User:Andrew Linden/Office Hours/2009 10 22

From Second Life Wiki
< User:Andrew Linden‎ | Office Hours
Revision as of 21:22, 14 November 2009 by Ardy Lay (talk | contribs) (Created page with '<!-- Transcript generated with [http://slog.whiz-kids.de SLog Wikifier] -->{{#if: }} <div id='box'> == Transcript == <div style='padding: 0.5em'> *<span id="msg_0"></span>[[#ms...')
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Transcript

  • [17:00] Sebastean Steamweaver: I haven't actually seen what LL had to say about that in the aftermath.
  • [17:00] Sebastean Steamweaver: There's been a huge discussion about it in the SL dev groups though.
  • [17:00] Sebastean Steamweaver: Hey Simon
  • [17:00] Simon Linden: hello
  • [17:00] Sebastean Steamweaver: Is Andrew back?
  • [17:00] Kaluura Boa: Hello everybody
  • [17:00] Fred Rookstown: Yeah, all I see is a couple of Lindens jumping around in the blogs basically asking everyone to continue a discussion that is really going nowhere
  • [17:00] Psi Merlin: Hello Simon
  • [17:00] Welcome to Linden office hours
  • [17:00] Simon Linden: I'm not sure .... haven't heard from him today, I'll go see...
  • [17:01] Ardy Lay: Hi Simon
  • [17:01] Kaluura Boa: Rez, rez, rez...
  • [17:01] Moon Metty: hellooo :)
  • [17:01] Fred Rookstown: Hi Morg
  • [17:01] Rex Cronon: hello everybody
  • [17:01] Simon Linden: hmm, I don't see Andrew around, looks like he's still out
  • [17:02] Sebastean Steamweaver: Heya Moon, Morg, and all.
  • [17:02] Morgaine Dinova: Hi Fred, Seb, Moon :-)
  • [17:02] Moon Metty: hi hi
  • [17:02] Morgaine Dinova: Hi Simon
  • [17:02] Latif Khalifa: hello all :)
  • [17:02] Rex Cronon: hiii
  • [17:02] Morgaine Dinova: Hi Latif
  • [17:02] Simon Linden: Let's see, annoucments... I think there's the next server (1.32) pilot roll coming in a week or two
  • [17:03] Moon Metty: the bugfixes look good
  • [17:03] Sebastean Steamweaver: Anything exciting in it?
  • [17:03] Fred Rookstown: Does that actually fix the recent notecard exploits or are they still broken?
  • [17:03] Dilbert Dilweg: Hopefuly Sim performance fixes
  • [17:03] Moon Metty: http://wiki.secondlife.com/wiki/Release_Notes/Second_Life_Beta_Server/1.32
  • [17:03] Simon Linden: I'm pretty clueless on the big items ... some bug fixes did go in
  • [17:03] Morgaine Dinova: I hope that the fact that there's a dog's head on the table has nothing to do with my katana.
  • [17:04] Moon Metty: svc-3819 is in 1.32 too
  • [17:04] in estate tools is inaccurate
  • [17:04] Arawn Spitteler: So, we aren't tryingto finish 1.30.7?
  • [17:04] Simon Linden: The notecard exploit should be fixed, I think
  • [17:04] Fred Rookstown: FINALLY
  • [17:04] Simon Linden: Hello Andrew
  • [17:04] Fred Rookstown: @ SVC 3819
  • [17:04] Moon Metty: hi andrew :)
  • [17:04] Andrew Linden: Hi everyone.
  • [17:04] Arawn Spitteler: Thisis a blue kitty friendly Service Animal
  • [17:04] Rex Cronon: hellor andrew
  • [17:04] Morgaine Dinova: Hi Andrew
  • [17:04] Sebastean Steamweaver: Hey Andrew
  • [17:05] GG Office Hours HUD v1.5: 5:00 PM: Andrew Linden: Technical issues, SL architecture, physics engine
 http://slurl.com/secondlife/Denby/213/45/34  (Started 5 minutes ago)
  • [17:05] Psi Merlin: Hi Andrew
  • [17:05] Simon Linden: That sit position was hurting my head
  • [17:06] Arawn Spitteler: Hi, AdrewI'm prbably in a freeze crash. Want to mention http://jira.secondlife.com/browse/SVC-4944 which was improved by a Burning La Restart, but remains a problem in my own sim
  • [17:06] Sit Action Objects, without Sit Targets, aren't working in my build.
  • [17:06] Andrew Linden: I think Phoenix did most of the work to fix the notecard upload exploit, but there was a bug. Babbage and Scouse fixed the bug, and Joe Linden helped push the priority at the end so that it got out in 1.30 rather than 1.32.
  • [17:06] Arawn Spitteler: If I culd ge agent target, I could do a lot better than this table
  • [17:07] Moon Metty: now we are left with no-trans notes not saving
  • [17:07] Fred Rookstown: llSitTarget needs to be set before the changed() event will be triggered, if that's what you're asking
  • [17:07] Moon Metty: SVC-4937
  • [17:07] Server is experiencing unexpected difficulties error when saving a notecard inside a prim.
  • [17:08] Arawn Spitteler: Changed event works on changed links, so sit targete never had to be set before
  • [17:08] Morgaine Dinova: Oh, is that what's caused the sit breakage all across SL, a change in timing of changed() event?
  • [17:09] Morgaine Dinova: I'm still deeply embedded everywhere
  • [17:10] Andrew Linden: There is a bug that causes scripted object links to fail... sometimes.
  • [17:10] Andrew Linden: That one is also intermitent and can vanish when the region is restarted.
  • [17:10] Andrew Linden: I wonder if that is related to SVC-4944
  • [17:11] Sebastean Steamweaver: The JIRA helper is silent o.0
  • [17:11] Fred Rookstown: BTW, any news on fixing the damn region freeze bug caused by mono?
  • [17:11] Arawn Spitteler: Scripted Object inks? Mine is ust an inability to sit, on a basic cube
  • [17:11] Simon Linden: No, but Babbage and another dev were looking at it
  • [17:11] Arawn Spitteler simply opts not to use Mono, and his scripts compile faster
  • [17:11] Andrew Linden: No, the jira helper will not re-mention an item if it was brought up not too long ago. Dunno what the timeout is on it.
  • [17:12] Fred Rookstown: Oh, Babbage is alive? He never shows his face anymore at officehours...
  • [17:12] Arawn Spitteler: I think it'sabout half a hour
  • [17:12] Moon Metty: one hour, or 12 other jira issues, Andrew
  • [17:12] Andrew Linden: yes, looks like Babbage was making some progress tracking down the lag on object rez, relating to MONO scripts
  • [17:13] Dilbert Dilweg: Looks Like that one is being ignored Fred
  • [17:13] Simon Linden: Well, he is alive, I can confirm that
  • [17:13] Andrew Linden: Babbage sent out an email talking about it. I skimmed it. He had found two or three possible hot spots, but wasn't done tracking it down.
  • [17:13] Rex Cronon: sai. did u link to images on your harddrive?
  • [17:13] Rex Cronon: oopes. wrong chat
  • [17:14] Psi Merlin: Anything to help the situation will be most welcome.
  • [17:14] Ardy Lay: Hello Ann
  • [17:15] Sebastean Steamweaver: Andrew, I had some questions about attachment permissions?
  • [17:15] Andrew Linden: I'm out of touch with the current state of server-1.32. Dunno what they're doing there. Simon, you're closer to the 1.32 dev business these days right? or is your main project slated for 1.34?
  • [17:15] Andrew Linden: Go ahead Sebastean.
  • [17:15] Simon Linden: Well, 1.32 is lurching towards a pilot roll in a week or two
  • [17:16] Simon Linden: Most of the stuff I'm working on is in 1.34
  • [17:16] Sebastean Steamweaver: When wearing an attachment, if the attachment is transferrable, you aren't able to put no-transfer inventory into them. I heard this was done because when you detached the attachment, it retained its transferpermission. Is there any way to change it so an attachment's permission can update, and re-enable putting no-trans objects into transferable items?
  • [17:16] Kaluura Boa: Ho yeah, please!
  • [17:17] Moon Metty: scary
  • [17:17] reddot99 Republic: modify is even more important there
  • [17:17] Andrew Linden: I hadn't heard about that Sebastian, but I'm not surprised.
  • [17:18] Andrew Linden: I think yes, it would be possible in theory to fix it -- probably a medium bug to fix.
  • [17:18] reddot99 Republic: it used to be ok,
  • [17:18] Moon Metty: it would mean slamming an unrezzed object
  • [17:19] Fred Rookstown: In other words, it'll be addressed in about 4 years.
  • [17:19] Sebastean Steamweaver: It makes it difficult for some products, especially scripts, that one person would product and another person would put into their attachments, which may or may not be transfer.
  • [17:19] Arawn Spitteler: Hihi, Hayleigh, Empty seat tends to be in the sothwest
  • [17:19] Arawn Spitteler: southeast
  • [17:20] Fred Rookstown: Next to the guy with the red hat
  • [17:20] Fred Rookstown: :P
  • [17:20] Simon Linden: Between Andrew and Fred
  • [17:20] Hayleigh Firehawk: thank you
  • [17:20] Sebastean Steamweaver: He who sits First, sits Last.
  • [17:20] reddot99 Republic: andrew, i have a purely physics related inquiry, is there any wa y to ensure a prim stops objects that are fired at it, like bullets and barriers
  • [17:20] Arawn Spitteler: If an inventory item is placed in another, the object should get a changed event, which migh check content perms
  • [17:20] Rex Cronon: ions are useless if the chair is not visible
  • [17:20] Fred Rookstown: red: That's a problem in FPSes
  • [17:21] reddot99 Republic: exactly, where does it look like i just came from?
  • [17:21] Andrew Linden: reddot99, could you elaborate? You want the object that is hit to stop the bullets?
  • [17:21] Fred Rookstown: From what I've seen in the source engine, objects with physical velocitiees greater than sometimes will go right through an object without even being slightly deflected
  • [17:22] reddot99 Republic: what fred said
  • [17:22] Simon Linden: That's the classic 'bullet through paper' problem with simulations
  • [17:22] Fred Rookstown: Think it has to do with the objet going too fast for the physics system to track collisions
  • [17:22] Arawn Spitteler: Longer bullets?
  • [17:22] Andrew Linden: correct Fred, moving objects don't move smoothly -- they pop to their next position
  • [17:22] Moon Metty: lol arawn
  • [17:22] reddot99 Republic: except, its meant to be a wall, not a peice of paper
  • [17:22] Andrew Linden: the physics engine runs at 45 fps max
  • [17:22] Moon Metty: that might work
  • [17:22] Fred Rookstown: Source engine solves this problem by firing rays with a tracer particle, instead of an actual bullet
  • [17:23] Andrew Linden: er... 45fps period, as far as the engine is concerned.
  • [17:23] Simon Linden: It comes down to how it calculates the move to the next position
  • [17:23] Andrew Linden: each engine step forward is 1/45th of a second
  • [17:23] Andrew Linden: so you can calculate the distance of the "jump" of a moving object if you know its velocity
  • [17:23] reddot99 Republic: longer bullets, messes up because it causes them to collide with each other
  • [17:23] Fred Rookstown: Maybe a speed cap on objects?
  • [17:23] Morgaine Dinova: Arawn's actually right. A "bullet" is actually two things, a short object plus a long effect. Let the physics handle the long effect, not the short object.
  • [17:24] Andrew Linden: yes, you need to get fancy to make reliable+fast bullets
  • [17:24] Fred Rookstown: We already have a 250m/s speed cap on avs.
  • [17:24] Arawn Spitteler: I'd think the Smulato would elongate the bullets, to check for collisions
  • [17:24] reddot99 Republic: 250 applies to everything
  • [17:24] Andrew Linden: a few gun creators have worked around the problem for some improvements
  • [17:24] Rex Cronon: r bullets supposed to go 5mps
  • [17:24] reddot99 Republic: faster rex,
  • [17:25] reddot99 Republic: many rifles break the sound barrier
  • [17:25] Fred Rookstown: But what's the terminal velocity, red?
  • [17:25] reddot99 Republic: 250,
  • [17:25] Andrew Linden: so what you really want reddot99 is for the physics engine to take more steps per second
  • [17:25] reddot99 Republic: here,
  • [17:25] Rex Cronon: i forgot to put the smilly face at the end of my previous sentence:)
  • [17:25] Fred Rookstown: One idea may be to factor in air resistance/drag?
  • [17:26] Fred Rookstown: that would slow the bullet in-flight
  • [17:26] Andrew Linden: also... the modern Havok engines can do what is called "continuous collisions" which handles the "jump too far each step" problem
  • [17:26] Andrew Linden: however, the continuous collisions feature was CPU-expensive to turn on, so we never did. We're not using it.
  • [17:26] reddot99 Republic: no, i thnk a sorta effect borrowed from relativity is relavent, as things accelorate, they get shorrter, or in this case, longer,
  • [17:27] Arawn Spitteler: I'd expect, you don't hve to calculate each step, if you extend the trajectory as a solid
  • [17:27] Fred Rookstown: Andrew: Have you checked that feature's performance on H6?
  • [17:27] Andrew Linden: No Fred
  • [17:28] reddot99 Republic: yes, please look at better physics engines ,.<
  • [17:28] You decline Club Zero Gravity from A group member named Ranchan Weidman.
  • [17:28] Arawn Spitteler: We could proably wite a better physics engine
  • [17:28] Fred Rookstown: No, Arawn, just no.
  • [17:28] Fred Rookstown: ;_;
  • [17:28] Andrew Linden: We might hire a new dev who has some experience using the Havok phys engine in previosu work. It would be nice to get some more Havok experience in the lab.
  • [17:29] reddot99 Republic: not unless you have a degree in physics,
  • [17:29] Fred Rookstown: You fired a guy who had H4 experience.
  • [17:29] Arawn Spitteler: Could you et such a person past psychological screening?
  • [17:29] Moon Metty: lol
  • [17:30] Andrew Linden: Arawn, it is a lot of work to write a good physics engine.
  • [17:30] Moon Metty: i can only imagine, Andrew
  • [17:30] Rex Cronon: gee reddot99. u don't need a degree in physics to compute the trajectory of an object:)
  • [17:30] office hours is half over
  • [17:30] Fred Rookstown: And seeing as how someone at the lab put too much trust in the client, I wouldn't trust them that much...
  • [17:31] Arawn Spitteler: Whoever gave us SVC-93 didn't hve a degree of rotation
  • [17:31] ROTATION and llSetRot incorrectly implemented for child prims
  • [17:31] Andrew Linden: computing the trajector is not hard. Computing the collisions and their results, accurately and quickly, is the hard part.
  • [17:32] reddot99 Republic: also, i just remembered, could we ge a restitution coefficient? if i spelled that right?
  • [17:32] Rex Cronon: isn't that what havok is supposed to do:)
  • [17:32] reddot99 Republic: get,
  • [17:32] Andrew Linden: reddot99, you want to be able to set the restitution?
  • [17:32] Arawn Spitteler: Restitute?
  • [17:32] Simon Linden: For collision 'bounciness' and such?
  • [17:32] Andrew Linden: right now the various materials have different restitutions, but you can only chose the materials we have available.
  • [17:33] reddot99 Republic: exactly
  • [17:33] Andrew Linden: yes, 'restitution' refers to how much energy is concerved on "bounce"
  • [17:33] reddot99 Republic: oh they do andrew?
  • [17:33] reddot99 Republic: could we get a list?
  • [17:33] Andrew Linden: The materials have distinct restitution, however you don't have much choice... just the few materials available.
  • [17:34] Andrew Linden: Also, the restitution of the materials might not be realistic -- there was some tweaking to help keep the physics engine stable.
  • [17:34] reddot99 Republic: yeah, but it'd be nice to be able to tell what it is,
  • [17:34] Andrew Linden: I think most of the values are a bit low (more lost energy) than real materials.
  • [17:34] Andrew Linden: Oh hrm...
  • [17:34] Andrew Linden: hang on...
  • [17:35] reddot99 Republic: tweaking meaning making sure it doesnt go over 1 or under 0?
  • [17:36] reddot99 Republic: i would like to know if theres a material with a value of zero,
  • [17:36] Fred Rookstown: Tweaking as in making sure that 250 rubber objects being rezzed in one spot != crashed server
  • [17:36] Simon Linden: yeah, anything above or below that would get interesting
  • [17:36] Arawn Spitteler: BAM Ceramic?
  • [17:37] reddot99 Republic: above would be a ball that bounce hieght increasing with each bounce
  • [17:37] Arawn Spitteler: Flubber
  • [17:37] Rex Cronon: it would be nice if when we set the material for a prim the physics engine would take it int consideration:)
  • [17:37] Andrew Linden: Hrm... I think the physics restituions are in the open-source bit of code
  • [17:37] Andrew Linden: F32 const LLMaterialTable::RESTITUTION_MIN = 0.02f;

F32 const LLMaterialTable::RESTITUTION_LAND = LLMaterialTable::RESTITUTION_MIN;

  • [17:37] Fred Rookstown gasps
  • [17:38] reddot99 Republic: theres a material max?
  • [17:38] Rex Cronon: ?
  • [17:38] Morgaine Dinova: No dilithium?
  • [17:38] Andrew Linden: yeah the MAX is used in the ::setRestitution() method to ensure that someone doesn't accidentally set it too high
  • [17:39] Fred Rookstown: Setting the land restitution to rubber would be hilarious.
  • [17:39] Rex Cronon: why doesn't matter what matterial the bullets r made of? all of them bounce the same:(
  • [17:39] Andrew Linden: if you set it too high then you can have instability in the physics engine
  • [17:39] reddot99 Republic: and also,i'd like to ask for a restitution of zero material
  • [17:39] Fred Rookstown: yeah
  • [17:39] reddot99 Republic: ie, "glue"
  • [17:39] Fred Rookstown: 0 bounce would be great, my grappling hook keeps ricogheting off of stuff
  • [17:39] Andrew Linden: ok, I'll make a note. Zero restitution wouldn't be too hard.
  • [17:39] reddot99 Republic: exactly
  • [17:40] Moon Metty: good idea, glue
  • [17:40] Rex Cronon: bullets would benefit from a zero value:)
  • [17:40] Fred Rookstown saves list
  • [17:40] Arawn Spitteler: Glue mightstick long enough, to collide at the point of collision
  • [17:40] Andrew Linden: not "glue" so much as ... half-filled bean bag
  • [17:40] Moon Metty: lol right
  • [17:41] Moon Metty: real glue may have side effects
  • [17:41] Fred Rookstown: Wait
  • [17:41] reddot99 Republic: i want an easier way to do sticky grenades
  • [17:41] Simon Linden: Right, something hitting a wall would stop, then fall to the ground
  • [17:41] Moon Metty: :)
  • [17:41] Fred Rookstown: Wouldn't glue be simulated with a negative restitution?
  • [17:42] Simon Linden: Negative would push you in the direction you were going, not bouncing back
  • [17:42] Andrew Linden: uh... negative restitution would be a bad idea
  • [17:42] Rex Cronon: nally with glue u could play some practical jokes infolving chairs and the eople sitting on them:)
  • [17:42] Twisted Laws: collision_start(integer n) { llSetStatus(STATUS_PHYSICS,FALSE); }
  • [17:42] Rex Cronon: finally with*
  • [17:42] Arawn Spitteler: Try Ethyl Btyrate
  • [17:42] reddot99 Republic: its not fast enough twisted
  • [17:42] Fred Rookstown: Twisted: You forgot lag
  • [17:42] reddot99 Republic: and fails in laggy sims
  • [17:43] Arawn Spitteler: Yeah, the problem with cllisions, is that we don't get historic data
  • [17:43] Twisted Laws: if the llDetectedPos is right, you can set pos back to that pos
  • [17:43] Fred Rookstown: it rarely is.
  • [17:43] reddot99 Republic: no, that sets center of detected object
  • [17:44] Moon Metty: Andrew, i had an idea about encroachment
  • [17:44] reddot99 Republic: quite useless if you have a linked object
  • [17:44] Fred Rookstown: even worse is a lack of hitnormals
  • [17:44] Andrew Linden: Go ahead Moon.
  • [17:44] Moon Metty: it seems a bit silly at first
  • [17:44] Moon Metty: you have been trying to make server code
  • [17:44] Fred Rookstown: so that bulletholes or blood can be rotated to match the prim's face
  • [17:45] Moon Metty: but the server doesn't always know where everything is
  • [17:45] Moon Metty: a client does, though
  • [17:45] Moon Metty: so i was thinking
  • [17:45] Fred Rookstown: No, don't trust the client with ANYTHING.
  • [17:45] Moon Metty: what if you make a Linden Bot to do the detecting?
  • [17:45] Moon Metty: it would be like mobile server code
  • [17:45] Fred Rookstown: Never, ever trust a client.
  • [17:45] Andrew Linden: actually, the server does know where everything is.
  • [17:46] Fred Rookstown: And the physics engine is supposed to keep track of positions.
  • [17:46] Andrew Linden: Give an example where the server does not know what the client does.
  • [17:46] Andrew Linden: Hrm... particles may be.
  • [17:46] Moon Metty: sculpties
  • [17:46] reddot99 Republic: particles
  • [17:46] reddot99 Republic: animations
  • [17:46] Moon Metty: client side rotating objects
  • [17:46] Fred Rookstown: Clouds (arguably)
  • [17:46] reddot99 Republic: lltarget mega
  • [17:46] Andrew Linden: Well, I'd like to give sculpties better physics shapes
  • [17:46] reddot99 Republic: omega
  • [17:46] Andrew Linden: then the server woudl know
  • [17:47] Moon Metty: alright, but a bot could work, in theory
  • [17:47] Andrew Linden: yeah you're right, client-side rotating objects
  • [17:47] Simon Linden: yeah, lltargetomega can be misleading or misused
  • [17:47] Fred Rookstown: Steal the code from OpenMetaverse, they have decent sculpties last I checked...
  • [17:47] Rex Cronon: it might be easier if we could select the physics shape that havak sees for a sculpties
  • [17:47] Moon Metty: and the benefit is, that you can update it without changing the server
  • [17:47] Fred Rookstown: Entire walkable landscapes made of gigantic sculpties
  • [17:47] Simon Linden: yep, and other objects too
  • [17:47] Kaluura Boa: About physics shapes, will the meshes also be spheres for Havok?
  • [17:47] Moon Metty: a resident could call the bot
  • [17:47] Fred Rookstown: No no no
  • [17:48] Moon Metty: point out the encroaching objects
  • [17:48] Fred Rookstown: No Linden bots.
  • [17:48] Fred Rookstown: If you have to use bots, you're doing it wrong
  • [17:48] Moon Metty: and deal with it
  • [17:48] Andrew Linden: Kaluura, ask Qarl Linden about mesh physics. I think he's working on the mesh project.
  • [17:48] Rex Cronon: some sculpties r better as balls, others as cubes, and even as torus:)
  • [17:48] Kaluura Boa: OK... I'll do
  • [17:48] Andrew Linden: Yes Rex, it would be nice if the collision approximations could be selected.
  • [17:49] Andrew Linden: Or perhaps if there were an algorithm that would find best-fit against a set of simple convex shapes.
  • [17:50] Andrew Linden: I would love to work on such a project. I have little hope of getting to it until we hire many more devs.
  • [17:50] Arawn Spitteler: Then, youll be developing the devs
  • [17:50] Moon Metty: hehe
  • [17:50] Fred Rookstown: LL seems deadset on losing as much money as possible with some of the recent projects they're doing.
  • [17:50] Fred Rookstown: So you may be seeing pinkslips instead
  • [17:50] Arawn Spitteler: Yes, but that's under the NDA
  • [17:51] Moon Metty: mmm, it's strictly viewer-side
  • [17:52] Moon Metty: but i had this bright idea too: VWR-15852
  • [17:52]
  • [17:52] Simon Linden: Hmm, interesting. Sort of like the no-entry walls?
  • [17:52] Moon Metty: yeah Simon
  • [17:53] Arawn Spitteler: Encroachment walls would detect objects, at their centers?
  • [17:53] Andrew Linden: Hrm... Moon for encroachment solutions we talked about the aggresive anti-encroachment per-parcel setting
  • [17:53] Simon Linden: This is just a building tool
  • [17:53] Moon Metty: it's just a visual aid
  • [17:53] Moon Metty: for builders
  • [17:53] Andrew Linden: if such a setting were available, and the viewr knew about it, then it could enforce the llTargetOmega() objects
  • [17:53] Fred Rookstown: I'm liking that encroachment wall idea...
  • [17:54] Moon Metty: let the residents know what they're doing, and you get a lot less load on support
  • [17:54] Arawn Spitteler: So, what we design, would stay inside our walls, rather than running off?
  • [17:54] Moon Metty: :)
  • [17:54] Fred Rookstown: Especially since the sims I administer require users to build in the sky due to frequent terraforming by the owner.
  • [17:54] reddot99 Republic: would also like a set hp function for objects, wandering back to combat related physiics issues <.<
  • [17:55] Simon Linden: If we had that, there'd be less excuse for accidentally encroaching onto another parcel
  • [17:55] Andrew Linden: yes, we talked about a visual indicator in the UI when people try to place objects where the encroachment rules clearly would prevent them -- it would need to be written eventually
  • [17:55] Andrew Linden: er... we've kicked that one around the lab
  • [17:55] Simon Linden: This could be it
  • [17:55] Arawn Spitteler: Oh, before th hour,where arre transcripts going? I wanted to show someone how seriously Linen Labs was taking her honey drop
  • [17:55] Ann Otoole grumbles or if people knew how to use a calculator they would not have encroachment issues...
  • [17:55] office hours is almost over
  • [17:55] Andrew Linden: Arawn, I've been collecting transcripts but have fallen out of habit of publishing them.
  • [17:56] Andrew Linden: It is my fault, and now I've got four months backlog.
  • [17:56] Fred Rookstown: Ann: Have you seen how irregular some parcels on the mainland can be?!
  • [17:56] reddot99 Republic: so, what happened to a sethp?
  • [17:56] Ann Otoole: yep. and it is always some variance on squares due to the way the land tool works.
  • [17:56] Andrew Linden: sethp? llSetHitPoints()?
  • [17:57] Fred Rookstown: Ann: My first parcel was basically a narrow L, so even then it wouldn't work.
  • [17:57] Opensource Obscure: Andrew: some residents may be willing to help you with transcripts, those are not sensitive transcripts. how could we help you?
  • [17:57] reddot99 Republic: yes, for objects, or vehicles, with a 10,000 limit perhaps,
  • [17:57] Opensource Obscure apologizes for his unclear-3AM-English
  • [17:57] Simon Linden: Parcels can also be disjointed, which makes it even more fun
  • [17:58] Ann Otoole: still all squares
  • [17:58] Moon Metty: the walls could also make supporst's life easier if they get screenshots, btw
  • [17:58] Ann Otoole: an L is 2 rectangles
  • [17:58] Andrew Linden: Opensource, that's a good idea, however I could probably write a script to process most of them if I just started trying.
  • [17:58] Fred Rookstown: lol, they're made up of tiny 4x4 squares, yes, but it's not as easy as you claim
  • [17:58] Andrew Linden: I'll try to update them soon.
  • [17:58] Arawn Spitteler: If we can't teah linear algebra, to lindens, how can we teach lawyers to use calculators?
  • [17:58] Ann Otoole: rofl
  • [17:58] Ann Otoole: call it dollars instead of meters?
  • [17:58] Arawn Spitteler: Oh, svc-22
  • [17:59] Vehicles crossing region borders aren't always treated as vehicles and can get incorrectly returned if the destination parcel is no-entry or parcel-full
  • [17:59] Sebastean Steamweaver: SVC-2885
  • [17:59] llSetObjectScale and llGetObjectScale
  • [17:59] Moon Metty: just in time, Arawn o_O
  • [17:59] Sebastean Steamweaver: MISC-3077
  • [17:59] List of Linden-Confirmed Easy Changes/Additions With Large Returns
  • [17:59] reddot99 Republic: they dont want to know, so they can overbill you,
  • [17:59] Fred Rookstown: I dunno, my lawyer's decent. I pay him in beer.
  • [18:00] Arawn Spitteler: Speakng of Beer, visit my Burning Life Build, Second Sumeran Bank of Decommmodification, where Beeris holy
  • [18:00] Arawn Spitteler: svc-2931
  • [18:00] Opensource Obscure: :D
  • [18:00] llSetLinkPrimitiveParams content breakage, only moves agents 54 meters now
  • [18:00] Rex Cronon: fred. r u sure u did't hire a con artist:)
  • [18:00] Thank you for coming to Linden office hours
  • [18:00] Fred Rookstown: SVC-3088 blah blah blah IP protection
  • [18:00] Expand group features to reduce need for more groups
  • [18:00] Fred Rookstown: what what
  • [18:00] Moon Metty: lol
  • [18:00] Fred Rookstown: wrong one
  • [18:00] Sebastean Steamweaver: SVC-2105
  • [18:01] for setting multiple, specific linked items
  • [18:01] Fred Rookstown: MISC-3088
  • [18:01] Block future exploits - Save script local
  • [18:01] Fred Rookstown: there we go
  • [18:01] Moon Metty: it's a good idea, but maybe not, and I'm not being indecisive
  • [18:01] Fred Rookstown: Rex: He likes cheap beer
  • [18:01] Andrew Linden: alright, I've got to go. Thanks for coming everyone.
  • [18:02] Moon Metty: thank you andrew :)
  • [18:02] Sebastean Steamweaver: Take care Andrew
  • [18:02] Simon Linden: I have to run too, nice to see everyone
  • [18:02] Sebastean Steamweaver: Later Simon
  • [18:02] Simon Linden: Thanks for coming
  • [18:02] Latif Khalifa: take care Andrew
  • [18:02] Rex Cronon: lol
  • [18:02] Latif Khalifa: laters Simon
  • [18:02] Moon Metty: and thanks Simon :)
  • [18:02] Ann Otoole: for a short time there was some beer sold here with 10% alcohol lol. till a county commissionor spotted it.
  • [18:02] Opensource Obscure: bye everybody
  • [18:02] Arawn Spitteler: Have we come? That'll be on the oft to the left
  • [18:02] Rex Cronon: tc andrew, simon
  • [18:02] Simon Linden: Bye all, see you next time

Generated with SLog Wikifier