a few weeks |
[11:04] | Gaius Goodliffe: | Crayons are not meant to be a weapon. :p |
[11:04] | Gaius Goodliffe: | And only rarely a food item. |
[11:04] | Arawn Spitteler: | The Crayon is mightier than the tax form |
[11:04] | Gaius Goodliffe: | hehe |
[11:04] | Andrew Linden: | every once in a while when I'm alone I shoot it around and build neat structures of crayon streams |
[11:05] | Crayon Katyusha (trail) whispers: Use Mouselook (press 'M') to shoot me. | |
[11:05] | Crayon Katyusha (trail) whispers: Choose 'Detach' from my menu to take me off. | |
[11:05] | Gaius Goodliffe: | Look out! He's got a crayon! :o |
[11:05] | Simon Linden: | The Crayoninator |
[11:05] | Arawn Spitteler needs Crayon Houses, at his Burning Life Build: I started a new group, to contemplate how Second Life could be open to Helen Keller | |
[11:05] | Rex Cronon: | hi andrew , arawn, gauis |
[11:05] | Gaius Goodliffe: | nifty! |
[11:05] | Andrew Linden: | doesn't work as well on openspace regions |
[11:06] | Gaius Goodliffe nods. | |
[11:06] | Simon Linden: | Gonna re-log, be right back... |
[11:06] | Andrew Linden: | on normal regions there usually aren't any gaps in the streams |
[11:06] | Arawn Spitteler: | Anyone know where I can find Collosal Cave? |
[11:06] | Rex Cronon: | don't make me take out my gun:) |
[11:06] | Andrew Linden: | unfortunately, or fortunately, as the case may be, the streams are temp-on-rez |
[11:06] | Gaius Goodliffe: | Yeah, openspace performance went through the floor when H4 came out, and never quite recovered. |
[11:07] | Andrew Linden: | There is a team that was working on some simulator performance stuff |
[11:07] | Andrew Linden: | in fact, Simon was one of the main guys on that team |
[11:07] | Gaius Goodliffe: | Speak of the Devil... :) |
[11:07] | Rex Cronon: | wb |
[11:08] | Simon Linden: | Yeah, performance is not great, particularly with Open Spaces. |
[11:08] | Simon Linden: | It's a variety of problems, there's no one-fix magic cure |
[11:08] | Arawn Spitteler: | It sounds like more avies to the prim right |
[11:08] | Gaius Goodliffe: | FWIW, on the topic of lingering bugs, I can confirm as of yesterday that bad rez attempts can still leave ghost objects (no prims, but their physics are there anyhow). |
[11:09] | Simon Linden: | There are some improvements in the pipeline, such as a speed-up in our internal parser (I talked about that last week) |
[11:09] | Arawn Spitteler: | I thought that was bad delete atempts? |
[11:09] | Andrew Linden: | so as to not be ambibuous between what we used to call "ghost prims" -- objects that you could see but were no longer on the region, I've been calling them "ghost colliders" |
[11:09] | Gaius Goodliffe: | Possibly... |
[11:09] | Andrew Linden: | er... ambiguous |
[11:10] | Gaius Goodliffe: | In this case, it was invisible objects that you can still collide with. |
[11:10] | Andrew Linden: | right |
[11:10] | Gaius Goodliffe: | Big invisible jet on my runway, quite a hazzard... |
[11:10] | Arawn Spitteler: | Ambibuous, to be Ambitious or Ambiguous, but eager not to distinguish |
[11:10] | Gaius Goodliffe: | I blame Wonder Woman. |
[11:10] | Andrew Linden: | hrm... Gaius, do you have a reliable repro recipe for that bug? |
[11:11] | Gaius Goodliffe: | Nope. Seems to only happen when I try rezzing/deleting things during high lag. |
[11:11] | Gaius Goodliffe: | The "ghost" remains until the sim is rebooted. |
[11:11] | Arawn Spitteler: | Maybe it could be used with a Griefer's Toy? |
[11:12] | Gaius Goodliffe: | Nope. |
[11:12] | Rex Cronon: | u can't view it, nor select it? |
[11:12] | Rex Cronon: | does a scan detect it? |
[11:12] | Gaius Goodliffe: | It's not there. But its physics are. |
[11:12] | Andrew Linden: | i thought the "collision body cleanup" code was really tight, but apparently there is still a bug there |
[11:12] | Gaius Goodliffe: | Not sure about a scan... |
[11:12] | Arawn Spitteler: | Have you seen what data llVolumeDetect returns? |
[11:13] | Gaius Goodliffe: | Hmm. No, I |
[11:13] | Gaius Goodliffe: | 'l have to try that next time... |
[11:13] | Arawn Spitteler: | Key, Mass, Owner? |
[11:13] | Andrew Linden: | hrm... that is an interesting idea, llVolumeDetect() |
[11:13] | Gaius Goodliffe: | It only seems to happen about once every other month for me though... |
[11:13] | Arawn Spitteler: | It should collide |
[11:14] | Gaius Goodliffe: | Just my luck, it won't happen again this year... |
[11:14] | Arawn Spitteler: | Readiness is your best preention |
[11:14] | Gaius Goodliffe nods. | |
[11:14] | Andrew Linden: | Well, I'll try to do some work on that problem... someday soon |
[11:14] | Arawn Spitteler: | Preventive Pre-emption |
[11:14] | Gaius Goodliffe: | I'll make a volume detector so I have it ready for next time. |
[11:14] | Simon Linden: | Someone last Thursday made a box you could drop on something, and it would spit out info ... I didn't get a copy. |
[11:15] | Andrew Linden: | I've just got so much stuff pliling up lately, it is hard to estimate a timeline on it |
[11:15] | Gaius Goodliffe: | Yeah not too hard to do... |
[11:15] | Gaius Goodliffe nods. | |
[11:15] | Andrew Linden: | I'd be interested to know if an LLVolumeDetect() collision with the invisible collider returns valid info |
[11:16] | Andrew Linden: | if it does, then it means there is some particular type of data that really is lost |
[11:16] | Andrew Linden: | and that might offer a clue on how to track it down |
[11:16] | Arawn Spitteler: | Can you flag the support requests, for Ghost Colliders? |
[11:17] | Andrew Linden: | that is... if the owner or other browsable info is valid on the detected object |
[11:17] | Gaius Goodliffe: | FWIW, what happened to me the other day ways, I tried to rez a jet, nothing happened, then tried two more times. Finally, about 15 seconds later, two jets appears at the location. I deleted one, then hopped in the other and it became immediately obvious I was colliding with something, so I deleted the second and found there was still an invisible third that couldn't be seen or selected in any way, but the physical shape was there anyhow. |
[11:17] | Rex Cronon: | gaius, that is what happens when there is lag |
[11:18] | Gaius Goodliffe nods. | |
[11:18] | Andrew Linden: | so someone new showing up would also not see the object, and it didn't show up in the parcel prim counts? |
[11:18] | Gaius Goodliffe: | I neglected to check the parcel prim counts, unfortunately... |
[11:18] | Gaius Goodliffe: | But it was definitely still there after I logged out and back in. |
[11:18] | Andrew Linden: | and a relog for yourself did not reveal the invisible object? |
[11:18] | Andrew Linden: | oh ok |
[11:18] | Gaius Goodliffe: | Right |
[11:19] | Rex Cronon: | if something doesn't show up , i have to rez boxes until finally all the boxes and the object show up all selected |
[11:19] | Gaius Goodliffe: | In fact it lingered there for an hour or so, then I finally rebooted the sim and th ghost went away. |
[11:19] | Andrew Linden: | Well, if anyone notices the ghost collider, I'd be curious to know if it shows up on parcel counts too. |
[11:20] | Arawn Spitteler: | About Land Objects? |
[11:20] | Gaius Goodliffe nods. | |
[11:20] | Andrew Linden: | Yes Arwan |
[11:20] | Gaius Goodliffe: | Last spring I had something similar happen while rezzing abuilding, and the prims then definintely did show up in ABout Land, but were unreturnable. |
[11:21] | Arawn Spitteler: | If a ghost collider is set physical, does it still drop? |
[11:21] | Andrew Linden: | Good question Arwan |
[11:21] | Andrew Linden: | I would guess "yes" |
[11:21] | Gaius Goodliffe: | Good question. Alas, my jets only turn physical after you sit on them, and when it's not there, you can't, even though you can collide with it. |
[11:22] | Arawn Spitteler: | So, we could rezz enought ping-pong balls, to lag the sim, and drop them on a collision detector |
[11:22] | Andrew Linden: | My desktop linux machine died last week, and I think the last office hour transcript is on the HD, so I haven't published it to the wiki yet |
[11:23] | Andrew Linden: | the motherboard died, so the HD is ok |
[11:23] | Arawn Spitteler: | Are Motherboards still available for U$25? |
[11:23] | Gaius Goodliffe: | If you want the kind that die quickly... :p |
[11:24] | Andrew Linden: | Well, there was a spare demo motherboard sitting around the office, so I'm going to try to use that |
[11:24] | Andrew Linden: | unfortunately... the case I got for it had a bad power supply |
[11:24] | Gaius Goodliffe: | (A decent motherboard runs around US$120-160 last time I looked.) |
[11:24] | Andrew Linden: | I'm almost cured of wanting to fight with computer hardware |
[11:25] | Arawn Spitteler: | Get one that was decent two years ago, the basic maintenance is to unplug and plug back in, and SL is supposed to be accessible |
[11:25] | Andrew Linden: | I've been out of touch on the latest MONO patch schedule |
[11:25] | Andrew Linden: | I think there was a release at the end of last week, and it got delayed |
[11:26] | Simon Linden: | I think there's an update this week |
[11:26] | Simon Linden: | The one at the end of last week didn't go out |
[11:26] | Andrew Linden: | ah yes, my email says a server deploy started yesterday |
[11:26] | Andrew Linden: | 1.24.6 |
[11:27] | Gaius Goodliffe: | Oh, on the subject of random feature requests -- I wish there was a way I could tell an object to not return to me upon going offworld without it having to be temp-on-rez. But not a big deal. (Just mention it because I just got a demo airship in my lost and found a moment ago.) |
[11:27] | Arawn Spitteler: | I thought that existed |
[11:27] | Rex Cronon: | die at edge? |
[11:28] | Gaius Goodliffe: | If you turn on temp-on-rez it'll silently delete, but it'll also disappear after a minute, and on this particular model it's nice for people ot be able to walk around inside looking at it for 10 minutes or so before taking it on a test flight |
[11:28] | Andrew Linden: | die at edge sounds familiar |
[11:28] | Andrew Linden: | that isn't a scriptable setting? hrm.... |
[11:28] | Gaius Goodliffe: | Hmm, I'll have to check... that might work (if it works, lol gotta love truisms) |
[11:29] | Glitch Braess: | Hello everyone :) |
[11:29] | Simon Linden: | Hi |
[11:29] | Rex Cronon: | hi |
[11:29] | Andrew Linden: | A quick scan of the code suggests that "die_at_edge" is not a feature yet |
[11:30] | Andrew Linden: | oh wait... |
[11:30] | Object: Hello, Avatar! | |
[11:30] | Rex Cronon: | what? |
[11:30] | Gaius Goodliffe: | Interesting... STATUS_DIE_AT_EDGE -- how long has that been there? |
[11:31] | Gaius Goodliffe never noticed before. | |
[11:31] | Arawn Spitteler: | http://www.lslwiki.net/lslwiki/wakka.php?wakka=llSetStatus |
[11:31] | Andrew Linden: | maybe it is a feature, there is one mention of it in the main script code |
[11:31] | Glitch Braess: | Ha, that is interesting. I've never noticed it myself. |
[11:31] | Gaius Goodliffe: | That may very well do exactly what I want, I'll have to experiement. |
[11:31] | Gaius Goodliffe: | Thanks :) |
[11:31] | Andrew Linden: | whew! for once we had the feature in place already |
[11:31] | Gaius Goodliffe: | hehe |
[11:32] | Glitch Braess: | lol |
[11:32] | Arawn Spitteler: | I was having a volume-detect follower dropping through the ground, but just set that to MoveToTarget where it was |
[11:32] | Andrew Linden: | Welcome Glitch, we're just talking about whatever. If you have a question or topic just throw it out on the table. |
[11:32] | Arawn Spitteler: | Resetting the things was cancelling MoveToTarget |
[11:32] | Glitch Braess: | Alright. :) |
[11:33] | Arawn Spitteler: | Gypsy Paz was having some problem with Vehicle Rezzers, for Burning Life, but he doesn't seem to be back from town, yet. |
[11:33] | Andrew Linden: | One of the LL developers was quizzing me the other day to see how hard it would be to add per-prim phantom |
[11:33] | Arawn Spitteler: | It was something about having to reset the VEHICLE_LINEAR_FRICTION |
[11:34] | Andrew Linden: | however, he was a little startled to realize that phantom actually means "collides with terrain only" |
[11:34] | Andrew Linden: | and he was thinking about a truely collisionless property |
[11:34] | Arawn Spitteler: | Volume Detect wasn't colliding with Terrain, when last I tried |
[11:35] | Andrew Linden: | VEHICLE_LINEAR_FRICTION was breaking the rezzability of his vehicles? I'm skeptical, but open minded. |
[11:35] | Andrew Linden: | VolumeDetect is truly collisionless, but with enter/exit callbacks |
[11:35] | Gaius Goodliffe: | (That don't work reliably) |
[11:35] | Arawn Spitteler: | Something the vehicles were doing, when rezzed, needed a setting in some cases, but not others. He's loacding donated vehicles |
[11:35] | Andrew Linden: | except against terrain, I don't think a callback happens there |
[11:36] | Andrew Linden: | Arawn, he has access to the scripts of the donated vehicles? |
[11:36] | Andrew Linden: | Well, this would be a good place to talk about new vehicle bugs. |
[11:36] | Arawn Spitteler: | He should, but isn't on line, at the moment |
[11:36] | Andrew Linden: | ok |
[11:37] | Andrew Linden: | In order to make sub-prims truly collisionless, we'd have to add a new feature |
[11:37] | Gaius Goodliffe: | (Notes https://jira.secondlife.com/browse/SVC-3106 regarding llVolumeDetect and collsion_start/end calls.) |
[11:37] | Andrew Linden: | [X] Collisionless Checkbox |
[11:39] | Andrew Linden: | Right Gaius, thanks for that link |
[11:39] | Andrew Linden: | that is just for the avatar right? |
[11:40] | Gaius Goodliffe: | Not sure. |
[11:40] | Andrew Linden: | The avatar is a special object... I wonder if SVC-3106 is related to how we sometimes pin the avatar body in place under the hood |
[11:41] | Gaius Goodliffe: | I also note that, although the bug says it's getting extra end events, it could be it's instead losing start events. |
[11:41] | Andrew Linden: | Hrm... this sounds familiar |
[11:42] | Andrew Linden: | There was a bug where avatars would collect too many collision start/end events, and some content was relying on this buggy behavior |
[11:42] | Andrew Linden: | I fixed the bug, and then the content broke |
[11:42] | Andrew Linden: | but I can't remember the jira #'s for that |
[11:42] | Gaius Goodliffe: | XD a misfeatre, then |
[11:42] | Gaius Goodliffe: | *misfeature |
[11:43] | Arawn Spitteler: | MisFeature or MisCreature |
[11:43] | Gaius Goodliffe: | It would seem to me that regardless of the amount, the number of starts and ends ought to be the same.. |
[11:43] | Andrew Linden: | SVC-3106 was created recently, so it may be new |
[11:44] | Andrew Linden: | I wonder if the bug was introduced in MONO instead of Havok4 project |
[11:44] | Gaius Goodliffe: | Hmmm. Could be... |
[11:45] | Andrew Linden: | well, it has a nice repro recipe, including the script, so it should be easy to track down |
[11:45] | Arawn Spitteler: | Stick an avatar on a spindle, between Mono and older version |
[11:45] | Gaius Goodliffe: | But as McCabe notes, these events have long been observed to be unreliable... |
[11:46] | Andrew Linden: | indeed |
[11:46] | Gaius Goodliffe: | Might have been around since H1... |
[11:46] | Andrew Linden: | also, there were some fixes for collision event handling... related to SVC-2511, that is in the 1.25-Server branch |
[11:46] | Arawn Spitteler: | A nice feature, if we could view collision forms, as with Sculpties, could we also see the Primatars? |
[11:46] | Andrew Linden: | there is a small chance that the spurious VolumeDetect events are related to that, but not sure. |
[11:47] | Gaius Goodliffe resists the urge to swich to his primatar. | |
[11:47] | Andrew Linden: | Arwan, that would be a nice feature. |
[11:47] | Andrew Linden: | I suspect the pressure for that feature to build until it must get done |
[11:48] | Andrew Linden: | the subject of mesh primitives came up again internally in LL last week or the week before |
[11:48] | Andrew Linden: | that is, providing a way to import meshes via some 3D design tool (maya, blender, whatever) |
[11:49] | Andrew Linden: | but if we support mesh objects then the collision shapes will also have to be supported |
[11:49] | Arawn Spitteler: | Sculpties are Viewer Side Textures, I'd expect Collision Mesh to burden the Sim |
[11:49] | Andrew Linden: | and sculpties need to start colliding correctly |
[11:49] | Gaius Goodliffe: | I've been told OpenSim supports that now (accurate collision shapes for sculties). |
[11:49] | Andrew Linden: | and builders are going to need to be able to visualize the true collision shapes |
[11:49] | Gaius Goodliffe: | *sculpties |
[11:50] | Rex Cronon: | so, is ll going to allow users to upoad .obj files? |
[11:50] | Andrew Linden: | Yeah, I'm not surprised that sculpties would already collide correctly in OpenSim |
[11:50] | Andrew Linden: | dunno Rex |
[11:50] | Arawn Spitteler: | Builders shouold be able to see collisibility, as for Hugger-Trains and Sculpties, and the Sitting Challenges, but Sculpty Colliders sound Sim-Costly |
[11:50] | Andrew Linden: | the topic pops up again and again over the years |
[11:51] | Gaius Goodliffe: | (To the extent anything does -- lol -- last time I looked at opensim the physics seemed a bit wonky). |
[11:51] | Andrew Linden: | I think it will eventually get done, and mabye *.obj files will be supported |
[11:51] | Rex Cronon: | if that is allowed, what r sculpties good for? |
[11:51] | Andrew Linden: | I'm unfamiliar with that format |
[11:51] | Rex Cronon: | very simple format, even a human can read and understand it:) |
[11:52] | Simon Linden: | I seem to recall it wasn't .obj files, but the format itself isn't the real problem, as someone (if not LL) will write an obj converter quickly |
[11:52] | Simon Linden: | The complication coming up was specifying all the levels of detail, which are needed for both rendering and physics |
[11:53] | Simon Linden: | Plus any SL specific attributes |
[11:53] | Andrew Linden: | right, the feature needed is mesh primitives, and some import/export path |
[11:53] | Andrew Linden: | two features |
[11:53] | Andrew Linden: | the two features are mesh primitives, some import/export path, and correct collisions |
[11:53] | Andrew Linden: | ... three features |
[11:53] | Arawn Spitteler: | Like specifying the levels of detail for Second Life, or Natural Philosophy. Levels of Detail might be something to be optimized on an ad-hoc basis |
[11:54] | Simon Linden: | Yeah, there's all sorts of triangle reduction methods out there. Some of it can be automated, but a builder is likely to want to see the results and tweak it |
[11:54] | Andrew Linden: | the three features are mesh primitives, some import/export path, correct collisions, and multi-group chat sessions |
[11:54] | Andrew Linden: | ... four features! |
[11:54] | Gaius Goodliffe: | BTW, this there a good reason why sculpty landing gear would be affected by prim seam collision problems more than non-sculpty landing gear (cut spheres, for example)? I made some sculpty landing gear, and the planes I made using them seem to hit the seams between prims on my runway a lot harder. |
[11:55] | Arawn Spitteler's not expected by the Spanish Inquisition: "Our three main weapons are: Suprise; Fear; A fanatical devotion to the Pope; these nifty red uniforms." | |
[11:55] | Gaius Goodliffe: | hehe |
[11:55] | Simon Linden: | Yes, a cut sphere will have a more accurate shape in the physics engine |
[11:55] | Andrew Linden: | Gaius, sculpties collide as convex hull's of torii, so I guess yeah they might be affected by prim seams more... they have more triangles that some other shapes |
[11:56] | Gaius Goodliffe: | mmm |
[11:56] | Gaius Goodliffe: | Makes sense. About time I replaced the runway with a big megaprim anyhow... |
[11:56] | Andrew Linden: | although, it may also just have to do with other things... the anti-prim-seams-magic is kinda complicated |
[11:57] | Arawn Spitteler: | I thought we had that one beat, by scripting the taper |
[11:57] | Andrew Linden: | points go to Arwan for catching the Monty Python reference I was going for. |
[11:58] | Gaius Goodliffe: | That could be, Arawn. It never seemed like a big problem before so I never bothered to play with the taper on the runway prims. At this point, it'll be simpler to just replace the series of prims with one long one. |
[11:59] | Gaius Goodliffe has a 120m long prim around here somewhere, should be perfect for the job.) | |
[11:59] | Andrew Linden: | Any last minute topics? This hour is almost up. |
[11:59] | Arawn Spitteler: | Should probably give it a try, with the taper trick, then just replace with the mega prim |
[12:00] | Arawn Spitteler: | Is the Primitar centered at Agent Target? |
[12:00] | Andrew Linden: | AgentTarget? |
[12:00] | Gaius Goodliffe: | The problem with the taper trick is, you then need something on the sides to disguise the holes created by the tapering |
[12:00] | Arawn Spitteler: | I understand you can taper it zero, as long as the prim thinks itelf tapered. |
[12:01] | Gaius Goodliffe: | o.O |
[12:01] | Gaius Goodliffe: | Okay, Ill have to look into this... |
[12:01] | Arawn Spitteler: | When we define seating, we set the target location for agent target. advanced -> Character ->.. |
[12:01] | Andrew Linden: | I've heard of that zero-taper oddness. I wonder if this is actually some quantization error thing that is actually introducing a non-zero positive taper |
[12:01] | Arawn Spitteler: | llDetectedPos returns the Pelvis, but the Displayed Target is elsewhere. |
[12:02] | Andrew Linden: | but I haven't been able to get to looking at that |
[12:02] | Andrew Linden: | Ah... the seated primitar is a different shape |
[12:02] | Andrew Linden: | and is not exactly centered at AvatarTarget |
[12:02] | Arawn Spitteler: | It's a bother, if you just want to say "/0Set Teleport", nd hte teleport then send people into the floor |
[12:03] | Arawn Spitteler: | Primitar isn't effected by Animation? |
[12:03] | Andrew Linden: | that is, the center of mass of the shape does not conincide with the avatar's center... there is some slight offset, as I recall |
[12:04] | Andrew Linden: | No, the avatar's collision shape (primitar, in this context) is not affected by animation |
[12:04] | Arawn Spitteler: | Stairs should teleport climbers on click, but we can't jsut position the builder, and touch to set |
[12:04] | Andrew Linden: | you could have an animation that moves the avatar around in some fancy dance, but the collidable shape will remain in place |
[12:05] | Arawn Spitteler: | So, when linked, the Primitar has a shape, different from when standing, by neither are where we're able to see, or where llDetectedPos is going to see |
[12:05] | Rex Cronon: | like this:) |
[12:05] | Gaius Goodliffe: | Right! :) |
[12:05] | Gaius Goodliffe: | Four different positions for your convenience. :) |
[12:05] | Andrew Linden: | yes Arwan, the actual collidable shape of the seated avatar is different from the standing avatar shape |
[12:06] | Andrew Linden: | and the center of the collidable is not necessarily the center of the avatar as seen on the client |
[12:06] | Andrew Linden: | or as reported by some script detector |
[12:06] | Andrew Linden: | however, the shape and offset are tweaked to attempt to provide an ok collision proxy for the appearent seated avatar shape |
[12:06] | Rex Cronon: | is there any way to get the actuall pos of an av, taking in consideration the animations? |
[12:07] | Arawn Spitteler: | Advanced -> Character -> Show Agent Target, that has no more to do with Primitar, than does llDetectedPos? |
[12:07] | Andrew Linden: | a script detector should report the avatar's true position |
[12:07] | Gaius Goodliffe: | "Where am I?" has many different answers. "Actual pos" is a point of view. |
[12:07] | Rex Cronon: | no, it doesn't |
[12:07] | Andrew Linden: | that is, the true position as used by the simulator code to figure out which parcel the avatar is standing on, which region it is in, etc |
[12:07] | Rex Cronon: | only particles can point to an avatar |
[12:07] | Arawn Spitteler: | llDetectedPos goes to Center of Bounding Box, from my observation |
[12:07] | Rex Cronon: | but u can't ask them for that pos |
[12:08] | Andrew Linden: | "center of bounding box" and distinct from... what? Arwan |
[12:08] | Gaius Goodliffe: | Particles show apparent position (i.e. visible position), not "actual" position. |
[12:08] | Andrew Linden: | that is, "center of bounding box" is not what you expeted? |
[12:08] | Arawn Spitteler: | Agent Target tends to be in front. |
[12:09] | Rex Cronon: | yes, visible pos, that is what i am interested in obtaining |
[12:09] | Gaius Goodliffe: | The problem is that visible pos is client dependent. |
[12:09] | Andrew Linden: | no Rex, the visible position is a client-side adjustment |
[12:09] | Rex Cronon: | there is no way right now to get that:( |
[12:09] | Andrew Linden: | the simulator does not know about visible position |
[12:09] | Arawn Spitteler: | About here, for me |
[12:09] | Gaius Goodliffe: | And it will be a different answer for different people. |
[12:09] | Rex Cronon: | the viewer(s) do know |
[12:10] | Andrew Linden: | Gaius is correct. |
[12:10] | Gaius Goodliffe: | SL Relativity. :) Like Special Relatively, but even more special. ;) |
[12:10] | Arawn Spitteler: | I built a follower, that hugs my hips, as myu position |
[12:10] | Andrew Linden: | alright, thanks for showing up. I've got to go soon. |
[12:11] | Rex Cronon: | bye everybody |
[12:11] | Simon Linden: | Bye everyone |
[12:11] | Rex Cronon: | have fun:) |
[12:11] | Glitch Braess: | Nice seeing you all. :) |
[12:11] | Gaius Goodliffe: | see ya later |
[12:11] | Rex Cronon: | nice . sim down:( |
[12:11] | Gaius Goodliffe: | Be excellent to each other. :) |
[12:11] | Arawn Spitteler: | File a support ticket? |