Open Source Meeting/2009-10-29

• paste transcript here[13:58] Ardy Lay: Hello Merov
• [13:59] Robin Cornelius: Dam any ideas about SLVoice and jackd and why its so unhappy?
• [13:59] Carjay McGinnis: Hello Merov
• [13:59] Merov Linden: hi all
• [13:59] Aimee Trescothick: hi :)
• [14:00] Ardy Lay: Hello Rob
• [14:00] Thickbrick Sleaford: hello everybody
• [14:00] Rob Linden: Hi everyone!
• [14:00] Carjay McGinnis: Ello Rob
• [14:00] Robin Cornelius: Hi
• [14:01] Rob Linden: agenda is here: https://wiki.secondlife.com/wiki/Open_Source_Meeting/Agenda
• [14:02] Rob Linden: thanks once again for the agenda Ardy!
• [14:02] Merov Linden: ducking the nerve foam bullets flying around
• [14:02] Ardy Lay: You are welcome.
• [14:02] Rob Linden: looks like exactly the stuff we should be talking about, so let's get to it
• [14:02] Rob Linden: first up:
• [14:02] Rob Linden: # SNOW-108[c
• [14:02] JIRA-helper: http//jira.secondlife.com/browse/SNOW-108:
• [#SNOW-108] timers fail: on non i386 systems
• [14:03] Merov Linden: checks that bug
• [14:03] Robin Cornelius: I though the patch only fixed Linux 64bit native builds
• [14:03] Robin Cornelius: i did not touch the windows paths
• [14:04] Fox Oxbar: I just added a comment. FastTimers is definitely working now for me.
• [14:04] Fox Oxbar: I've gone in and out of "standby" power mode, and had the machine on battery power (throttled clock speed) since I saw the problem.
• [14:04] Robin Cornelius: Fasttimers may not work correctly on multicore win systems or even 64 bit
• [14:04] Merov Linden: haven't tried/tested
• [14:05] Carjay McGinnis: they work acceptable on my dual core but that's not a windows system, 64 bit Linux
• [14:05] Carjay McGinnis: "acceptable" since I never measured the drift or compared to a different method
• [14:06] Rob Linden: hmm....ok, well, I think we maybe need to have a separate issue opened up for Win64 fast timers, so that we can re-resolve this one
• [14:06] Robin Cornelius: the linux paths use what i believe to be the best avaiable method anyway
• [14:06] Fox Oxbar: This is Vista32 on a 2x64 AMD chip
• [14:06] Rob Linden: we probably shouldn't piggyback on the old bug because it's going to get confusing describing which problem we're talking about
• [14:06] Robin Cornelius: windows uses additional code to scale by CPU frequency etc so maybe that lags sometimes/
• [14:06] Merov Linden: looks like Robin's code can be extended / adapted to Win 64
• [14:07] Rob Linden: sounds like it. maybe a subtask of this one
• [14:07] Merov Linden: you added the required member (LLFastTimer::sClockResolution) to tweak in all sorts of cases
• [14:07] Merov Linden: so +1 in closing that bug and opening a new one to cover Windows 64
• [14:08] Fox Oxbar: If it helps, the issue I was seeing was that the FastTimers display totally ground to a stop after a second or two of enabling it. The values displayed were WAY out of bounds, like in that JIRA's screenshot
• [14:08] Robin Cornelius: the win API for timing is more complex
• [14:09] Rob Linden: ok...I've opened SNOW-326 for this
• [14:09] JIRA-helper: http//jira.secondlife.com/browse/SNOW-326:
• [#SNOW-326] SNOW-108 patch: for Windows
• [14:09] Merov Linden: kewl
• [14:09] Carjay McGinnis: hm, I usually simpy use QueryPerformanceCounter
• [14:09] Rob Linden: k....let's move on
• [14:09] Carjay McGinnis: it's even available in the kernel
• [14:09] lonetorus Habilis: hey morg
• [14:09] lonetorus Habilis: talk of the sun
• [14:09] Fox Oxbar: Would you like me to copy over my comment with my environment info to the new page?
• [14:10] lonetorus Habilis:  ;)
• [14:10] Rob Linden: SNOW-272 is next up
• [14:10] JIRA-helper: http//jira.secondlife.com/browse/SNOW-272:
• [#SNOW-272] message: media_plugin_webkit: occurs on startup
• [14:10] Rob Linden: Fox: that would be wonderful, thanks!
• [14:10] Merov Linden: ok
• [14:10] Merov Linden: so I have a working solution... at last!...
• [14:10] Morgaine Dinova: Hi lone :-)
• [14:10] Merov Linden: I need to do 2 things before posting a patch for review and committing
• [14:11] Merov Linden: 1- upload ares, curl and boost recompiled with the correct compile options and MVCRT
• [14:12] Merov Linden: 2- create and post a patch for the few files (install.xml, some cmake, some CMakeLists) we need to change
• [14:12] Merov Linden: I'm planning to get to that by tomorrow
• [14:13] Merov Linden: so we can get feedback on the patch before I commit Monday
• [14:13] Merov Linden: at which time we'll create an RC I think
• [14:13] Merov Linden: s/we'll/we should/
• [14:13] Merov Linden: that's the plan
• [14:14] Rob Linden: well, this is the most important thing blocking, so I'm wondering if maybe we go ahead and commit before the weekend and generate a build
• [14:14] Morgaine Dinova: Hi Adiu :-)
• [14:15] Merov Linden: We can do that
• [14:15] Merov Linden: though, as I'm tweaking some files used in build
• [14:15] Merov Linden: I'd like to get some feedback from people building on other platforms
• [14:16] Merov Linden: I'll test on Mac but I don't on Linux
• [14:16] Merov Linden: I don't see any reason why those changes would break build anywhere else... but still...
• [14:16] Rob Linden: ah, ok....I can certainly try a LInux build. I also think we may be best off just being ready for a quick revert if we've got a build buster
• [14:17] Merov Linden: +1 on that
• [14:17] Adiuvous Reinoir: hi morgaine :)
• [14:17] Merov Linden: happened in the past IIRC
• [14:17] Pixel Gausman: stumbles in late
• [14:18] Rob Linden: yeah, if we had a lot of people checking in a lot of colliding stuff, I'd be more nervous about this strategy, but we're coming to what should be the end of the process
• [14:18] Morgaine Dinova: hands Pix a .... erm, what do dragonflies eat?
• [14:18] Rob Linden: Soft has a few patches we looked at earlier today that he's possibly going to slip in, but nothing major
• [14:19] Merov Linden: other insects... catch them in fly...
• [14:19] Thickbrick Sleaford: I thought they eat dragons
• [14:19] Morgaine Dinova: Aha! Part of the debugging team then, cool
• [14:19] Pixel Gausman: don't put SNOW-157 in. i'm trying to figure out what the *real* issue is, and i haven't figured it out yet..... despite all the help in irc
• [14:19] JIRA-helper: http//jira.secondlife.com/browse/SNOW-157:
• [#SNOW-157] [[User:: most|: most]]: textures don't load logged on an OpenSim
• [14:20] Rob Linden: (btw, Soft is planning on incorporating SNOW-319 , VWR-8726 and VWR-4232 if you're curious
• [14:20] JIRA-helper: http//jira.secondlife.com/browse/SNOW-319:
• [#SNOW-319] Search All: overflows floater when UI Size is smaller than 1.0
• [14:20] JIRA-helper: http//jira.secondlife.com/browse/VWR-8726:
• [#VWR-8726] off swirling: lights for scripted objects
• [14:20] JIRA-helper: http//jira.secondlife.com/browse/VWR-4232:
• [#VWR-4232] particles don't: disappear when UI is hidden
• [14:20] Rob Linden: Thickbrick: if I recall correctly, I think one or more of them is from you
• [14:21] Thickbrick Sleaford: yes, SNOW-319
• [14:21] Rob Linden: Pixel, thanks for the warning on SNOW-157
• [14:22] Rob Linden: Thickbrick, if you just want to commit SNOW-319 yourself, feel free
• [14:22] Thickbrick Sleaford: it's reviewed?
• [14:22] Rob Linden: yup
• [14:22] Thickbrick Sleaford: ok
• [14:22] Pixel Gausman: yeah, it's in a nasty part of the code, that non-OGP users also use, so i'm scared to touch anything in there
• [14:23] Rob Linden: Thickbrick...oh, wait, I think Soft may have been planning a closer look first
• [14:23] Thickbrick Sleaford: oh, ok
• [14:24] Rob Linden: that's one we all looked at and said "it looks correct" on first inspection, but Soft wanted to check how that member variable was used elsewhere
• [14:24] Robin Cornelius: Merlov can i quickly jump back a moment, i was afk. Are you going to add the (one) vcredist to the installer so it stops installing private assemblies?
• [14:25] Merov Linden: Robin: I was not planning to add that
• [14:25] Merov Linden: no
• [14:25] Robin Cornelius: ok i will open a jira for that as i think its a good idea long term
• [14:26] Robin Cornelius: CRTs get updates from MS security fixes too
• [14:26] Merov Linden: true
• [14:26] Merov Linden: that what got us into that mess in the first place
• [14:26] Thickbrick Sleaford: could the CRT problem be causing SNOW-324 too?
• [14:26] JIRA-helper: http//jira.secondlife.com/browse/SNOW-324:
• [#SNOW-324] video media: no longer works in Snowglobe. It does in the SL Viewer. Youtube videos and quicktime movies do not work at all, button is toggled and does not respond.
• [14:26] Robin Cornelius: no using 3 winSxS crts and 1 private crt got you there ;-)
• [14:27] Thickbrick Sleaford: (that one is about streams that worked in 1.1.2 not working in 1.1.3)
• [14:27] Robin Cornelius: sorry rob, i'm done for the moment on that
• [14:27] Rob Linden: urgh...Thickbrick, you're probably right
• [14:28] Rob Linden: Robin: no prob...thanks for bringing that up
• [14:28] Jacob Urvile: How are you guys?
• [14:28] Merov Linden: Thickbrick: could be but I was confused reading the bug report really...
• [14:28] Rob Linden: on 324: I failed to notice that the difference was between 1.1.2 and 1.1.3...that's important
• [14:29] Rob Linden: well, foo
• [14:30] Merov Linden: I'll add a comment to 324 once we get a build of 272 to see if that change things for that resident with that media
• [14:31] Rob Linden: yeah, there's probably not much sense in spending time trying to fix up 1.1....let's just get it right for 1.2
• [14:31] Rob Linden: ok....let's talk about the bug sprint ( SNOW-280 )
• [14:31] JIRA-helper: http//jira.secondlife.com/browse/SNOW-280:
• [#SNOW-280] 1.2 test: sprint #1
• [14:32] Rob Linden: I've been postponing for lots of reasons, but I'll blame SNOW-272 only because it's fun to make Merov feel guilty
• [14:32] Rob Linden:  :-P
• [14:33] Rob Linden: anyway....what sort of interest is there in participating?
• [14:33] Merov Linden: that's your Karma you're hurting Rob, not mine...
• [14:33] Rob Linden: lol
• [14:34] Twisted Laws: when are you planning to do it, rob?
• [14:34] Merov Linden: may be we should explain what a test sprint is?
• [14:35] Rob Linden: Twisted: well, we could hijack our weekly meeting, but...
• [14:35] Rob Linden: first, Merov has a point
• [14:35] Rob Linden: so....about half of you have done this before
• [14:36] Rob Linden: it's basically an inworld gathering where we meet and try to get through a bunch of test plans by dividing up the duties
• [14:36] Jacob Urvile: Rob, your the coolest linden worker ever
• [14:37] Rob Linden: the idea is to make sure that Snowglobe 1.2 doesn't have any embarressing issues before it goes live
• [14:38] Rob Linden: ...which, given that it goes on our main download page, it's pretty important that we don't have any whoppers there
• [14:38] Pixel Gausman: Rob: depends on how u define embarrassing. :)
• [14:39] Fox Oxbar: I could be interested in participating, depending of course, on the time it happened
• [14:39] Rob Linden: Pixel: I guess I put that back on you all :) what do you want us to represent as being a great community built viewer?
• [14:40] Rob Linden: Last time, we tried to have it on 2pm on Tuesday, and didn't get more than a couple of people showing up
• [14:40] Rob Linden: is that a bad time, or was that just a one time problem?
• [14:40] Merov Linden: I'm afraid that "great community built viewer" is a far off target... may be always moving out of grasp
• [14:40] Twisted Laws: i missed cause of the dentist, sorry :p
• [14:40] Pixel Gausman: apologies if i came across as snarky. i will say that for the OGP testing, the focus should be on making sure i didn't break the non-OGP path
• [14:40] Rob Linden: (2pm PST)
• [14:41] Fox Oxbar: That time would work for me, anyway
• [14:41] Rob Linden: Pixel: no offense taken, and I agree about the non-OGP path...I'm pretty confident that part is working pretty well, but it's good to know for sure
• [14:41] Carjay McGinnis: I missed it since I didn't know what the "test sprint" was about, guess I missed the announcement on the list
• [14:42] Morgaine Dinova: There's a brownbag meeting on open sourc e viewer policy in half an hour. Apparently what makes a great community built viewer is under some threat.
• [14:42] Rob Linden: Carjay: ok....I'll send a better email this time around (and sooner)
• [14:43] Pixel Gausman: Morgaine: sizzle down until we know what they actually intend to do
• [14:43] Rob Linden: I'll also try to remind folks on IRC sooner. If you all could help recruit for that, that would be wonderful
• [14:43] Carjay McGinnis: not saying it was badly announced, just telling why I missed it
• [14:43] Geneko Nemeth: Wouldn't it be a bit late to actually know for sure of the policymakers' intention?
• [14:43] Rob Linden: Carjay: I suspect you weren't alone....it was poor marketing on my part
• [14:43] Geneko Nemeth: Cat already out of the bag and all.
• [14:43] Morgaine Dinova: Pixel: it would have been better if it had been done more logically to avoid all sizzling.
• [14:44] Rob Linden: Geneko and Morgaine: we're still very much in listening mode
• [14:45] Pixel Gausman: so next tuesday at 2pm was what we settled on?
• [14:45] Rob Linden: Pixel: I think so....is there anyone here interested in coming who can't make that time?
• [14:45] Pixel Gausman: oh poo that conflicts with a weekly meting
• [14:46] Morgaine Dinova: That's the worry of course. Previously the "mode" was open source mode. Now the "mode" is "listening mode" ... which implies change.
• [14:46] Bronson Blackadder: morgaine where is that brownbag at?
• [14:46] Thickbrick Sleaford: I'm guessing if they can't make it 2pm tuesday, they aren't ehre either
• [14:46] Rob Linden: Pixel: is there a better time for you? 3pm? 1pm? Next day?
• [14:46] Pixel Gausman: 1pm
• [14:46] Morgaine Dinova: I don't know, wasn't invited. But Robin is carrying FOSS interests there.
• [14:46] Pixel Gausman: Morgaine: as am i
• [14:47] Rob Linden: Morgaine: FOSS is pretty well represented
• [14:47] Merov Linden: will be there too
• [14:47] Thickbrick Sleaford: too...
• [14:47] Carjay McGinnis: sounds great, Merov, one Linden I would really like to see there
• [14:47] Pixel Gausman: oh cool.
• [14:47] Rob Linden: back on the bug sprint.....how does Tuesday, 1pm work
• [14:47] Robin Cornelius: Blondin announced this one will not be recorded due to technical issues, can anyone rig a recorder some how?
• [14:47] Pixel Gausman: safety in numbers. :)
• [14:47] Robin Cornelius: my jackd system has failed badly
• [14:47] Twisted Laws: ok with me
• [14:48] Morgaine Dinova: Rob: I don't see it as a battle of representations. The conflict was unnecessary.
• [14:48] Fox Oxbar: i know i'm a stranger to these meetings, but I would be interested in helping with the bug sprint. However, I wouldn't be able to help at 1pm. 1:30 or later
• [14:48] Ardy Lay: Rob, either is fine with me.
• [14:49] Rob Linden: Fox...ok, maybe you can join us late. These tend to take 90 min or so
• [14:49] Fox Oxbar: Sure, that sounds good
• [14:50] Rob Linden: cool....so, 1pm on Tuesday. I'll send out the mail about it later
• [14:51] Robin Cornelius: brb
• [14:51] Rob Linden: alrighty...well, I'd like to wrap this meeting up a little early if you all don't mind, since several of us need to get to the brownbag
• [14:51] Carjay McGinnis: no problem, just represent our views well :)
• [14:52] Rob Linden: sorry that not everyone is invited, but we think we've got a pretty good mix of folks for a good conversation, and I'm sure there's going to be at least a summary
• [14:52] Morgaine Dinova: I commend Merov's post on SLdev as input to that meeting.
• [14:52] Pixel Gausman: yeah, Merov, excellent post
• [14:52] Merov Linden: is getting extra Karma points
• [14:52] Rob Linden: indeed
• [14:53] Rob Linden: alrighty, thanks everyone for coming!
• [14:53] Pixel Gausman: Merov: u need karma points if u r getting a root canal. :)
• [14:53] Morgaine Dinova: Thanks Rob :-)
• [14:53] Merov Linden: I will :(
• [14:53] Merov Linden: knnws Pixel is following his Twitter thread
• [14:54] Pixel Gausman: d'oh!
• [14:54] Merov Linden: needs a brb too before the brown bag
• [14:54] Rob Linden: hehe
• [14:54] Rob Linden: see ya
• [14:54] Merov Linden: cheers
• [14:54] Thickbrick Sleaford: there's another BB scheduled for a couple of weeks from now, so I hope that means they at least allowing for revising their policy
• [14:54] Twisted Laws: thanks, bye