User:Andrew Linden/Office Hours/2008 08 28

Transcript of Andrew Linden's office hours:

[17:26]	Gypsy Paz:	careful what you wish for ;)
[17:26]	Gypsy Paz:	hey Andrew ;)
[17:26]	Arawn Spitteler:	Did you just sit inthe same seat as I?
[17:26]	Bad Mojo Hat: No I didn't sit inthe same seat as you. What's it like?
[17:26]	Joulie Jewell gave you very important.
[17:26]	Fnordian Link:	Ah, Andrew Linden I presume. We were just talking about you. :)
[17:26]	Arawn Spitteler:	1.24 disaster?
[17:26]	Bad Mojo Hat: That's more precision than I need to know.
[17:26]	Andrew Linden:	hi
[17:26]	Andrew Linden:	I'm late because I totally forgot
[17:26]	Fnordian Link:	Greepings. :)
[17:27]	Andrew Linden:	There was some other meeting I had this evening (after this one) which was cancelled
[17:27]	Gypsy Paz:	no worries, we had a blast while we were waiting
[17:27]	Fnordian Link:	That's always a good excuse. Hard to argue with too.
[17:27]	Andrew Linden:	and somehow I got it stuck in my mind that I was totally free this afternoon
[17:27]	Fnordian Link:	Yes, we came up with several new theories.
[17:27]	Bad Mojo Hat: Thanks for chatting, Arawn.
[17:27]	Fnordian Link:	And Simon. Good to meet you too. :)
[17:27]	Sindy Tsure:	heya andrew & simon!
[17:28]	Andrew Linden:	No 1.24 disater that I know of Arawn.
[17:28]	Andrew Linden:	disaster
[17:28]	Gypsy Paz:	;)
[17:28]	Simon Linden:	Hi everyone
[17:28]	Andrew Linden:	ok, I have a little bit of news...
[17:28]	Arawn Spitteler:	The range of slpp seems to've been reigned in.
[17:28]	Gypsy Paz:	oh?
[17:29]	Andrew Linden:	all the bug fixes I've been talking about that have been waiting for 1.24 to complete "have the merge lock"
[17:29]	Sindy Tsure w00ts
[17:29]	Andrew Linden:	which means they are going through a final test, after which they will get merged back into the 'release' codebase
[17:30]	Gypsy Paz:	nice
[17:30]	Andrew Linden:	which means they will get into the 1.25 branch when it freezes
[17:30]	Fnordian Link:	Cool. Any really yummy fixes?
[17:30]	Sindy Tsure:	know the plan for 1.25? is it back to 2-weeks-per-sim-release or will there be a pause to let mono settle?
[17:30]	Andrew Linden:	It also means that whenever the 1.25 code goes up on the test grids you'll be able to test there
[17:30]	Andrew Linden:	well... I don't have a list of all the bugs handy but...
[17:31]	Andrew Linden:	SVC-2511 is in there
[17:31]	Sindy Tsure:	oops..
[17:31]	Andrew Linden:	and also the camera oscillation bug
[17:31]	Sindy Tsure:	that one will be nice!
[17:31]	Arawn Spitteler:	That'll be nice
[17:31]	Gypsy Paz:	oh nicer!
[17:31]	Sindy Tsure:	is that a viewer bug??
[17:31]	Arawn Spitteler:	H4 bug
[17:32]	Sindy Tsure:	ah...
[17:32]	Gypsy Paz:	I saw something about that in the release notes for the RC, not sure if we are talking about the same thing?
[17:32]	Andrew Linden:	No the camera oscillation is a server bug -- the server currently supplies a "hint" as to where the camera should NOT go
[17:32]	Andrew Linden:	and that hint is oscillating
[17:32]	Andrew Linden:	There are some other camera motion bugs that are indeed client-side...
[17:33]	Andrew Linden:	any glitches during camera mode changes (select vs focus on click point, for example) would be a client-side bug
[17:33]	Andrew Linden:	also, incorrect response to mouse-camera movements would be client-side
[17:33]	Andrew Linden:	another little bit of news...
[17:34]	Andrew Linden:	talking about server-side camera hints for "anti avatar oclusion"...
[17:34]	Andrew Linden:	I've been lobbying internally that we move ALL of the camera controls to the client, for a while now
[17:35]	Andrew Linden:	and someone mentioned that it is probably going to go onto the schedule
[17:35]	Andrew Linden:	on the short list (next few months or so)
[17:35]	Sindy Tsure:	scripts will still have access tho, right?
[17:35]	Arawn Spitteler:	Would camera data still be available script side?
[17:35]	Andrew Linden:	I'm sure the scripts will be able to contro the camera.
[17:36]	Sindy Tsure:	some watch where the camera is looking, too
[17:36]	Fnordian Link:	Will any of these changes/fixes have any impact on the issue of camera position not updating when using a teleporter?
[17:36]	Andrew Linden:	Really what happens there is the server supplies a hint to the client as to where its camera should be, and the camera obeys
[17:36]	Andrew Linden:	really this just means the ray-tracing done to make sure the avatar stays in view will move to the client
[17:37]	Andrew Linden:	No Fnordian, I don't think that this stuff will impact that teleporter camera bug
[17:37]	Arawn Spitteler has watched the occassional dissappearning environment, around his TPs
[17:37]	Sindy Tsure:	like when you sit and end in limbo for a minute, arawn?
[17:37]	Fnordian Link:	Ah, that's too bad. But perhaps some day...
[17:37]	Gypsy Paz:	with your leg in the air ;)
[17:37]	Arawn Spitteler:	That's a different bug, going by way of 0,0,0
[17:37]	Sindy Tsure nods
[17:38]	Sindy Tsure has a camera hat.. just as long as you don't break that, andrew! :)
[17:38]	Fnordian Link:	I get that one too.
[17:38]	Sindy Tsure:	heya rex
[17:38]	Andrew Linden:	Simon, any news on the mono or simulator performance front?
[17:38]	Arawn Spitteler:	I'm sure we'lll also have Lag Orbitting, for some time.
[17:38]	Fnordian Link:	Hi Rex. :)
[17:38]	Rex Cronon:	hi sindy, fnordian
[17:38]	Rex Cronon:	hi everybody
[17:38]	Andrew Linden:	Simon, the instrumentation of the simulator for more performance metrics... that project is supposed to complete soon right?
[17:38]	Andrew Linden:	Is the ETA still on target?
[17:39]	Gypsy Paz:	speaking of teleporters, the bug that let llSetLinkParamters() on avatars go long distances was a really nice alternative to warpPos, is that going to be permantly nerfed?
[17:39]	Rex Cronon:	oops. am i late or what:)
[17:39]	Andrew Linden:	Yes Gypsy, that misfeature is actually a bug and will be fixed
[17:39]	Arawn Spitteler:	Yes, Tardy to be more precise, but so was Andrew.
[17:39]	Simon Linden:	yes, tomorrow is supposed to be our last day. We have some regions on the main grid running with it right now - two high load ones, and are getting some good info
[17:39]	Andrew Linden:	the fix is already in place somewhere... either deploying now, or it will be in 1.25. Can't remember exactly where it is.
[17:39]	Arawn Spitteler:	Fixed in which way?
[17:39]	Sindy Tsure:	any surprises, simon?
[17:40]	Simon Linden:	The one surprise was that the ray-tracing with the camera positioning seems to be costly, so if we can move that to the client, or slow it down a bit, it would help
[17:40]	Sindy Tsure:	oh!
[17:40]	Andrew Linden:	Arawn, llSetLinkParameters() could be used to move a seated avatar out beyond the linkable distance. That is fixed -- the avatar must be within a linkabile distance like any other prim.
[17:41]	Rex Cronon:	so, no more mega teleporters:(
[17:41]	Andrew Linden:	the llSitOffset() or whatever that feature is called, will still be usable to violate the linkability rules.
[17:41]	Arawn Spitteler:	I know it got pretty wonky, over a kilometer.
[17:42]	Sindy Tsure:	not once they're sitting tho, i think
[17:42]	Fnordian Link:	Why was it deemed a bug and not a feature?
[17:42]	Gypsy Paz:	well, thats not all that useful with the new altitides
[17:42]	Gypsy Paz:	although, I tested warpPos on the beta grid yeasterday and it works very well in mono
[17:42]	Arawn Spitteler:	I think WarpPos would use an extra prim.
[17:42]	Fnordian Link:	WarpPos works fine in a single prim.
[17:43]	Gypsy Paz:	and with it sped up, it can return from long disances fast enough
[17:43]	Andrew Linden:	lemme find the internal jira on that llSetLinkParameters() problem...
[17:43]	Gypsy Paz:	I have it handy
[17:43]	Gypsy Paz:	oh, the internal one, nevermind
[17:44]	Fnordian Link:	I read the public one but I don't think it mentioned a specific reason for it to be considered as a bug.
[17:44]	Andrew Linden:	Ah, I see... it was filed by one of our testers because the functionality had changed
[17:44]	Sindy Tsure:	there's always the suggestions to add llTeleport, too.. that'd sorta eliminate the need for warppos and moving sit targets
[17:44]	Gypsy Paz:	its a shame though, that was one of those really cool bugs ;)
[17:45]	Arawn Spitteler thought it a fine feature.
[17:45]	Fnordian Link:	I think llTeleport has been "in the works" for a couple of years now, hasn't it?
[17:45]	Rex Cronon:	using that "bug" was possible to tp from here up to 4k high:)
[17:45]	Andrew Linden:	yes, llTeleport() would be handy. My memory as to why that was dropped is fuzzy -- I wasn't the one working on it...
[17:45]	Sindy Tsure:	maybe grief potential?
[17:45]	Arawn Spitteler recalls Simon was specificaly working out its reasonable moderation
[17:45]	Sindy Tsure:	don't see that it's more than what you get with warppos and all that, tho
[17:46]	Fnordian Link:	You could grief just as easily with a warpPos script.
[17:46]	Fnordian Link:	Right.
[17:46]	Andrew Linden:	but I think it was concluded that there were too many problems -- a lot of work would have to be done to catch all of the parcel permissions problems and other corner cases
[17:46]	Gypsy Paz:	well, I believe warpPos started out as a bug too ;)
[17:46]	Rex Cronon:	so if somebody uses llsettext to grief, than that function will also be "fixed"?
[17:46]	Sindy Tsure:	warpPos is sortofa hack, tho.. if they wanted to add llTeleport, it'd need thinking about first.. like does it need a permissions bit, etc
[17:46]	Fnordian Link:	For llTeleport or for the linkparams bug?
[17:47]	Sindy Tsure:	hi sundhi
[17:47]	Fnordian Link:	True....probably it would have to be like animations...request permission.
[17:47]	sundhi Joubert:	hey
[17:47]	Fnordian Link:	And sitting would grant permission automatically.
[17:47]	Andrew Linden:	No Rex, it isn't grief modes in general, but grief mode that violate existing features... such as parcel access lists
[17:47]	Andrew Linden:	I think...
[17:47]	Arawn Spitteler:	llSitTarget is a standard for violating ban lines.
[17:47]	Fnordian Link:	Oh I see. So it's a bug because it could be used in its current form to get around existing prohibitions on entering a parcel.
[17:48]	Andrew Linden:	Fnordian, no I'm talking about why the llTeleport() feature idea was put on hold
[17:48]	Fnordian Link:	Whereas warpPos is blocked by ban lines and object entry restrictions.
[17:48]	Fnordian Link:	Oh okay.
[17:48]	Fnordian Link:	Hmm...I'm sure llTeleport could be set up with similar permissions to those used with landmarks.
[17:49]	Arawn Spitteler:	llTeleportAgent could work with LandMarks
[17:49]	Andrew Linden:	yes, it is software so we can make it do whatever we want, however I guess it was determined to be a bigger project than we could take on at the time
[17:49]	Gypsy Paz:	I rember reading about llTeleport once upon a time, and I think they were talking about the ability to teleport objects into other sims was a big griefer risk
[17:49]	Fnordian Link nods. So it got bumped in priority.
[17:49]	Andrew Linden:	now it needs to trickle back up the ranks of "projects to do"
[17:50]	Andrew Linden:	Simon, do you think your sim performance instrumentation is going to make the 1.25 code freeze?
[17:51]	Fnordian Link:	BTW, I recompiled my Multiporter IV scripts a couple of hours ago...and was surprised to find that their idle time was higher under mono than the LSL versions.
[17:51]	Simon Linden:	I'd guess so - it's based on the current Mono code, and we haven't heard complaints from the busy sims running on it
[17:51]	Sindy Tsure:	just curiousity.. how big is the scope of the performance stuff, simon?
[17:52]	Simon Linden:	Most of it is actually instrumentation - we added a system to get more data, and find out more detail where the sim is spending it's time
[17:52]	Gypsy Paz:	will that be availble in the public viewers?
[17:52]	Simon Linden:	There are some changes in our main region loop,however, that should give up more CPU time on an empty/idle sim to the others on the same computer
[17:53]	Simon Linden:	No, it's internal (and our tools to get the data are pretty ugly :)
[17:53]	Sindy Tsure:	oh.. that's going in?? you talked about that a while ago.. seemed like a big win on openspace sims in particular
[17:53]	Sindy Tsure:	internal tools always are :)
[17:53]	Gypsy Paz:	;)
[17:53]	Simon Linden:	We're hoping to be able to see an improvement in our test regions
[17:53]	Fnordian Link:	However, even though they're using more script time when idle (according to the estate Debug listing anyway) than their LSL counterparts, they were about twice as fast leading up to the actual teleport (they were in a class IV open space sim with most prims used and a bunch of scripts running, so they were slow).
[17:53]	Sindy Tsure:	and you're lucky if the docs are more than "read the code!"
[17:54]	Simon Linden:	I do have one other bit of news ... I worked on code to allow setting a simple shape for the physics engine via llSetPrimitiveParams(), so it might use a box or sphere instead of a complicated mesh
[17:55]	Rex Cronon:	for the whole linked set?
[17:55]	Simon Linden:	I found that I could optimize the Kart 1.0 vehicle in the library - in the current state, it's 24 prims, and over 3000 verticies
[17:55]	Simon Linden:	By stripping out the small parts, and using boxes for wheels (!) it's 16 prims, and about 800 verticies
[17:56]	Sindy Tsure:	wow!
[17:56]	Gypsy Paz:	ohh that sounds cool, espically if you added drag handles to the bounding box ;)
[17:56]	Sindy Tsure:	hi moon
[17:56]	Gypsy Paz:	hey moon
[17:56]	Fnordian Link:	Hi Sweety. :)
[17:56]	Andrew Linden:	3k vertices server-side? That seems kinda high
[17:56]	Moon Fairymeadow:	smiles softly helo hugz all
[17:56]	Simon Linden:	I need to get some numbers to see how much that really improves things - it doesn't use that detail unless you're colliding with it
[17:56]	Simon Linden:	That's the number Havok told me in their debugger
[17:56]	Andrew Linden:	ah
[17:57]	Simon Linden:	Yeah, I suspect the wheels are worst - they are torii
[17:57]	Arawn Spitteler:	I was surprised when my Waterslide worked, at Lummerland. That thing uses a hollow trough
[17:57]	Andrew Linden:	When troubleshooting vehicles waaay back in the middle of the havok4 project I rebuilt the kart with spheres for wheels and couldn't measure an impact
[17:58]	Andrew Linden:	that is, no change to the performance that I could measure
[17:58]	Simon Linden:	I'm not sure the detail comes into use unless there are collisions - like someone is walking on top of it
[17:59]	Simon Linden:	It wasn't with vehicles, but I could definitely measure a difference when using a simple shape like a sphere on a torus, which is to be expected
[18:00]	Simon Linden:	There are some objects, like a 'wave generator', that use cut/twisted objects and thus a mesh, which can be simplified. We know they slow down regions now, and don't need the detailed shape
[18:00]	Andrew Linden:	So Simon, the simulator performance stuff... what you've done is instrument the simulator so that accurate measurements can be made of where the CPU cycles are being spent, right?
[18:00]	Simon Linden:	Right
[18:00]	Andrew Linden:	The metrics can be enabled/disabled real-time?
[18:01]	Simon Linden:	Yes, they are off by default, and we can drop in on a region that's slow and turn them on
[18:01]	Andrew Linden:	How are the metrics browsed?
[18:01]	Simon Linden:	We also made an interface so we can turn on code profiling easier
[18:01]	Andrew Linden:	Are they just in the logs?
[18:01]	Andrew Linden:	Or can you view them from the client UI?
[18:01]	Simon Linden:	They make their own log file, and we have some scripts to extract info and make graphs
[18:02]	Simon Linden:	No viewer access now (nor any planned)
[18:02]	Andrew Linden:	Ok, so the Residents won't be able to view them.
[18:02]	Sindy Tsure:	:(
[18:02]	Arawn Spitteler:	No Viewrr Access as in Hackable or no data downoad?
[18:02]	Simon Linden:	We also cleaned up the times shown in the ctrl-shift-1 panel ... they weren't fully accurate
[18:02]	Andrew Linden:	Sorta makes sense... just get it working first. Maybe someday we'll add UI and data streaming
[18:02]	Simon Linden:	No data sent to the viewers
[18:02]	Andrew Linden:	(in our copius free time)
[18:02]	Sindy Tsure:	lol
[18:03]	Sindy Tsure:	that's something that appeals to a lot of residents - better ways to track down problems without having to call LL for help
[18:03]	Sindy Tsure:	http://jira.secondlife.com/browse/SVC-835 for example.. 130 votes
[18:03]	Sindy Tsure:	"Allow "Top Scripts on Parcel" for mainland land owners"
[18:03]	Simon Linden:	Yes, we wanted a system where we might hear of a region having problems, then go turn on a few switches and get more information about what's the slowdown
[18:03]	Andrew Linden:	I guess at the moment it will be a special tool that the team will then use to analyze problem regions and then try to find fixes
[18:04]	Sindy Tsure nods
[18:04]	Simon Linden:	Yes - one thing that's obvious now is the avatar-owned scripts (attachments) aren't really accounted for or throttled very well
[18:05]	Simon Linden:	So someone can have a really heavy attachment load and slow everyone else down
[18:05]	Sindy Tsure isn't an estate manager but has heard they aren't really reported unless the person is sitting
[18:05]	Simon Linden:	Correct - the accounting of that time hasn't been very good
[18:05]	Andrew Linden:	supposedly Babbage wants to work on script-resource scheduling. Do you know of any tentative timeline on that Simon?
[18:06]	Andrew Linden:	I know Babbage is actually the guy to ask.
[18:06]	Simon Linden:	No - he was talking today about another problem, spreading long jobs out over several frames rather than trying to do them all at once
[18:06]	Andrew Linden:	I don't think attached scripts are reported even if the avatar is sitting.
[18:06]	Andrew Linden:	Attached script data lives in a special data structure, away from most other scripts
[18:07]	Gypsy Paz:	not directly, but the objec they are sitting on inherits the avatars script time in the estate toools
[18:07]	Gypsy Paz:	if not accurate
[18:07]	Sindy Tsure:	i think it may be, andrew.. i had a vteam tell me a couch at my place was eating 90ms..
[18:07]	Andrew Linden:	hrm... I suppose that might be possible Gypsy
[18:07]	Sindy Tsure:	couch with a bunch of people sitting on it.. turned out one had some attachment that was just horrible
[18:08]	Andrew Linden:	huh... I wonder how that could happen.
[18:08]	Sindy Tsure:	poor scripting?
[18:08]	Sindy Tsure:	:P
[18:08]	Gypsy Paz:	yeah, I've seen a simple poseball report 400ms
[18:08]	Simon Linden:	Hmmm - wonder if we have a bug and are running the sitting AV's scripts twice, once as a AV, once as part of the object
[18:08]	Simon Linden:	That would be bad :)
[18:08]	Andrew Linden:	That's what I was just thinking Simon
[18:09]	Glitch Braess:	lol
[18:09]	Arawn Spitteler wonders if the Pose Balls could screen avatars for such data
[18:09]	Simon Linden:	We need to report the AVs script time properly, and perhaps make it somehow available, like the render time or such so people can understand the cost of all the HUDs and AOs
[18:10]	Gypsy Paz:	that would be very nice to have in the estate tools, or even the land tools
[18:10]	Sindy Tsure:	that can be tricky.. i've seen people going the ARC witch hunts..
[18:10]	Gypsy Paz:	hahaa
[18:10]	Sindy Tsure:	"zomg! your arc is over 1000! ban!!"
[18:11]	Andrew Linden:	"ARC" what is that acronym/
[18:11]	Arawn Spitteler:	Avatar Render Cost.
[18:11]	Sindy Tsure:	avatar rendering cost
[18:11]	Andrew Linden:	oh right
[18:11]	Arawn Spitteler:	A party with no ARC over 1000, wouldn't experince much lag, due to lack of attendance.
[18:12]	Simon Linden:	yes - it's a tricky balance. We don't want to punish the more advanced users who are using more features, but at the same time, don't want to let a few people slow down a region if the owner doesn't want it
[18:12]	Sindy Tsure:	just saying it can be tricky when giving numbers to the masses.. :)
[18:12]	Sindy Tsure:	right! :)
[18:12]	Simon Linden:	Frankly, the people who have more AV control, animiations, expressions, etc are more interesting
[18:12]	Andrew Linden:	Ideally we could allow the client to modulate avatar detail/costs more effectively
[18:13]	Sindy Tsure has heard similar things suggested
[18:13]	Sindy Tsure:	it wasn't very fine control, tho. wsa more like a suggestion to visually mute anybody over some ARC value
[18:14]	Sindy Tsure:	to help folks with older machines
[18:14]	Arawn Spitteler:	And Australians
[18:14]	Sindy Tsure:	of course
[18:14]	Gypsy Paz:	well, doesnt imposters put a good LOD over that?
[18:14]	Andrew Linden:	I wonder if it would be possible, eventually, to provide feedback to the offender somehow... letting them know that everyone has decided to render them as an uninteresting sphere or something
[18:14]	Andrew Linden:	Yes, I think imposters help
[18:14]	Sindy Tsure:	i don't think they count ARC, though
[18:14]	Andrew Linden:	but what we find is that people get comfortable with low FPS
[18:15]	Arawn Spitteler:	Last I had imposters on, Avatars wuld dissappear.
[18:15]	Sindy Tsure:	think imposters is just distance-based
[18:15]	Andrew Linden:	and improvements to render FPS in an area just encourages many to dial their detail up until they are at the same low FPS
[18:15]	Rex Cronon:	the gfx has decided to reder u as a blob. take some thing OFF. would u?
[18:15]	Gypsy Paz:	hahaa
[18:15]	Sindy Tsure:	we're all carp.. just gonna keep eating if you feed us more!
[18:15]	Fnordian Link:	Perhaps we could even take it further and choose how we would like to view any given avatar...choose from a selection of other avatars (or strange shapes) to render them as (on our viewer only).
[18:16]	Fnordian Link:	So if I want to see someone else as a giant 6-headed monster avatar I can.
[18:16]	Arawn Spitteler:	that'd be a user optomized viewer.
[18:16]	Andrew Linden:	seriously... the graphics programmers make optimizations and then check the metrics to see if they made a difference
[18:16]	Gypsy Paz:	lol, that would also be a good way to deal with naken n00bs
[18:16]	Fnordian Link:	Or perhaps I want to view everyone as tiny vegetables. :)
[18:16]	Arawn Spitteler:	I suppose it would be more consistent with First Laught
[18:17]	Andrew Linden:	often there is a little blip upwards on release and then it creeps down again
[18:17]	Sindy Tsure:	that's ok as long as the steps back <= steps forward
[18:17]	Fnordian Link:	The steps sideways are the most interesting.
[18:17]	Sindy Tsure:	:)
[18:17]	Andrew Linden:	Spoken like a true Discordian
[18:18]	Fnordian Link chuckles.
[18:18]	Arawn Spitteler:	I find I actually had Impopsters turned on, yet people were visible.
[18:18]	Andrew Linden suspects Arawn is also a Discordian
[18:18]	Sindy Tsure:	they still are visible, just rendered at much lower quality
[18:18]	Arawn Spitteler:	this new viewer didn't adopt my traditional settings, so I spent some time in Blind Staggers, before I realised Camera Transition Time had been turned on
[18:18]	Fnordian Link:	(They're everywhere.)
[18:19]	Andrew Linden:	yes, imposters will update much slower, and may appear more pixelated
[18:19]	Arawn Spitteler:	Early on, I was talking to a lot of invisible people.
[18:19]	Sindy Tsure:	were they really there, though?
[18:19]	Sindy Tsure:	if so, it's ok
[18:19]	Sindy Tsure smiles
[18:20]	Arawn Spitteler:	They showed up on Minimap, buit no Mini Me
[18:20]	Andrew Linden:	Alright, well sorry I was late, but I've got to go now.
[18:20]	Rex Cronon:	bye andrew
[18:20]	Sindy Tsure:	ok.. cya, andrew!
[18:20]	Fnordian Link:	Thanks for coming and answering our questions. :)
[18:20]	Gypsy Paz:	kk, thanks andrew
[18:20]	Rex Cronon:	it can happens to anybody:)
[18:20]	Arawn Spitteler:	Hey, your next meeting's been canceled, come to Matara for Coffee.
[18:20]	Moon Fairymeadow:	may i have a teddy bear please
[18:20]	Andrew Linden takes a moment to save the transcript...
[18:20]	Sindy Tsure hears lindens are coming back to the forums.. looking forward to the Ask Andrew forum..
[18:21]	Sindy Tsure ducks
[18:21]	Andrew Linden:	I haven't read the forums in a year or so
[18:21]	Sindy Tsure:	and you're a happier person for it? :)
[18:21]	Sindy Tsure:	it can be a bit of a zoo..
[18:21]	Sindy Tsure runs, too
[18:21]	Sindy Tsure:	ty to simon, too!
[18:21]	Sindy Tsure:	cya, all!
[18:22]	Rex Cronon:	bye everybody
[18:22]	Rex Cronon:	bye sindy
[18:22]	Arawn Spitteler:	So, on the SLPP issue, should it be downgraded, as nice to have, or has it the votes to upgrade?
[18:22]	Andrew Linden:	SLPP?
[18:22]	Fnordian Link:	Simon, I have a technical question that I've been wondering about..not sure who to ask really.
[18:22]	Arawn Spitteler:	The large teleport distances, that were quickly desired as content
[18:22]	Andrew Linden:	I think I asked what SLPP was last office hour
[18:22]	Bad Mojo Hat: How are you, Fnordian.
[18:23]	Simon Linden:	sure, go ahead
[18:23]	Arawn Spitteler:	llSetLinkPrimitiveParam
[18:23]	Moon Fairymeadow:	smiles softly pokes andrew may i have a teddy bear please
[18:23]	Bad Mojo Hat: Bye bye.
[18:24]	Bad Mojo Hat: I am very pleased to meet you Simon.
[18:24]	Fnordian Link:	Thanks. My question is about object inventories and their loading/updating times. As you add more to an object its inventory updates/loads much slower. Is this updating/loading time for large object inventories something that is likely to ever be improved?
[18:24]	Moon Fairymeadow:	smiles softly thank you both
[18:24]	Fnordian Link:	For vendors with a hundred or so objects it makes updating them a teeth-gnashing experience.
[18:25]	Simon Linden:	I hope so, but I don't think there's going to be a quick solution