User:Andrew Linden/Office Hours/2010 09 07
Jump to navigation
Jump to search
Transcript
[11:00] | Welcome to | Linden office hours |
[11:00] | Moon Metty: | ahhh, the music already started |
[11:00] | reddot99 Republic: | linden office hours, sans the lindens, lol |
[11:00] | reddot99 Republic: | somebody try to remind andrew |
[11:01] | Jonathan Yap: | The new jira seems to finally be up |
[11:01] | Latif Khalifa: | not for everyone |
[11:01] | Jonathan Yap: | You do not see the new dashboard Latif? |
[11:01] | Latif Khalifa: | nope |
[11:02] | Imaze Rhiano: | new dashboard? |
[11:02] | Moon Metty: | me neither |
[11:02] | Jonathan Yap: | http://jira.secondlife.com/secure/Dashboard.jspa |
[11:02] | Jonathan Yap: | That is what I get |
[11:02] | Jonathan Yap: | get redirected to, that is |
[11:02] | Moon Metty: | oh |
[11:02] | Moon Metty: | vwr-123 |
[11:02] | ||
[11:02] | Moon Metty: | i see |
[11:03] | Moon Metty: | the jira helper doen't have access |
[11:03] | Jonathan Yap: | The will probably need to be reprogrammed |
[11:03] | Imaze Rhiano: | ah... jira dashboard - I can see that |
[11:03] | Imaze Rhiano: | at least it isn't spamming like in last meeting |
[11:04] | reddot99 Republic: | lol |
[11:04] | reddot99 Republic: | cant get to jira |
[11:04] | Ardy Lay: | New JIRA up yet? |
[11:04] | Moundsa Mayo: | Hi Andrew |
[11:04] | Moon Metty: | hi Andrew |
[11:04] | Jonathan Yap: | It is up for some of us |
[11:04] | Moon Metty: | ca va bien? |
[11:04] | Andrew Linden: | hello, sorry I"m late... |
[11:04] | Imaze Rhiano: | hi Andrew |
[11:04] | Moundsa Mayo: | We kept yer seat warm |
[11:04] | Andrew Linden: | I forgot to boot my Windows computer until right before office hours |
[11:04] | reddot99 Republic: | lol |
[11:04] | Jonathan Yap: | Latif said try 63.246.25.177 |
[11:04] | Andrew Linden: | and it takes several minutes to this thing to boot up |
[11:05] | Mojito Sorbet: | SL works on Linux too, ya know :) |
[11:05] | Andrew Linden: | yeah, but I sometimes crash out on linux (macpro laptop + Ubuntu 10.04) |
[11:05] | Andrew Linden: | and sometimes it is a hard crash -- something overheats and I have to reboot |
[11:06] | Andrew Linden: | Simon is on vacation today. |
[11:06] | Moon Metty: | oh yeah |
[11:06] | Moon Metty: | well deserved |
[11:06] | Andrew Linden: | Of course, I have very little news since Friday afternoon. |
[11:06] | Andrew Linden: | This morning I've been working on math utils for overhauling the linkability rules. |
[11:07] | reddot99 Republic: | i've got a question for you, one that i wanted to get in on friday |
[11:07] | Andrew Linden: | My plan is to implement a simple bounding sphere check for the collection of prim centers |
[11:08] | Andrew Linden: | The new algorithm would be prim-scale agnostic |
[11:08] | Andrew Linden: | but would be invarient under link-order |
[11:08] | Andrew Linden: | and should be much faster than the old system |
[11:08] | Mojito Sorbet: | Wouldn't that break long thin chains of linked things? |
[11:08] | Moon Metty: | the smallest sphere possible around all bounding boxes? |
[11:08] | Rex Cronon: | hello everybody |
[11:09] | Andrew Linden: | I'd compute th smallest bounding sphere for the set of prim centers (no boxes involved). |
[11:09] | Moon Metty: | oh |
[11:09] | Andrew Linden: | No, this would not break "long chains". |
[11:09] | Moon Metty: | the smallest sphere possible around all prim centers? |
[11:09] | Moon Metty: | yes |
[11:09] | Andrew Linden: | There would be just one bounding sphere limit (probably ~64 meter diameter). |
[11:09] | Imaze Rhiano: | why we need linking limits anyway? |
[11:09] | Latif Khalifa: | what number did you end with, 64m? |
[11:10] | reddot99 Republic: | i would ask that you make it smallest cube/sphere that fits arounds the objects, using a set max volume, that'd be more usefull i think |
[11:10] | Rex Cronon: | 64d dimater? not radius:( |
[11:10] | Moon Metty: | so no two centers will be further apart than 64 m |
[11:10] | Moon Metty: | or 128 m |
[11:10] | Andrew Linden: | yeah, 64m -- it has to be bigger than the current 54 m limit. |
[11:11] | Rex Cronon: | 128 is also bigger:) |
[11:11] | Andrew Linden: | plus some tolerance because the old bounding sphere calculation has some "unpradictable" inaccuracies |
[11:11] | Andrew Linden: | but I'd have to support all legacy linkable content. |
[11:11] | reddot99 Republic: | like spelling? |
[11:11] | Mojito Sorbet: | What is bad about overly large link sets? |
[11:12] | Andrew Linden: | The main reasons for linkability limits IMO are: |
[11:12] | Andrew Linden: | (1) objects don't collide across region boundaries (yet) |
[11:12] | reddot99 Republic: | can we get joints back at any point? |
[11:13] | Andrew Linden: | (2) prim-encroachment, where your neighbor's objects extend over into your parcel, and you can't do anything about it |
[11:13] | Andrew Linden: | yes, both of those problems exist right now |
[11:13] | Andrew Linden: | however to expand the limits just makes them worse. |
[11:13] | Andrew Linden: | I'd like to get those two done before expanding the linkability (if possible). |
[11:13] | Rex Cronon: | u do something about it. u file an AR:) |
[11:14] | reddot99 Republic: | couldnt raycasting be used to figure out encroachment? or a method to "show only prims on my parcel" |
[11:14] | Andrew Linden: | Yeah, well the support for AR's does not scale well. |
[11:14] | Latif Khalifa: | enroachment needs fixinf for megprim liberation too |
[11:14] | Andrew Linden: | We know how to solve prim encroachment... just need to schedule that work. |
[11:14] | Andrew Linden: | ATM the team I'm on is focusing on performance |
[11:15] | Rex Cronon: | even if implements prim-encroachement. people will still AR |
[11:15] | reddot99 Republic: | like how somebody got an autobanned during a tournement because a competetior had an entire group AR the guy at once? |
[11:15] | Rex Cronon: | if u implement* |
[11:15] | Andrew Linden: | and the current cost of linkability calculations is one of the current hot spots |
[11:15] | LOM Runner: | I just wanted to know about the face. It was mentioned at SLCC.. so just following up |
[11:16] | Imaze Rhiano: | just remove linkability calculations :P and you will totally eliminate that hotspot :P |
[11:16] | Andrew Linden: | /me reads back... |
[11:16] | Andrew Linden: | no jointed objects for a while |
[11:17] | reddot99 Republic: | aww, wait, 'for a while'. you mean they |
[11:17] | Andrew Linden: | What was the "face" + "SLCC" thing about? That statement does not compute. |
[11:17] | reddot99 Republic: | 'll come back? |
[11:17] | Moon Metty: | hehe |
[11:17] | LOM Runner: | sorry.. like Face AO |
[11:17] | LOM Runner: | facial expressions |
[11:17] | Moon Metty: | are you saying NI to that old lady? |
[11:18] | Andrew Linden: | We'd love to have jointed objects. We just have way too much work that will have to get done before then -- I can't even speculate on when joints will get done. |
[11:18] | Latif Khalifa: | yeah performance on busy sims is pretty bad, that is deservably the first prio |
[11:19] | reddot99 Republic: | how about raycasts, how far along the line is that? |
[11:19] | Imaze Rhiano: | what about script memory limits? I heard that they are move to back log... |
[11:19] | Andrew Linden: | Did someone talk about facial expressions + Animation Overides at SLCC? |
[11:19] | LOM Runner: | yeah.. enabling that ability |
[11:19] | LOM Runner: | to create animations for our faces... finally |
[11:19] | Andrew Linden: | Script memory limits are on hold I think... mostly done but waiting for something. |
[11:20] | reddot99 Republic: | would mean adding parts to the skeleton |
[11:20] | Andrew Linden: | I can't remember what the holdup for script limits was. |
[11:20] | reddot99 Republic: | and adjusting skeletons can do a lot |
[11:20] | Andrew Linden: | Huh. I don't remember hearing that we had ambitions short-term plans for facial animations. |
[11:21] | reddot99 Republic: | that reminds me though, way to change priority of existing animations |
[11:21] | reddot99 Republic: | would be nice |
[11:21] | Arawn Spitteler: | /me has very slow text, but IE is slower, today: Wouldn't facial expressions be a mere texture game? We can't script our avatar textures, would be the only problem. |
[11:21] | Andrew Linden: | So, I don't know anything about it. We might be planning/working on it -- I'm a little out of touch when it comes to viewer features. |
[11:21] | Moon Metty: | facial animations become interesting with augmented reality + webcam |
[11:21] | reddot99 Republic: | theres lip sync beta in advanced menu |
[11:22] | LOM Runner: | naw.. not a lecture... it was just talked about due to... awww Moon got it |
[11:22] | Moon Metty: | :D |
[11:22] | Andrew Linden: | Arawn, try disabling the water layer in your debug settings. I'm wondring how much your FPS changes when you do that. |
[11:22] | Andrew Linden: | Arawn, you're currently at about 3FPS, right? |
[11:23] | reddot99 Republic: | thats microsoft for ya, |
[11:23] | Andrew Linden: | Arawn, disabling the water did not help? |
[11:24] | Arawn Spitteler: | Water? I'm not using Windlaugh at all. |
[11:24] | reddot99 Republic: | turning the alpha water layer does a lot regardless |
[11:24] | reddot99 Republic: | turning off rather |
[11:25] | LOM Runner: | well, it has to be put on the table soon, as we decrease the need for input devices |
[11:25] | Andrew Linden: | Arawn, I meant... disable the rendering of the water completely: AdvancedMenu --> RenderingTypesOption --> Water |
[11:25] | Andrew Linden: | aka CTRL + ALT + SHIFT + 7 |
[11:25] | reddot99 Republic: | go ahead and removes all other alpha layers as well |
[11:25] | reddot99 Republic: | remove |
[11:26] | Arawn Spitteler: | testing - testingg, still a little slow, but my system is having issues. |
[11:26] | Andrew Linden: | Hrm... disabling the water and alpha types does not help my FPS at all. |
[11:27] | Andrew Linden: | I'm currently at 11.5 fps today. |
[11:27] | Rex Cronon: | try ctrl+alt+shift+9 |
[11:27] | Andrew Linden: | I was asking you to try it Arawn because another LL dev showed a demo where he had disabled the transparency of water for a slight gain in FPS. |
[11:28] | Arawn Spitteler: | Turning off Water and characters broght me up to 17 fps |
[11:28] | Andrew Linden: | I figured turning it completely off would help. |
[11:28] | reddot99 Republic: | charecters probably did more, |
[11:28] | Latif Khalifa: | i don't see any difference either |
[11:28] | Moundsa Mayo: | 11.5 Sim FPS, Andrew? |
[11:28] | reddot99 Republic: | veiwer |
[11:28] | Andrew Linden: | Yeah, I guess my hardware is getting old. |
[11:28] | Vista Questi: | sims not dilataed much, so no |
[11:29] | Cummere Mayo: | if theres a differrance for me its about 1-2 frames andrew |
[11:29] | reddot99 Republic: | also, andrew, any news on rez queueing? |
[11:29] | Arawn Spitteler: | Diusabling water only has me at 2.6, which is better than 1.9 |
[11:29] | reddot99 Republic: | probably graphics card is choking on sculpts, |
[11:29] | Andrew Linden: | Oh I see, my graphics were set for "Mid". |
[11:29] | Andrew Linden: | At "Low" I get 27 FPS. |
[11:29] | Vista Questi: | Andrew I'm suprised at you! surely you can afford better hardware than that! <smile> |
[11:30] | office hours | is half over |
[11:30] | Moundsa Mayo: | Ah. I'm at 17.1 at High. Meh. |
[11:30] | reddot99 Republic: | he's on a laptop, |
[11:30] | LOM Runner: | why does Chris Pirillo echo in my brain |
[11:30] | Andrew Linden: | Meh. This is my windows machine. I only use it for LL and to make sure I didn't break the Windows build. |
[11:30] | Vista Questi: | Its good to run what the little folk have from time to time, then. ;) |
[11:30] | Andrew Linden: | No, somtimes I'm on a laptop, but then I crash out 50% of the time. |
[11:31] | Andrew Linden: | Anyway, what I'm working on this week... |
[11:31] | reddot99 Republic: | even on my old macbook, only crashs i've had are from teleport failures |
[11:31] | Andrew Linden: | finishing up the new linkability rules, and trying to get my latest work up for test. |
[11:31] | reddot99 Republic: | the rotation fixs? |
[11:31] | Andrew Linden: | My main dev branch has some build problems that were merged in from someone elses work. |
[11:32] | Andrew Linden: | And that someone is fixing it. Might be done today, there was progress on Friday. |
[11:32] | Andrew Linden: | So hopefully up this week. |
[11:32] | reddot99 Republic: | reuse of a variable? |
[11:32] | Moundsa Mayo: | Pizza toppings in source file. |
[11:33] | Arawn Spitteler: | Code Reuse, I'm sure. |
[11:33] | Andrew Linden: | No, someone tried to use some new "tr1 C++" standards, but those don't yet work on the old version of Visual Studio we're using. |
[11:33] | reddot99 Republic: | lol |
[11:34] | Ardy Lay: | Oops. New programmer? |
[11:34] | Arawn Spitteler: | That should have shown up on Alpha Test, on trying to build |
[11:34] | Andrew Linden: | so he's converting to use boost:: instead (boost is a big lib of C++ "might become standards" stuff) |
[11:35] | reddot99 Republic: | perhaps he tried with a newer version of visual studio |
[11:35] | Jonathan Yap: | Andrew, now that the new jira is up can we see what is on your backlog? |
[11:35] | Andrew Linden: | Hrm... Jonathan you want to see what is on my personal backlog? the backlog of my team? or LL's backlog? |
[11:36] | Jonathan Yap: | The team in general |
[11:36] | reddot99 Republic: | and rez queues still trigger, and without getting everybody to stop rezzing at once, they stay stopped for a while, |
[11:36] | Cummere Mayo: | all three andrew |
[11:36] | Andrew Linden: | Hrm. |
[11:36] | Cummere Mayo: | it would be nice to see the same sort of transparency for the server stuff as the snowstorm team is trying to have |
[11:37] | Jonathan Yap: | (Unless it is Top Secrekt) |
[11:37] | Jonathan Yap: | I have an item for that backlog, but first would like to see if it is already there |
[11:37] | Andrew Linden: | I'd like to hope that it would eventually be easy to compile and publish those backlogs. |
[11:38] | Andrew Linden: | ATM I spend some of my time at these office hours talking about my own backlog, and some of what my team is working on or planning to work on |
[11:38] | Andrew Linden: | however, the format it isn't in a nice sorted list |
[11:39] | Andrew Linden: | it would be nice to just have a status page up somewhere |
[11:39] | reddot99 Republic: | ton of sticky notes on moniter? |
[11:39] | Cummere Mayo: | /me nods. if you could publish like snowstorm does you could give us a link |
[11:39] | Vista Questi: | nobody writes with a pen anymore! =^-^= |
[11:39] | reddot99 Republic: | i've printed on stickynotes before, heh |
[11:40] | Andrew Linden: | When Simon gets back I'll ask him if we have plans to publish our team's backlog. |
[11:40] | Imaze Rhiano: | pfftt... pen and paper have still superior resolution and more colors than you can imagine.... and low energy consumption |
[11:40] | Andrew Linden: | I wouldn't want to compile the backlog of LL. I'd rather just point you at a web page. |
[11:40] | Moon Metty: | :) |
[11:41] | Andrew Linden: | I don't have time to keep up with all that we're working on. |
[11:41] | Cummere Mayo: | any luck with getting the lag time on tp or crossings down? |
[11:41] | Jonathan Yap: | I think that is what this new jira system is supposed to make easier |
[11:41] | reddot99 Republic: | how long till mono upgrade? |
[11:41] | Arawn Spitteler: | Compiled Log of LL is called a product |
[11:41] | Andrew Linden: | No luck yet Cummere. We've got a few project to pursue on that front. |
[11:41] | Andrew Linden: | One of which is faster linkability rules (not complete). |
[11:41] | Cummere Mayo: | /me nod. |
[11:41] | NullSubset Burner: | hmm- was the, Included boost libraries prevent VC90 from compiling-removed from Jira? |
[11:42] | Cummere Mayo: | just asking andrew. i know it is a long project :) |
[11:42] | Andrew Linden: | Don't know the ETA for mono upgrade. I'll make a note to ask Kelly. |
[11:43] | reddot99 Republic: | can the abuse email address estate feature be enabled? do estate owners/managers request it? or does linden labs keep that disabled |
[11:43] | Andrew Linden: | Ok, I'm going to have to leave very shortly for an 11:45 meeting. |
[11:43] | Moon Metty: | yup |
[11:43] | Cummere Mayo: | ty andrew :) |
[11:43] | Andrew Linden: | The abuse email feature was disabled and I don't think there are any plans on re-enabling it. |
[11:44] | Rex Cronon: | tc andrew |
[11:44] | Jonathan Yap: | Thank you for your time Andrew |
[11:44] | Andrew Linden: | I wonder who knows about that feature. |
[11:44] | Moundsa Mayo: | Thanks for your time, Andrew. |
[11:44] | Qie Niangao: | Thanks Andrew. |
[11:44] | reddot99 Republic: | can it be reenabled on a special server roll like ones you mentioned friday |
[11:44] | Latif Khalifa: | Andrew, thanks for your time :) |
[11:44] | Moon Metty: | thanks Andrew, see you soon :) |
[11:44] | Andrew Linden: | dunno reddot99. I'll make a note to investigate, but ask again next hour. |
[11:44] | reddot99 Republic: | k, |
[11:46] | Andrew Linden: | Gotta go. Thanks for coming. |
[11:46] | reddot99 Republic: | would ask kelly, but i've got classes all morning during that time, |
[11:46] | Moon Metty: | i get "untrusted connection" when i try to reach the jira now |
[11:46] | reddot99 Republic: | see ya |
[11:46] | Vista Questi: | Yip! |
[11:46] | NullSubset Burner: | thank you Mr Andrew |
[11:46] | reddot99 Republic: | i got that already moon |
[11:46] | Imaze Rhiano: | thank you Andrew |
Generated with SLog Wikifier