Open Source Meeting/2010-03-18

[14:01] Merov Linden: ok, it's 2pm
[14:01] Tillie Ariantho: DId I mention that I dont crash when I take snapshots with FRAPS instead of using the snapshot feature? ^^
[14:01] Merov Linden: so time to start
[14:01] Ardy Lay: Guess that's what happens when a group arrives at about the same time.
[14:01] Eddi Decosta: hello/bonjour
[14:01] Merov Linden: grabs the url of the agenda
[14:02] Techwolf Lupindo: What is FRAPS?
[14:02] Merov Linden: salut Eddi
[14:02] Eddi Decosta: a va ?
[14:02] Robin Cornelius: that might be out of date, not sure if it got updated
[14:02] Tillie Ariantho: [1]
[14:02] Data Landau: screen capture i believe
[14:02] Merov Linden: https://wiki.secondlife.com/wiki/Open_Source_Meeting/Agenda
[14:02] Tillie Ariantho: Yes, benchmarking, snapshots and movie recording. :)
[14:02] Eddi Decosta: thanks merov, i need to close some tab before read lol
[14:03] Merov Linden: OK, it was a busy week again...
[14:03] Eddi Decosta: hey Latif
[14:03] Tillie Ariantho: Eddi: you just need a bigger display. :P
[14:03] Ardy Lay: Merov, thanks for the binary installers for Snowglobe 1.4.0, some friends were asking for that.
[14:03] Latif Khalifa: Hello :)
[14:03] Merov Linden: I feel we're moving in the right direction but I've still a bunch of work to do before everything lines up as we want
[14:03] Eddi Decosta: Tillie, i have a 23' of screen for now but with the all jira open , its bordelic
[14:04] Eddi Decosta: LOL
[14:04] Merov Linden: thanks Ardy :)
[14:04] Merov Linden: "bordelic"? that's English ;)
[14:04] Tillie Ariantho: 1.4.0 works well so far. :)
[14:04] Eddi Decosta: loll , i dont think so Merov :p
[14:05] Eddi Decosta: french word , sorry lol
[14:05] Morgaine Dinova: Merov: is Soft around? I replied to her request for details of the Firefly work, to the extent that we know them. Very interested in her reply.
[14:05] Merov Linden: ok so, first point of the agenda: legal progres on Snowglobe (tm)
[14:05] Techwolf Lupindo: merov, I got llqtwebkit going. But still have gcc/boost problems due to () casing wieerd linkage problems at the end of the build.
[14:05] Techwolf Lupindo: causing that is
[14:05] Tillie Ariantho: thinks Robin drinks far too much. :P
[14:05] Merov Linden: Morgaine: Soft (he's a "he" BTW) is still around but busy on security stuff these days
[14:06] Merov Linden: so he's been very active in the list...
[14:06] Merov Linden: back to the agenda: I've made zero notable progree on this
[14:06] Morgaine Dinova: Merov: righto, thanks. "He" noted, was just repeated the pronoun used by others :-)
[14:06] Morgaine Dinova: repeating*
[14:06] Merov Linden: so, I think we need to keep it there s
[14:07] Robin Cornelius: oh i'll mention it now, if there is time i have 2 jira's to insert into proceedings later
[14:07] Merov Linden: this branding issue needs to come to a resolution
[14:08] Merov Linden: Robnin: sure, at the end of the agenda, please jump in
[14:08] Merov Linden: point 2: SG 1.4
[14:09] Merov Linden: not much to add to what I wrote in the agenda: binaries produced
[14:09] Thickbrick Sleaford: There's a few patches in the pipeline for 1.4
[14:09] Morgaine Dinova: Is "vendor branch" *any* vendor (with branding separate), or specifically vendor == LL?
[14:09] Merov Linden: it was long 'cause there were changes on the parabuild machines here at LL and that broke the Mac and Windows build
[14:09] Merov Linden: now that's fixed
[14:10] Merov Linden: so next here is to switch the "official" release to 1.3.2 and point trunk to 1.4 as Thickbrick mentioned in IRC
[14:10] Merov Linden: that's the logical step
[14:10] Merov Linden: Morgaine: hold that thought a minute
[14:11] Merov Linden: Point 3: SG 2.0: llqtwebkit
[14:11] Morgaine Dinova: kk, leave to end
[14:11] Merov Linden: Techwolf said something about this at the beginning of the meeting
[14:12] Merov Linden: see you're having problem building... hmm... did you log a JIRA?
[14:12] Merov Linden: that's the best to pass the problem to the relevant guys (with patch if applicable)
[14:12] Thickbrick Sleaford: We're talking about the boost problems?
[14:12] Merov Linden: yes
[14:13] Merov Linden: whatever problem you have building this baby
[14:13] Techwolf Lupindo: Its a known bug with boost. I just don't have the boost jira/bugzilla number handy.
[14:13] Thickbrick Sleaford: boost 1.34 can't be build with gcc 4.4
[14:13] Thickbrick Sleaford: is that the one?
[14:13] Merov Linden: wait: they're still using 1.34?
[14:13] Merov Linden: we switched to 1.39 for all viewer 2.0
[14:13] Geneko Nemeth: Snowglobe is already on 1.39...
[14:14] Thickbrick Sleaford: I think sg2.0 had the same problem of misnamed boost packages
[14:14] Merov Linden: so I'm surprised
[14:14] Thickbrick Sleaford: the actual files in the package were from 1.34
[14:14] Thickbrick Sleaford: let me check
[14:14] Techwolf Lupindo: gcc 4.4.3 and boost 1.39 on my sytem.
[14:14] Merov Linden: yeah but that was fixed for SG 1.3 remember?
[14:14] Merov Linden: I hold the release for a long time because of that bug
[14:15] Merov Linden: you know what? I bet viewer 2.0 beta 3 doesn't have the fix we got for SG 1.3.2
[14:16] Merov Linden: actually, I'm pretty sure of this as I remember dscussing with Brad
[14:16] Thickbrick Sleaford: I think the same problem is present in the boost in SG2, but not sure anymore. Downloading the boost tarball now.
[14:16] Dzonatas Sol: Question about Snowglobe 2.0: Do we want to make sure 64 bit works for Snowglobe 2.0 and not worry about 64 bit on the official Viewer 2.0 Beta? (where do I insert this question)
[14:16] Techwolf Lupindo: 1.34 and 1.39 has some changes that require version check in llmath/llmath?????.cpp/h. I was working on it last night. But that is a seperate bug. I need to check snowglobe to see if its hard coded or #if. If not, 1.39 or higher is requried.
[14:17] Merov Linden: Dzonatas: I would say so though I'll push to get those patches merged into the official trunk, even if we don't build for linux 64 yet
[14:17] Techwolf Lupindo: boost is turning out to be a real PITA. :-)
[14:17] Merov Linden: the thing is: we'll eventually have to...
[14:17] Robin Cornelius: Dzontas, have you got 64bit breakage?
[14:18] Merov Linden: anyone else building llqtwebkit from source?
[14:18] Merov Linden: I'm curious...
[14:18] Thickbrick Sleaford: I jsut checked a new downlaod. The boost version in SG2 trunk has: #define BOOST_LIB_VERSION "1_34_1"
[14:18] Techwolf Lupindo: I've got a 64-bit build of snowglobe2.0 going, but with failing media.
[14:18] Dzonatas Sol: Viewer 2.0 beta bombs at launch with assertion failures, so i left it at that
[14:18] Morgaine Dinova: Techwolf: I'm on boost-1.35.0-r3 in Gentoo, but Gentoo stable is at1.41.0-r3 now. Should I get into sync, or will that cause trouble with Snowglobe?
[14:18] Dzonatas Sol: (64 bit being th reason)
[14:18] Latif Khalifa: flash doesn't work in 64bit browsers either
[14:18] Robin Cornelius: I've not built latest qtwebkit yet Merov
[14:18] Techwolf Lupindo: Morgaine, i'me still on 1.39 but have 1.41 slotted also.
[14:19] Morgaine Dinova: kk
[14:19] Techwolf Lupindo: I need to eselect boost set 2 and emerge -v world.
[14:19] Robin Cornelius: my 64bit builds of snowglobe2 trunk work fine apart from media
[14:19] Zorin Frobozz: Is there any performance benefit building 64bit?
[14:19] Dzonatas Sol: memory
[14:19] Techwolf Lupindo: Not so much as performance, but LOT more stable.
[14:20] Latif Khalifa: buiding 64 bit will slighty increase moemory usage
[14:20] Robin Cornelius: a native build using system upto date libs is more stable
[14:20] Merov Linden: ok ok: I *just* grep beta 4 and it says BOOST_LIB_VERSION 1_39
[14:20] Cylindrical Cage: whispers:
So you want to play with Psi's collar?
Think you can control her?
Think again!
Don't try to escape and you may regain your freedom ...
[14:20] Zorin Frobozz: Would be interesting to try 64 bit on OS X
[14:21] Geneko Nemeth: whispers: ...wait a second!
[14:21] Tillie Ariantho: I want SL 64bit on Win7. :-/
[14:21] Merov Linden: so, that will be fixed as soon as I turn this vendor branch stuff on and merge it to Snowglobe2 trunk
[14:21] Thickbrick Sleaford: so no need to re-open SNOW-431 ?
[14:21] JIRA-helper: http//jira.secondlife.com/browse/SNOW-431:
[#SNOW-431] downloads for: Snowglobe 1.3 aren't really 1.39.0
[14:21] Merov Linden: well, that really makes the point about finishing those scripts...
[14:22] Merov Linden: well, you can reopen it : technically it *is* broken right now
[14:22] Thickbrick Sleaford: ok
[14:22] Merov Linden: 'm glad I know where that comes from and that a solution is coming :)
[14:22] Thickbrick Sleaford: at least so people know they're no halucinating.
[14:22] Merov Linden: +1
[14:23] Techwolf Lupindo: I found a link. This is the problem I'me having with boost now. And due to gcc/boost interactions, ie:gcc build with boost version has to stay the same for the bianry to run, gcc version and boost version. https://svn.boost.org/trac/boost/attachment/ticket/2069/boost_mpl_gcc-4_4.patch
[14:23] Zorin Frobozz: Caging someone because they clicked on you might net you an AR, I'd be careful with that Psi
[14:24] Merov Linden: enters god mode
[14:24] Merov Linden: ;)
[14:24] Tillie Ariantho: :P
[14:24] Robin Cornelius: you should glow or somthing when you do that
[14:24] Tillie Ariantho: ^^
[14:24] Latif Khalifa: lol
[14:24] Merov Linden: ok, no one answered my question about building llqrwebkit? or trying to?...
[14:24] Techwolf Lupindo: So the boost bugs I have is local to my system, but its going to haunt LL someday when upgrading to gcc 4.4.x
[14:24] Latif Khalifa: display halo ;)
[14:25] Robin Cornelius: Merov i said i've not quite got upto date on that, i was keeping it built
[14:25] Robin Cornelius: i cloned the latest source
[14:25] Techwolf Lupindo: Merov, I have built llqtwebkit.
[14:25] Robin Cornelius: then the merge with my changes failed and i had to get on with some paid work
[14:25] Merov Linden: I know you 2 guys asked for all of this to come out so I'm not surprised :)
[14:25] Techwolf Lupindo: I removed all the hacks I had to do to get to use the system qtwebket and used the VENDER define.
[14:26] Merov Linden: ok, well, moving on to the next item
[14:26] Merov Linden: SG2.0: vendor branch and tag
[14:26] Techwolf Lupindo: Then had to fugout out wietehr ro not to still use .so or go with default of .a lib.
[14:26] Merov Linden: OK so, all of this will need a good wiki but, basically, we now have an exported trunk as we had in the old days
[14:26] Dzonatas Sol: I'm happy with a 32 biot llqtwebkit... so no hurry on a 64bit version
[14:27] Merov Linden: that the viewer-2-0 trunk
[14:27] Merov Linden: or rather: that *will be* the trunk once I turn the commit script on
[14:28] Techwolf Lupindo: I need to update the gentoo overlay to use the new trunks sometimes....I did find the oss-viewer was updated last week.
[14:28] Merov Linden: I tagged the beta3 while I was at it so we now have something that parallels what we did before
[14:28] Merov Linden: Morgaine: what was your question again?
[14:28] Latif Khalifa: which branches do you plan to export?
[14:28] Thickbrick Sleaford: Merov, does that mean that every commit done to trunk internally gets exported (with appropriate filtering)?
[14:29] Eddi Decosta: Merov, the beta3 without crash on FPS?
[14:29] Morgaine Dinova: Is "vendor branch" *any* vendor (with branding separate), or specifically vendor == LL?
[14:29] Merov Linden: right now the mapping is : vendor == LL
[14:30] Merov Linden: that's an export of our internal dev trunk
[14:30] Techwolf Lupindo: Is vender == oss-viewer in svn.secondlife.com?
[14:30] Merov Linden: filtered of the "not appropriate for GPL use" stuff (kdu, vivox)
[14:30] Merov Linden: Techwolf: yes
[14:31] Morgaine Dinova: oss-viewer is not Snowglobe?
[14:31] Eddi Decosta: ok, cause i have some serious drop of FPS, and this cause me some bad crash in beta3 or 4 *furstrating*
[14:31] Merov Linden: I have to say though that I don't really see which other vendor that could be Morgaine
[14:31] Merov Linden: trying to understand what else it could have been
[14:32] Thickbrick Sleaford: So that means that instead of big source drops there will be individual commits with the original commit message?
[14:32] Latif Khalifa: how is sg2 trunk going to be synced with viewer2 trunk? someone going to do it manual merges?
[14:32] Tillie Ariantho: You never know, maybe IBM branches out at some time, and then contributes back? Things like that happen elsewhere, too. ^^
[14:32] Merov Linden: another thing is: we will eventually "brand" that export "Snowglobe" directly so that noone uses the SL brand by mistake
[14:32] Morgaine Dinova: Merov: when you created "vendor branch", I was assuming that you're aiming for rebrandability, so you could do a deal with some other outfit to rebrand vendor branch to their brand.
[14:33] Merov Linden: but that's just concerning a couple of strings and icons
[14:33] Morgaine Dinova: Hence the question.
[14:33] Merov Linden: I see Morgaine
[14:33] Merov Linden: so may be "vendor" branch is a misnomer then
[14:34] Dzonatas Sol: I'll assume the move to hg is pushed to after Snowglobe 2.0 is in distribution
[14:34] Merov Linden: the idea is that, for 3rd party viewers makers, LL is a "vendor" (provides something)
[14:34] Merov Linden: that's the way I saw it
[14:34] Merov Linden: but that's a confusing term
[14:34] Robin Cornelius: upstream?
[14:34] Latif Khalifa: i think it's clear now that LL is the vendor mentioned, let's move on :)
[14:35] Merov Linden: may be I should just refer to it as "oss-viewer" , i.e. "the oss exported viewer"
[14:35] Merov Linden: that's less confusing (possibly)
[14:36] Dzonatas Sol: less confusing and opens up the option to cherry pick patches from oss-viewer to put into snowglobe
[14:36] Merov Linden: then, if someone wants to rebrand it under their name (3rd party viewer Foo), up to them
[14:36] Morgaine Dinova: LL isn't a vendor for TPVs though, it's not even "upstream" --- there is no more "upstream" once one moves to distributed SCM's like Hg --- it's distributed, everyone pulls from change sets from everyone else :-)
[14:36] Merov Linden: then there's the question of merging or syncing that branch with Snowglobe
[14:36] Merov Linden: Snowglobe trunk I mean
[14:36] Latif Khalifa: yeah, that's far more interesteing for devs
[14:37] Latif Khalifa: sg2.0 needs to keep up with oss-viewer
[14:37] Merov Linden: exact Latif
[14:37] Merov Linden: for the moment, I'm planning to do regular merges (weekly I think)
[14:37] Merov Linden: so that we don't stray too far off
[14:38] Thickbrick Sleaford: That sounds good
[14:38] Robin Cornelius: did you answer t-bricks question? are we going to see commit by commit in the logs?
[14:38] Morgaine Dinova: There's not even a trunk anymore once everything moves to Hg ... there's just branches, spread out across the community. It needs a rather differrent mindset :-)
[14:38] Techwolf Lupindo: I've heard some merges took up an entire month for one devoloper to do.
[14:38] Merov Linden: other opitions are on the table, including fully automatic but I think that could lead to problems (unstability)
[14:38] Merov Linden: what do folks here think?
[14:39] Robin Cornelius: I think if we can see the incomming commits it makes it far far easier, but sooner or later they will collide with snowglobe changes so a hand merge is probably needed to at least oversee the merge
[14:39] Merov Linden: Techwolf: a month? wow... that's a sea of changes...
[14:39] Dzonatas Sol: the history of the commits are significant
[14:39] Latif Khalifa: a weekly change to start with is probably good enough
[14:40] Latif Khalifa: but perhaps automatic merge into sg2 trunk is not a bad idea
[14:40] Thickbrick Sleaford: merov you mean weekly oss-view --> snowglobe, or weekly internal --> oss-viewer ?
[14:40] Merov Linden: I'm expecting some conflict resolution to be done for sure
[14:40] Morgaine Dinova: Merges will take next to no time once you've gone to Hg. If someone's change set is so large that it causes global trouble, refuse it. Bad factoring.
[14:40] Latif Khalifa: or apoint a merge meister
[14:40] Merov Linden: Thikbrick: weekly oss-viewer->snowglobe
[14:41] Thickbrick Sleaford: and contious internal --> oss-viewer?
[14:41] Merov Linden: internal->oss-viewer will be daily or more
[14:41] Latif Khalifa: oss-viewer is supposed to go to automatic daily or even per commit
[14:41] Thickbrick Sleaford: *continous
[14:41] Dzonatas Sol: if it is for the history, i'd vote for hg and learn it despite my bias for git
[14:41] Merov Linden: yeah, continuous ideally
[14:41] Thickbrick Sleaford: very cool
[14:41] Techwolf Lupindo: trunk is meant to change often. Branches, like RC, are ment to be seperate for testing and eventelly final relised. So automated into truck would be nice, but must have commint comment so we can see what was changed in the last few hours.
[14:42] Morgaine Dinova: There won't be a trunk, Tech. Just change sets on many branches.
[14:42] Robin Cornelius: #opensl-svn will suddenly be very useful if its commit by commit sync
[14:42] Latif Khalifa: Robine, yeah
[14:42] Latif Khalifa: -e
[14:42] Dzonatas Sol: i think svn is gonna be the main bottleneck no matter how we merge
[14:42] Geneko Nemeth: There probably will still be "mainstream" branches that new developers want to pull from and from which releases are made.
[14:42] Merov Linden: Morgaine: it's not a trunk that's true but there's always a reference version that's used as a rallying one
[14:43] Merov Linden: it's call "trunk" by analogy
[14:43] Robin Cornelius: or HEAD on git
[14:43] Geneko Nemeth: HEAD is not a branch, it's the top rev of the branch you're working on.
[14:43] Robin Cornelius: sorry Master
[14:43] Morgaine Dinova: HEAD is only one's own head, it's relative
[14:43] Geneko Nemeth: You might be thinking of "master"...
[14:43] Robin Cornelius: i'm not thinking ;(
[14:43] Morgaine Dinova: Hehe
[14:43] Geneko Nemeth: but I never work in the master branch.
[14:44] Merov Linden: anyway, I think the plan is rather clear here now
[14:44] Geneko Nemeth: And master is only one's own master, too.
[14:44] Latif Khalifa: merov, i think that once you sync sg2 to some rev of oss-vewer, you can do automatic merges from then on with mergin revision range memrge
[14:44] Merov Linden: just need to execute it (action: me)
[14:45] Latif Khalifa: seems like most people want very frequent if not automatic oss-viewer -> sg2 trunk merges
[14:45] Morgaine Dinova: This very readable article is worth absorbing. It really points out the change in mindset that has to come when working with Hg and Git -- [2]
[14:45] Techwolf Lupindo: Six months of nothing from LL is a very long time to wait for code drops and put many developers in a super long holding patteren.
[14:45] Merov Linden: Beeing the one on the hook if build is broken, I'm a little warry to do automatic merges
[14:46] Merov Linden: that's my biais here
[14:46] Dzonatas Sol: automate svn, thumbs up
[14:46] Merov Linden: internally, whoever push crap on the main and break it has to wear the "shreck ears" and fix immediately
[14:47] Morgaine Dinova: Svn is yesterday, needs to disappear fast for us as soon as all LL is Hg internally.
[14:47] Merov Linden: but if that breakage flows to us, I'm the only one on hook
[14:47] Latif Khalifa: Morgaine, i think we got your point about 4 post earlier ;)
[14:47] Techwolf Lupindo: Merov, do you have a furry on devolopement team? Thay may backfire due to wanted to wear them ears all the time. :-D
[14:47] Thickbrick Sleaford: lol
[14:47] Merov Linden: Morgaine: LL is hg internally *but* the internal repo is a mess with everyting mixed (server, viewer, whatever)
[14:48] Merov Linden: so we can't "hg clone" and be done
[14:48] Latif Khalifa: Merov, the only worring bit is that if merges are manual and your're bogged down with some work it doesn't happen
[14:48] Morgaine Dinova: Latif: yeah I know, was just repeating it for your benefit ;-)
[14:48] Latif Khalifa: perhaps more than one person can be merge meister, one from the oss community
[14:48] Merov Linden: Latif: you're right , I have a choice of 2 evils
[14:48] Latif Khalifa: merov, you and robin :D
[14:49] Latif Khalifa: or some other voluteer :)
[14:49] Merov Linden: Latif: you have a point indeed
[14:49] Latif Khalifa: just to avoid single-person-of-failure scenariou
[14:49] Merov Linden: we could rotate that merge job among the ones with commit provileges
[14:49] Merov Linden: *privileges*
[14:49] Latif Khalifa: would make it easer on all methinks
[14:49] Robin Cornelius: commit monkey
[14:49] Robin Cornelius: ;-p
[14:49] Thickbrick Sleaford: hides behind his sit
[14:50] Morgaine Dinova: Maybe "privilege" should be renamed to "burden" ;-)
[14:50] Techwolf Lupindo: Don't look at me. I would eat the monkey.
[14:50] Merov Linden: and a policy that's : roll back if break, don't try to fix oss-viewer
[14:50] Techwolf Lupindo: :-)
[14:50] Latif Khalifa: yeah oss-viewer should be pretty much readonly on the public side
[14:50] Dzonatas Sol: or drop if community then works fully on hg
[14:50] Merov Linden: so we don't have shreck ears :)
[14:51] Latif Khalifa: hehe
[14:51] Merov Linden: that's a really good suggestion (especially since it's less work for me :) )
[14:51] Latif Khalifa: i knew you'd like it :)
[14:52] Merov Linden: still. I don't want to impose a burden on volunteers
[14:52] Robin Cornelius: merov can you mask some branches to read only in the svn? would ensure that and prevent any stupid accidents
[14:52] Merov Linden: I mean: it's not the most fun stuff to do
[14:52] Robin Cornelius: oss-viewer should only be updated by your scripts
[14:52] Merov Linden: Robin: I actually don't think I can
[14:52] Merov Linden: need to check that with CG
[14:52] Latif Khalifa: well you could start with getting oss-viewer auto sync going, the try doing initial oss-viewer -> sg2 trunk merge, then we can talk how to proceed from then on
[14:53] Merov Linden: Latif: that's the reasonable plan
[14:53] Latif Khalifa: the community committers will ahve time to think about the merge meister thing in the meanwhile
[14:53] Merov Linden: I've my work cut out for this week for sure
[14:53] Dzonatas Sol: so now, we just need a volunteer to handle the hg repos for snowglobe
[14:53] Latif Khalifa: Merov, you looks like you do :)
[14:54] Merov Linden: hehe
[14:54] Dzonatas Sol: forgets the name of the hg website...
[14:54] Thickbrick Sleaford: hs.secondlife.com
[14:54] Merov Linden: ok, I think we actually covered all the other points while discussing this
[14:54] Thickbrick Sleaford: or bitbucket.org/lindenlab
[14:54] Merov Linden: i.e. next steps etc...
[14:55] Merov Linden: Robin: we have 5 minutes for your JIRA issues
[14:55] Robin Cornelius: SNOW-577
[14:55] JIRA-helper: http//jira.secondlife.com/browse/SNOW-577:
[#SNOW-577] star maps:
[14:55] Merov Linden: want to jump in?
[14:55] Robin Cornelius: a fun one for a change
[14:55] Latif Khalifa: hehe
[14:55] Robin Cornelius: and SNOW-581
[14:55] JIRA-helper: http//jira.secondlife.com/browse/SNOW-581:
[#SNOW-581] media_plugin_base undefined:
[14:55] Jacek Antonelli: Neat :D
[14:56] Thickbrick Sleaford: Robin, did you think about doing the stars db in llsd xml, so people can make their own star maps more easily?
[14:56] Merov Linden: Beautiful!
[14:56] Robin Cornelius: no i just used the existing astronment formats from the yale BSC
[14:56] Thickbrick Sleaford: that would also simplify the reading code
[14:57] Robin Cornelius: well i could write a parser YBSC->LLSD in 5 mins then change the viewer code to read LLSD stars
[14:57] Thickbrick Sleaford: I guess it doesn't matter if the YBSC format is well defined
[14:57] Jacek Antonelli: Hrm, personally I think using the existing standard makes sense. Are there tools out there that work with YBSC?
[14:57] Merov Linden: nice things is that it's not pulling GB of data from Google Earth or somthing
[14:57] Morgaine Dinova: Merov: I'm a little uncertain now about the GPL "corresponding source" that will be provided for LL Viewer-2.* releases, once you've completed your work. Will that be a tagged version living in oss-viewer branch? Or will it be something else?
[14:58] Techwolf Lupindo: re:llqtwebkit. One problem that I have to fix on my sytem is this error: libmedia_plugin_webkit.so: undefined symbol: _ZN7QObject5eventEP6QEvent
[14:58] Robin Cornelius: yes, but there are a binary and text format of that, the binary format has less information, but there are a handful of programs on lnux that use those formats
[14:58] Merov Linden: Morgaine: I think we (I) should create a relevant tag or each release as I just did for beta 3
[14:58] Robin Cornelius: bear in mind YBSC is a 1970 format
[14:58] Jacek Antonelli: Okay. It's not important to me which format it's in, so *shrug* :)
[14:59] Robin Cornelius: its messy parsing it
[14:59] Morgaine Dinova: Merov: aha, cool. But it comes from oss-viewer branch, right? (Will come I mean, eventually)
[14:59] Robin Cornelius: anyway please add ideas to the jira
[14:59] Robin Cornelius: any comments on SNOW-581
[14:59] Robin Cornelius: its more a discussion point than an actual patch
[14:59] Thickbrick Sleaford: I was thinking that having the abilty to create non-realistic starscapes is very useful for machinma, etc.
[15:00] Merov Linden: Yes... which is an export from the internal dev trunk
[15:00] Morgaine Dinova: Gotcha. Thanks :-)
[15:00] Merov Linden: so it's the same for all intent and purposes, no?
[15:00] Merov Linden: checks 581
[15:01] Latif Khalifa: robin my only worry for SNOW-581 is the merging problems it might cause
[15:01] Robin Cornelius: yea thats why i want to get discussion going and get something acceptable in side LL
[15:01] Merov Linden: yeah: that's one where we need the relevant dev engaged
[15:01] Robin Cornelius: yea they might have some grand plan
[15:02] Merov Linden: I can chase him down (only a couple of suspects... )
[15:02] Morgaine Dinova: Merov: it's a fine point which I'll let the lawyers worry about. :-) As long as it's functionally indistinguishable from Viewer-2.0 (minus KDU and Vivox), I don't see how FOSS peeps can complain.
[15:02] Jacek Antonelli: waves and scoots over to UXIG. Bye all!
[15:02] Latif Khalifa: yeah and perhaps wait for it come via oss-viewer if they decide to adopt it
[15:02] Thickbrick Sleaford: wonders how long before there's a message template parser for plugin messages
[15:03] Merov Linden: ok, it's 3pm and I have another meeting starting in 5 min...
[15:03] Robin Cornelius: Merov, Monroe was media team and I dragged him in to openSL, never got Asi there
[15:03] Merov Linden: <grin/>
[15:03] Robin Cornelius: or llsam either
[15:03] Latif Khalifa: ok Merov, good luck with all the work we scheduled for you for the next week :)
[15:03] Merov Linden: thanks for coming today

Open Source Meeting/2010-03-18

Navigation menu

Search