User:Andrew Linden/Office Hours/2007 10 23

From Second Life Wiki
< User:Andrew Linden‎ | Office Hours
Revision as of 09:43, 25 October 2007 by Andrew Linden (talk | contribs) (User:Andrew Linden/Office Hours/2007 10 22 moved to User:Andrew Linden/Office Hours/2007 10 23: Woops, created this page with the wrong date.)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Transcript of Andrew Linden's office hours:

[16:32] Sidewinder Linden: oh yes... i think there is a setting to have it announce the time every once in a while, but i leave it off
[16:32] Kitto Flora: Ah
[16:32] Object: Hello, Avatar!
[16:32] Sidewinder Linden: hi andrew
[16:33] Andrew Linden: Hello
[16:33] Ryozu Yamamoto: o_o
[16:33] Kitto Flora: Plaing sound, Vol = 0 all the time :(
[16:33] PulseBurst Flow: cool, the second hand turns..
[16:33] Kitto Flora: Hi Andrew
[16:33] Kooky Jetaime: what am I up to? about 6ft
[16:33] Ryozu Yamamoto: Wait, is llDetectedLinkNumber even supposed to work on collisions?
[16:33] Sidewinder Linden: hehe i guess that's their way of "turning it off" ;)
[16:33] Sidewinder Linden: by the way, i wanted to say thanks for the great work and bug reports you guys have been putting in
[16:33] Ryozu Yamamoto: Works fine for touch events, but doesn't work on collision events even here
[16:34] Sidewinder Linden: i was talking to a few other pm's and they're all jealous of the quality of issue reporting and follow-up that is happening on the havok4 beta ;)
[16:34] Andrew Linden: The llDetectedLinkNumber() doesn't work for collision events?
[16:34] Ryozu Yamamoto: Doesn't appear to
[16:34] Kooky Jetaime: you know
[16:34] PulseBurst Flow: that is a noisy watch. ;-D
[16:34] Sidewinder Linden: but i suppose this particular beta attracts folks who are maybe a bit more technical and serious about getting things to work
[16:34] Sidewinder Linden: hehe
[16:34] Sidewinder Linden: ok ok ok ok
[16:34] Sidewinder Linden: i'll take it off
[16:34] Ryozu Yamamoto: Well
[16:34] Kooky Jetaime: I sooooo think that we need it written somewhere how things are supposed to behave so we know if they do or not
[16:35] Ryozu Yamamoto: This is a serious thing, it could make or break the entirety of SL
[16:35] Sidewinder Linden: yes...
[16:35] Andrew Linden: Someone once made a large piano keyboard that you could walk on and it would trigger sound and color changes on the keys you hit.
[16:35] Andrew Linden: Anyone have a copy of that?
[16:35] Kitto Flora: Well - we been there doen that and hav a lot of fancy scripts
[16:35] Sidewinder Linden: on llDetectedLinkNumber() - i'll have to check with andrew...
[16:35] Kitto Flora: Easy to spot things that are not right
[16:35] Kitto Flora: And also trace cause
[16:35] Sidewinder Linden: yes i've been wandering around in-world and enouraging all the vendors i can think of to try their products
[16:36] Kooky Jetaime: Nope, but I have about 30 collision detecters at my house set up as a alternating static / phantom(Voldetc)
[16:36] Kooky Jetaime: Have you hit Grey over at BlackOps?
[16:36] Sidewinder Linden: anything that does scriptinga nd relies of setpos or collissions....
[16:36] Kooky Jetaime: DCS
[16:36] Sidewinder Linden: hmm nope....
[16:36] Kooky Jetaime: Carlos Inc
[16:36] Ryozu Yamamoto: It's just on child prims, etc
[16:36] Kitto Flora: My train uses setpos some, and lots of colisions.
[16:36] Kooky Jetaime: DCS might be a good one since it depends on collision with the avatar
[16:36] Ryozu Yamamoto: Err, nevermind that
[16:37] Kitto Flora: The collide on the vehicles all appears to be working nominally
[16:37] Sidewinder Linden: btw - feel free to ping anyone you know and encourage them too - i'll follow up with those guys
[16:37] Ryozu Yamamoto: Collisions seem to work, it's just detecting which link number the collision happened with
[16:37] Sidewinder Linden: that's good kitto
[16:37] Sidewinder Linden: ahh ok
[16:37] Andrew Linden: ok that is good to know
[16:37] Sidewinder Linden: i seem to remember a bug about linkage order...
[16:37] Kitto Flora: But its hard to do more detailed testing because of other problems
[16:37] Ryozu Yamamoto: Aye, there was some commenting on that
[16:37] Ryozu Yamamoto: The new way is more desireable, but might break things
[16:37] Kitto Flora: The SetPos() damage to root prim is/was a killer
[16:38] Ryozu Yamamoto: Not familiar with that one Kitto?
[16:38] Sidewinder Linden: oops - heh - when you moved andrew i thought you'd relogged - did you see the question about llDetectedLinkNumber()?
[16:38] Kooky Jetaime: I im'ed Dimentox Travanti (maker DCS) Sidewinder
[16:38] Andrew Linden: I think maybe Kelly Linden fixed that llSetPos() problem on roots.
[16:38] Sidewinder Linden: [16:33] Ryozu Yamamoto: Wait, is llDetectedLinkNumber even supposed to work on collisions?
[16:38] Andrew Linden: No, what is the question about llDetectedLinkNumber()?
[16:38] Kitto Flora: https://jira.secondlife.com/browse/SVC-855 <- setpos problem .. fixed
[16:38] Andrew Linden: Is this a jira item you're talking about?
[16:39] Ryozu Yamamoto: Should it work for collision events at all? http://jira.secondlife.com/browse/SVC-841 reports it as returning 0 for collisions with any link number.
[16:39] Ryozu Yamamoto: I personally expected that if a collision happens with a child prim, it would tell you which link number it came from
[16:39] Ryozu Yamamoto: But testing it here on the main grid, the it also returns 0
[16:39] Andrew Linden: Yes, I think it is supposed to report the link number of the child prim that was hit.
[16:40] Ryozu Yamamoto: Ah, in which case, it's broken here too, I think
[16:40] Sidewinder Linden: hmm well in that case, we might break content by changing that, but i see what you mean about the expectation... :)
[16:40] Andrew Linden: In any case, when we are in doubt we see what the current behavior is here and try to replicate it.
[16:40] Ryozu Yamamoto: Or my repro is broken =|
[16:40] Sidewinder Linden: do you think anyone would rely on it returning zero for child prims?
[16:40] Ryozu Yamamoto: I doubt it
[16:40] Sidewinder Linden: <- knows that's a hard one to know for sure...
[16:40] Ryozu Yamamoto: It returns 0 no matter what prim it's in
[16:40] Sidewinder Linden: oh so it's just not working at all?
[16:40] Ryozu Yamamoto: So I'd be kind of curious who'd use it instead of just using 0 ;)
[16:41] Ryozu Yamamoto: Correct
[16:41] Object: Hello, Avatar!
[16:41] Andrew Linden: Sounds like a real bug.
[16:42] Kitto Flora: I have collision detecting scripts in multiple child prims. ON MG they work as expected
[16:42] Andrew Linden: The current status of Havok4 is best revealed by looking at the recently closed bugs on jira.secondlife.com
[16:42] Andrew Linden: so I'm not going to try to remember everything that has changed recently.
[16:42] Kooky Jetaime: yep, again, need to write out what's supposed to happen so you know when somethings broke..
[16:42] Kitto Flora: On H4 - dont know yet - needs stable train to test on
[16:42] Andrew Linden: We're going to try to update the preview tomorrow.
[16:42] Ryozu Yamamoto: Hmm
[16:43] Sidewinder Linden: ryozu - what are you doing?
[16:43] Kooky Jetaime: like today, the memory console window, we had a bug filed regarding an "grey pita window blocking mouse clicks" when in actuallality, the window wasn't a bug, but the fact that it is empty is.
[16:43] Ryozu Yamamoto: I'm making myself very very confused
[16:43] Kooky Jetaime: I know that feeling
[16:44] Andrew Linden: Kooky, are you talking about a Havok4 bug or a bug in SL proper?
[16:44] Sidewinder Linden: kooky?
[16:45] Andrew Linden: I should remind everyone that Thursday's office hours will be somewhere in the Havok4 preview.
[16:45] Andrew Linden: Anyone have a suggested location?
[16:45] Ryozu Yamamoto: I like Bug Island myself
[16:45] Kooky Jetaime: Andrew - hit Ctrl-Shift-0
[16:45] Sidewinder Linden: thanks andrew... i'll put that in the blog post too as a reminder
[16:45] Kooky Jetaime: bug island works
[16:45] Kitto Flora: South end of Busy Bens?
[16:45] Kooky Jetaime: just as long as noones trying to crash it
[16:46] Ryozu Yamamoto: Erk
[16:46] Ryozu Yamamoto: My repro script is busted >_<
[16:46] Andrew Linden: Hrm... yeah I get the shadow of a window.
[16:46] Kooky Jetaime: yep
[16:46] Kooky Jetaime: someone reported that as a bug, not knowing what caused it
[16:46] Kooky Jetaime: or how to get rid of it
[16:46] Kooky Jetaime: thats not the bug, the bug is that its blank
[16:46] Sidewinder Linden: did we decide on a location for the thursday meeting?
[16:46] Andrew Linden: I'll just ignore that bug... someone reported it already and I"m trying to focus on Havok4.
[16:47] Andrew Linden: Bug Island sounds fine, let's try that.
[16:47] Kooky Jetaime: it comes down to the "do we know what is expected
[16:47] Sidewinder Linden: bug island or busy beans?
[16:47] Sidewinder Linden: does it matter?
[16:47] Kooky Jetaime: just put it on the wiki
[16:47] Sidewinder Linden: kk
[16:47] Kooky Jetaime: cuz I'll forget by then
[16:47] Ryozu Yamamoto bows head in shame.
[16:47] Andrew Linden: I don't have any favorite spots on the preview, so I'll just say Bug Island. I'll also post to the wiki.
[16:48] Sidewinder Linden: ok
[16:48] Andrew Linden: There was something I was going to bring up...
[16:48] Ryozu Yamamoto wonders off to retry the repro with a non broken script
[16:48] Sidewinder Linden: heh sok ryozu
[16:48] Andrew Linden: I should write these ideas down. Hopefully it will come to me.
[16:49] Sidewinder Linden: i am wondering if any of you are seeing any texture problems... i think there was one report of retexturing not behaving correctly...?
[16:49] Kooky Jetaime: Hi Sai
[16:49] Andrew Linden: Texture problems in Havok4?
[16:49] Kitto Flora: Not tried
[16:49] Saijanai Kuhn: hey
[16:49] Sidewinder Linden: i realize it's not a havok4 issue proper, but thought it would be worthwhile to double-check if it's in our branch for some reason
[16:49] Andrew Linden: Maybe we should roll the Havok4 client back to something with fewer bugs.
[16:50] Sidewinder Linden nods... maybe... but in a week or so hopefully these would be cleared anyway...
[16:50] Sidewinder Linden: as long as they're not blocking the testing we really need to be doing, it seems ok to stay with this build...
[16:51] Andrew Linden: Lessee... I've got a few tickets open with Havok.
[16:51] Ryozu Yamamoto: hahaha... >_> Nevermind SVC-841, it's all working just fine
[16:51] Andrew Linden: They're still working on some crash bugs.
[16:51] Andrew Linden: And I've got another ticket open... related to the avatar losing its inverse-inertia-hack
[16:52] Ryozu Yamamoto: You know Andrew, I've noticed that it's happening a lot less
[16:52] Andrew Linden: --> causes the avatar collision shape to tumble --> bent knees and difficulty controlling avatar.
[16:52] Ryozu Yamamoto: I can't get the avatar to sink by looking down, but haven't looked at the actual rotation values
[16:52] Ryozu Yamamoto: Not witht he mouse trick anyways
[16:52] Andrew Linden: Well, I fixed it for the avatar Ryozu.... on Monday (not deployed yet).
[16:53] Andrew Linden: However, it will still be a problem for objects that use llSetStatus(STATUS_ROTATE_X, FALSE).
[16:53] Andrew Linden: and I don't know why it is happening yet
[16:53] Andrew Linden: Oh I remember what I was going to ask you guys...
[16:54] Andrew Linden: At which point do you think a Havok4 simulator would be ready to test with an small deploy to the main SL world?
[16:54] Andrew Linden: Such as a sandbox and an estate or two?
[16:54] Ryozu Yamamoto: Oh, right, set status on rotates is indeed not working
[16:55] Kitto Flora: When crossing h1/h4 border is no longer a p[roblem.
[16:55] Sidewinder Linden: that's a good point kitto...
[16:55] Ryozu Yamamoto: That would be a good start ;
[16:55] Andrew Linden: I guess what I mean is... what kinds of bugs (perhaps the existing bugs as examples) woudl block that and what would not?
[16:55] Ryozu Yamamoto: If there are any bugs that appear to permanantly break an object, I'd say it's a no go
[16:55] Andrew Linden: Naturally, any bug that actually destroys content -- rezzes all messed up for example, is a blocker.
[16:55] Ryozu Yamamoto: Anything that would need a script recompile also a no go
[16:56] Sidewinder Linden: so for instance the llsetpos and rotate issues, such that if you pull to inventory borken - that's a no go, right ryozu?
[16:56] Ryozu Yamamoto: Aye
[16:56] Sidewinder Linden: hehe "borken" - umm broken even ;)
[16:56] Andrew Linden: What bugs require a "script recompile"?
[16:56] Kitto Flora: we konws borked
[16:56] Ryozu Yamamoto: But also, if running a script in H4 requires any kind of reset or recompile, there will be erroneous bug reports based on that
[16:56] Andrew Linden: I take it you're thinking aobut some example problems in the past.
[16:57] Kooky Jetaime: well
[16:57] Ryozu Yamamoto: I have issues with llMoveToTarget working differently after a recompile
[16:57] Ryozu Yamamoto: Not sure on it, it's pretty erratic
[16:57] Tmyclyk Dmytryk: sometimes resets or recompiles actually could be quite detrimental to normal running scripts
[16:57] Tmyclyk Dmytryk: depending on their nature
[16:57] Kooky Jetaime: Not recompile, but what about reattaching after recompile?
[16:57] Kitto Flora: There have been changes in the past that it was said 'you may need to recompile script'. That turns out to be a bad thing.
[16:57] Sidewinder Linden: tmychyk - what would be an example of a reset/recompile being detrimental?
[16:57] Ryozu Yamamoto: Definatly no automatic recompile of a script based on changing sims
[16:58] Andrew Linden: Ok, so you're thinking of bugs/changes in the past, but not particular Havok4 bugs?
[16:58] Tmyclyk Dmytryk: I couldn't say any direct examples but some scripts hold data over time, and if they're reset could lose that data
[16:58] Kooky Jetaime: And sometimes reset/recompile just isn't an option
[16:58] Ryozu Yamamoto: My design was bad, given, but I had a gift box system that relied on the gift boxes not being reset/recompiled
[16:58] Sidewinder Linden: ahh
[16:58] Sidewinder Linden: and what would happen if they were reset/recompiled?
[16:59] Tmyclyk Dmytryk: Say for example a user configures an object they're wearing via voice chat, a reset/recompile would completely wipe that out
[16:59] Tmyclyk Dmytryk: The default values would once again take place
[16:59] Ryozu Yamamoto: And yeah, depending on the the Perms of a script, recompile/reset may be disabled
[16:59] Kitto Flora: Hard to say yet... when one can move anything scripted across the h1/h4 border and it keeps on running normally - then ome believes that recompile is not necessary.
[16:59] Sidewinder Linden: andrew, would crossing versions force a reset or recompile?
[16:59] Andrew Linden: "crossing versions"? what do you mean by that?
[17:00] Sidewinder Linden: crossing from an H1 region to an H4 or vice versa
[17:00] Andrew Linden: crossing a region boundary between diff version simulators? No that should not require a recompile.
[17:00] Sidewinder Linden: and it certainly wouldn't transparently do a recompile on the fly, right?
[17:00] Andrew Linden: When a recompile is required it means that
[17:01] Andrew Linden: there was a bug that caused some scripts to compile or save incorrrectly
[17:01] Tmyclyk Dmytryk: also recompiling requires sending the source code to the viewer, i could see scripts being stolen through some sort of cache manipulation
[17:01] Andrew Linden: and so when they are loaded they come in wrong or bugged
[17:02] Andrew Linden: so it means that a script compile or save bug was deployed for a little while and was eventually caught and fixed
[17:02] Andrew Linden: Hence, any bug that would show different behavior when moving from a Havok1 to Havok4 region
[17:02] Andrew Linden: or back again
[17:02] Andrew Linden: could be considered such a bug
[17:02] Ryozu Yamamoto nods.
[17:02] Andrew Linden: that is... different behavior in the script
[17:03] Andrew Linden: so ... if you see a particular behavior in H4 (on an attachment for example)
[17:03] Andrew Linden: and see different behavior when going from that H4 to a H1 and back again to H4
[17:03] Andrew Linden: then that would be a blocker bug
[17:04] Kitto Flora: I agree with that
[17:04] Andrew Linden: if you see different behavior in H4 than H1, but the same behavior in H4 every time and the same diff behavior in H1 every time... that might not be a blocker.
[17:05] Andrew Linden: I just fixed a bug right before this office hour...
[17:05] Andrew Linden: SVC-821 I think.
[17:05] Kitto Flora: Depends what the difference is...
[17:05] Andrew Linden: It requires a recompile of some objects...
[17:05] Ryozu Yamamoto: Ah, that one
[17:05] Sidewinder Linden: the llgroundrepel one?
[17:05] Andrew Linden: just objects that were using llGroundRepel() that had been taken to inventory in H4
[17:06] Andrew Linden: so, the script corruption is confined to the H4 inventory
[17:06] Sidewinder Linden: ahh so this doesn't affect anything in-world on the main second life world...
[17:06] Ryozu Yamamoto: So no real problem there
[17:06] Andrew Linden: and anything that is sitting in your inventory from before H4 should rez just fine.
[17:06] Andrew Linden: Yes, but an example of a disasterous bug if it had been deployed to SL proper.
[17:06] Sidewinder Linden: yes
[17:06] Kooky Jetaime: yikes
[17:07] Andrew Linden: Well... "disasterous" is perhaps a heavy word for it
[17:07] Andrew Linden: but it would certainly have caused some pain.
[17:07] Kitto Flora: Something like the 'leaning' bug I would not consider to be blocking.
[17:07] Kooky Jetaime: lotsa pain
[17:07] Sidewinder Linden: for any script that uses that feature
[17:07] Sidewinder Linden: by the way, related to the "what you've done in h4 inventory" concept...
[17:08] Andrew Linden: Kitto, you're talking about the leaning vehicle bug?
[17:08] Kitto Flora: Yes
[17:08] Andrew Linden: Dan Linden just showed me a bug which suggests that the center of mass of objects is just wrong.
[17:08] Kitto Flora: Humm
[17:08] Ryozu Yamamoto: lol!
[17:08] Andrew Linden: He had an object that appeared to have its center of mass at the position of the root prim.
[17:08] Ryozu Yamamoto: I saw him working on that, had 3 linked cubes hanging off the edge of a box
[17:08] Andrew Linden: yeah
[17:09] Kitto Flora: Well - its to do with having a sphere as the root I think - change root o box and it straightens out
[17:09] Andrew Linden: that bug could cause all sorts of problems... especially with H4 ground vehicles
[17:09] Ryozu Yamamoto: Could be part of our problem with cars
[17:09] Ryozu Yamamoto: My Dominus has taken to leaning to one side
[17:09] Andrew Linden: I'm going to try to fix that tomorrow first thing, since it will probably fix more than one open bug.
[17:10] Kitto Flora: Despite leaning, the Swan Ride works well, and does cross h4/h4 boundary too
[17:10] Sidewinder Linden: cool
[17:10] Sidewinder Linden: h1/h4 kitto?
[17:10] Kitto Flora: h1/h4 I have not tried much on
[17:10] Andrew Linden: Anyone here from Europe?
[17:10] Kitto Flora: Have to find a border with long return times and building allowed both sides
[17:10] Andrew Linden: er... in Europe now.
[17:11] Kitto Flora: 1am or later in europe
[17:11] Sidewinder Linden: can i ask a quick "survey-style question" about inventory?
[17:11] Andrew Linden: yeah, that is why i was curious.
[17:11] Andrew Linden: I was thinking about moving my office hours around to different times
[17:12] Andrew Linden: perhaps one of them around 12:00 to work well for europeans.
[17:12] PulseBurst Flow: in US
[17:12] Ryozu Yamamoto: Any time is good for me
[17:12] Kitto Flora: Noon Pacific is a hot time for europeans
[17:12] Andrew Linden: Yeah, 12:00 SLT
[17:13] Andrew Linden: I won't change anything for another week or two
[17:13] Andrew Linden: but was just curious if anyone had opinions on what the best times for office hours would be.
[17:14] Kitto Flora: For me this is a bad time and so is 8AM - usually
[17:14] Andrew Linden: Kitto, you're a regular here. What time is good for you? :-)
[17:14] Kitto Flora: Noon is possible... otherwise after 5pm
[17:15] Kitto Flora: Pacific time...
[17:15] Kitto Flora: This is usually travel time for me
[17:15] Sidewinder Linden: what if we do one at 12 pdt and another at 5:00 pdt? is that workable?
[17:15] Andrew Linden: Ok, well I'll take it into consideration when I try to change the schedule.
[17:16] Kitto Flora: Lots better for me
[17:16] Andrew Linden: Anybody have any particular issues to talk about?
[17:16] Sidewinder Linden: i have one... general...
[17:16] Sidewinder Linden: how do you all feel about refreshing the beta preview database, to bring the accounts and inventory up to current... note that this will mean you will lose any scripts here that haven't been saved off to text somewhere...
[17:16] Kitto Flora: Noon is probably good for europeants - 8pm/9pm/10pm for them
[17:16] Andrew Linden: Oh right, that was what I was trying to remember.
[17:16] Sidewinder Linden: i've gotten several requests to refresh the db, but wasn't sure if that would create problems
[17:16] Kooky Jetaime: doesn't the inventory catch up every day or so?
[17:16] Ryozu Yamamoto: That's fine for me. My inventory's a huge mess anyways
[17:16] Sidewinder Linden: it would pull the preview grid offline for about 6-8 hours
[17:16] Sidewinder Linden: nope kooky
[17:17] Sidewinder Linden: this is a snapshot of at this point several weeks ago
[17:17] Kitto Flora: Refresh is good - will import latest product scripts for me
[17:17] Sidewinder Linden: ok...
[17:17] Kooky Jetaime: then whats with the "Syncronizing Inventory" message I see at login?
[17:17] Tmyclyk Dmytryk: never hurts toupdate it :D
[17:17] Sidewinder Linden: heh until you say "but you wiped out my special woozle i just built" ;)
[17:17] Andrew Linden: "Synchronizing Inventory"? You see that every time Kooky?
[17:17] Tmyclyk Dmytryk: haha well the primary idea behind the beta is everything is volatile.
[17:17] Kooky Jetaime: Yea
[17:18] Kitto Flora: Well right now - Beta grid testing is dead cause we waiting for all the bug fixes
[17:18] Kooky Jetaime: I mean I log in rarely but I've seen it twice now since H4
[17:18] Kooky Jetaime: give me a sec and I'll log in right now
[17:18] Andrew Linden: It is probably synching with your client's cached notion of what is in your db.
[17:18] Kooky Jetaime: ah
[17:18] Kitto Flora: How bout update before the new deply tomorrow?
[17:18] Sidewinder Linden: i believe, but am not 100% sure, kooky, that us a cache update locally
[17:18] Kooky Jetaime: I'd say put a note so people can grab anything they need to from beta
[17:18] Ryozu Yamamoto: What Kitto said. I'm kind of waiting for a few known bugs to clear up before I start pestering about other percieved bugs of mine.
[17:18] Kooky Jetaime: don't just up
[17:18] Andrew Linden: No... we were thinking more like next week right Sidewinder?
[17:18] Kooky Jetaime: update it cuz some people might be working on stuffs
[17:19] Sidewinder Linden: right - we were thinking about doing a new code deploy in the next day or so
[17:19] Sidewinder Linden: and then refreshing the database perhaps monday or tuesday
[17:19] Kitto Flora: True - best to give a warkning to save new scripts offline
[17:19] Sidewinder Linden: so taht we would not have the preview offline over the weekend or going into the weekend
[17:19] Ryozu Yamamoto nods.
[17:19] Sidewinder Linden: and right - it woudl give notice so that you can grab what you need first
[17:19] Sidewinder Linden: ok.. so we'll plan to do a db refresh early next week, and i'll be announced again in the blog post about this upcoming refresh
[17:20] Ryozu Yamamoto: I'd announce the new code on the grid and the db refresh coming up in one post, consolidate =D
[17:20] Kooky Jetaime: yep
[17:20] Sidewinder Linden: i had one more strategy question...
[17:21] Sidewinder Linden: at what point do you think i should do the blog post with the "impassioned pitch" for product vendors to "try everything that they can that is scripted" on the havok4 preview?
[17:21] Ryozu Yamamoto: Not quite yet
[17:21] Sidewinder Linden: from talking to ppl in-world i think the product vendors may be thinking this is "too technical" of a preview for them
[17:21] Ryozu Yamamoto: I still have some big concerns about scripted physics not working qutie right
[17:21] Sidewinder Linden: and at some point i want to push to be sure we get broad testing from them
[17:21] Sidewinder Linden: yes agreed ryozu...
[17:22] Sidewinder Linden: once the various llsetpos and rotation and linkage bugs are cleared, would that maybe seem like the right time?
[17:22] Kitto Flora: If all my toys work.. I'd call it good :)
[17:22] Sidewinder Linden: hehe ok kitto :)
[17:23] Ryozu Yamamoto: It'd be a good start, I'd also want llMoveToTarget looked at, that's a very widely used command and it's current functionality is just
[17:23] Ryozu Yamamoto: weird
[17:23] Andrew Linden: Yeah not yet. When we've solved most of the annoying little differences and have just a few obscure blockers.
[17:23] Sidewinder Linden: ok
[17:23] Ryozu Yamamoto: That's the only real major thing in my mind stopping it
[17:23] Kitto Flora: The few I've asked about if'n they have tested on H$ .. they are remarkably uninterested.
[17:23] Ryozu Yamamoto: Heh
[17:23] Sidewinder Linden: i just wanted to have us all thinking about that question, since i suspect we'll have "one good shot" at getting a bunch of product vendors in here testing. if i ask them mult times they may burn out quickly.
[17:23] Kooky Jetaime: they'll be uninterested until all their products break
[17:23] Ryozu Yamamoto: I think you're right Sidewinder
[17:24] Ryozu Yamamoto: Yeah =/
[17:24] Sidewinder Linden: thanks ryozu...
[17:24] Kooky Jetaime: then it'll be "Why didn't you test more" blah blah blah
[17:24] Sidewinder Linden: i've found in talking directly to folks that they are interested - once they understand that their products might break if they don't test them...
[17:24] Kitto Flora: The way to get people interested in testing is to use their input - ie. fix the bugs.
[17:24] Sidewinder Linden: the new bugs from valradica vale cme in this way - they make the rendesvous widget...
[17:25] Sidewinder Linden: once they heard about the connection betwen our work and theirs, they were in here in a matter of hours trying things
[17:25] Ryozu Yamamoto nods.
[17:25] Sidewinder Linden: ok thanks... i'll hold on the big push, but hope that we're only a few weeks away ;)
[17:25] Kooky Jetaime: whoo hoo
[17:25] Kooky Jetaime: havok 4 by xmas
[17:25] Ryozu Yamamoto: I think part of the problem is, people don't understand just what all is affected by physics: "Linking? But I didn't check the 'physical' box so why's that matter?"
[17:25] Andrew Linden: Oh yeah, that reminds me... I need to remind Seifert that "The Future" should be up on the preview grid.
[17:25] Sidewinder Linden: shhhhhhhh - hehe :)
[17:25] Sidewinder Linden: we'll see
[17:26] Kooky Jetaime: then hello mono
[17:26] Sidewinder Linden: hi spectrum
[17:26] Kooky Jetaime: did someone just kick the sub-sonic dude
[17:26] Sidewinder Linden: not me...
[17:26] Ryozu Yamamoto: Familiar looking....
[17:26] Kooky Jetaime: Nope here he is
[17:26] Sidewinder Linden: so do we have other issues
[17:26] Kitto Flora: Its also now VERY complex - the difference between 'new bug' and 'exisiting old bug' and 'its a feature' and 'thats an upgrade'.
[17:26] Kooky Jetaime: Spectrum is gonna give me a headache though with that low sound wave
[17:27] Sidewinder Linden: spectrum - please behave or you will be banned
[17:27] Kooky Jetaime: Thanks :)
[17:27] Spectrum Flow: what did i do
[17:27] Spectrum Flow: ?
[17:27] Spectrum Flow: i took the sub off
[17:27] Sidewinder Linden: do we have other issues from teh beta preview to discuss?
[17:27] Sidewinder Linden: ok thanks spectrum
[17:27] Ryozu Yamamoto: 's all good then ;)
[17:27] Kitto Flora: 59 minutes and no mention of megaprims?
[17:28] Sidewinder Linden: LMAO
[17:28] Andrew Linden: Thank you! I don't have to talk about megaprims today!
[17:28] Saijanai Kuhn: so what IS up with the unmnetionables?
[17:28] Sidewinder Linden: well kitto - go for it...
[17:28] Tmyclyk Dmytryk: lol well that's up in debate in the forums
[17:28] Sidewinder Linden: "the prims that cannot be named", eh?
[17:28] Kitto Flora: Strictly my dear, I dont GAD.
[17:28] Saijanai Kuhn: too big for the name box
[17:28] Kitto Flora: I never use em
[17:28] Ryozu Yamamoto: Bwhahaha
[17:29] Kitto Flora: Just hav customers who bump into them sometimes
[17:29] Saijanai Kuhn: I can see uses for them as long as they don't mess things up. Otherwise...
[17:30] Sidewinder Linden: andrew... did we end up with some conclusion about this, or is it better to just say "yes, we've heard a lot of reasons to do various things and haven't made a final decision"?
[17:30] Ryozu Yamamoto chuckles
[17:30] Andrew Linden: Sidewinder, I've ben telling everyone "This is what i'm going to do about megaprims..."
[17:31] Saijanai Kuhn: and .... is ...
[17:31] Saijanai Kuhn: ?
[17:31] Andrew Linden: Others have been saying "We're thinking about doing something about megaprims... what do you think?"
[17:31] Kitto Flora gets a stone tablet and chisle ready...
[17:31] Andrew Linden: MY PLAN for megaprims is as follows...
[17:31] Saijanai Kuhn: the elipses that cann't be named?
[17:32] Andrew Linden: Nerf megaprims larger than 256 meters on a side (clamp to 256 m)
[17:32] Spectrum Flow: do physical megaprims cause an issue
[17:32] Andrew Linden: then leave them in their current semi-broken semi-supported state
[17:32] Andrew Linden: until I can fix the parcel-object collision problem (my neighbor's obejcts overlap my parcel)
[17:32] Saijanai Kuhn: will that cause problems later on when you try to add new features?
[17:32] Ryozu Yamamoto: Then once bounding box collisions with parcel borders are enabled to prevent encroachment, prim sizes shouldn't be nearly as much of an issue?
[17:33] Saijanai Kuhn pleads with the Lindens to keep teh gigaprim. 65K on a side isn't THAT big
[17:33] Ryozu Yamamoto: lol
[17:33] Andrew Linden: once that is solved then I would open up prim sizes to any size... as long as it does not violate any parcel owner's preferences
[17:33] Tmyclyk Dmytryk: What controls will the victim of an overlapping issue get anyway?
[17:33] Andrew Linden: that means... we may need to add a "ZOMG don't ever let anyone put overlap objects on my parcel" checkbox for parcel owners... off by default.
[17:34] PulseBurst Flow: so is that definitive policy, oor just your view?
[17:34] Saijanai Kuhn: would any size include Qarl's call for 4x256 vertex sculpties? A rope that stretchs throughout a sim would be interesting
[17:34] Sidewinder Linden: andrew... once this is done, what will happen to builds already rezzed in-world with larger than 256^2 megaprims?
[17:34] Tmyclyk Dmytryk: What if that option wasn't selected then they select it? Will it return or just push it off back onto their parcel?
[17:34] Sidewinder Linden: would they be removed and never rez again?
[17:34] Ryozu Yamamoto: Andrew's one of the big devs afaik
[17:34] Andrew Linden: That is ANDREW LINDEN's MEGAPRIM PLAN.
[17:34] Andrew Linden: I haven't heard a better plan yet.
[17:34] Kooky Jetaime: heheh
[17:35] Kitto Flora: That sounds like a good plan Andrew. Do you plan to work on resolving the collision issue?
[17:35] Kooky Jetaime: oh I'm gonna go shoot someone I think :P
[17:35] PulseBurst Flow: Oh...will there be an offical LL policy announced?
[17:35] Andrew Linden: Some plans are impossible to do.
[17:35] Andrew Linden: Micheal was asking me for input so he could form an official plan.
[17:35] Andrew Linden: I should get back to him on that...
[17:35] PulseBurst Flow: well, sounds ok...people need word so they can plan...
[17:36] Saijanai Kuhn: as long as they don't interfere with expansion of physics, sculpties, flexies, etc., I personally don't care at this point
[17:36] Sidewinder Linden: andrew - so that would remove existing megaprims larger than 256x256 once havok4 update is put on a sim?
[17:36] Kooky Jetaime: I'll see ya'll later, got things I need to do
[17:36] Sidewinder Linden: would existing builds using the monstrous ones end up with "missing pieces"?
[17:36] Tmyclyk Dmytryk: Andrew, what about some way for sim owners to control max prim size?
[17:36] Kooky Jetaime: Andrew - what about these ejection standups?
[17:36] Tmyclyk Dmytryk: in addition to that
[17:36] Tmyclyk Dmytryk: Would be useful for say, sandboxes
[17:36] Andrew Linden: Yes, clamping to 256 meters on a side is something I'll try to push through on Havok4.
[17:37] Andrew Linden: what are "ejection standups"/
[17:37] Andrew Linden: ?
[17:37] Kooky Jetaime: where you stand up
[17:37] Kooky Jetaime: and are literally ejected
[17:37] Sidewinder Linden: sounds ok to me... i also haven't heard a better plan... does everyone here agree that this makes sense? if not, now is the time to speak ;)
[17:37] Kooky Jetaime: like just a sec ago when I stood up
[17:37] Kooky Jetaime: and got pushed
[17:37] Ryozu Yamamoto: 's fine by me
[17:37] Tmyclyk Dmytryk: works for me :)
[17:37] Kooky Jetaime: they're random in occurence
[17:37] Kooky Jetaime: but I've stood up before and got thrown 50m
[17:37] Andrew Linden: oh... you mean in Havok4? They can be much more violent in Havok4
[17:38] Ryozu Yamamoto: Though I'd look at the possibility of preventing 256^2 prims being rezzed on ANY land you don't own or aren't in the group of
[17:38] Andrew Linden: some of the standups in Havok1 actually block the simulator long enough to crash it.
[17:38] Ryozu Yamamoto: Especially sandboxes
[17:38] Kooky Jetaime: well, they're happening in H1/MG but will they be better in H4?
[17:38] Kitto Flora: If it was stranding up from a physical object, then its a normal thing.
[17:38] Ryozu Yamamoto: Actually Andrew, do you think you can do that?
[17:38] Andrew Linden: Kooky, I have a plan for fixing those in H4, but I'm not sure if I'll be able to get to that *before* H4 is considered "done" or *after*.
[17:38] Kooky Jetaime: ok
[17:39] Ryozu Yamamoto: Auto-return of megaprims > 100m in sandboxes?
[17:39] Kooky Jetaime: pdr
[17:39] Andrew Linden: Ack, I've got to run.
[17:39] Saijanai Kuhn wants to build a version of Qarls' ant with megaprim sculpties...
[17:39] Ryozu Yamamoto: because honestly, that's the biggest method of greifing with them
[17:39] Ryozu Yamamoto: lol
[17:39] PulseBurst Flow: well, if there is eventual support for large prims, presuming you have permission on land,..how about some creation tools, and also, spheres, etc..
[17:39] Andrew Linden: Hrm... no megaprims in sandboxen is possible.
[17:40] Sidewinder Linden: pulseburst... perhaps later - after havok4 is released... :)
[17:40] Saijanai Kuhn: because little runty ants are boring
[17:40] Andrew Linden: Thanks for coming. See you all later.
[17:40] Sidewinder Linden: but that will depend on futures and we're not there yet...
[17:40] Kitto Flora: How bout a general fix in the system - set a max size allowed in a parcel
[17:40] Ryozu Yamamoto: Seeya guys
[17:40] PulseBurst Flow: ok..something to look froward too.
[17:40] Sidewinder Linden: thanks everyone
[17:40] PulseBurst Flow: thanks..see you..
[17:40] Sidewinder Linden: keep up the good work, and we'll try to have a refresh up before the weekend
[17:41] Kitto Flora: Owww h ope so
[17:41] Saijanai Kuhn: was quite interesting. THanks for hosting us
[17:41] Ryozu Yamamoto: Thanks for the hard work, as always ;)
[17:41] Sidewinder Linden: thanks for the great bug reports and testing again btw - really helps a huge amount
[17:41] Kitto Flora: Update Beta soon!! :)
[17:41] Opensource Obscure: thanks, guys
[17:41] Tmyclyk Dmytryk: :D can't wait, this entire deal sounds promising
[17:41] Kitto Flora: Bye Andrew
[17:41] Sidewinder Linden: whooo
[17:41] Sidewinder Linden: like that ey? ;)
[17:42] Kitto Flora: Bye Sidewinder
[17:42] Sidewinder Linden: bye kitto ... and all
[17:42] PulseBurst Flow: bye all