Simulator User Group/Transcripts/2012.03.20

From Second Life Wiki
< Simulator User Group
Revision as of 10:56, 21 March 2012 by Andrew Linden (talk | contribs) (Created page with "Simulator_User_Group {| | Prev 2012.03.16 | Next 2012.03.23 |} == List of Spe…")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search


Prev 2012.03.16 Next 2012.03.23

List of Speakers

Chieron Tenk Davido Chrome Draconis Neurocam
Fancy Detector Flip Idlemind Ima Mechanique
Kallista Destiny Kaluura Boa Kelly Linden
Lorca Linden MartinRJ Nalates Urriah
Qie Niangao Rex Cronon Sahkolihaa Contepomi
Simon Linden Talarus Luan Vincent Nacon


[12:00] Meeter: Welcome to the Server User Group

[12:00] Vincent Nacon: but the physic shape is basic enough to sit

[12:00] Sahkolihaa Contepomi: Now to toy with this setting that adjusts when the viewer automatically rerenders objects that cause long stalls on the render thread.

[12:00] Sahkolihaa Contepomi: derenders*

[12:01] Simon Linden: ok, le's get rolling ... Andrew is on vacation today, so I'll be the Master of Ceremonies

[12:01] Rex Cronon: hello everybody

[12:01] Vincent Nacon: aww ok, I'm going home

[12:01] Simon Linden: The rollout news is that there isn't a main rollout this week

[12:01] Vincent Nacon: :P (go ahead, just kidding)

[12:01] MartinRJ: hehe

[12:02] Simon Linden: The RC channels will all get the same region crossing code ... AFAIK this is still the under-the-hood re-work and won't have any noticiable change

[12:02] Chieron Tenk: hi all

[12:02] Rex Cronon: hi

[12:02] MartinRJ: hello Rex

[12:03] Rex Cronon: hello martinrj

[12:03] Fancy Detector: Kelly Linden has arrived!

[12:03] Sahkolihaa Contepomi: Hey Kelly.

[12:04] MartinRJ: Hello Kelly

[12:04] Simon Linden: Andrew told me he had no progress to report on the viewer / LSL syntax work

[12:04] Fancy Detector: Lorca Linden has arrived!

[12:04] Rex Cronon: hello kelly

[12:04] Vincent Nacon: ah nice of ya to come, Lorca

[12:04] Lorca Linden: hey all!

[12:04] Rex Cronon: hi lorca

[12:04] Simon Linden: And ... (drumroll please) Lorca is here to talk a bit about a developent viewer

[12:05] Lorca Linden: so for those of you who have been folliwing pathfind

[12:05] Lorca Linden: ing

[12:05] Vincent Nacon: oh it's a viewer meeting now?

[12:05] Lorca Linden: you'll know that there are many things you have been unable to do on the test grid without the viewre tools

[12:05] Lorca Linden: well it affects server too

[12:05] Lorca Linden: we need server side testing

[12:05] Lorca Linden: i'll be quick with my update

[12:05] Vincent Nacon: no rush

[12:05] Lorca Linden: i just wanted you all to know that you can now get the project viewer which will allow you to generate your own server side nav meshes

[12:05] Lorca Linden:

[12:06] Rex Cronon: nice:)

[12:06] Lorca Linden: so please help us test on aditi with these tools

[12:06] Lorca Linden: it also has the ABCD material settings that allow for different pathfinding speeds

[12:06] Davido Chrome: Lorca, Nyx mentioned Physics changes. I have set up a simball court to test. what did he mean by that?

[12:07] Lorca Linden: well we've had to make changes to the physics librareis and updgrade the havok AI rversion

[12:07] Vincent Nacon: can you tell us a bit more about ABCD? how are they different and how it formed

[12:07] Lorca Linden: to get pathfinding working

[12:07] Lorca Linden: so it will be helpful to test physics heavy builds on the pathfinding regions

[12:07] Davido Chrome: What about a physics heavy game?

[12:07] Lorca Linden: ABCD basicaly allows you to set up to 4 different speeds for characters

[12:07] Lorca Linden: most definitely alvekatt

[12:07] Lorca Linden: in the next few days we'll be putting up some more interesting regions

[12:08] Vincent Nacon: but it's not about shortest path and the longest, path, right?

[12:08] Lorca Linden: but the ABCD basically will let you create 4 different types of characters

[12:08] Lorca Linden: that will follow different pathfing speeds

[12:08] Lorca Linden: so cars might go slowly on grass (and try to avoid that in pathing, all things being equal)

[12:08] Davido Chrome: Lorca, I would kind of need a parcel in those sim, on the one that has good Sim FPS, where I can build a court that won't get returned on me. They take a while to set up.

[12:08] Lorca Linden: but goats could go quickly on grass

[12:09] Lorca Linden: alvekatt, we will be creating a sandbox in the next day or two

[12:09] Lorca Linden: where things do not get returned regularly (on aditi)

[12:09] Lorca Linden: next to the existing pathfidning test regions

[12:09] Lorca Linden: so if you have any other questions

[12:09] Rex Cronon: can a lsl script determine over what kind of terrain it is over? grass, sand, snow...

[12:09] Lorca Linden: feel free to email me at

[12:09] Lorca Linden: it can with the ABCD settings cron

[12:10] Simon Linden: Who should people contact if they have regions or builds that would be good to test?

[12:10] Davido Chrome: Well, where do I report problems I run intò?

[12:10] Vincent Nacon: I seconded Rex's quesiton

[12:10] Lorca Linden: but you would have to set grass to A and then sand to B for example

[12:10] Lorca Linden: and then different characteer types have a pathing speed (represented as a %) on each type of terrian

[12:10] Lorca Linden: more info on that is here :

[12:10] Lorca Linden: and problems should be reported in PATHBUG public JIRA

[12:11] Vincent Nacon: and how does one apply those setting on a non-pathfinding object?

[12:11] Lorca Linden: I don't want to hijack this meeting for too long, so best thing to do is read over our FAQ

[12:11] Lorca Linden: we will also be starting a pathfinding office hours

[12:11] Lorca Linden: test out the new project viewer, and file any bugs in pathbugs

[12:11] Lorca Linden: next week

[12:11] Vincent Nacon: ah ok good

[12:11] Lorca Linden: more info on that will be posted on the main pathfinding wiki page

[12:12] Davido Chrome: Including problems that arise that has nothing to do with pathfinding? I am testing the game not because it uses path finding, but because it relies heavily on the SL physics engine.

[12:12] Lorca Linden: good question alve : yes

[12:12] Lorca Linden: physics problems you see on pathfinding regions but not on standard ones

[12:12] Lorca Linden: are DEFINITELY something we want to know about

[12:12] Lorca Linden: even if they are not directly pathfinding related

[12:13] Lorca Linden: thanks for your interest all!

[12:13] Davido Chrome: I have allready noticed that objects with complex physics shapes seeeem to be worse at script collision detection.

[12:13] Nalates Urriah: Are u going to have an in-world group for chat?

[12:13] Lorca Linden: please file details for that on PATHBUG

[12:13] Qie Niangao: Thanks, Lorca. Very interesting stuff.

[12:13] Lorca Linden: alve

[12:13] Davido Chrome: Gotcha, Thank you.

[12:13] Lorca Linden: no problem, thank you all!

[12:13] Davido Chrome: Uh wait

[12:13] Vincent Nacon: take care

[12:13] Simon Linden: Thanks for coming by

[12:14] Davido Chrome: PATHBUG, is that just a tag in the regular Jira?

[12:14] Vincent Nacon: save it for later, Alve

[12:14] Lorca Linden: np, yup, if you have any issues finding it alvekatt or other questions just email me directly

[12:14] Davido Chrome: Ok.

[12:14] Lorca Linden: bye all

[12:14] Vincent Nacon: muhaha!

[12:14] Rex Cronon: tc lorca

[12:14] Simon Linden: It's a project you select when making the new jira

[12:14] Davido Chrome: Ok. :)

[12:14] Simon Linden: Like VWR or SVC or such

[12:14] Davido Chrome: Ah. Then I know.

[12:15] Talarus Luan: There is no PATHBUG project... yet

[12:15] Vincent Nacon: it's like watching a man trying to speed things up cause he missed his change to join a restroom break

[12:15] Talarus Luan: Unless it is private

[12:15] MartinRJ: I think there is

[12:15] Vincent Nacon: chance*

[12:15] Talarus Luan: Looking at the projects page rightnow

[12:15] MartinRJ: I've seen a pathbug Jira

[12:15] Talarus Luan: Oh wait there it is

[12:15] MartinRJ: If I'm not totally nuts

[12:15] Nalates Urriah: Me too.

[12:15] Vincent Nacon: yeah it's there

[12:15] Simon Linden: Kelly - did you have any news or topic?

[12:15] Talarus Luan: Was looking at the name field

[12:16] Nalates Urriah: try starting a new JIRA and se if it is in the drop down..

[12:16] Kelly Linden: I have one brief item I could discuss

[12:16] Talarus Luan:

[12:16] Vincent Nacon: same way as you report for havok related or mesh

[12:16] Vincent Nacon: shoot Kelly

[12:16] Kelly Linden: I mentioned at the scripting UG yesterday that I had completed llGetAgentList for an upcoming maint-server: not the next maint-server but probably the one after that.

[12:17] Talarus Luan: Ahh cool.

[12:17] Vincent Nacon: hear that, Draconis?

[12:17] Talarus Luan: There was some question in one of the scripting groups about the meaning of the constants.

[12:17] Kaluura Boa: Wow! (I missed that UG... Whyyyyyy?)

[12:17] Talarus Luan: Is there a wiki entry for it yet?

[12:17] Draconis Neurocam: yes

[12:17] Kelly Linden: list llGetAgentList(integer max_agents, integer flags): max_agents is just for script memory safety - there is no guaranteed order or randomness if you get less than the number of people there


[12:18] Flip Idlemind: ^ This made my day yesterday

[12:18] Kelly Linden: If you choose too high a max_agents ( > 200 ) or invalid flags the returned list will contain a single error string instead of any IDs.

[12:18] Talarus Luan: Where's the wiki entry, Draconis?

[12:18] Draconis Neurocam: no the yes was to vincent asking me

[12:18] Draconis Neurocam: i dont know if there is a wiki page

[12:18] Talarus Luan: Oh OK

[12:19] Vincent Nacon: sorry

[12:19] Kelly Linden: There is no wiki page for it yet. This is still quite a ways out - like I said, maint-server *after* next maint-server at best.

[12:19] Kaluura Boa: Do the returned Id include the AVs 10m or so beyond the border?

[12:19] Kelly Linden: Kaluura: no. same region only.

[12:19] Rex Cronon: we can have 200 avatars in a sim?

[12:19] Talarus Luan: Well, nothing wrong with starting a wiki page for it with a prelim warning. :D

[12:19] Vincent Nacon: no Rex

[12:19] Vincent Nacon: muhaha!

[12:20] Talarus Luan: I'd do it, but I don't remember how to start the template. :P

[12:20] Vincent Nacon: maybe on paper but not in world

[12:20] Kelly Linden: Rex: no, I could drop the limit to 100 but there is a hard coded 200 limit that no region supports.

[12:20] Ima Mechanique: Kelly what does AGENT_LIST_PARCEL_OWNER return?

[12:20] Simon Linden: We'll happily work on the 200 limit if we can ever hit it

[12:20] Flip Idlemind: Everyone on any parcel you own?

[12:20] Kelly Linden: Ima: a list of all agents on parcels owned by the same person or group that owns the parcel under the object.

[12:20] Ima Mechanique: ahh, not parcel owners then ;-)

[12:21] Rex Cronon: for a sec i thought that sims were suddenly able to support 200 ave:)

[12:21] Draconis Neurocam: haha, we have a way of calculating our own traffic now

[12:21] Talarus Luan: So, single parcel, all owned parcels in region, or entire region.

[12:21] Talarus Luan: Any restrictions on who can use it?

[12:21] Vincent Nacon: yup, advertising market is going to be happy

[12:22] Ima Mechanique: any throttle on it?

[12:22] Kelly Linden: To re-iterate: The limit is for (distant) future proofing so that in a happy future that allows a lot of agents in a single region wouldn't crash these scripts. You should use 100 or the highest number you can handle and reasonably expect to be present.

[12:22] Kelly Linden: No throttle.

[12:22] Vincent Nacon: at least have 0.2 delay for pre-caution

[12:22] Kelly Linden: nope, no delay.

[12:22] Vincent Nacon: but you sure it can't be abused?

[12:23] Ima Mechanique: delays don't work

[12:23] Simon Linden: There's no particular reason to slow this one down

[12:23] Talarus Luan: Yeah, it's all in-memory info

[12:23] Rex Cronon: so fun. robo adds jumping from sm to sim and trying to sell u things:)

[12:23] Kelly Linden: It only uses the script's time.

[12:23] Kallista Destiny: Lols

[12:23] Draconis Neurocam: it has to have no delay, because theres no way to use it fully without lots of polling

[12:23] Vincent Nacon: I meant flood of parcel request to grief sim's resource

[12:23] Vincent Nacon: hmm ok

[12:24] Flip Idlemind: Im afraid to ask this, since it may just give you ideas, but will "admin mode" Lindens be invisible to this?

[12:24] Rex Cronon: would make a nice way to kill everybody in a fighting sim:) no more searching for avatars:)

[12:24] Kelly Linden: Unlike calls like llRezObject which can effect simulator performance outside of script time, this one only takes script time.

[12:24] Simon Linden: This won't add a lot of functionality that you can't do already with sensors, but will be much, much easier

[12:24] Vincent Nacon: you want to touch Linden's god power, don't ya?

[12:24] Draconis Neurocam: even if you were to use llgetregionagentcount, to check updates, say someone entered and left at the same time, the only way you would know of the difference is to poll it again

[12:24] Kelly Linden: Admin mode linden's will not be visible - same as how llSensor works

[12:25] Flip Idlemind: Aww but then my greeter won't work

[12:25] Kelly Linden: :)

[12:25] Davido Chrome: Can you get agent positions?

[12:25] Flip Idlemind: Yeah with llGetObjectDetails

[12:25] Qie Niangao: right

[12:25] Draconis Neurocam: what flip said

[12:25] Vincent Nacon: but yeah, don't even bother hoping to bug Lindens with stuff like that. some poor linden get harassed everyday

[12:26] Talarus Luan: Kelly, is there a JIRA feature request link for it?

[12:26] Kelly Linden: There is an old jira request but the feature described there doesn't match the implementation

[12:27] Talarus Luan: OK, was just going to link any related feature requests.

[12:27] Qie Niangao: maybe

[12:27] JIRA-helper: [#MISC-243] Create llAgentSensor to list agents in whole region

[12:28] Qie Niangao: or

[12:28] JIRA-helper: [#SVC-5488] llGetRegionAgents() -- returns list of agent keys in the current region

[12:28] Kelly Linden: The SVC one is the one I imported / worked off of.

[12:28] Kelly Linden: Consider it 'official'

[12:29] Vincent Nacon: llGetRegionAgent does sounds like a better name for it

[12:29] Flip Idlemind: Well I hate to get greedy but as long as Kelly's giving us stuff

[12:29] Flip Idlemind: SCR-91 ?

[12:29] JIRA-helper:

[#SCR-91] llRequestAgentKey (llName2Key)

[12:29] Vincent Nacon: oy you had to ask?

[12:30] Meeter: Timecheck : User Group is half over

[12:30] Davido Chrome: This function hasn't hit the wiki yet?

[12:30] Kaluura Boa: One of the Jiras ask for some sorting... Is there some sorting in llGetAgentList()?

[12:31] Kelly Linden: There is no human-usable sorting applied to the results of llGetAgentList

[12:31] Vincent Nacon: it'd have to be llUsername2Key() btw

[12:31] Flip Idlemind: Would it?

[12:31] Flip Idlemind: Could it be Legacy Name to key?

[12:31] Vincent Nacon: but yeah.... name is shorter I guess

[12:31] Vincent Nacon: can't use display name anyway

[12:32] Draconis Neurocam: legacy name to key would be longer, username is the only one that won't return resident and is not temporary

[12:32] Qie Niangao: (personally, in place of SCR-91, I'd settle for a dataserver event API to Search)

[12:32] Flip Idlemind: In other words could you get my key with "Flip Idlemind" or would it have to be "flip.idlemind"?

[12:32] Flip Idlemind: (I get why "F L I P" won't work)

[12:33] Vincent Nacon: I seriously doubt we're gonna have last name options at sign up ever again

[12:33] Draconis Neurocam: yeah qie i agree that makes the most sense

[12:33] Simon Linden: That's interesting Qie ... I'm not sure how things would be returned, however. LSL is pretty messy with complex data

[12:33] Qie Niangao: true

[12:33] Vincent Nacon: hmm interesting for sure

[12:34] Simon Linden: Vincent - someone told me there was a follow-up blog post or something about the last name options, and the future didn't look good

[12:34] Simon Linden: but my info is second hand

[12:34] Vincent Nacon: I know, that's why I have a doubt about it

[12:35] Vincent Nacon: so i think user name is here to stay

[12:35] Draconis Neurocam: i still would like to know what the second and third last name ever were

[12:35] Draconis Neurocam: oskar would only say the first and fourth

[12:36] Vincent Nacon: must be a joke name or something

[12:36] Ima Mechanique: or someone's actual name, like Rosedale

[12:36] Vincent Nacon: muhaha!

[12:37] Nalates Urriah: Simon, Rod basically said its not going to happen, last names...

[12:37] Simon Linden: Right, the user name is definitely here to stay. The product guys and execs know people are asking for the old-style accounts, so hopefully they'll get it onto a work list at some point. But the news is "don't hold your breath"

[12:37] Simon Linden: oh well

[12:37] Rex Cronon: or batman, joker...

[12:37] Kallista Destiny: lols

[12:37] Vincent Nacon: well simon, the only reason to bring back last name options is to start dumping non-active accounts

[12:37] Nalates Urriah: We are going to get - and _ for user names...

[12:38] Simon Linden: ok ... well, the table is definitely open at this point ... any topics, bugs, etc?

[12:38] MartinRJ: yes

[12:38] MartinRJ: VWR-28609

[12:38] JIRA-helper:

[#VWR-28609] Please add PRIM_NAME and PRIM_DESC to keywords.ini

[12:38] MartinRJ: (it's on the agenda: )

[12:38] Draconis Neurocam: haha

[12:39] Simon Linden: I think that falls under Andrew's work to make the viewer LSL definition files available for downloading

[12:39] Draconis Neurocam: it does

[12:39] MartinRJ: and VWR-8761

[12:39] JIRA-helper:

[#VWR-8761] Cannot delete object description.

[12:39] Simon Linden: I'm not sure where exactly he is, but he told me yesterday he did not have any progress to report on it

[12:39] MartinRJ: which is a server-side problem I think

[12:40] Flip Idlemind: Oh and you also can't change the description of an attachment's root prim

[12:40] Flip Idlemind: It resets when you detach and reattach it

[12:40] Kallista Destiny: One of the problems that I see is that people only btch if they are unhappy. Example V2 came out with the sidebar, lots of b=people bitched loudly about it. Firestorm just came out with their FUI release with no Sidebar, guess what a non trivial number of people started bitching about no Sidebar.

[12:40] Rex Cronon: i had similar problems with description but now recently

[12:41] Rex Cronon: but not*

[12:41] Object: Hello, Avatar!

[12:41] Flip Idlemind: My problem with the removal of the sidebar was the same as my problem with the addition of the sidebar: sudden, abrupt change

[12:41] Simon Linden: One part of VWR-8761 is a Pandora's Box: "If you remove the description via script llSetObjectDesc("") it doesn't update the description while it's still selected - so it stays at the old value until you re-select it inworld."

[12:41] Qie Niangao: Kallista: so we should start bitching about things that make us happy?

[12:42] Talarus Luan: <.<

[12:42] Simon Linden: I discussed a similar bug yesterday with Maestro - basically we don't have support for simultaneous multiple edits on objects.

[12:42] Kallista Destiny: No... but perhaps we should say thank you , or this is a really good idea. Every once in a while.

[12:42] Davido Chrome: Oh yeah, is that bug with link number not updating unless you take an object into inventory fixed?

[12:42] Flip Idlemind: I still say the best idea would have been an optional sidebar

[12:42] MartinRJ: the main problem it seems, I believe is that you cannot delete it manually at all, imho

[12:42] Vincent Nacon: when pandora happens... make a rule

[12:42] Flip Idlemind: Like there is in my viewer ;D

[12:43] Simon Linden: For example, if you started changing the description in normal editing, and a script changes it, which one wins? It's a tough problem that goes pretty deep

[12:44] Rex Cronon: u get race condition

[12:44] Simon Linden: But the change to null seems like something that should work

[12:44] Chieron Tenk: one cannot delete it by script, either

[12:44] Rex Cronon: u can make it atomic

[12:44] Simon Linden: I wonder if the viewer is sending up the data or not

[12:44] Davido Chrome: Simon, a lot of bike builders use separate scripts for wheels because New objects you add to linksets doesn't get their number added at the end.

[12:44] Object: Hello, Avatar!

[12:45] Davido Chrome: Which means, every time you cahnge something, you need to go through the script and update the link numbers, or something else than the wheels will start spinning...

[12:45] Draconis Neurocam: alvekatt, why cant they just loop through and find the proper name or description for both wheels?

[12:46] Draconis Neurocam: and use those links

[12:46] Davido Chrome: Draconis, some do that. But shouldn't it be more efficient if you don't need to do that everytime you change the number of prims?

[12:46] Simon Linden: That sounds ugly, AlveKatt, but changing the link order there seems scary ... I can imagine other things are built expecting the current behavior

[12:47] Vincent Nacon: most vehicle use another set of scripts for each wheels because of collision handle and omega target

[12:47] Vincent Nacon: and some other things that can't be set by Link function

[12:47] Davido Chrome: Well, yeah. omega in llSetLinkParams isn't as Local as it should be I think...

[12:48] Draconis Neurocam: you can set link target omega now vincent

[12:48] Draconis Neurocam: ah

[12:48] Vincent Nacon: yeah I remembered now but I meant there are others function as well

[12:48] Vincent Nacon: but yeah

[12:48] Qie Niangao: I don't think it's any more or less local in a child-script prim, is it?

[12:48] Flip Idlemind: So, I had an idea for a function that I think could be useful, but I'm not 100% sure it would be useful enough to actually implement :/

[12:49] Davido Chrome: I tried to use Omega for spinning, and link rotation for turning once, wasn't pretty.

[12:49] Simon Linden: What's that, Flip?

[12:49] Vincent Nacon: as well different Timer event

[12:49] Flip Idlemind: And Im not sure it would be worth making a JIRA for

[12:49] Draconis Neurocam: if its as finicky as particle omegas i could see an issue

[12:49] Flip Idlemind: AND, Im not even 100% sure it's possible (without server-side haxxy ness)

[12:50] Davido Chrome: Still, i think the bug whee you have to take the object into inventory to update links affects the loop scripts too.

[12:50] Draconis Neurocam: not if you are going by some prim parameter to define them alvekatt

[12:50] Vincent Nacon: we still need object link in hierarchy system

[12:51] Davido Chrome: I guess you can do it with some Fancy math, but I can't remember that part from school. :(

[12:51] Flip Idlemind: Ok, this is what I would have put: "llSetObjectVisibility"

[12:51] Flip Idlemind: "Currently there is no easy way to make a linkset completely transparent, and then return it to its previous visibility, without first checking and saving each prim's alpha value. llSetObjectVisibility(0) would make the linkset completely transparent, and llSetObjectVisibility(1) would bring each prim back to its original alpha value."

[12:52] Vincent Nacon: huh?

[12:52] Rex Cronon: u set a script for each prim...

[12:52] Draconis Neurocam: flip it sounds like it would just be a list of floats in the script , or an array of floats on the server, and which is more efficient

[12:52] Kallista Destiny: llSetObjectVisibility( bool value)

[12:52] Kelly Linden: The work required is the same, somewhere. Why should the sim itself be tracking previous alpha values for the few cases that need the feature, instead of offloading that work to the few scripts that need it?

[12:52] Davido Chrome: Baically wants the script to save the initial transparancy settings.

[12:53] Simon Linden: People would ask for more than a boolean transparency value too

[12:53] Vincent Nacon: I think I'm confused... don't we have llGetAlpha or the like of it?

[12:53] Draconis Neurocam: he wants there to be extra storage space vincent to save a preset value, via a function

[12:53] Kelly Linden: There comes a point, I think, where the feature request is to have the sim do the script for you and while that may make sense in some cases where the work is significant I don't think at first glance that this is suck a case.

[12:53] Vincent Nacon: ...

[12:53] Davido Chrome: Can't you use LLGetLinkPrimParams?

[12:54] Kelly Linden: er. such a case.

[12:54] Davido Chrome: For this?

[12:54] Flip Idlemind: You can use it, but the point is you actually have to use it

[12:54] Talarus Luan: Yeah, but you will have to fiddle with separating the color from it, and saving it.

[12:54] Talarus Luan: Very messy

[12:54] Draconis Neurocam: i agree with kelly, especially with the speed in list operations now it would be failry trivial and usable everywhere once you made your own function

[12:55] Davido Chrome: Talarus, write a custom function and put it in the Wiki?

[12:55] Meeter: Timecheck : User Group is almost over

[12:55] Kallista Destiny: well treat it as a ROT instead of a triplet.

[12:55] Vincent Nacon: well I got better idea than storing some data in alpha.... give a prim 3 value slots that only accessed by LSL?

[12:55] Talarus Luan: Then when you want to set it back, you have to read it again, and set the color and alpha again.

[12:55] Talarus Luan: Unless we have... dare I ask it? :P

[12:56] Flip Idlemind: However...what I had in mind for "llSetObjectVisibility" wouldn't even actually modify the prim's alpha

[12:56] Kallista Destiny: llSetObjectColorAlpha( rot value);

[12:56] Flip Idlemind: They would all stay the same, the object would just "pretend" to be completely invisible

[12:56] Talarus Luan: If all you are wanting is a global alpha for a linkset, I think that would have to be a big change that would go all the way to the viewer.

[12:56] Flip Idlemind: So no storing of initial values would be required

[12:57] Draconis Neurocam: llGetLinkAlpha would help surely in flip's case, because grabbing it is a different value than setting it

[12:57] Vincent Nacon: it seem like a silly way to deal with data storage like that

[12:57] Simon Linden: If you're thinking about a new display parameter like that, it ends up requiring changes in the data storage, messaging, the sim and viewer. It's fairly costly so the bar is pretty high

[12:58] Flip Idlemind: Hmm maybe

[12:58] Vincent Nacon: so why not just add 3 value input in a prim?

[12:58] Vincent Nacon: same way like description

[12:59] Vincent Nacon: but only accessible by LSL

[12:59] Flip Idlemind: But if you could do it, basically you could make the sim "lie" to me and tell me all the prims are invisible, even though their actual alpha values wouldn't change

[12:59] Draconis Neurocam: you could use the key space for collisionsprite :P

[12:59] Flip Idlemind: Like I said, not sure it would be worth it, but it's an idea

[13:00] Simon Linden: Right -- it sounds like a lot of work compared to what we'd get, that's all

[13:00] Meeter: Thank you for coming to the Server User Group

[13:00] Draconis Neurocam: flip I think you make a valid case for llGetLinkAlpha

[13:00] Vincent Nacon: I know what you're trying to do and I support it but not with that method

[13:00] Kallista Destiny: Changes to the database are a big deal...

[13:00] Vincent Nacon: yeah, one way or another

[13:01] Davido Chrome: Are we talking about not streaming graphics of invisible link objects here?

[13:02] Flip Idlemind: No I think that would cause problems

[13:02] Simon Linden: I have to get going ... thanks everyone for coming today

[13:02] Flip Idlemind: More like, streaming a "fake" alpha value

[13:02] Flip Idlemind: Baiii

[13:02] Davido Chrome: Thank you!

[13:02] Rex Cronon: tc simon

[13:02] Vincent Nacon: ah alrighty

[13:02] Qie Niangao: Thanks Simon, Kelly.


Prev 2012.03.16 Next 2012.03.23