Content Creation/Scripting User Group/Transcripts/2011 11 14

From Second Life Wiki
< Content Creation‎ | Scripting User Group‎ | Transcripts
Revision as of 10:57, 10 January 2012 by Kelly Linden (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

List of Speakers

Ardy Lay Fancy Detector
flexi campfire Ima Mechanique
Kopilo Hallard Liisa Runo
Motor Loon Susie Chaffe
TheBlack Box Vincent Nacon

Transcript

[09:01] Motor Loon: Yo Kelly man

[09:01] Kopilo Hallard: greetings kelly

[09:02] F L I P (flip.idlemind): May I ask "why" as well?

[09:02] Kaluura (kaluura.boa): Hello, hello...

[09:02] Nimh (nimh20.vandeverre): https://jira.secondlife.com/browse/SVC-7251?

[09:02] flexi campfire: [#SVC-7251] Remove the automatic logout of agents that are Away too long

[09:02] F L I P (flip.idlemind): Why are you upset about not getting force-logged-out

[09:02] Koli Contepomi (sahkolihaa.contepomi): MAny people didn't like it, so it was removed.

[09:02] Kaluura (kaluura.boa): There is setting for that in all the viewers that I know...

[09:02] Koli Contepomi (sahkolihaa.contepomi): I pay nothing.

[09:02] Motor Loon: Most viewers can disable that anyway Nimh

[09:03] Motor Loon: whats the reason for keeping it?

[09:03] Kaluura (kaluura.boa): And it's off-topic for this UG...

[09:03] Nimh (nimh20.vandeverre): Now you turn voice off

[09:03] tehKellz (kelly.linden): We don't keep a list of every person that requests a feature

[09:03] Nimh (nimh20.vandeverre): jeez

[09:03] tehKellz (kelly.linden): This is a chat-only UG

[09:03] tehKellz (kelly.linden): And the focus is scripting.

[09:04] tehKellz (kelly.linden): There is a strong desire *by many* to be able to be in world AND be AFK. I can understand the desire to get auto logged off and if you feel strongly about it (it appears you do) then please file a jira in the SVC project.

[09:04] F L I P (flip.idlemind): Theoretically you could modify a viewer so it does log you off for being idle too long. It's not that you're no longer "allowed" to auto log off, you're no longer "forced" to auto log off

[09:04] tehKellz (kelly.linden): Good morning everyone. :)

[09:04] Kopilo Hallard: :)

[09:04] Koli Contepomi (sahkolihaa.contepomi): Afternoon!

[09:04] Nal (nalates.urriah): Hi

[09:05] Nimh (nimh20.vandeverre): Ok so now that they have disabled it server side--how would you modify the viewer to override that?

[09:05] tehKellz (kelly.linden): I don't know how much news I am going to have. I've been working from home to take care of a sick family pretty much the entire last week.

[09:05] Motor Loon: viewer can close the connection anytime Nimh

[09:05] Koli Contepomi (sahkolihaa.contepomi): :(

[09:05] Kaluura (kaluura.boa): ...Even commit suicide...

[09:05] F L I P (flip.idlemind): Awww

[09:05] F L I P (flip.idlemind): Guess its cold and flu season and whatnot

[09:06] tehKellz (kelly.linden): yeah. just a cold but a long lasting one.

[09:06] Kaluura (kaluura.boa): To stay in SL and avoid people prevents the cold and flu...

[09:06] Kaluura (kaluura.boa): :P

[09:06] Nal (nalates.urriah): Airborn and Zicam...

[09:07] tehKellz (kelly.linden): The server maintenance project on LeTigre (I think) is not getting promoted this week due to a bug with objects (non-vehicle) crossing region borders.

[09:08] tehKellz (kelly.linden): The llTransferLindenDollars and transaction_result are on a few regions on aditi now.

[09:08] Motor Loon: Wunderbar

[09:08] Motor Loon: Thanx for doing that one

[09:08] tehKellz (kelly.linden): I haven't heard any complaints, but I just got in this morning.

[09:08] tehKellz (kelly.linden): :)

[09:08] Motor Loon: Mucho needed

[09:09] Fancy Detector: Charlar Linden has arrived!

[09:09] Ima Mechanique: working fine for me so far

[09:09] tehKellz (kelly.linden): And I think that is all I have for news this morning.

[09:10] Charlie (charlar.linden): Hey all

[09:10] Kopilo Hallard: hey

[09:10] Kaluura (kaluura.boa): If that's all, just tell us all you know about the LF functions...

[09:10] tehKellz (kelly.linden): :)

[09:11] tehKellz (kelly.linden): Anyone else have any topics or questions?

[09:11] Kaluura (kaluura.boa): Since I learn that it's not lTeleportAgent() which is used... The used function seems to require a LM...

[09:11] Nal (nalates.urriah): Anything on framed motion drift?

[09:11] tehKellz (kelly.linden): I'm afraid I don't have anything on that Nal

[09:11] Kaluura (kaluura.boa): In the forums, Simon explained a bit...

[09:12] Nal (nalates.urriah): Thx Kal, I'll find it

[09:12] Kaluura (kaluura.boa): You must use times which are multiples of 1/45 because llSKfM is based on frames

[09:12] Kaluura (kaluura.boa): *integer multiples of 1/45

[09:13] Kaluura (kaluura.boa): And I added this on the wiki already.

[09:13] Susie Chaffe: http://community.secondlife.com/t5/LSL-Scripting/llSetKeyframedMotion-turning-a-corner/td-p/1225219 is the thread

[09:13] tehKellz (kelly.linden): ah interesting

[09:13] Kaluura (kaluura.boa): It really improves things...

[09:17] tehKellz (kelly.linden): Hrm.

[09:17] tehKellz (kelly.linden): No other questions? I think today is topic day but we can jump to jira triage if it is this quiet.

[09:18] Kopilo Hallard: seems rather quiet

[09:18] Kaluura (kaluura.boa): Monday morning...

[09:18] Ardy Lay: I haven't been doing any scripting lately so have nothing new.

[09:18] Ima Mechanique: triage is good, I don't recall we did any last week any how

[09:18] Kopilo Hallard: same

[09:19] tehKellz (kelly.linden): Yeah I think we just did discussion last week. All right. https://jira.secondlife.com/secure/IssueNavigator.jspa?requestId=14052&mode=hide

[09:19] tehKellz (kelly.linden): Skip the first three. I need to tell jira not to put subtasks first.

[09:19] tehKellz (kelly.linden): SCR-31

[09:19] flexi campfire: http://jira.secondlife.com/browse/SCR-31

[#SCR-31] New LSL function: llGetParcelMusicURL()

[09:20] Ima Mechanique: oohh one of my favourite entries

[09:20] Kopilo Hallard: scripted parcle controls mmm

[09:20] tehKellz (kelly.linden): SCR-199

[09:20] flexi campfire: http://jira.secondlife.com/browse/SCR-199

[#SCR-199] Reduction/removal of llCastRay limitations

[09:21] tehKellz (kelly.linden): I imported 31.

[09:21] tehKellz (kelly.linden): I think 199 is done already?

[09:22] tehKellz (kelly.linden): At the very least it is already acknowledged, it even has falcon assigned to it.

[09:22] tehKellz (kelly.linden): I know he made changes and I'm pretty sure those were released.

[09:22] tehKellz (kelly.linden): But maybe he plans more? I'll follow up with him later today.

[09:23] tehKellz (kelly.linden): SCR-154

[09:23] flexi campfire: http://jira.secondlife.com/browse/SCR-154

[#SCR-154] llRezObjectFast() and llRezAtRootFast(), rezzing functions with zero delay

[09:23] Kaluura (kaluura.boa): Not a chance...

[09:23] Kaluura (kaluura.boa): Somebody must tell it...

[09:24] Motor Loon: Dont really see a reason why not to.... since people just use multiple scripts to do it anyway

[09:24] tehKellz (kelly.linden): The problem I see with rez is we have a lot of 'grey goo fence' logic that may rely on the current behavior to be effective.

[09:25] tehKellz (kelly.linden): Rez also is one of the few (only?) script actions that still uses "energy" which further rate limits.

[09:25] Ima Mechanique: throttle it and limit it to grey goo as a maximum?

[09:25] Kaluura (kaluura.boa): And such a function is a door widely open to abuse...

[09:25] Motor Loon: Wouldnt it be enough to have the goofence work on the object thats rezzing... while the scripts itself didnt have a limitation

[09:26] tehKellz (kelly.linden): When I talked to andrew about this last the conversation quickly turned to a bit more radical an idea - llNoFunctionSleeps to disable function sleeps entirely instead of continuing the piecemeal function at a time approach.

[09:27] Liisa Runo: limit/owner rather than by objct

[09:27] tehKellz (kelly.linden): The goo fence is a complex system that takes a LOT of factors into account to attempt to differentiate good goo (ALife) from bad goo (grief attack)

[09:28] Kaluura (kaluura.boa): Go ahead! Nobody will complain if delays disappear... Well not many ppl will...

[09:28] Squirrely Wrath (squirrel.wood): dem griefers will be happy happy

[09:28] tehKellz (kelly.linden): Some of the factors used by the goo fence is the number of different rezzers and the generation of the objects. Reducing the need to split scripts will have an effect on the fence.

[09:29] tehKellz (kelly.linden): so it isn't that we can't do it, it is just that it really needs to be well thought out and there may be unintended consequences.

[09:29] Ima Mechanique: so unsurprisingly, the goo is a sticky subject

[09:30] Squirrely Wrath (squirrel.wood): Alert community ground protection units, local unrest structure detected. Assemble, administer, pacify.

[09:30] F L I P (flip.idlemind): Since we're kinda sorta starting to use Aditi for things again, try it out over there?

[09:30] tehKellz (kelly.linden): That fence has evolved over many many years in a constant battle between "ack they took out my sim" and "but my fish need to breed!"

[09:30] Motor Loon: Personally, I see nothing bad about a removal of function sleeps

[09:30] tehKellz (kelly.linden): It would be opt-in so it wouldn't break legacy content.

[09:30] Motor Loon: Wont give greifers "tools" they dont always have (via workarounds)

[09:31] tehKellz (kelly.linden): The exception is things like the grey goo fence that take the number of scripts involved into accoubnt

[09:31] Squirrely Wrath (squirrel.wood): perhaps a per script setting ?

[09:31] tehKellz (kelly.linden): yeah, it would be a new function call.

[09:31] Squirrely Wrath (squirrel.wood): like you can throttle the events

[09:31] Kopilo Hallard: so much concern over "griefers" oh for the grid to have such attention again :p

[09:32] Kaluura (kaluura.boa): Griefers always ruin everything.... Even when they're not here...

[09:32] tehKellz (kelly.linden): It is still very much an active issue.

[09:33] Squirrely Wrath (squirrel.wood): (off topic: any chance we'll ever see an increase on the per sim prim limit?)

[09:33] Kaluura (kaluura.boa): Why not a tier reduction?

[09:33] tehKellz (kelly.linden): I'm going to close this one as Deferred, it is complex and we don't have time to look into it now unfortunately

[09:35] tehKellz (kelly.linden): I don't know squirrely. I obviously don't want to get into 'prim equivalence' but I think we had kind of hoped a smarter accounting system would allow us to do better on that front. There are lots of ways that 15k prims is already too many, but a reevaluation of that is probably not on the table near or mid term

[09:35] Squirrely Wrath (squirrel.wood): okay

[09:36] tehKellz (kelly.linden): Upping it to 18k or 20k would cause problems, but being smarter about what kind of content there was to better balance it would be great.

[09:37] Liisa Runo: 10 trees for 1 prim

[09:38] Liisa Runo: 1 ugly house cost 9000 prims even if it looks like 50 prims

[09:38] tehKellz (kelly.linden): heh

[09:39] Ardy Lay: Is there a set limit to the number of scripts a task can contain?

[09:39] tehKellz (kelly.linden): No

[09:39] Kaluura (kaluura.boa): In theory... Over 4 billions...

[09:40] Ardy Lay: I ask because somebody entered my place, created a prim, then created scrips in it until the region went down.

[09:40] tehKellz (kelly.linden): I'm pretty sure you'd hit some memory limits in the region before then.

[09:40] Kaluura (kaluura.boa): =^_^=

[09:40] tehKellz (kelly.linden): Ardy: How recently?

[09:40] Ardy Lay: About 8 months ago.

[09:41] tehKellz (kelly.linden): I recall a bug about that being fixed. Yes, definitely since then.

[09:41] Ardy Lay: Okay

[09:41] tehKellz (kelly.linden): I think if you have more than 1k scripts you won't be able to rez the object.

[09:42] Motor Loon: dont matter if its already rezzed °?°

[09:42] Ardy Lay: He created the scripts in the object while it existed in the simulator.

[09:42] tehKellz (kelly.linden): yup. We blocked that as well.

[09:42] Ardy Lay: Thanks

[09:43] TheBlack Box: i am a bit worried that with all this tweaking of the current scripting system, many "radical" improvements to the SL architecture are not on the table for a long long time ... stuff like hierarchical linking, which seems to be essential for a 3d environment, for example ... i wonder how much better of a virtual world we could get if Lindens, with all their experience, could come up with if they would start over from scratch

[09:44] Kopilo Hallard: passing out

[09:44] Ima Mechanique: TheBlack Box, that worked SOOO well with viewer 2 ;-)

[09:44] Motor Loon: lol

[09:44] Kaluura (kaluura.boa): Yeah...

[09:44] tehKellz (kelly.linden): :) There is a great article about starting over from scratch and some of the fallacies involved.

[09:45] tehKellz (kelly.linden): But I can't find it right now.

[09:45] Nal (nalates.urriah): O.O I would expect a start over project to take years to reach a useable point.

[09:45] Motor Loon: at best

[09:45] TheBlack Box: i know it is just a dream :) but still some improvements seem to be impossible as long as it all has to be small incremental additions to the existing system

[09:46] tehKellz (kelly.linden): Heirarchical linking is not impossible in the current system.

[09:46] tehKellz (kelly.linden): It isn't trivial, but it isn't impossible.

[09:46] TheBlack Box: yes ... Andrew mentioned that he looked into hierarchies when he started to help on mesh ... nbut i think he gave up again

[09:46] Nal (nalates.urriah): What does H-Linking do for a 3D world, if you'll answer a sort of newbie question?

[09:47] tehKellz (kelly.linden): Nal: I usually think of a robot.

[09:47] Motor Loon: just really makes life easyer for soo many things

[09:47] TheBlack Box: Nalates: Joints

[09:47] Nal (nalates.urriah): Ah! ok Thx

[09:47] tehKellz (kelly.linden): Nal: You want to move the arm and have the hand move with it. Right now the hand and arm are both linked to the 'root'.

[09:47] Ima Mechanique: the thigh bone is connect to the knee bone...

[09:47] Vincent Nacon: sorry, little late

[09:48] Nal (nalates.urriah): I was thinking date and objects rather movement

[09:48] Nal (nalates.urriah): data^

[09:48] TheBlack Box: object-animation in general is and will remain yucky until we have hierarchies

[09:48] tehKellz (kelly.linden): It would also be nice to have say a dining room set and be able to unlink a chair from the table and have it still be a chair (kind of a bad example but that kind of idea - take a piece off without taking it all to pieces)

[09:49] Kaluura (kaluura.boa): That too!

[09:49] TheBlack Box: the best analogy i can come up with is that the current system is like doing 2D image-manipulation without having layers

[09:49] Kaluura (kaluura.boa): And the other way around too..

[09:49] Motor Loon: some form of "grouping" might do too

[09:49] tehKellz (kelly.linden): Basically Heirarchy allows/enables grouping.

[09:50] Motor Loon: but we'd need better pivot controls too then

[09:50] Motor Loon: for it to be really useful

[09:50] tehKellz (kelly.linden): Havok has a lot of pivot controls we could hook into.

[09:51] Motor Loon: It does... problem is... not breaking all the old shit °?°

[09:51] tehKellz (kelly.linden): As far as I understand at least.

[09:51] TheBlack Box: yes Motor ... a real hierarchy would involve relative pos, rot and scale at each sub-node .. just like it is in a collada-scene-tree

[09:51] tehKellz (kelly.linden): You would have to build new things to use the new features of course

[09:52] Motor Loon: ooh... and lets have it used that wonderful new mesh accounting too then

[09:52] Kaluura (kaluura.boa): hehehe

[09:52] tehKellz (kelly.linden): 13 PE per joint! Motor requested it!

[09:52] Motor Loon: </sarcasm>

[09:52] Vincent Nacon: Falcon and Andrew still want to add heierarchy for vehicle and other things

[09:52] TheBlack Box: hehe

[09:52] Nal (nalates.urriah): :)

[09:52] Vincent Nacon: ....not sure if anyone asked but anything on llSetKeyFramedMotion()?

[09:53] Nal (nalates.urriah): I did early on... want a note card

[09:53] Nal (nalates.urriah): ?

[09:53] tehKellz (kelly.linden): http://community.secondlife.com/t5/LSL-Scripting/llSetKeyframedMotion-turning-a-corner/td-p/1225219

[09:53] tehKellz (kelly.linden): that has the comment from Falcon

[09:53] Motor Loon: vehicles with working suspension is a longtime dream in sl.... sighs...

[09:53] Vincent Nacon: patience, Motor

[09:53] TheBlack Box: maybe once Qarl is done with his current mesh-deformer project we could let him come up with a master-plan for hierarchies ? :) he understands the subject really well

[09:53] Kaluura (kaluura.boa): And look at the caveats on the wiki page... The part about time...

[09:54] Vincent Nacon: I want it too but we can't bug them into it yet

[09:54] Motor Loon: Sure we can afford that Black? ,.-)

[09:54] Vincent Nacon: not till we have hierarchy

[09:55] tehKellz (kelly.linden): 5 minute warning

[09:55] Kaluura (kaluura.boa): Before it's too late... SCR-79

[09:55] flexi campfire: http://jira.secondlife.com/browse/SCR-79

[#SCR-79] llMatchGroup() - Checks if an object or agent is active in a specified group.

[09:55] Motor Loon: gonna take more than 5 mins to get thru that list

[09:55] Motor Loon: that function would be nice

[09:56] Ima Mechanique: SCR-30 is before that ;-)

[09:56] flexi campfire: http://jira.secondlife.com/browse/SCR-30

[#SCR-30] llGive* feedback and transaction event

[09:56] Vincent Nacon: don't we have something like that, Flexi?

[09:56] tehKellz (kelly.linden): SCR-30 I need to decide how to deal with since I've done part of it (on aditi now)

[09:56] Ima Mechanique: and it's half acknolwedged ;-)

[09:56] tehKellz (kelly.linden): And there are duplicates

[09:56] Kaluura (kaluura.boa): Yep... Half-done already.... That's why I skipped it.

[09:57] Kaluura (kaluura.boa): THis group thing has been bugging me for years.

[09:58] Motor Loon: Would atleast be nice to be able for a script to check if an avatar is in a certain group

[09:58] Ima Mechanique: hehe, I'd like a llInGroup(), which just checks if the Avatar is a member of the objects group

[09:58] Kaluura (kaluura.boa): *IN* a group... Not if the group tag is active...

[09:58] Motor Loon: workaround needs bots and crap... and really a pain

[09:58] tehKellz (kelly.linden): Ima: Don't we have that already?

[09:58] tehKellz (kelly.linden): llSameGroup or something?

[09:58] Object: Hello, Avatar!

[09:58] Liisa Runo: most easy way would be to just unblock llGetObjectDetails to give group key also for agents

[09:59] Kaluura (kaluura.boa): Same group than the object...

[09:59] Ima Mechanique: not that I'm aware, only SameGroup which require you be ACTIVE

[09:59] Motor Loon: llsamegroup is useless

[09:59] tehKellz (kelly.linden): Ah the difference would be active vs. all groups they are in.

[09:59] Vincent Nacon: I don't see what's wrong with it? maybe bad coder's logic in handling that?

[10:00] Ima Mechanique: well only the one the object is set to, but yes it would check against ANY group they're members of

[10:00] Kaluura (kaluura.boa): The sim already knows that... It knows where we can rez, where we can enter... Even with not tag over our head... That's what a script needs to know too.

[10:00] Ima Mechanique: the MatchGroup function would mak eit redundant though

[10:00] Motor Loon: why not take it a step further and let us check against any groupkey and not just what the object is currently set to

[10:00] tehKellz (kelly.linden): SCR-79 wouldn't solve that as described since it only checks the active group as well

[10:01] tehKellz (kelly.linden): Just because the sim knows doesn't mean other people should know.

[10:01] Ima Mechanique: Motor, there are privacy concerns about that extra step ;-)

[10:02] tehKellz (kelly.linden): I think this is a thorny issue because of the strong desire for privacy around SL and things in SL. I can definitely see people not wanting others to know some of the groups they are in.

[10:02] Motor Loon: Nothing I'd be concerned about °?°

[10:02] Kaluura (kaluura.boa): Still, that a bit stupid that an object in the group to which ppl belong can't know that they belong...

[10:02] tehKellz (kelly.linden): And there is stricter requirements for an object being in a group than just knowing the UUID of the group.

[10:02] Kaluura (kaluura.boa): No privacy concern here... You're belong or not.

[10:03] Kaluura (kaluura.boa): -re

[10:03] Ima Mechanique: yes Kaluura, which is why I'd propose there isn't a privacy issue as they 'both' belong to the same group ;-)

[10:03] tehKellz (kelly.linden): The check for whether or not they are in (as opposed to have active) a group that matches the objects group seems a bit easier.

[10:03] tehKellz (kelly.linden): But that isn't SCR-79

[10:03] Vincent Nacon: what I don't understand what do you need it for? not securiy check, right? because you should have copy of those UUID check anyway

[10:04] Motor Loon: can we edit SCR-79 to fit our "needs" ? °?°

[10:04] Kaluura (kaluura.boa): All the jira about groups turn around the same problem... This is the right answer, whatever this jira says...

[10:04] Ima Mechanique: Vincent, because that list of UUIDs can get VERY long and out of date

[10:04] Vincent Nacon: ...and?

[10:04] Vincent Nacon: that's.... good?

[10:05] Vincent Nacon: I mean group don't change

[10:05] tehKellz (kelly.linden): All right, 5 minutes past time!

[10:05] tehKellz (kelly.linden): Thanks for coming out everyone.

[10:05] Vincent Nacon: later Kelly!

[10:05] Motor Loon: damn he noticed

[10:05] Koli Contepomi (sahkolihaa.contepomi): See you kelly & Charlie.

[10:05] TheBlack Box: see you

[10:05] Ima Mechanique: group membership does change though

[10:05] Vincent Nacon: oh come on now, he's only human

[10:05] TheBlack Box: thanks for the oh

[10:05] tehKellz (kelly.linden): Next week is Turkey week in the US. I think I'll still be here monday though.

[10:05] Ima Mechanique: both people joining and leaving

[10:05] Koli Contepomi (sahkolihaa.contepomi): Turkey week, nice way to put it. xP

[10:06] tehKellz (kelly.linden): And Cranberry Sauce week.

[10:06] Vincent Nacon: Ima, I just don't understand what you're trying to do or gain from it