User:Andrew Linden/Office Hours/2008 01 29

From Second Life Wiki
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Transcript of Andrew Linden's office hours:

[11:03] Teravus Ousley: :D
[11:03] Chalice Yao grins. "SL makes you develop split personalities, eh?"
[11:03] Teravus Ousley: Greetings Sir Andrew
[11:03] Chalice Yao: Heya Andrew :)
[11:03] April Heaney: Hello, Andrew :)
[11:03] Andrew Linden: hello
[11:03] Anders Falworth: Hi, Andrew.
[11:03] Rex Cronon: hello andrew
[11:04] Rex Cronon: any new developments on the h4 front:)
[11:04] Andrew Linden: Yeah, a little bit. Lemme summarize the latest news...
[11:04] Rex Cronon: cool
[11:05] Andrew Linden: No vehicle bugs fixed... I've been kinda busy moving my desk and settling into a new building
[11:05] Andrew Linden: so I've only been working on one thing... trying to optimize a few things
[11:05] Andrew Linden: people have been complaining about performance issues
[11:05] Andrew Linden: vehicles in particular, but also crowds of people
[11:06] Andrew Linden: So I've found one small bottleneck
[11:06] Andrew Linden: and one collision shape optimization
[11:07] Andrew Linden: that recovered 50% performance (running twice as fast) for one particular vehicle lag problem
[11:07] Andrew Linden: dunno how the changes will translate for general content
[11:07] Chalice Yao: Heya Simon, Barney :>
[11:08] Andrew Linden: A 2x speed improvement isn't as good as I was hoping for, since the complaints were that some vehicles were 4x too slow
[11:08] Chalice Yao: Hi Kelly.
[11:08] Simon Linden: hi everyone
[11:08] Andrew Linden: but it is progress
[11:08] Andrew Linden: Welcome Simon.
[11:08] Kelly Linden: Hi
[11:08] Teravus Ousley: Hello
[11:08] Andrew Linden: Lessee... what else...
[11:08] Rex Cronon: hello simon, kelly
[11:08] Andrew Linden: there were a few bugs fixed that we want to get out soonish
[11:08] Saijanai Dagger: hey kelly
[11:08] Andrew Linden: since there are a bunch of regions running Havok4 on the main grid now
[11:08] Saijanai Dagger: ulti-tasking meeting at Groupies
[11:09] Andrew Linden: Kelly and Simon, do you recall what the most important bug fixes were that are pending the next update?
[11:09] Andrew Linden: I'm just giving a quick summary of what has changed in development since the last office hour.
[11:09] tx Oh: huhu
[11:10] Kelly Linden: Some estates had objects lots of objects returned to owners when converted to Havok4 - Simon fixed those. TargetOmega on children fixed, I just fixed a bug in llGetStartParameter and ...
[11:10] Simon Linden: I'm not sure what went into the last update ... but the objects getting improperly returned has been fixed
[11:10] Kelly Linden: And a llSetPrimitiveParams bug I fixed yesterday.
[11:10] Creem Pye: primitive params for position of avatars, kelly?
[11:10] Andrew Linden: Ah yes, there were a few regions that tried Havok4 and had to go back to Havok1 for the returned objects problem.
[11:11] Simon Linden: FWIW I'm working on flying (with attachments) now - there are still some bugs in there
[11:11] Kelly Linden: Creem: No, this one had to do with setting textures / colors on different sides.
[11:11] Creem Pye: ah ok
[11:11] Kelly Linden: SVC-1146
[11:11] Andrew Linden: Simon, are you talking about the buoyancy problem? or something else?
[11:12] Andrew Linden: Simon, I dunno if you noticed, but I moved the attachment force application to the beginning of the agent action ::apply() method, instead of the end
[11:12] Andrew Linden: I think that makes for more correct force handling from the attachment's perspective
[11:13] Simon Linden: Buoyancy was the starting point, and that's fixed. Now it's a matter of slowing/stopping when appropriate - the AV keeps coasting forever at high altitudes. You and I should discuss after this - it has to do with decreasing efficency when up high, which prevents the slowing impulse from stopping the AV
[11:13] Andrew Linden: ah yes, ok, after this hour
[11:14] Andrew Linden: Ok, I guess that is most of the update.
[11:14] Andrew Linden: We wanted to try to start testing a new version today, so we would be ready for the Wed update for the H1 regions
[11:14] Andrew Linden: dunno if we're going to make that
[11:15] Andrew Linden: So, I guess the topic of conversation is now open, should anyone have something they want to discuss.
[11:15] Chalice Yao nods and has a few questions, actually.
[11:15] Chalice Yao: If I may? :>
[11:16] Andrew Linden: Yes, please ask.
[11:17] Chalice Yao: Simon, this one might be for you in particular. The avatar speed limit..currently still at 64m/s to my knowledge..has been met with overwhelmingly negative reactions. Not all, but the very most of people I talked to about it disliked it. Especially now that the object limit seems to have finally raised to 4k in general, people need a fast way to get up there, with flight enhancers and whatsoever.
[11:17] Chalice Yao: Are there plans to finally raise it again?
[11:17] Andrew Linden: I already expanded the speed limit, it will be in the next update.
[11:17] Andrew Linden: The new hard limit is 256 for all modes (fly and walk)
[11:18] Andrew Linden: and there is a soft transition when the avatar is flying... an attempt to have more like Havok1 response
[11:18] Chalice Yao nods. "That sounds good indeed. Thank you lots. It's appreciated. :)"
[11:18] Rex Cronon: is it just me, or right now is impossible to tunnel through objects?
[11:18] Chalice Yao: It's very possible, just like in Havok 1.
[11:18] Andrew Linden: basically I tweaked it so that it was hard at 256, but would provide 64 m/sec for this particular rocket pack I had on hand to test
[11:19] Andrew Linden: I haven't tested anything else except for some versions of the OmniPhaze, which seemed to work fine.
[11:19] Andrew Linden: It should still be possible to tunnel through objects with the appropriate attachment.
[11:19] Barney Boomslang: simon/andrew - no idea wether you know about the case, but when H4 was deployed to Caledon SteamSkyCity last time, there was some weird disconnect of the sim from the grid - any idea wether the next code drop will allow SteamSkyCity go back to H4 (it was reverted to H1 by Sidewinder I think)?
[11:20] Andrew Linden: Sidewinder was trying to learn more about that yesterday, dunno what he found.
[11:20] Andrew Linden: That problem appears to be a network connectivity problem between the host servers dedicated to H4 and some H1 servers
[11:21] Andrew Linden: That is, it sounds like there is a firewall or routing problem.
[11:21] Andrew Linden: Joel will have to find someone in the Ops team to help him figure out what is wrong, but I assumed he was having those discussions
[11:21] Andrew Linden: since he has someone in ops helping him acquire hosts to deicate to the H4 regions.
[11:22] Arawn Spitteler had a thought, that HIP2 could go to Havoc 4, for sandbox comparison
[11:22] Andrew Linden: Er sorry, I meant Sidewinder (aka Joel)
[11:22] Andrew Linden: HIP2? Could you please expand that acronym for me?
[11:23] Arawn Spitteler: Help Island Public 2, is next to Help Island Public, and they might do for side by side sandboxes
[11:23] Andrew Linden: Ah.
[11:23] Arawn Spitteler: HIP has too much traffic, and HIP2 has too little
[11:23] April Heaney: Not a question per se, but just some notes I've been keeping since I'm one of the early adopters... The friction between ground and avatars seems high and rather "sticky", though prim and avatar friction seems low. Ground vehicle and ground friction seems very low. Air vehicles have some very sudden and violent responses to forces. And last but not least, the download queue. It frequently stacks up with pending downloads for several minutes before clearing, which I've heard other H4 sims report also.
[11:23] Andrew Linden: Yeah, I'm letting Sidewinder and others coordinate where to deploy Havok4.
[11:24] Andrew Linden: I'm just trying to focus on fixing the bugs.
[11:24] Creem Pye: will there be another vehicle meeting this Thursday?
[11:25] Andrew Linden: Yeah, if I could fix some of the current vehicle bugs for which I have demos then I would like to have another vehicle session
[11:25] Andrew Linden: however, I think most of the vehicles I've got on hand will still be broken on any H4 region that is deployed on Thursday.
[11:25] Andrew Linden: April, most of those bugs you mention are known.
[11:25] April Heaney nods.
[11:26] Andrew Linden: Some of them might not every get fixed... in particular the avatar's stickyness to the ground.
[11:26] April Heaney: That's why I didn't JIRA them, I looked and saw similar ones there.
[11:26] Andrew Linden: There is a tradeoff between the avatar staying on the ground when standing on a hill
[11:27] Andrew Linden: and when it is being pushed from external objects
[11:27] April Heaney: The stickiness itself isn't an issue, more the consistency between motion on a prim versus ground.
[11:27] Andrew Linden: also... the avatar tends to stop walking a little faster than it used to
[11:28] Andrew Linden: Yeah, I'd like to fix that. I think there is a bug in Havok4 that is causing that, but I'll have to look into it again to be sure.
[11:28] Andrew Linden: that is... a bug deep in the Havok4 physics engine
[11:28] Teravus Ousley: Havok4 doesn't give you a 'mu' friction option does it? :D
[11:28] Andrew Linden: mu friction? I'm unfamiliar with that term.
[11:29] Teravus Ousley: well, in ODE, for example.. it's a float value that drives the amount of friction.. and also the amount of time it takes the CPU to calculate it
[11:29] April Heaney: Differing friction coefficients for different material types would be lovely, but I think we can wait. ;)
[11:29] Gaius Goodliffe: llSetFriction(float) :)
[11:30] Andrew Linden: Havok only has one coefficient of friction (and I think it is what is typically known as "static friction")
[11:30] Andrew Linden: actually, different materials in SL do have different coefficients of friction
[11:30] April Heaney: I was under the impression that was being discarded.
[11:30] Andrew Linden: I'll leave it as an excersize to the attendees to actually measure those values ;-)
[11:30] Creem Pye: it would be nice to get some quantitative data on those coefficients =)
[11:30] Arawn Spitteler recalls that Surface Friction, as a variable, is a popular wish
[11:30] April Heaney: Agreed, Creem.
[11:30] Gaius Goodliffe: Sounds like a good project for you, Creem. ;)
[11:31] Chalice Yao nods. "A settable on on an object or prim would be wonderful, I think :>"
[11:31] Creem Pye: hehe I guess I can try...
[11:31] Chalice Yao: *one
[11:31] Andrew Linden: No, I tried to tweak the material friction values to get approximately similar sliding behavior as Havok1
[11:31] Andrew Linden: per-prim friction settings in Havok4 is not possible (or at least, not easy)
[11:32] Andrew Linden: however per-object settings can be done... I think you need to set the material of the root prim for it to take effect on the entire object
[11:32] Chalice Yao: yes, I think object would be way enough :>
[11:32] Creem Pye: speaking of friction, it also seems super-high on my vehicles (enough to prevent them from moving)
[11:32] Andrew Linden: Well, at the moment you have the varios materials to chose from: rubber, stone, metal, glass
[11:33] Creem Pye: I could give you a sample, but maybe you already have enough of a backlog with vehicles, Andrew
[11:33] Andrew Linden: Creem, that sounds like a bug that I thought was fixed.
[11:33] Andrew Linden: I don't currently have any vehicles that suffer from it, so yes I'd like a copy.
[11:33] Creem Pye: ok, let me just write up a notecard describing the general issues for you..
[11:34] April Heaney: I only noted land vehicles turning like they were on greased tires. I have a rather complex aircraft that's mostly attachments that suffers a violent death if you look at it funny in H4 :)
[11:34] Gaius Goodliffe: (Join the Lindens, get free vehicles. :D )
[11:34] Andrew Linden: The vehicle friction problem was a big one when moving to Havok4.
[11:34] Andrew Linden: I had to add some magic stuff to the vehicle code to compensate for it
[11:34] Teravus Ousley: hehe, 'shiny' stuff :D
[11:35] Andrew Linden: however that magic tended to break hover vehicles, so if you have non-zero hover on your vehicle the "anti-friction magic' is disabled
[11:35] Andrew Linden: that is, using non-zero vehicle hover, not the llSetHover() stuff
[11:35] Andrew Linden: that is totally independent behavior to the vehicle
[11:36] Andrew Linden: This office hour appears to be better attended than most. I'm curious what draws the new attendees?
[11:37] Andrew Linden: Is it the bugs in the many regions that are now experimenting with Havok4?
[11:37] Barney Boomslang: in my case it was the direct experience with the H4 code drop, yes
[11:37] Creem Pye: for me it's mostly just my vehicle issue =)
[11:37] Teravus Ousley: one particular item on this meeting was the AW Groupies announcement by Rex
[11:37] Gaius Goodliffe: They sense the impending doom if they don't get their bugs addressed soon. :)
[11:37] April Heaney: I own one of the H4 regions, and had corresponded with Sidewinder, who suggested I come here. :)
[11:37] Chalice Yao: I also assume the awareness of Havok 4 is rising in general.
[11:37] Chalice Yao: So :> more people in the know and curious.
[11:37] Teravus Ousley: .. probably increased attendence somewhat
[11:37] Shirley Marquez: I'm just curious about all the code changes being worked on -- the underpinnings of SL are about to change radically
[11:37] Andrew Linden: Yes, that is what I thought. Good, I guess. This is the right place to bring up important bugs.
[11:38] Chalice Yao: With mono on the 30th, we're in for a fun ride :>
[11:38] Barney Boomslang: The focus on H4 code drops of course is H4 - but it seems to include other changes, too. For example we got some object-returns of bigprims encroaching other parcels due to parcel autoreturn - but those where out forever in SteamSkyCity. Maybe a short comment on other changes in the H4 code branch?
[11:38] April Heaney: 2008 will be a fun year.
[11:38] Barney Boomslang nods and looks _so_ forward to Mono, even if it breaks lots of his scripts at first ;)=
[11:38] Rex Cronon: i thought that people might be interested
[11:39] Saijanai Dagger: Shirley, AW Groupies is discussing how to reduce the pain for libsl and opensim when the breakages of protocols starts
[11:39] Andrew Linden: Simon might have already fixed that megaprim autoreturn thing. There was a bug with objects that were out of bounds.
[11:39] Chalice Yao: I have a question /bug that is not directly related to Havok 4, but fits into the whole sim behavior change process that I'd like to mention..I will JIRA it as well, depending on consensus here.
[11:39] Barney Boomslang: andrew: ah, that would be good. because I hate to get the landing gear of staemskycity in my lost+found ;)
[11:40] Andrew Linden: Most of the Havok4 changes were to the physics, however there was also some cleanup around how our code interfaces with the physics, and that is where apparently non-physics bugs would show up.
[11:40] Chalice Yao: Okay:
[11:40] Chalice Yao: Currently, when you have, say, 1 prim left on a parcel..you still have the normal 500 temp prim limit. However, in this example you cannot rez an object made of more than 1 temp prim. Basically, even if the object is temp in inventory, the rez process checks for temp -after- rezzing. Thus, products that rez temp prims are affected by the normal prim limit on rez, but afterwards I can copy them like mad, of course.
[11:40] Andrew Linden: We overhauled how objects are actually moved, at least in our internal API for changing object positions
[11:40] Simon Linden: I fixed two problems - there was a bug we introduced a while back that caused some prims to not link correctly, and they ended up off-world and were thus returned. The 2nd was the distance we allowed stuff to go underground was shortened, and so more stuff was returned. Both are fixed for the next push
[11:40] Andrew Linden: so we're still tracking down the various bugs that were introduced there.
[11:41] Barney Boomslang: sounds like I should try some of my non-phys vehicles, too, with H4, then
[11:41] Andrew Linden: Oh right, that linking bug was an ownership mismatch problem.
[11:42] Andrew Linden: The megaprim objects also suffer from a linking bug that is not fixed yet, and might not be, since we don't officially support megaprims yet
[11:42] Barney Boomslang: well, the problem with the steamskycity returns seem to have been megaprims that are on group land, but cross small parts of non-group parcels. at least all those prims that where returned had that problem.
[11:42] Andrew Linden: Some megaprim objects could link in Havok1 but now violate the simplified link rules.
[11:42] Kelly Linden: I think the cause of that particular one is relatively rare and related to mass ownership change, for example an island owner who buys an island with content on it. I've been talking with Don / Blacklight some about what could cause it or how to more generally fix the underlying bug.
[11:42] Gaius Goodliffe nods.
[11:42] Barney Boomslang: (and yes, I had spontuous delinking of megaprim based stuff, but that could be fixed on my side)
[11:42] Gaius Goodliffe: I have a blimp that won't link anymore.
[11:42] Kelly Linden: "that particular bug" = the ownership issue that caused things to unlink and be returned.
[11:43] Andrew Linden: Really Gaius? No megaprims in that blimp?
[11:43] Gaius Goodliffe: Actually, a few. That's why... it's hit by that bug.
[11:43] Andrew Linden: Ah.
[11:43] Chalice Yao: The new calculations don't even allow megaprims to link that are at the very same spot, interestingly.
[11:44] Andrew Linden: yes, the linking rules pivot on the absolute dimensions of the object
[11:44] Andrew Linden: rather than just distances between centers
[11:44] Rex Cronon: shouldn't 2 prims at same pos be able to be linked?
[11:45] Rex Cronon: maybe a new set of linking rules are needed?
[11:45] Saijanai Dagger: if megaprims past a certain size cause problems with links, I think the answer has to be "no."
[11:45] Gaius Goodliffe wonders if you could take MIN(10.0, dimension) when doing those calcs.
[11:45] Andrew Linden: perhaps. It makes sense that they should however it isn't clear to me yet how to special case those in.
[11:46] Andrew Linden: Did you know that the new linking rules have been published in the public wiki somewhere?
[11:46] Andrew Linden: lemme find the link
[11:46] tx Oh: yes
[11:46] Saijanai Dagger: /meisghs. More documentation issues.
[11:46] Rex Cronon: don't think the problem lies with the megaprims, more likely with the rules
[11:46] Andrew Linden: perhaps some of you could go through them and suggest some changes
[11:46] Chalice Yao: Mh. Megaprims are tolerated, not supported, Rex.
[11:47] Gaius Goodliffe: https://wiki.secondlife.com/wiki/Linkability_Rules
[11:47] Andrew Linden: Thanks Gaius.
[11:47] Gaius Goodliffe: *In general* it's actually a little bit more permissive than before.
[11:47] Andrew Linden: Kelly, you had an idea for changing the rules to accomodate megaprims. Perhaps you could state it here.
[11:48] Rex Cronon: i think i read that, but should the rules be modifyied, if the need to? or are the rules set in stone?
[11:48] Kelly Linden: basically what Gaius also said
[11:48] Kelly Linden: here is a random hack idea for megaprim linking ..... if the single prim has a radius larger than a 10x10x10 object we could just use the 10x10x10 radius for the linking equations ....
[11:48] Andrew Linden: The rules are not yet set in stone, but the cement is hardening.
[11:48] Gaius Goodliffe roots for the MIN(10.0, dimension) rule. :)
[11:49] Barney Boomslang: kelly: so you propose to handle a megaprim as a 10x10x10 just for the linking equations?
[11:49] Chalice Yao: Seems like the simplest solution, yes.
[11:49] Kelly Linden: Aye
[11:49] Kelly Linden: It is a hack though, and it could end up one we would have to support forever.
[11:50] Kelly Linden says that in his scariest and most ominous voice
[11:50] Chalice Yao: Not quite, if I remember what was said about prims one day being free-scale on your parcels ;)
[11:50] Barney Boomslang: sounds good to me, since it would all ow link in bigger megaprims than the 60m limit normally would allow (I have some flashes in wellsian that are 100m in one dimension and linked in a triple)
[11:50] Chalice Yao: it will need to change then, anyways.
[11:50] April Heaney: Are we not getting > 10m prims in the future, once the encroachment issues are worked out?
[11:50] Andrew Linden: Well... maybe not. We could never decrease the minimum dimension
[11:50] Andrew Linden: but we could increase it later (after megaprim liberation)
[11:50] Teravus Ousley: heh
[11:50] Rex Cronon: eventually, shouldn't the servers fully spport megaprims?
[11:50] Kelly Linden: Ah, right.
[11:51] Chalice Yao: Yes, they will, once parcel/prim overlaps are handled right. AFAIK.
[11:51] Andrew Linden: Sounds like a good idea Kelly. Why don't you try it out.
[11:51] Barney Boomslang dreads the day that parcel/prim overlaps are handled "right" because possibly SteamSkyCity will blow up in his face (uh, lost+found more specificially) ...
[11:52] Barney Boomslang: (obviously not LLs fault, but desmonds for cutting rather odd resident parcels in that sim)
[11:52] April Heaney: I told you about the dangers of steampunk, Barney.
[11:52] Kelly Linden: Anyone have an object with megaprims linked they could give me, that links in H1 but not H4?
[11:52] Arawn Spitteler has an idea, which would also be useful for the Merchant Groups Problem, if it's any good: Do I understand that the foundation of SteamSkyCity is getting returned, as a megaprim should be, that encroaches on another's space?
[11:52] Gaius Goodliffe: I suspect "handling right" would be allowing owners to return manually objects that overlap their land, but exempt that from auto-return.
[11:53] Andrew Linden: Barney, the plan is to initially not enforce the encroachment rules automatically, but allow parcel owners to invoke them manually
[11:53] Gaius Goodliffe: What he said. :)
[11:53] Barney Boomslang: andrew: sounds good
[11:53] Gaius Goodliffe: So if you neighbor doesn't mind, SL doesn't either. But if they do, they can do something about it.
[11:53] Chalice Yao: Sounds fair.
[11:53] Barney Boomslang: uh, kelly, which one of the ppl are you LOL? can't find you, but would have an object that shows the symptoms
[11:53] April Heaney nods.
[11:54] Creem Pye gave you P-40 rev.20.
[11:54] Chalice Yao: Kelly's the silver-haired, red-eyed Linden wonder over there :> With the beard.
[11:54] Chalice Yao winks.
[11:54] Gaius Goodliffe: Did any of you guys want a sample airship that breaks under the new linkability rules for testing, if you're going to try to fi that?
[11:54] Barney Boomslang: yeah, spotted him finally :)
[11:54] Rex Cronon: although, shouldn't new linked prims be checked, before they r rezzed?
[11:54] Gaius Goodliffe: *fi
[11:54] Gaius Goodliffe: *fix
[11:54] Simon Linden: Gaius - sure, send one to me
[11:54] Gaius Goodliffe: (ugh, sticky X key)
[11:55] Barney Boomslang: the coerced object has a propeller - that uses two megaprims linked to the center, links fine with H1 but fails with H4
[11:55] Kelly Linden: Barney, the one you gave me looks like multiple unlinked items taken to inventory (according the icon) - was it last taken from a H4 (and thus currently broken) or a H1 (and thus currently not broken)?
[11:56] Barney Boomslang: kelly: taken from H1 with a propeller and the thing it is resting on
[11:56] Simon Linden: Gaius - is there a Jira for that? If not, I'll make an internal one
[11:56] Gaius Goodliffe: No JIRA that I know of.
[11:56] Barney Boomslang: should be two linksets, the top linkset with the link problem under H4
[11:56] Gaius Goodliffe: I figured since it involved a megaprim it'd just get marked "won't fix" so I didn't bother to JIRA it.
[11:56] Kelly Linden: great, thanks barney.
[11:57] Rex Cronon: i think that a land owner should be able to set the parcel, so that it can allow only prims belonging to a certain person/group to rezz partially in the parcel
[11:57] Rex Cronon: just a thought
[11:58] Gaius Goodliffe: @Rex: Sounds like a good idea long term, I'm guessing we'll see the simpler implementation first, though.
[11:58] Andrew Linden: Rex, that sounds like a good idea. It would have to wait for the second or third iteration on the enchroachment problem.
[11:58] Arawn Spitteler: My thought, is that Merchants would like an alternative to group membership. If Rez-Rights could be accorded to Groups and Indivicuals, by roster, as with Access
[11:58] Chalice Yao: About rezzing, did you catch my question (or statement) about the temp prims and the normal prim limit? :>
[11:58] Andrew Linden: Chalice, please restate the question.
[11:58] Chalice Yao: Currently, when you have, say, 1 prim left on a parcel..you still have the normal 500 temp prim limit. However, in this example you cannot rez an object made of more than 1 temp prim. Basically, even if the object is temp in inventory, the rez process checks for temp -after- rezzing. Thus, products that rez temp prims are affected by the normal prim limit on rez, but afterwards I can copy them like mad, of course.
[11:59] Chalice Yao: unless the prim limit is 0, of course.
[11:59] Chalice Yao: you can still rez temp then, I -think-
[12:00] Andrew Linden: Yes Chalice, that sounds like a bug. I think someone else was telling me about that problem a week ago or so
[12:00] Chalice Yao: It actually also affects Havok 1 sims, hence my comment about it not quite being Havok 4 related :>
[12:00] Andrew Linden: It was Torley, he had pointed me at a foum thread.
[12:00] Chalice Yao: Ahh. Okay.
[12:01] Gaius Goodliffe: What? Lindens actually read the threads? :o
[12:01] Gaius Goodliffe: ;)
[12:01] Barney Boomslang: gaius: they must be masochistic ;)
[12:01] Gaius Goodliffe: Haha! Probably a job requirement...
[12:01] Teravus Ousley: 'thick skinned'
[12:01] Andrew Linden: Yes, since it is non Havok4 I'm having trouble prioritizing it (despite the fact that I did most of the recent coding on parcel temp prim limits)
[12:02] Chalice Yao noddles. "Should I JIRA it?"
[12:02] Andrew Linden: No, I think there already is a jira
[12:02] Chalice Yao: Ah. Okay.
[12:02] Andrew Linden: However, a really clear recipe for reproducing it would be nice, if there isn't one already.
[12:03] Andrew Linden: Your description of the problem seemed simpler than the one I read last.
[12:03] Chalice Yao nods. "Simply rez a temp object or prim cluster that has more prims than the current leftover parcel prims you have.
[12:03] Chalice Yao: It will fail.
[12:03] Chalice Yao: so will copying those once in-world, unless you copy an amount below the leftover prim limit.
[12:03] Andrew Linden: Ok, perhaps I'll look into it in my spare time, sounds simple enough.
[12:04] Chalice Yao smiles. Thank you :>
[12:04] Andrew Linden: Ok, I'm going to have to head out soon, and the hour is over.
[12:04] tx Oh: as well when you select a temp rez link set, the selected link set will count against the normal prims
[12:04] Andrew Linden: Thank you all for coming.
[12:04] Teravus Ousley: Thanks for having us
[12:04] Creem Pye: thanks for your time =)
[12:04] April Heaney: Thank you for the updates.
[12:05] Andrew Linden: Yes, selected is counted different from temp rez.
[12:05] Andrew Linden: Mostly so that the temp on rez objects you have selected are not reaped before you let go.
[12:05] tx Oh: yep.
[12:05] tx Oh: also thanks. bye
[12:05] Andrew Linden: hrm... there is a clue
[12:06] Andrew Linden: when you rez something, does it show up selected or not?
[12:06] tx Oh: oh?Q
[12:06] Simon Linden: Thanks everyone ... I have to go ... bye
[12:06] Kelly Linden waves good bye. Lunch calls and my tummy rumbles!
[12:06] Chalice Yao: if you are in build mode, it does.
[12:06] Gaius Goodliffe: Dpeends on if the edit window is open or not.
[12:06] Teravus Ousley: take care Kelly
[12:06] Andrew Linden: So they would count against the full prim limit.
[12:06] Rex Cronon: bye kelly
[12:06] Barney Boomslang is back to his lab - needs to build a torture device to make desmond recut those stupid parcels in SSC. it's a dirty job, but someone has to do it ...
[12:06] Andrew Linden: I wonder if that is why they refuse to rez.
[12:06] Gaius Goodliffe: I thought "selected/sat upon" didn't.
[12:06] Arawn Spitteler has this in fromk the Metaphysics eEngine, of SL: You can see the video here: http://www.cbsnews.com/sections/i_video/main500251.shtml?source=nav_video
[12:07] Chalice Yao: Hmmh. I guess it should give them selected status after temp status on rez, in a way, then.
[12:07] Andrew Linden: Yeah, I can't remember. I'll have to look at the code.
[12:07] Chalice Yao: I am not sure if I was in build mode.
[12:07] tx Oh: hmm..
[12:07] Chalice Yao: I don't think so.
[12:07] Chalice Yao: well, it's a simple test either way :>
[12:07] Rex Cronon: bye everybody. lets see if i can tp to SWT
[12:07] tx Oh waves
[12:07] Alpha Prim: nooo im not gonna dance on the table... smiles
[12:08] Chalice Yao: Aw.
[12:08] Gaius Goodliffe: Thank you, gentlemen. Always a pleasure...
[12:08] Teravus Ousley: take care Rex
[12:08] Teravus Ousley: :D
[12:08] Alpha Prim: see you around people
[12:08] Teravus Ousley: thanks for the announcement
[12:08] Chalice Yao: Take good care, Andrew. And thanks again for listening :>