Open Source Meeting/2010-05-18

From Second Life Wiki
Jump to: navigation, search

Agenda Tuesday, 18 May 2010

  1. Weekly Snowglobe update - Merov Linden
    1. See the SG 1.x to 2.0 spreadsheet
    2. Roll back to 2.0.1: yes? no?
    3. Major features port tasks: SNOW-666 to SNOW-670
    4. lltexturefetch.cpp: lots of Aleric's patches half way there in 2.0 and can't be safely ported due to other changes. Should we start an audit of that code with stability and perf in mind?
  2. Left over topics from Thursday, May 13 (feel free to move to next Thursday):
    1. Question: What on earth, or in Second Life, is the 'feature' I am killing in SNOW-652 good for?
    2. Regular posting of snowglobe usage stats. Will this help guide the direction of development?
  3. Can there be a way for folks to donate $L to non LL paid developers that work on snowglobe?
  4. llrobla V2.0/replacment update
  5. Progress on SL7B snowglobe display?

Jira Issues

  1. Looks like r3349 will be reverted. Do these still need discussion (e.g. for viewer-external)?:
    1. SNOW-649 Changeset 3349 added unit test to llplugin but forgot LL_TESTS, resulting in a build failure in lltut.cpp
    2. SNOW-650 Tries to build pulseaudio when pulseaudio not found.
    3. SNOW-651 Standalone build failure r3349, cmake error on SLPlugin
    4. SNOW-654 Bulding of integration_tests when not needed
    5. SNOW-658 Notifier Issue Since [sldev-commits] r3349
  2. Unscheduled SNOW issues triage: would be good to go with the community through those and decide for some what we should take immediately (2.0 or 1.4) or later.


[13:58] Boroondas Gupte: hi Ardy
[13:59] Ardy Lay: Hi
[13:59] Boroondas Gupte: hi Celierra
[13:59] Celierra Darling: Hey Ardy, and Boroondas :)
[13:59] Ardy Lay: Hi
[14:00] Celierra Darling: What's been going on lately?
[14:01] Robin Cornelius: Yay you made it
[14:01] Techwolf Lupindo: make it
[14:01] Boroondas Gupte: let's see ... the viewer-external merge will be rolled back, I think (maybe a topic today)
[14:01] Boroondas Gupte: other than that, dunno
[14:01] Celierra Darling: Mm, did you see the e-mail just sent by Oz introducing himself?
[14:01] Techwolf Lupindo: Lucky I got here, i'me doing a compile now for you Robin. gcc 4.4.3 for debuggin.
[14:01] Merov Linden: well, I don't know what happened the 1st time
[14:02] Techwolf Lupindo: Oz?
[14:02] Boroondas Gupte: oh, yeah
[14:03] Merov Linden: ok, I see that the great Oz is making my agenda easier :)
[14:03] Boroondas Gupte: this just in:
[14:03] Merov Linden: Other than that:
[14:04] Merov Linden: k, I think we should start
[14:04] Aimee Linden sees many "yellow prim road" jokes in her future
[14:04] Boroondas Gupte: ?
[14:05] Merov Linden: OK, first on the agenda, a small list of SG2.0 actions
[14:05] Jonit Ivory: Huh, that's inspired Aimee
[14:05] Merov Linden: so I made that darn spreadsheet public
[14:05] Merov Linden: that was a piece of work! I went through every single commit and checked the code...
[14:05] Merov Linden: glad it's done
[14:06] Techwolf Lupindo: Merov, how long did you take you to whip up that google cod?
[14:06] Techwolf Lupindo: doc
[14:06] Merov Linden: basically, the other items on the agenda derive from it
[14:06] Merov Linden: Techwolf: I didn't count, couple of days but spread out over a week I'd say
[14:07] Merov Linden: anyway, it needed to be done
[14:07] Merov Linden: so first, the proposed roll back
[14:07] Techwolf Lupindo: Wow....i'me having trouble keeping trakc of my own bugs and you did that...."we are not worthy" *bows*
[14:07] Merov Linden: I emailed the list on this as we discussed at the last Hippo meeting (Thursday)
[14:07] Techwolf Lupindo: :-)
[14:08] Merov Linden: the consensus in the meeting was : yeah, lets roll back so we work on stable 2.0.1 for a while
[14:08] Ardy Lay: I am not having build issues so will just sit this topic out.
[14:08] Merov Linden: now, the 2 feedback email on the list were more nuanced
[14:09] Merov Linden: so I'm asking the question again to you guys
[14:09] Robin Cornelius: i've also got SNOW-659 on latest SG2 trunk
[14:09] JIRA-helper:
[#SNOW-659] Standalone SG2 fails to start, sigsegv in LLFastTimer
[14:09] Merov Linden: especially the heavy lifters (Aleric, Techwolf :) )
[14:10] Merov Linden: do you think we should roll back that 2.0.2 commit merge?
[14:10] Merov Linden: and sit on 2.0.1 code base for a while
[14:10] Aleric Inglewood: I think internal should freeze their changes until they caught up with our 1.x patches.
[14:10] Merov Linden: (till we're done with SG1.x merge back)
[14:10] Boroondas Gupte: :-P
[14:10] Aleric Inglewood: I don't see why WE have to redo it for 2.0
[14:11] Techwolf Lupindo: With the snowglobe usage up, we need to get out at least one 2.x.
[14:11] Aleric Inglewood: especially when it's unclear if THEN the changes will make it upstream or not
[14:11] Robin Cornelius: it will be a lot easier to try to catch up to a fixed version, we all seem to be spending way too much time chasing non stop obscure build bugs
[14:11] Merov Linden: Aleric: there are a lot of people moving things full speed in all sorts of direction, we have no choice but do things in parallel
[14:12] Robin Cornelius: and every merge seems to pull more in diverting attention from the porting the patches
[14:12] Merov Linden: we know we'd have to port
[14:12] Robin Cornelius: so +1 freeze for a short while, merge patches, get a release out there
[14:13] Merov Linden: OK, the issue *I*'m having is battling viewer-external merges all the time and, consequently, not helping on 1.x merge back
[14:13] Techwolf Lupindo: From what i've heard, too many showstopping like bugs in current 2.x trunk.
[14:13] Merov Linden: I know there are a lot of then I could expedite swiftly but I can't do that while doing merges all the time
[14:14] Robin Cornelius: it would seem a rollback to a fixed release then give ourselves a chance to regroup would be the logical answer
[14:14] Merov Linden: Yes, my point exactly Robin
[14:15] Techwolf Lupindo: At least tag a 2.0.1<?> branch so we can start working on a viable reliease and have something we can port 1.x code to.
[14:15] Merov Linden: K, silence is not necessarily consent so: who does *not* like that roll back plan?
[14:16] Merov Linden: well now, I have to believe that this silence is consent
[14:16] Merov Linden: :)
[14:16] Techwolf Lupindo: text chat takes a while
[14:16] Ardy Lay: Either way is fine with me.
[14:16] Boroondas Gupte: Nyx, some of that that 2.0.2 stuff in work is your multi-wearables, isn't it? Would it be possible for you to branch from snowglobe instead of from an internal branch?
[14:16] Aleric Inglewood: I'm trying to be silent.. cause I don't have much constructive to say :/
[14:16] Ardy Lay looks at the revision graph.
[14:17] Merov Linden: my goal is to get that 1. port in there as fast as possible
[14:17] Merov Linden: so we can move to more interesting things...
[14:18] Merov Linden: and have a chance to get some of those patches in 2.1 (main viewer)
[14:18] Merov Linden: OK, it's settled then: I'll do the rollback
[14:18] Aleric Inglewood: What I'd WANT is to make huge changes (I'd say, improvements) to the code... but I can't, I feel, which is very frustrating. And I can't because I stopped believing that such improvements would ever make it upstream, thus only increasing the merge load for you.
[14:19] Merov Linden: We're starting to get more and more people involved in Open Source now Aleric
[14:20] Merov Linden: see: 3 Linden here today :)
[14:20] Merov Linden: and we have Oz coming on board
[14:20] Merov Linden: I'm really hoping things will change
[14:20] Techwolf Lupindo: Is it possiabel to tag a 2.0.1 branch and commint/port to that?
[14:20] Merov Linden: but your point is well taken Aleric: we need to show with tangible actions that things are changing
[14:21] Aleric Inglewood: (And since it seems that what comes from internal is breaking things more than otherwise, or horrible new "features" like 2.0... I believe the best thing for the open source project 'SG' would be to just fork and forget about "upstream"... but even that is not an option because some, like merov, are paid by LL )
[14:21] Merov Linden: Techwolf: we could but I don't see that as a big advantage. I'll still have to do a 2.0.2 merge later
[14:23] Merov Linden: "paid by LL" doesn't mean I can't contribute to GPL code
[14:23] Merov Linden: anyway, there are certainyl other ways to accelerate the rate of change in SG
[14:23] Merov Linden: agree
[14:24] Merov Linden: I guess Oz will be working on all those aspects
[14:24] Techwolf Lupindo: Where is Oz?
[14:24] Merov Linden: next on the agenda: new SNOW 666 to 670 tasks
[14:24] Aleric Inglewood: If we truely fork, and start with major changes (for example.. we start with a source wide indent... we truely become independent from 'upstream'), then LL will say this project failed and you'd be taken off the SG project merov.
[14:25] Boroondas Gupte: Techwolf: in Boston
[14:25] Techwolf Lupindo: I though the new linden be hrere today
[14:26] Celierra Darling: Second day might be too soon, probably still struggling to get the whole dev environment up and running... :P
[14:26] Merov Linden: Techwolf: he told me he had issues with his LL avatar
[14:26] Merov Linden: "wardrobe malfunction" type may be :)
[14:26] Calcite Serendipity: Has he been in SL before? Or does he have to learn the viewer?
[14:27] Techwolf Lupindo: If he is truely a newbie...viewer 2.0 is giving him/her a headace now.
[14:27] Techwolf Lupindo: :-)
[14:27] Merov Linden: He's been in SL before
[14:27] Merov Linden: k, on the 666 to 670 tasks, is there any takers for some of them?
[14:28] Merov Linden: I'm hoping Pixel will take 666 (she said she will already)
[14:28] Merov Linden: and Robin SNOW-669
[14:28] JIRA-helper:
[#SNOW-669] Port of SNOW-238 to SG 2.0 : Add Socks 5 Proxy Support
[14:28] Thickbrick Sleaford: I can take SNOW-667
[14:28] JIRA-helper:
[#SNOW-667] Port of SNOW-93 to SG 2.0 : Automatic language translation of chat messages
[14:28] Merov Linden: \0/
[14:28] Merov Linden: thanks Thickbrick!
[14:29] Boroondas Gupte: I don't know whether we'd have to re-do the accunt selection from scratch (SNOW-670). OpenSim and OGP users haven't been too happy with our solution, IIRC.
[14:29] JIRA-helper:
[#SNOW-670] Port of SNOW-129 to SG 2.0 : Select account from list at login
[14:29] Techwolf Lupindo: What I was planning on doing is going though the google doc list and snagging the simple stuff to get out of the way once a branch is desided on. (Currrnetly a rollback now)
[14:30] Merov Linden: Techwolf: if you do that *please update the spreadsheet* and *log a JIRA*!
[14:30] Techwolf Lupindo: Hahahah, jira? What that? :-X
[14:30] Merov Linden: I'm planning to do the same so we should avoid working on the same thing
[14:30] Techwolf Lupindo has a bad habit of no changelog....:-/
[14:31] Merov Linden feels he won't escape creating JIRA for all those little ones :/
[14:31] Merov Linden: k, I'll reate JIRA for those I'll take for sure
[14:32] Boroondas Gupte: same here
[14:32] Robin Cornelius: Merov, i will need some input from others for UI layout for 669
[14:32] Techwolf Lupindo: Once I get my workflow set up on the correct branch, should be able to create and post patch for jira at once.
[14:33] Ardy Lay: Look like the login information handling is changed in later revesions.
[14:33] Merov Linden: Robin: sure, someone is cleaning some wiki doc on the new XUI BTW
[14:33] Merov Linden: should be avialable soon
[14:33] Boroondas Gupte: cool
[14:33] Robin Cornelius: great
[14:34] Merov Linden: k, so we have 3 of those "big" one assigned, I can take SNOW-668
[14:34] JIRA-helper:
[#SNOW-668] Port of SNOW-352 to SG 2.0 : Add optional double-click teleport
[14:34] Merov Linden: (easy one I think)
[14:34] Merov Linden: SNOW-670, may be wait till Pixel is done with 666
[14:34] Merov Linden: how's that for a plan?
[14:35] Thickbrick Sleaford: shouldn't be much change in double-click teleport between 1.x and 2.x, except that it might use some improvments...
[14:35] Calcite Serendipity: Is there any chance of having ported?
[14:35] JIRA-helper: [#SNOW-392] llTextBox in client Non-Functional - Functionality Patch Attached
[14:35] Merov Linden: indeed, some XUI hacking though
[14:36] Merov Linden: Calcite: I need to commit to 1.4 before porting
[14:36] Merov Linden: that one
[14:37] Merov Linden: k, next on the agenda: lltexturefetch.cpp
[14:37] Merov Linden: that's a special tribute to Aleric :)
[14:37] Merov Linden: there was a *lot* of commits and patches by Aleric, Thickbrick and a little bit of myself in that code in 1.x
[14:38] Boroondas Gupte: how comes patches get in partially, anyway?
[14:38] Boroondas Gupte: merges gone wrong internally?
[14:38] Robin Cornelius: meh, i spend ages de-deadlocking texturefetch as well -;p
[14:38] Merov Linden: some have been taken in, some not, some were done in a different way
[14:38] Boroondas Gupte: oh, this is about several patches, then?
[14:38] Merov Linden: sorry about forgetting you Robin
[14:39] Robin Cornelius smiles
[14:39] Boroondas Gupte remembers
[14:39] Merov Linden: I looked into th patches and v2.0 code and clearly, we can't apply the patches blindly and hope it'll work
[14:39] Robin Cornelius: certainly not with the mutex patches, they will need reassessing
[14:40] Merov Linden: this need some deep analysis, what I wrote as "audit" (may be not a good word but you see what I mean)
[14:40] Robin Cornelius: helgrind is the wayforward
[14:40] Thickbrick Sleaford: there's also a new codepatch to experiment with: http-textures
[14:40] Boroondas Gupte: dunno if helgrind finds everything, but it might help for sure
[14:40] Thickbrick Sleaford: *code path
[14:40] Robin Cornelius: thread/mutex/concurancy debugging on valgrind
[14:41] Merov Linden: actually, following that work Aleric made a proposal some weeks ago to make mutex dead locking less of an issue
[14:41] Merov Linden: Aleric, have you been working some more on that idea?
[14:42] Aleric Inglewood: You mean the automatic detection of the order in which mutexes are locked?
[14:42] Robin Cornelius: thats also a helgrind function
[14:42] Boroondas Gupte: I think he means the class template.
[14:42] Merov Linden: that and the new set of classes you proposed
[14:43] Merov Linden: IIRC you proposed to put the shared data and mutex into a single class, maing clear what mutexes control
[14:43] Aleric Inglewood: Oh yes that... no that got interrupted by the TPV policy debacle... and ever since I've worked on other things that now need my attention (want to finish what I started)
[14:43] Merov Linden: understood
[14:44] Merov Linden: well, looking at the old and new code, I see SG 1.x did a binch of changes to make things cleaner / less dead lock prone and that has not all been ported
[14:45] Merov Linden: I agreed though with Aleric's point that that code needed more than just "debug/fixes" but some serious rework
[14:45] Robin Cornelius: with http-texture on beta thats a good target for falling over deadlocks in texture fetch
[14:45] Merov Linden: so I was thinking about starting something around it, keeping as an objective to make that code more stable and perform better
[14:46] Merov Linden: k, I'll unearth your email Aleric and use it as a starting point for a new conversation on opensource-dev
[14:46] Aleric Inglewood: Merov: ... that sounds great... but what sense does it make to make *major* changes like that if it's not going to be used by upstream? It will become a maintenance nightmare
[14:48] Merov Linden: you're right... so I'll go an pull the devs in so that we can have that conversation with them, or some of them, directly...
[14:48] Merov Linden: anyway, if we show that SG performs better, they'll get an incentive to take our patches... :)
[14:48] WolfPup Lowenhar: well at least i made it for the last part of the meeting
[14:48] Robin Cornelius: i think its time for some more of those graphs i did ;-)
[14:49] Merov Linden: yes, that was a question I had Robin
[14:49] Merov Linden: how did you do them?
[14:49] Merov Linden: could we get some metrics like that automatically extracted?
[14:49] Robin Cornelius: monitor viewer stats and save them every second to lists, then just write them to a file
[14:49] Merov Linden: that would be great to measure progress
[14:50] Robin Cornelius: in fact it was a patch i stole out of imprudence then hacked up a little
[14:50] Merov Linden: cool!
[14:50] Techwolf Lupindo: heh, I could drag out my last year project that did some client side performace testing and mesurments.
[14:51] Techwolf Lupindo: I was told that it was being done internially, but I havn't seen or heard a peep sence then.
[14:51] Merov Linden: well, having some measurement on perf we all agree on would sure help focusing perf improvement efforts
[14:51] Robin Cornelius: but as i said simply saving viewer stats such as texture queue FPS, bandwdth etc was all i did
[14:51] Techwolf Lupindo: So I dropped the project and went onto other things.
[14:52] Robin Cornelius: Zen was asking me to repeat on latest 2.X as the pipe line was improved
[14:52] Merov Linden: interesting
[14:53] Merov Linden: if we could get such numbers for each iteration that would be great
[14:53] Robin Cornelius: repeatability and reproducability are the biggest issues
[14:54] Merov Linden: as always with overall perf measurement
[14:54] Techwolf Lupindo: Really need some test regines setup just for this.
[14:54] Merov Linden: better than hear say though...
[14:54] Robin Cornelius: well they say there are lies dam lies and statistics
[14:55] Robin Cornelius: but they could be useful but we do need to be careful with the results
[14:55] Aleric Inglewood: Real improvements come from change of algorithms, not (manual) optimization. That means in our case radiacal changes in how things are achieved, and those can be thought up without measuring too.
[14:55] Techwolf Lupindo: Lag spikes are worse then slow but steady framerate.
[14:55] WolfPup Lowenhar: looks like im going to have to read the log of this meeting
[14:55] Merov Linden: could you make the results you gathered so far public Robin?
[14:55] Ardy Lay: Speaking of which, any way we can see viewer usage statistics?
[14:56] Boroondas Gupte: That's topic 2.2 on the agenda
[14:56] Merov Linden: Ardy: I need to ask internally, not sure
[14:56] Ardy Lay: I expect the answer to be "HELL NO!"
[14:57] Aleric Inglewood: (For example... I'd think of a ccache-like system between prim-level data and ready-to-send-to-the graphics card "compilation" there off. That would speed up animations (which repeat themselves), but also, for example, Latha's tail movement. It would be pre-compiled say.)
[14:57] Techwolf Lupindo: Hah, I still have the results I did last year up.
[14:59] Merov Linden: there are certainly more than one place for perf improvement for sure
[14:59] Merov Linden: I agree that often, a radical change in the algo is required (case in point: the map changes I did...)
[15:00] Celierra Darling: (Aleric, that sounds like display lists in OpenGL... Might want to think in groups bigger than prims, but it gets hard to activate many textures simultaneously on the graphics card, need to palletize them somehow...)
[15:00] Ardy Lay: Somebody has started work on texture atlasing.
[15:00] Aleric Inglewood: I think that right now, we the easiest way to get a really big improvement, is to concentrate on using the other cores :p ... But then I ignore that LL wants to support the apparently 90% of the users with a 486 on 50 Mhz.
[15:00] Aleric Inglewood: s/we//
[15:01] Merov Linden: he
[15:01] Thickbrick Sleaford: the autmatic atlasing in 2.0 kind-of-almost works
[15:01] Merov Linden: yeah: class 0 machines can't ne brushed aside... that's biting us...
[15:01] Merov Linden: time check: 3pm already
[15:01] Techwolf Lupindo: Merov, any docs on what class machine is what number?
[15:02] Merov Linden: I think that's an internal classification, not of much value
[15:02] Merov Linden: just saying that class 0 are the bottom of the barrel and there are *a lot* of them out there...
[15:03] Ardy Lay: Deferred rendering makes them all feel like class-0 machines.
[15:03] Techwolf Lupindo: I know a great many SL users using laptops.
[15:03] Robin Cornelius: laptops don't need to be class 0
[15:03] Merov Linden has a MacBookPro... not class 0... :)
[15:03] Robin Cornelius: some of the nvidia mobile GFX cards are pretty good
[15:04] Techwolf Lupindo: hardware imporovments have slowed down in the past years. Years ago, a new system was usally 5 to 10 times faster. Today, a three year old system is less then 2x slower then a new system.
[15:04] Celierra Darling: Intel's still the biggest supplier of graphics last I checked, though, thanks to their Geforce 2-level cheap integrated stuff... :(
[15:04] Merov Linden: anyway: throwing hardware at the problem is not something we can do: we have to work on what resi have
[15:05] Merov Linden waiting for memristors architecture to become main stream...
[15:05] Thickbrick Sleaford: heh
[15:05] Merov Linden: that'll take a while :)
[15:06] Merov Linden: k, I think we reached time limit
[15:06] Techwolf Lupindo: The current imporvments in hardware are more engery effect cores.
[15:06] Ardy Lay: What are Pastrami and Runitai up to lately?
[15:06] Merov Linden: we didn't address point 3 which is very interresting so need to table that for Thursday
[15:07] Boroondas Gupte: 2.1 wasn't adressed, either (partially discussed last Thursday, I think)
[15:07] Merov Linden: Ardy: I really prefer having people talk about their stuff themselves, I hate to misrepresent what people do
[15:08] Ardy Lay: I was thinking the were from Windward Mark and worked on rendering.
[15:09] Merov Linden: yes, they do work on rendering, that's not saying much though :)
[15:09] Ardy Lay: Yeah, okay.
[15:09] Celierra Darling: There aren't many graphics-related things lately in the blog posts though, so it does narrow things down a bit :P
[15:09] Merov Linden: 652: I'll ping the securoty minded folks on this one
[15:10] Ardy Lay: Meh, revision graphing is broken on SVN stuff starting with all the V2 work.
[15:10] Thickbrick Sleaford: I think there's also other app/ uris that might be a problem
[15:10] Ardy Lay: Totally useless.
[15:10] Merov Linden: k, need to wrap up and go
[15:11] Ardy Lay: All branches are displayed as seperate projects.
[15:11] Celierra Darling: Thanks Merov :)
[15:11] Merov Linden: I do hope that Oz will have his avatar fixed by Thrusday and will be here next time
[15:11] Boroondas Gupte: \o/
[15:12] Boroondas Gupte: We're getting a backlog of meeting topics, it seems
[15:12] Merov Linden: some, true
[15:12] Ardy Lay tries to find the real head of the V2 repository.
[15:12] Merov Linden: I do feel we discuss the important things though
[15:13] Merov Linden: at least, I feel those meetings are useful to me
[15:13] Merov Linden: what about you guys?
[15:13] Ardy Lay: Yes
[15:14] Jonit Ivory: Indeed, I am merely observing though, but it is insightful
[15:14] Robin Cornelius: Oh one thing Merov, we are on a deadline for the SL7B submission
[15:14] Thickbrick Sleaford: they're certainly useful, it's just that I'm too much of a sloth to do stuff outside the meeting
[15:14] WolfPup Lowenhar: im here to try and help and to learn as well
[15:14] Merov Linden: eeew... what's the deadline?
[15:14] Merov Linden: when are we dead?
[15:15] Ardy Lay: Ah, found it. Was renamed. Drat. SG2 is still seperate. Oh well.
[15:15] Robin Cornelius: 20th
[15:15] Robin Cornelius: meh
[15:15] Boroondas Gupte: oh
[15:15] Boroondas Gupte: that's soon
[15:15] Merov Linden: Yahuza!
[15:15] Techwolf Lupindo: I though they start taking apps on the 20th?
[15:16] Robin Cornelius: There was only a short opening for applications
[15:16] Merov Linden: I'll email Oz now
[15:16] Boroondas Gupte: what do we need till the deadline? only the application?
[15:16] Robin Cornelius: # LL will start accepting applications - 12th May 2010
  1. Deadline for applications – 20th May 2010
[15:16] Robin Cornelius: yes just an application 100 words or less saying the kind of thing we are about
[15:16] Robin Cornelius: and a name or too
[15:16] Merov Linden: I talked to Howard about it BTW and he thought it was a good idea
[15:17] Robin Cornelius: Shall i just put an application in?
[15:17] Merov Linden: just want to make sure Oz is on board too
[15:17] Merov Linden: let me email him and cc you, ok?
[15:17] Merov Linden: if all aligns, we'll make a wider announcement on the list
[15:17] Robin Cornelius: sure
[15:17] Merov Linden: cool!
[15:18] Robin Cornelius wonders why i'm just not on the payroll the number of LL emails i get day
[15:18] Merov Linden: OK, need to go now
[15:18] Aleric Inglewood: what application?
[15:18] Merov Linden: see you Thursday, same time
[15:18] Aleric Inglewood: bye!
[15:18] Boroondas Gupte: SL7B display for Snowglobe project
[15:18] Robin Cornelius: for a stand at the 7th birthday exihibition
[15:19] Thickbrick Sleaford: see you
[15:19] Aleric Inglewood: What would you want to show on a stand in-world??
[15:19] Boroondas Gupte: we don't know, yet, I guess
[15:19] Robin Cornelius: well other viewers have stands, open sim had a stand
[15:19] Celierra Darling: Think of it as marketing :P
[15:20] Robin Cornelius: its more about awareness and presence there will be lots of people there who know nothing of snowglobe or ud
[15:20] Robin Cornelius: *us
[15:20] Aleric Inglewood: marketing... okeeey

Generated with SLog Wikifier