Simulator User Group/Transcripts/2014.01.14

From Second Life Wiki
Jump to navigation Jump to search


Prev 2014.01.07 Next 2014.01.21

List of Speakers

Dante Spectre Duckie Dickins Eddi Decosta
Ima Mechanique Jenna Felton Jonathan Yap
Kallista Destiny Kelly Linden Lexbot Sinister
Lucia Nightfire Marianne McCann Mona Eberhardt
Motor Loon Nalates Urriah Qie Niangao
Rex Cronon Simon Linden Tankmaster Finesmith
Tiberious Neruda


[12:00] Simon Linden: Let's see ... news this week is pretty short, since we didn't update anything

[12:00] Simon Linden: From what I can tell, we'll promote the code in the RC channels next week to the full grid

[12:01] Simon Linden: We're like to get new code next week on Wednesday but we haven't nailed down what yet

[12:01] Simon Linden: ... and that's about it for news

[12:02] Simon Linden: So the table's open for questions, topics, ideas, and if it gets dull, rants

[12:03] Mona Eberhardt: Are there any plans for a method to store persistent information (such as a script's settings) in-world using something else instead of an object description or hovertext?

[12:04] Mona Eberhardt: Hi Mari

[12:04] Marianne McCann waves

[12:05] Simon Linden: Not really, Mona, but we've discussed it before .... we never sorted out a design that seemed good without needing more work than we could take on

[12:05] Mona Eberhardt: Also, are there any plans for a standardised method for menu creation that will make things easier for scripters?

[12:05] Jenna Felton: i support that request too, and with json we had a nice format of how the information could be structured :)

[12:05] rcdsQueChatLog: Rex Cronon, you can read the log here:

[12:06] Rex Cronon: hello everybody

[12:06] Mona Eberhardt: Hi Rex

[12:06] Tankmaster Finesmith: your late :P

[12:06] Lucia Nightfire: yes, late Rex is late

[12:06] Kelly Linden: The json feature set was born from a desire to control / spawn viewer UI from LSL. I don't think progress has been made on that front however.

[12:06] Rex Cronon: netw

[12:06] Rex Cronon: network problems:(

[12:07] Rex Cronon: hi mona, tankmaster, lucy

[12:07] Ima Mechanique: Any chance of this getting implemented?

[12:07] JIRA-helper: [#STORM-2006] Values for brush constants in call to llModifyLand are off by 1 - Second Life Bug Tracker

[12:07] Jenna Felton: with json you can create keyword = value pairs to store it serialized in a string field, or keep it in a sql database.. this is why i mention json here :)

[12:07] Kelly Linden: I haven't figured out why that jira is in STORM and not BUG

[12:08] Ima Mechanique: probably just habit ;-)

[12:08] Qie Niangao: One like it was in the jira since, like, 2006

[12:08] Kelly Linden: Well there is always a chance.

[12:08] Lucia Nightfire: how about 1m - 10m brush and the ability to specify a land height for output instead of a blind strength, heh

[12:08] Jenna Felton: BTW, About the menus, i had created a jira before .)

[12:08] Jenna Felton:

[12:09] Jonathan Yap: Ima, why can't you do that as part of your changes?

[12:09] Ima Mechanique: even if it is a snowball in hell's chance

[12:09] Ima Mechanique: it's server side Jonathan

[12:09] Ima Mechanique: those are the values that LSL engine uses

[12:09] Simon Linden: ugh, I wonder how much would break when we fix that

[12:09] Jonathan Yap: Not if you just use 0 1 and 2 for the new values of the constants

[12:09] Kelly Linden: Script compilation, including turning constants into their values, is server side.

[12:09] Jonathan Yap: ahhh

[12:09] Ima Mechanique: that's why the request for NEW constants Simon

[12:10] Kelly Linden: Yes, it would require new constants.

[12:10] Jonathan Yap: are depreciate the current ones

[12:10] Jonathan Yap: *and

[12:10] Simon Linden: doh, yes ... yeah, that really should be a BUG and get imported

[12:10] Kelly Linden: probably LAND_BRUSH_2x2, LAND_BRUSH_4x4 and LAND_BRUSH_8x8

[12:10] Simon Linden: it doesn't belong in STORM

[12:10] Lucia Nightfire: llModifyLand() depending on the land owner being online or in the region also sucks

[12:10] Ima Mechanique: the old constants are LAN_size_BRUSH, the request is for new LAND_BRUSH_size

[12:10] Jonathan Yap: We thought it belonged in storm yesterday

[12:10] Lucia Nightfire: esp if its the land owner's script doign teh call

[12:11] Mona Eberhardt: Have you guys started considering adding LSL capability for controlling normal and specular maps?

[12:12] Tankmaster Finesmith: ask alexia to move it...

[12:12] Simon Linden: Yes Mona, it's on my list of things to check out

[12:12] Mona Eberhardt: it could be PRIM_NORMAL and PRIM_SPECULAR

[12:12] Mona Eberhardt: The rest should function in a similar manner to PRIM_TEXTURE.

[12:12] Qie Niangao: let's don't forget alpha-masking vs alpha-blending of the diffusemap, too.

[12:12] Qie Niangao: speaking of which... (This may be more viewer than server) Would it make sense for particle textures to be able to use alpha-masking?

[12:13] Kelly Linden: I don't like LAND_size_BRUSH -> LAND_BRUSH_SIZE - it is too opaque about what the change is and half of scripters will guess wrong.

[12:13] Simon Linden: right, it matches up pretty well to existing functions, so my initial impression was it's do-able ... given the time to get to it

[12:13] Jonathan Yap: Kelly, will you move that jira and edit it to what you think is a good plan?

[12:14] Kelly Linden: Yap: I've already talked with Oz about it this morning, and simon just now, it will get sorted.

[12:14] Jonathan Yap: Ima is working on the viewer changes for importing the lsl syntax

[12:14] Mona Eberhardt: Also, Simon - and Marianne just reminded me... Have you noticed that there's something wrong with's security certificate? It says it's invalid and that it's only valid for and *

[12:14] Kelly Linden: this is really, really, unrelated to the lsl syntax file changes

[12:14] Mona Eberhardt: *^

[12:14] Ima Mechanique: Kelly, I prefer your suggestion of using the actual size. Make it clear what the effect should be

[12:15] Simon Linden: No Mona, I haven't but I'll pass that along to someone

[12:15] Mona Eberhardt: It started happening a while ago. Must have been an hour or two. Perhaps less.

[12:16] Nalates Urriah: Chrome is blocking because of https and the domain redirect that is happening now... no profile info available in viewer or via web site.

[12:16] Tankmaster Finesmith: an hr or two is a while ago??

[12:16] Kelly Linden: Qie: Have you played with the PSYS_PART_BLEND_FUNC_* parameters?

[12:17] Tiberious Neruda: you know, those blending funcions are damn cool... any chance we may get to use those on prims too?

[12:17] Qie Niangao: Oh, I have a bit, actually, but that's a good point... that might cover it, maybe... I hadn't thought about that.

[12:17] Mona Eberhardt: Well, Tank, I noticed the issue twenty minutes ago, but I hadn't checked the site for about 6 hours. So, I'm guessing here. :/

[12:17] Ima Mechanique: Tank, if you NEED to shop, a couple hours is a long while

[12:17] Kelly Linden: Tiberious I don't know.

[12:17] Mona Eberhardt: So, I'm not sure if I'm one of the first users to notice it or if it's been around for longer.

[12:18] Marianne McCann: isn't the shopping, but the social media. The spot with the busted snapshot gateway

[12:19] Tiberious Neruda: can imagine prims that could invert all the colors behind it

[12:20] Qie Niangao: Tiberious, you haven't found a magic water-hiding BLEND_FUNC have you??

[12:20] Kelly Linden: The cert issue has been forwarded on, thanks for bringing it up.

[12:20] Tiberious Neruda: nope

[12:20] Mona Eberhardt: You're welcome, Kelly.

[12:20] Tiberious Neruda: I've only played with it a little bit

[12:21] Qie Niangao: I don't think it exists, but we need "anti-water", now that invisiprims are gone.

[12:21] Motor Loon: hear hear

[12:21] Lexbot Sinister: ^hear hear!

[12:21] Tiberious Neruda: agreed

[12:21] Lexbot Sinister: new boats are starting to look ridiculous :(

[12:21] Marianne McCann: Yes, Qie, agreed

[12:21] Mona Eberhardt: Hiding water... For the part of ship hulls that's under the waterline?

[12:21] Qie Niangao: and tunnels

[12:22] Marianne McCann: Also would be good for caves, etc

[12:22] Lexbot Sinister: For anything thats under water, more or less. Underwater builds.

[12:22] Marianne McCann: drydocks

[12:22] Mona Eberhardt nods

[12:22] Jenna Felton: i would miss the invisprims :)

[12:22] Lexbot Sinister: Dungeons and Abyss bubbles!

[12:23] Eddi Decosta: hi there

[12:23] Tiberious Neruda: now that we can do translucent and shiny with materials, I don't see any more uses for invisiprims

[12:23] Simon Linden: hmmm ... that would be cool but would require some interesting viewer rendering code to exclude it from a volume

[12:23] Lexbot Sinister: Would there be any way to implement that? " Do not render water and water effects here" ?

[12:23] Rex Cronon: hi

[12:24] Tiberious Neruda: that's fine. Being able to exclude and include water is something we would really like and would make great use of

[12:24] Marianne McCann: Agreed, Tiberous. I can think of more than a couple good uses for both

[12:25] Lexbot Sinister: The prim volume is only a way to give position and parameters.

[12:25] Tiberious Neruda: right

[12:25] Motor Loon: Just put the old invisi texture back into the viewer, and fix the problem is was for casting shadows (which was why it was disabled as I understand)

[12:26] Kelly Linden: profile cert issue should be fixed now.

[12:26] Marianne McCann: The ability to fill a pond, river, or swimming pool at higher than ground elevation would be lovely, andbeing able to subtract for tunnels under water level would be a wonderful addition

[12:26] Tiberious Neruda: and if you REALLY wanted to go whole-hog on the idea, change it so that it throws scriptable events when an object collides with water

[12:26] Simon Linden: Right, it seems the best way to define that space would be large phantom objects that don't really do anything else. So there's the question of how to mark things like that, all the UI and editing, then the ability to render it appropriately

[12:26] Tiberious Neruda: or stops colliding with water

[12:26] Marianne McCann: Tiberious - for a potential "splash," an AO "swim" trigger, etc?

[12:26] Tiberious Neruda: yes

[12:26] Tankmaster Finesmith: invisprims have other issues esides shadows

[12:27] Tiberious Neruda: currently, water checks are on timers usually, and kinda intensive

[12:28] Tiberious Neruda: if you had a 'water_in' and 'water_out' event... you could move swimming or buoyancy-related functions to those events and save sim strain

[12:28] Lexbot Sinister: i'm not entirely sure what water in SL is >.< i mean, in comparison to all the other things... Does it have any other properties even than height/depth?

[12:28] Tankmaster Finesmith: different windlight settings

[12:28] Nalates Urriah: And there is that thing of what casts a shadow and how far... So, a skybox roof could shade the skybox but nor the grounds hundreds of meters below. Seems water render and shadow render may have some things common. And then there is particle rain/snow.... lots of things that are volume related...

[12:29] Jenna Felton: that are actually interesting events Tiberius

[12:29] Jonathan Yap: Does the physics engine have any water properties that are currently unused?

[12:29] Jenna Felton: 'water_in' and 'water_out'

[12:30] Tiberious Neruda: well, those were just abstract names to show when something has entered/exited the water

[12:30] Lexbot Sinister: Thats what i wonder though... if water only has the property of "waterline happens on x meters" which is static for mainland and can be adjusted on estates

[12:30] Nalates Urriah: Profiles are working again... is back....

[12:30] Marianne McCann: yay

[12:30] Kelly Linden: Yeah, I said that above but it scrolled by rather quickly

[12:31] Marianne McCann: Still not acceoting snapshots to feed, thoguh that is a likely unrelated issue.

[12:31] Mona Eberhardt: Yup, the feeds are working once again. Thanks Kelly!.

[12:31] Lexbot Sinister: So water_in and water_out could be just reading at what altitude water happens

[12:31] Marianne McCann: Thanks indeed

[12:31] Simon Linden: yeah, snapshots are still broken

[12:31] Tiberious Neruda: maybe

[12:32] Tiberious Neruda: but if we get the ability to exclude or include water with prim properties, they'd probably fire on those too

[12:32] Mona Eberhardt: There's another issue with the feeds, Simon.

[12:32] Marianne McCann: I'm curious, Simon, as to what appears to be the issue there. Seems awful sporadic and quite the headache all around.

[12:32] Duckie Dickins: as I understand it, water is part of the nav mesh and has properties as well which stop pathfinding characters as well

[12:32] Marianne McCann: I'm sure it;s a bit of "whack a mole" (but not Mole) for you

[12:32] Mona Eberhardt: The links in feed posts now get this in front of them:

[12:32] Lexbot Sinister: Btw, if we could volumetrically exclude water, the extension would be to volumetrically include water. " within thiese parameters, water DOES happen"

[12:33] Duckie Dickins: It would be nice to replace prim water with actual water. :)

[12:33] Simon Linden: the snapshots? It has to do with authentication under the hood when getting the credentials and then using them to fetch the snapshot config info

[12:33] Mona Eberhardt: So, when someone posts a link to an external website (a blog post or something), they are confronted with an error message. Namely, this:

The page you were looking for doesn't exist

If you continue to receive this message after following a link from within this site, please use the options under the viewer's "Help" menu or visit Second Life Support.

Go to

[12:33] Kallista Destiny: Even simple rectilinear volumes would be great.

[12:34] Marianne McCann: That does sound messy, Simon. A shame it's being so persnickety.

[12:35] Mona Eberhardt: Maybe it's a good opportunity for the Lab to redesign the feeds from the ground up and beef them up on all fronts (including integration with the dashboard etc).

[12:35] Simon Linden: Mona - do you have a link to a broken example? The web site is outside of where Kelly or myself work, so we're really just passing along information to others

[12:35] Mona Eberhardt: Yes Simon, I have. Go to this post:

[12:35] Marianne McCann: Indeed, Mona. Would take a few resources to do so, but a better product could potentially be crafted.

[12:36] Jenna Felton: I can see the page without erors

[12:36] Mona Eberhardt: The link given (the thingy) ends up like this:

[12:37] Mona Eberhardt: The post gives a link to a blog post - and it becomes as I mentioned... :/

[12:37] Jenna Felton: ahh, sorry,

[12:38] Jenna Felton: the link behind gives eror

[12:38] Duckie Dickins: maybe because a minus sign was included as part of the link so the site couldn't parse it correctly

[12:38] Duckie Dickins: at the beginning..

[12:38] Duckie Dickins: -http://etc etc

[12:38] Mona Eberhardt: Hm. Maybe this is the cause.

[12:39] Simon Linden: if you look at the page source, all the links include that hyphen pre-pended on "http"

[12:39] Simon Linden: That may be confusing it

[12:40] Duckie Dickins: on the other hand the web app should be smart enough to try and not parse malformed URLs

[12:40] Simon Linden: it may be a bug there in the url detection

[12:40] Simon Linden: yes, definitely

[12:41] Duckie Dickins: thanks for breaking the interweb. :D

[12:41] Duckie Dickins: we can all go home now. :P

[12:42] Simon Linden: I don't know what kind of framework or architecture it's written under ... but looking at the page in Firefox it's definitely highilghting and targeting using that extra hyphen and it really shouldn't be

[12:43] Qie Niangao: Oh, here's a puzzle for Kelly. In a SLU thread about something else, it's demonstrated ( that Mono script memory use is affected by the length of the *names* of states and functions (but not variables). I'm not sure how that's even possible, but wonder if it's somehow expected.

[12:45] Rex Cronon: i had that problem qie. each time my lsl obfuscator uses long names:(

[12:45] Simon Linden: Is it a big effect on memory? Or just the string sizes?

[12:45] Qie Niangao: I think just the symbol sizes. It's hard to tell because mem is mostly allocated in 512 b blocks.

[12:46] Lucia Nightfire: Rex, stop using obfuscation, heh

[12:46] Kelly Linden: hm.

[12:46] Rex Cronon: when i asked before about why longer names take more memory i was given a lecture of how lsl works. people didn't believe me:(

[12:47] Qie Niangao: or, heh, *optimize* the obfuscation to single-character symbols :p

[12:47] Simon Linden: yeah, down at that granularity there are all sorts of alignment and block allocation coming into effect ... it's not necessarily going to be nice and linear or predictable to the exact byte

[12:47] Tankmaster Finesmith: one_really_cool_and_long_string_name_for_rex

[12:47] Lucia Nightfire: using 48 - 64 byte local and global names is ridiculous

[12:47] Kelly Linden: lsl functions become essentially methods on a class in C# for mono scripts.

[12:48] Ima Mechanique: hehe I'd like to see that code ;-) That's hoe LSL Editor does it ;-)

[12:48] Rex Cronon: it can do that qie, but sometimes long ones can be interesting:)

[12:48] Kelly Linden: some of C#'s introspection features may come into play.

[12:49] Kelly Linden: but I'm not sure why variables would be different than functions/states.

[12:50] Qie Niangao: oh, hmm. obviously LSL can't access reflection / introspection, but that does sound a plausible reason for the symbols to be retained, behind the scenes.

[12:51] Kelly Linden: reflection/introspection is actually used extensively in supporting mono scripts internally, in the script engine.

[12:52] Qie Niangao: interesting.

[12:52] Lucia Nightfire: anythign comign down the pipe in the next few weeks?

[12:52] Tankmaster Finesmith: stuff

[12:52] Lucia Nightfire: yay

[12:52] Qie Niangao: (Thanks, Kelly.)

[12:54] Simon Linden: We don't have anything new to announce, Lucy

[12:54] Lucia Nightfire: how are group bans comign along, heh

[12:54] Tankmaster Finesmith: and whats baker up to these days? :D

[12:54] Simon Linden: The LSL syntax file work is chugging along, and the group ban work is in testing and getting close

[12:54] Rex Cronon: baker is baking code;)

[12:55] Simon Linden: Baker had some extra vacation time around the holidays so the last few weeks were not very productive, but I think I heard there was a new build today with a fix or two ... it's pretty close, I think

[12:56] Tankmaster Finesmith: good to hear

[12:57] Ima Mechanique: I have to go chug ;-)mlsl syntax merging :-(

[12:57] Ima Mechanique: bye all

[12:57] Rex Cronon: tc ima

[12:57] Jenna Felton: take care Ima

[12:58] Jenna Felton: when you have some space on TOdo list, i'd have a request for anti-grief hooks :) like the hook about taking controls, add a scripted hook to pushing the avatar

[12:58] Jenna Felton: so a script could prevent others scripts to push you around, so you would need no movelook

[12:58] Lucia Nightfire: hooks?

[12:58] Rex Cronon: btw. after this meeting i have something to say about the chat logger. so if u r interested don't leave right away...

[12:58] Simon Linden: I'm not sure what you mean ... a hook where?

[12:59] Dante Spectre: must run, take care all.

[12:59] Jenna Felton: when a script want to push you around, than it can force you. with a hook another script would get take control of that and prevent the physic engine from knowing that

[13:00] Jenna Felton: so there is no needs to push you back to position in order you stay in place

[13:00] Lucia Nightfire: but doesn't that counter what a land owner allows on a push parcel?

[13:00] Duckie Dickins: what if the grief script hooks the vehicle script moving you around with said hook.

[13:00] Simon Linden: That sounds really tough to do ... and get the API right so it can't be abused

[13:00] Jenna Felton: its about not being able to be orbited in sandboxes

[13:01] Jonathan Yap: If you are sitting, doesn't that stop that problem?

[13:01] Mona Eberhardt: Simon, has anyone looked again at SEC-1376?

[13:01] Lucia Nightfire: a flag that blocks pushes on an avatar would be nice, but that would still counter what the parcel allows

[13:01] Mona Eberhardt: The Vivox exploit that allows someone to follow you where you teleport after the session.

[13:02] Jenna Felton: that stops, but you not always want sit or it is too late to sit down

[13:02] Simon Linden: I haven't heard back on that one, Mona - I'll check if there's been any info

[13:02] Lucia Nightfire: illegal tpv's capitalize on that

[13:02] Jenna Felton: but as i said, when you have place on the todo list :)

[13:02] Lucia Nightfire: as well as other naughty things

[13:03] Simon Linden: I have to go ... thanks everyone for coming this week


Prev 2014.01.07 Next 2014.01.21