User:Zero Linden/Office Hours/2009 jun 02

[12:58] Saijanai Kuhn: hey ∞
[12:59] Infinity Linden: hola amigos y amigas
[12:59] xstorm Radek: eeeppp
[12:59] xstorm Radek: hi Infinity :-)
[12:59] Saijanai Kuhn: https://wiki.secondlife.com/wiki/AW_Groupies/Chat_Logs/AWGroupies-2009-06-02
[13:00] Mojito Sorbet: 今日は
[13:00] Infinity Linden: cool. thx again for being the note taker, sai
[13:00] Saijanai Kuhn: I have to go back and find a month's worth from may :-/
[13:00] Saijanai Kuhn: was there in virtuality but not in spirit
[13:01] ATechwolf Foxclaw: Any zeros here? :-)
[13:01] Infinity Linden: 明日
[13:01] xstorm Radek: ∆∏∑
[13:01] Infinity Linden: zero's on his way, i'm sure
[13:01] Patnad Babii: hello people :)
[13:01] Infinity Linden: i just saw him headed for his desk
[13:01] xstorm Radek: :-)
[13:01] Saijanai Kuhn: Yo O Cardinality of the Null Set
[13:01] Patnad Babii: Hi Zero
[13:01] Infinity Linden: see... infinity has zero summoning powers
[13:02] xstorm Radek: *GIGGLES* :)Saijanai 01:38, 3 June 2009 (UTC)
[13:02] Saijanai Kuhn: AH, this morning's meeting, zero: https://wiki.secondlife.com/wiki/AW_Groupies/Chat_Logs/AWGroupies-2009-06-02
[13:02] Rex Cronon: hello invisible people
[13:02] xstorm Radek: hi Rex
[13:02] Saijanai Kuhn likes showing off his 1337 regex skills
[13:03] Rex Cronon: hi xstorm
[13:03] Rex Cronon: i don't see anybody
[13:03] xstorm Radek: ĦĦĦ
[13:03] Infinity Linden: sai... http://xkcd.com/208/
[13:03] Imaze Rhiano: peekaboo
[13:04] Zero Linden: encode Ħ
[13:04] xstorm Radek: yep
[13:05] Saijanai Kuhn: cute
[13:05] Rex Cronon: how did u display that char?
[13:05] Rex Cronon: Ħ
[13:05] Rex Cronon: i can copy it
[13:05] Rex Cronon: and past it
[13:06] Saijanai Kuhn: use the extended atoi lib on the wiki
[13:06] Rex Cronon: paste*
[13:06] Rex Cronon: .
[13:06] Rex Cronon: ?
[13:06] Rex Cronon: link please
[13:06] xstorm Radek: same symble thats used in china and many other backrounds it has a good history
[13:06] Infinity Linden: hmm... no zha yet
[13:06] Saijanai Kuhn: though, I don't know how to convert utf-8 to utf-16 and so on
[13:06] Imaze Rhiano: OMG! What has happened for Master Sai's head?
[13:07] artoo Magneto: pardon are you using voice or only chat ?
[13:07] Infinity Linden: sai.. it involves sacrifices and dead chickens
[13:07] Zero Linden: we use text chat only
[13:07] Zero Linden: usually
[13:07] artoo Magneto: k
[13:07] xstorm Radek: *GIGGLES* :)Saijanai 01:38, 3 June 2009 (UTC)
[13:07] Zero Linden: ah - that symbol ISN'T a chinese symbol at all
[13:07] xstorm Radek: nope
[13:07] xstorm Radek: but its like one
[13:07] Mojito Sorbet: Good. This morning the Enterprise team used voice. Most annoying
[13:07] Zero Linden: it is an H with a stroke
[13:07] Saijanai Kuhn: rex: https://wiki.secondlife.com/wiki/Library_Combined_Library
[13:07] xstorm Radek: yep
[13:08] xstorm Radek: he he he he he
[13:08] Rex Cronon: u+0126
[13:08] Zero Linden: really, only in this poor excuse for a font! ha ha
[13:08] Rex Cronon: U+0126
[13:08] Saijanai Kuhn: I have a hat that sorta does this
[13:08] Infinity Linden: my favorite is U+203d
[13:08] Mojito Sorbet: It certainly confused my Japanese dictionary
[13:08] Infinity Linden: the interobang
[13:08] Zero Linden: U+203D
[13:08] Rex Cronon: thank you sai
[13:09] Infinity Linden: interrobang is the name of my favorite super hero
[13:09] Zero Linden: okay - it is almost 10 after
[13:09] Infinity Linden: he has the ability to make you question statements made while excited
[13:09] Skills Hak: hiyas
[13:09] Saijanai Kuhn: The most interesting advertising icon in the world?
[13:09] artoo Magneto: hey skills
[13:10] xstorm Radek: ψ
[13:10] Zero Linden: (say "Encode ψ" to make it show up on the object
[13:10] xstorm Radek: :-)
[13:10] xstorm Radek: ok
[13:10] Dahlia Trimble: so what's the agenda?
[13:10] Saijanai Kuhn: zha's online, presumably afk
[13:10] Zero Linden: SO -
[13:10] Zero Linden: The agenda is this: OGPX!
[13:10] xstorm Radek: Encode ψ
[13:10] Zero Linden: Wooooot!
[13:10] Dahlia Trimble: X?
[13:10] Zero Linden: encode ψ
[13:11] xstorm Radek: caps
[13:11] Zero Linden: Well - we needed a name to distinguish the group from the protocol
[13:11] xstorm Radek: greek
[13:11] Infinity Linden: maybe we could call it OGP chi
[13:11] Dahlia Trimble: if it's X should we move this meeting to the adult continent?
[13:11] xstorm Radek: that sounds great
[13:11] Zero Linden: I think, alas, that while the IETF is VERY aggressively pro-Unicode in its protocols
[13:11] xstorm Radek: *GIGGLES* :)Saijanai 01:38, 3 June 2009 (UTC)
[13:12] Rex Cronon: encode U+00E0
[13:12] Zero Linden: It is quite ASCII-phillic for its communications
[13:12] Zero Linden: U+00E)
[13:12] Zero Linden: U+00E)
[13:12] Zero Linden: dnag
[13:12] Zero Linden: U+00E0
[13:12] Zha Ewry: Remarkably 1970s centrick markup
[13:13] xstorm Radek: encode Ж
[13:13] Zero Linden: so - hoping to start with a brief status update from Infinity
[13:13] Patnad Babii: are you going to the stockholm IETF with OGPX ?
[13:13] Zero Linden: Infinity - (sorry to put you on the spot) - where's the OGPX effrot at?
[13:13] Infinity Linden: so
[13:13] Infinity Linden: we have the mailing list created
[13:13] Mojito Sorbet: encode 酒
[13:14] Rex Cronon: U+0415
[13:14] Infinity Linden: and we've put in our request for a bof
[13:14] Infinity Linden: and there's beginning to be discussion on the ogpx list
[13:14] Infinity Linden: that's it in a nutshell
[13:14] Infinity Linden: i can do more detail if peeps want
[13:14] Zero Linden: cool
[13:15] xstorm Radek: encode ≈
[13:15] Zero Linden: well - how about two / three sentences on the aim of OGPX, in your view
[13:15] Infinity Linden: OGPX is the group intended to sheperd the OGP RFCs through the standardization process
[13:16] Zero Linden thinks he built that widget just so any transcript of his office hours would have to be in a fully Unicode-compliant system!
[13:16] Infinity Linden: OGP as everyone recalls, is the next-gen VW protocol
[13:16] Infinity Linden: its distinguished from LESS by it's focus on "second life-like" worlds
[13:16] Infinity Linden: like LL, OpenSim, RealXtend,
[13:16] Dahlia Trimble: so if there is going to be a standardization, that implies that there is an implementation?
[13:17] Infinity Linden: there are several existng bits of implementation
[13:17] Patnad Babii: will there be a phase 2 to the OGP test, teleport from opensim to SL grid soon?
[13:17] xstorm Radek: and do not forget they are beta testing blue mars out there now
[13:18] Infinity Linden: part of the standardization process is for all of us (us being OpenSim, PyOGP, LL) to agree ons ome basics so we can work on more "intresting:" problems
[13:18] Infinity Linden: more intersting than simply logging in
[13:18] Zero Linden: We don't, at this point, have specific plans for phase 2 tests
[13:18] Pixel Gausman: :(
[13:18] Pixel Gausman: best way to test RFCs is thru implementation
[13:18] Zero Linden: Agreed
[13:19] xstorm Radek: is this do to standard problem ?
[13:19] Zero Linden: however, LL doesn't have a lot of spare engineering resources right now
[13:19] Infinity Linden: well... we learned a number of things from last year's interop
[13:19] Zero Linden: (anyone notice Wall St.....?)
[13:19] Infinity Linden: interop work..
[13:19] xstorm Radek: yes
[13:20] xstorm Radek: i lost just about all my money
[13:20] Zero Linden: and so, for now, we will be focusing on writing specs, implementing and testing things in pieces, and working through standards process
[13:20] SC.2.1.A.TL5: hit by Morgaine Dinova (Morgaine Dinova)
[13:20] Zero Linden: those things can take time anyway - and as much as I'd love to lead a team of six engineering implementing phase 2 right now.....
[13:20] Techwolf Lupindo: (Ys. Cisco got listed in dowjones and cause it to go up yesterday)
[13:20] Zero Linden: ... we aren't
[13:20] xstorm Radek: i have all of 2k of my life savings left from stock drop
[13:20] Infinity Linden: but before the specifications can move to be a "standard" there have to be multiple implementations
[13:20] Infinity Linden: so
[13:21] Pixel Gausman: does IETF like to see implemented specs before ratification?
[13:21] Saijanai Kuhn channels Morgain: https://wiki.secondlife.com/wiki/AW_Groupies_In-World_Meeting_Agenda
[13:21] Dahlia Trimble: so nobody is implementing?
[13:21] Morgaine Dinova: Thanks Sai
[13:21] Zero Linden: We've also seen that the IETF (and other standards bodies) prefers that small, clearly scoped peices be bitten off, chewed, digested, and put into
[13:21] Infinity Linden: nope. IBM's been working on some implementations
[13:21] Zero Linden: the standards hopper before large other-worldly visions are set forth
[13:21] xstorm Radek: is this the backseat type of thinking ?
[13:21] Zero Linden: :-)
[13:22] Holger Gilruth: /ao off
[13:22] Infinity Linden: we have what we did last summer
[13:22] Infinity Linden: our position at the moment is we want to consolidate our gains frmo last summer
[13:22] xstorm Radek: basing every thing on 2 year old data ?
[13:22] Infinity Linden: and agree to what the spec SHOULD look like
[13:22] Zero Linden: So - I think concentrating on getting the first, small things designed, agreed upon, implemented, and into the queue
[13:22] Zero Linden: will significantly grease the skids for the future
[13:23] Infinity Linden: how often does HTTP change?
[13:23] Morgaine Dinova: "Consolidate our gains from last summer"? Does that have a technical translation outside of manager speak?
[13:23] Infinity Linden: we goofed on a number of things
[13:23] Zero Linden: yeah - i means turn them into something concrete
[13:23] Infinity Linden: we;re fixing htem
[13:23] xstorm Radek: there has been updates yes
[13:23] Infinity Linden: rather than do a complete implementation and THEN do the standard
[13:24] Infinity Linden: we're doing a bit of the implementation, then a bit of the standard
[13:24] Infinity Linden: then we do a bit more implementation, then a bit more standard
[13:24] Dahlia Trimble: why not just document and standardize the existing sl protocols
[13:24] Pixel Gausman: so what's the next bit to implement?
[13:24] Mojito Sorbet: That would be MSFTs approach
[13:24] Saijanai Kuhn: Dahlia, 95% of them are still in UDP
[13:24] Rex Cronon: this has the potential to go on forever:(
[13:24] Dahlia Trimble: so?
[13:24] Saijanai Kuhn: and not necessarily well-designed packets at that
[13:25] Mojito Sorbet: UDP　is not exactly robust
[13:25] Zero Linden: really - they aren't something that is a viable, open, extensible framework
[13:25] Infinity Linden: rex. large software development projects _do_ tend to have long lifecycles
[13:25] Zero Linden: we could just write it up and declare it *done* -- but it wouldn't be the kind of protocol that would really support a wide ecosystem
[13:25] Infinity Linden: we don't want to limit our implementation in 2020 to our view of the virtual world in 2009
[13:25] Saijanai Kuhn looks at Java 1.6, er, 6
[13:25] Dahlia Trimble: long lifecycle projects are often rendered dead by disruptive technology
[13:26] Rex Cronon: i agree infinity
[13:26] Infinity Linden: and anyone remember when simonyi wrote the first version of word?
[13:26] Mojito Sorbet: A long-running process can spin off incremental steps along the way though
[13:26] xstorm Radek: is there any backup plans just incase a problem is found or a better steps have been found ?
[13:26] Infinity Linden: dahlia.. what are these disriuptive technologies? maybe we could plan now to avoid their ramifications?
[13:27] xstorm Radek: soo all the eggs are not in one basket ?
[13:27] Dahlia Trimble: how about hypergrid
[13:27] Morgaine Dinova: One incremental step per year isn't a useful rate. Any plans to speed up Gridnauts' successes to more than 1 per year?
[13:27] Infinity Linden: the sad/happy truth is... you deal with the world as it comes
[13:27] Dahlia Trimble: or blue mars
[13:27] Dahlia Trimble: or O3D
[13:27] Infinity Linden: blue mars is interesting
[13:27] Mojito Sorbet: BM claims they can scale up real big.
[13:27] Zero Linden: We'd love to have involvement from hypergrid
[13:27] Infinity Linden: but i don't see them opening their protocol stack anytime soon
[13:28] Infinity Linden: we did actually ask them to participate
[13:28] xstorm Radek: they have a good base to work on but i do not think they will last in the long run
[13:28] Infinity Linden: they declined
[13:28] Rex Cronon: but, as morgain said, the time interval between steps is too big
[13:28] Infinity Linden: @rex
[13:28] Infinity Linden: true
[13:28] xstorm Radek: they are basing it to much as a game style
[13:28] Infinity Linden: and this is the problem with the current economic situation
[13:29] Zero Linden: I think the first aim is a sucessful specification of low level messaging framework, and establishment of login and teleport (or agent auth, and presence establishment) would be a real accomplishment
[13:29] Zero Linden: for the next year
[13:29] Infinity Linden: linden is a big place, and it takes a while to "steer the ship"
[13:29] Pixel Gausman: eventqueue needs some fleshing out
[13:29] Infinity Linden: or turn the ship, as it were
[13:29] xstorm Radek: in any good company you have to have a great base to build on
[13:29] Infinity Linden: @pixel
[13:29] Infinity Linden: right
[13:29] Infinity Linden: we're going ot do that on the list
[13:29] Saijanai Kuhn mutters something about group IM and goes back to sleep
[13:29] Dahlia Trimble: the OGP carrot has been nibbled down to a stub and all that's left is the stick it was hanging from
[13:29] Zero Linden: this would mean that - using just that amount of base, many different kinds of projects could build on those base concepts and explore a pretty wide variety of virtual world situations
[13:29] xstorm Radek: is people start to see its not one the right path it will not last long
[13:29] Saijanai Kuhn: https://wiki.secondlife.com/wiki/AW_Groupies_In-World_Meeting_Agenda point 2
[13:30] TheBlack Box: someone needs to check out how to set the sails towards hierachical linking
[13:30] Pixel Gausman: we need to start toying with some implementations of inventory using OGP
[13:30] Infinity Linden: well... the benefit to large slow moving objects is that they do have inertia
[13:31] Geo Meek: networks are getting smaller
[13:31] Infinity Linden: so it's unlikely OGP will disappear.
[13:31] Zero Linden: Toy away, pixel!
[13:31] Infinity Linden: i think our executives have publicly announced we see an open future
[13:31] Zero Linden: no - really - I mean, this is how this will work - folks will explore, try out things
[13:32] Dahlia Trimble: if you think the rest of the world isnt developing more exciting virtual world technologies, you may be in for a rude surprise
[13:32] Zero Linden: if we have a common framework (OGP Base) then we can more easily share experiments
[13:32] Zero Linden: and then more easily refine them
[13:32] Saijanai Kuhn: Philip mentioned the Boz bit, er, interop flag, in Groupies a few weeks ago, Zero
[13:32] Zero Linden: and then move them to being std.
[13:32] Saijanai Kuhn: bozo*
[13:32] Infinity Linden: @dahlia... participating in OGP does not limit you from participating in HyperGrid efforts
[13:33] Infinity Linden: there's no reason an opensim instance can't run both OGP and HG
[13:33] Dahlia Trimble: infinity, I am participating in hypergrid efforts :)
[13:33] Zha Ewry: One of the bigger challanges, is the lack of a current OGP client which is beinf supported
[13:33] Infinity Linden: then what are you complaining about?
[13:33] Zero Linden is not sure what the "bozo bit" or "interop flag" is
[13:33] Pixel Gausman: the viewer is an issue
[13:33] Dahlia Trimble: I'm complaining about the total lack of any usable solutions from the OGP efforts
[13:34] Pixel Gausman: for OGP, the viewer needs lots of work
[13:34] Saijanai Kuhn: A flag to signal that permissions are granted for taking inventory off-world
[13:34] Infinity Linden: it was one of the proposed implementations for how content owners flag items that can be taken off the main grid and rezzed in a distant grid
[13:34] Latif Khalifa: Zha, how can you have OGP client when you don't have OGP?
[13:34] Zha Ewry: Well, even one that supports what's been done, and is up to speed with the rest of the tree
[13:34] Latif Khalifa: what has been done? you mean login? :P
[13:35] xstorm Radek: login ?
[13:35] Zha Ewry: Login and TP, and.. even those need the modded client.. as will future stuff
[13:35] Latif Khalifa: well that's pretty much what OGP is now as implemented, login and teleport
[13:35] Pixel Gausman: the viewer needs a forward merge, and there are some deep seated issues that need to be worked in viewer code.
[13:36] Pixel Gausman: like regionhandles
[13:36] xstorm Radek: ? will a data base we have now support coding like that ?
[13:37] Infinity Linden: yes. region handles as 64 bit ints need to die
[13:37] Zero Linden: We understand the issue about support for today's OGP in the viewer somehwere
[13:38] Zero Linden: And, I we are talking with the Snowglobe effort to see if that will be a good long term home for that code
[13:38] Zero Linden: I'm quite hopeful there
[13:38] Morgaine Dinova: Is Philip's Snowglobe a potential OGP client, or is yet another branch required? Zha's work is really hampered by this. We examined some of the issues in the AW Groupies meeting today.
[13:38] Pixel Gausman: careful with that "hope"
[13:38] Pixel Gausman: it'll get ya every time. :)
[13:39] Pixel Gausman: Snowglobe might be a good home for the OGP viewer code.
[13:39] xstorm Radek: we are back to the eggs in that basket again it seems
[13:39] Youri Ashton: snowglobe testing is not done yet aswell, im one of the testers and so is Philip Linden
[13:39] Zero Linden: Well, xstorm - please - build another OGP client!
[13:39] Zero Linden: :-)
[13:40] xstorm Radek: ya got me there ;-)
[13:40] Latif Khalifa: hmm perhaps we could build non graphical client
[13:40] xstorm Radek: may be in 12 years
[13:40] Morgaine Dinova: Zero: that's not a good answer to xstorm, because working on another client will not help you get where you want to go. And we have pyogp already.
[13:40] Pixel Gausman: xstorm: that's the trump card. a new viewer is a big amount of work.
[13:40] Latif Khalifa: for protocol testing
[13:40] Skills Hak: ASCII <3
[13:40] Youri Ashton: why a non graphical client? dont you want to see anything?
[13:41] xstorm Radek: ?
[13:41] Youri Ashton: you need graphics to see things
[13:41] Techwolf Lupindo: I would like to have a working non-3D client as sometimes one doen't need it.
[13:41] xstorm Radek: graphis are made up od code
[13:41] Saijanai Kuhn: well, a strategy for using pyogp is to use a normal client to get a 3D view of the pyogp client's actions
[13:41] Geo Meek: vary good point Techwolf
[13:42] Infinity Linden: and there are plenty of thigns that are "interesting" that don't have to do with 3d: inventory management, friends list, IM, group voice, etc
[13:42] Zero Linden: xstorm - as Morgain points out - there are multiple implementations of OGP, though indeed most are thin
[13:42] Morgaine Dinova: We're talking about protocols here anyway. How or whether the client renders graphics is interesting and useful, but doesn't directly impact on the protocol. Or shouldn't.
[13:42] Talin Sands: yes i aggree a non-Graph client will help make these little chunks easyer to standardize
[13:42] Zero Linden: but *I* can't directly help the "all in one basket" -- I'm involved in one effort: Linden Lab/Second Life
[13:42] xstorm Radek: true
[13:42] Techwolf Lupindo: And such a client can help a lot in the area of accessiabley. I saw a seeing eye dog the other day that blind folks use.
[13:42] Zero Linden: I see several other efforts here
[13:43] Youri Ashton: adding a viewer that asks very little of any pc would be a good adition for some people. not everyone can afford a really good pc
[13:43] Zero Linden: so - the baskets are really there to be woven (to really overuse a metaphor)
[13:43] Infinity Linden: and more peeps going out and implementing what's in the specs will help everyone... it'll find bugs in the spec and maybe new usecases
[13:43] Pixel Gausman: Youri: good point, but off topic.
[13:43] xstorm Radek: :-)
[13:43] Infinity Linden: @techwolf. pathfinder and i have you covered
[13:43] Dahlia Trimble: unfortunately nobody else has any incentive to implement OGP
[13:44] Youri Ashton: i was nearly replying to Latif Khalifa
[13:44] Mojito Sorbet: It is important for a nongfx client that there be a way to have the server NOT　send any object info.
[13:44] Infinity Linden: pathfinder is interfacing with the disabilities community and i talk with him to get updates on where accessibility is critical
[13:44] Pixel Gausman: Dahlia: i think OGP between OpenSim grids is a valid use case (as is hypergridding)
[13:44] Infinity Linden: so we can make sure there's nothing in the spec that makes accessibility hard
[13:44] Techwolf Lupindo: Infinity, thank you. Even though I have fur and don't need covering. ;-)
[13:44] Infinity Linden: +1 pixel
[13:44] Zero Linden: +1 pixel
[13:44] Saijanai Kuhn: we have a newish member I met at pathfinder's the other day. She's very much into the enabling clients usecase
[13:44] Pixel Gausman: linking OpenSim grids would be a good thing
[13:45] xstorm Radek: lets hope some young person on teengrid thinks of some thing new when they get in to MIT
[13:45] Dahlia Trimble: Pixel, unfortunately I dont know of any opensim users who are asking for OGP
[13:45] Pixel Gausman: Dahlia: /me is
[13:45] Saijanai Kuhn: Louise Later. I think she's one of the people Pathfinder deals with
[13:45] Youri Ashton: if I understood it correctly, something like that is goign to happen already, Pixel
[13:45] Dahlia Trimble: well we need an AD then
[13:45] Infinity Linden: right. i _think_ i've talked to louise as well
[13:46] Pixel Gausman: Dahlia: yes, we do
[13:46] Zero Linden: Okay - so, this conversation isn't heading in a valuable direction
[13:46] Latif Khalifa: well the prospect of Linden Lab implemeting OGP on their grid within the next 3 years is slim to none, hence no one is really interested in developing it
[13:46] Zha Ewry: And AD with a matching client, and some code sync
[13:46] Morgaine Dinova: Latif++
[13:46] Youri Ashton: Which Linden said something about that if i understood everything alright. including giving people their own offline sim if they wanted (correct me if im wrong about this @ Lindens)
[13:47] Zero Linden: Look- we are NOT here to debate the manner in which things should be done or if OGP is right, or if something about the way Linden is doing things
[13:47] xstorm Radek: is there going to be a way to let other companys link in to SL from there servers and there VR System down the line ?
[13:47] Mojito Sorbet: That is the inside-the-firewall project
[13:47] Infinity Linden: okay latif. so you're suggesting we tell JHurliman, AFrisby and Teravus that you think we shouldn't work with them? what?
[13:47] Zero Linden: OGP is of common interest - and Linden is supporting that effort with Infinity full time, and myself on it
[13:48] Zero Linden: we beleive that it is a good start - and we think that plenty of you do here too
[13:48] Zero Linden: SO - Here's the thing
[13:48] Dahlia Trimble: sorry it
[13:48] xstorm Radek: yes
[13:48] Morgaine Dinova: What latif says is the reality. At this rate, a fully specc'd OGP-based system would take half a decade, if not longer. It's no surprise at all that Opensim isn't interested, because the OGP progress curve tells them that it'll be obsolete before it ever does anything. Hence the pressure for Gridnauts step #2.
[13:48] Zero Linden: Let's each talk about what we are going to do
[13:49] Zero Linden: Not what we all think the others of us should do
[13:49] Dahlia Trimble: sorry, it's not my intention to disrupt any OGP efforts, I'd just like to see some incentives
[13:49] Infinity Linden: again. morgain. you're speaking for the entire OpenSim development community?
[13:49] Dahlia Trimble: nobody speaks for the entire opensim community
[13:49] Infinity Linden: dahlia... you could participate in the OGP discussions
[13:49] Dahlia Trimble: I have :)
[13:50] Rex Cronon: maybe the lab could hve something like the X-Prize:)
[13:50] Zero Linden: Well - If you don't see incentive enough in the visition we've all worked the last 20 months to co-develop - I don't have anything else to offer you
[13:50] Saijanai Kuhn would like to pull us back into practicalities if he may?
[13:50] Saijanai Kuhn: as I understand it, LL is working on OGPx documentation
[13:50] Morgaine Dinova: Infinity: no, I speak for nobody. But anybody, including yourself, should be able to see what the current OGP growth rate is. I'm sure you wouldn't deny it either, if you didn't spend all your time making straw men remarks like who I represent.
[13:50] Zero Linden: I'm sorry, Rex, we don't have any billionare men-of-leisure on our board
[13:50] Dahlia Trimble: the vision appears to be fading :(
[13:50] Techwolf Lupindo: The lack of visiable progress is what got a few folks a bit cenicial here I think. I keep hearning about this opengrid thing, but havn't seen anything usable or some sort of visiable progress on it. If there is, i'me just not seeing it.
[13:51] Saijanai Kuhn: Zha is working on an AD implementation in OpenSIm
[13:51] Youri Ashton: aight, im going again. i got a meeting with the rest of the translators and Danica Linden in a few minutes. thnx for the meeting Zero
[13:51] Youri Ashton: bye bye everyone
[13:51] Pixel Gausman: and i'm fussing with the viewer
[13:51] Rex Cronon: tc
[13:51] Zero Linden: welcome, Youri
[13:51] Dahlia Trimble: bye :)
[13:52] Saijanai Kuhn: from my perspective: we need some progress to a "phase 2" of some kind for OGP, and within/part of that phase 2, we need things like are mentioned in point 2 of the AW Groupies agenda
[13:52] Infinity Linden: morgaine, dahlia, there's no obligation that you participate. but. we do have a community forming around OGPX
[13:52] Zha Ewry: And we're trying to end up with somethign that's actually going to be usable by the whole community, not livbing on a 3 rev back client
[13:52] Zero Linden: And we are writing docs, and working on the rather complex set of interactions in identity and authentication and other developments in automated services, trying to see where they need to dovetail into existing work, and where they don't
[13:52] Saijanai Kuhn: not to mention Zha's need for a nimble client to work with for testing her code
[13:52] Rex Cronon: i was thinking something arund 10k:)
[13:52] Infinity Linden: we worked very hard internally to convince our management to shake loose funding for the current work
[13:52] Pixel Gausman: so Zero, is there a next chunk that we chew off and implement?
[13:52] Infinity Linden: and we're finally getting some traction
[13:52] Infinity Linden: it sounds like there's nothing any of us can do to convince you
[13:52] Infinity Linden: but
[13:52] Dahlia Trimble: well I will be participating in opensim interop efforts, and I like LL and am interested in them continuing to be successful players
[13:53] Infinity Linden: give it a couple months and see how far we get
[13:53] Infinity Linden: then another few months
[13:53] Zero Linden: I think experimenting with inventory - or at least what is needed for appearance would be an EXCELLENT thing to build and try
[13:53] Mojito Sorbet: Latest update to OGP I find is October. Anything more recent, or am I　looking in the wrong place?
[13:53] Pixel Gausman: ok, me too
[13:53] Saijanai Kuhn: are there things we can request to happen soonest?
[13:53] Morgaine Dinova: Zero: have you been designing the infrastructure "interop bit" that Philip intimated was on the critical path for interop with Opensim?
[13:53] Zero Linden: IETF Internet-Drafts, Mojito
[13:53] xstorm Radek: i feel one day a person with god powers will send me to the cornfield for the way i think and feel
[13:53] Infinity Linden: i think we'll be building momentum and you can jump in to participate whenever you feel the time is right
[13:53] Saijanai Kuhn: Like more ccordination with pyogp or some branch of openmv to get Zha's code tested?
[13:53] Techwolf Lupindo: I see the interop thing something that should be worked on by LL. If they don't, somewhere around 2020, there will be two enitties, LL and the other grid that is all interoped and tied together that in total may equel or suprass LL total users.
[13:54] Zero Linden: I am not going to comment on things the Chairman of the Board intimates on!
[13:54] Zero Linden: --- I like my job!
[13:54] Saijanai Kuhn: or the group IM/service discovery issues we've batted around for a good while now?
[13:54] Infinity Linden: @sai ... there's probably going to be smoe discussino fo that on the OGPX list soonish
[13:54] xstorm Radek: but unless there is a way LL can get better backing and income to bring in more programers it looks like this may take longer
[13:54] Zero Linden: I'm also not interested in speculation
[13:55] Infinity Linden: i had a breif chat with JHurliman about using URLs as cap names
[13:55] Zero Linden: about 2020 or any other date
[13:55] Zero Linden: SO -
[13:55] Infinity Linden: so you could use the URL to find info about the service
[13:55] Infinity Linden: but
[13:55] Zero Linden: we have some drafts - if you think they need work, or suck, or are not enough, write more!
[13:55] Infinity Linden: i don't want to put words in john's mouth
[13:55] Infinity Linden: and there's still some work to be dnoe
[13:55] Twisted Laws: is there a link to the OGPX list?
[13:55] Saijanai Kuhn: isn't a rationally named cap sorta contra-CAP?
[13:55] Zero Linden: if you think there are areas of OGP that need fleshing out (Inventory? IM?) then write up drafts, or at the very least put together some ideas, a straw man and put it on th elist
[13:55] Zero Linden: on the list
[13:55] Zero Linden: THAT is what we can do now
[13:56] Latif Khalifa: infinity, i think you overestimate hurli's enthusiasm for ogp ;)
[13:56] Morgaine Dinova: I couldn't get Zero to comment on distributed asset services despite putting in on formal agenda either. It seems that anything that isn't already in OGP is strictly "no comment" these day, and therefore the community cannot enter dialogue with LL on it. No surprise that Opensim isn't interested.
[13:56] Infinity Linden: (other than to say it seems to be bubbling up to the top of both of our agendas)
[13:56] Zero Linden: please please please don't wait for drafts to come from Infinity's and my fingers so you can then implement
[13:56] Infinity Linden: krunk. safari fail
[13:56] Infinity Linden: (i'm looking up the ogpx URL)
[13:56] Zero Linden: No Morgaine - I'm just not going to comment on abstract ideas -
[13:57] Zero Linden: we need concrete protocol proposals
[13:57] Zero Linden: I'm not going to sit in this channel and debate abstractions
[13:57] Saijanai Kuhn: Infinity had the start of one concerning group IM via the AD...
[13:57] Infinity Linden: https://www.ietf.org/mailman/listinfo/ogpx
[13:57] Saijanai Kuhn: and Zha has said they're implementing inventory via CAPs in her AD code. THose seem concrete
[13:57] Twisted Laws: cool, ty
[13:57] Infinity Linden: there's the URL for the OGPX mailing list info page
[13:58] Infinity Linden: there's a small archive of past postings
[13:58] Zero Linden: Please, write it up - even informally - say " this is how it could work - with actuall calls" -- "here's some actual protocol I mocked up and tested with pyogp" -- "here's a particular area I want to work on and interop op on - who want's to code the other side"
[13:58] Infinity Linden: including a draft agenda
[13:58] Infinity Linden: and i stress the word "DRAFT"
[13:58] Zero Linden: Excellent - let's see those written up and posted to the OGPX mailing linst
[13:58] Zero Linden: they don't need to be written up to the level of Internet Draft to get started
[13:59] Zero Linden: just some concrete technical discussion on the mailing list
[13:59] Infinity Linden: s/Draft Agenda/Draft Charter/
[13:59] Pixel Gausman: fyi on inventory.... we're making use of the WebFetchInventoryDescendents code that exists in the viewer but is unused on agni
[13:59] Pixel Gausman: off to another mtg
[13:59] Zero Linden: Think of them as "here's some ideas of what we've implemented and experimented with on inventory CAPs" woudl be great
[14:00] Zero Linden: yes -- actually, it is the end of the hour here
[14:00] Zero Linden: but a piece of administrivia:
[14:00] Zero Linden: I'm going to steer this meeting to be seriously focused on the OGPX activity and mailing list
[14:01] Zero Linden: here's a carrot (anemic though it is): If there is clear activity on the list that warrants more discussion time - then we should increase the frequency of this meeting
[14:01] Zero Linden: and - (gasp) - I think it should be in voice
[14:01] Morgaine Dinova: Zero: here's a concrete proposal so that you don't have to comment on speculation. (Wait for the transcript and it'll be in writing). --> Add 1 bit to all objects, call it "Interop Bit" or "Allow Copying Asset to Other Grids". Allow creators of assets to set the bit, nobody else. Set the default value to FALSE. Get the viewer peeps to add a bit toggle to build panes to set it." Job done to satisfy Philip's stated requirement.
[14:02] Zero Linden: Morgaine - that completely ignores the technical reality of composite objects
[14:02] Geo Meek: :-)
[14:02] Infinity Linden: hmm... getting close to pumpkin time for me
[14:02] Morgaine Dinova: Nope, this is per asset/
[14:02] Geo Meek: now your talking
[14:02] Zero Linden: I say it again - that completely ignores the reality of composite obejcts
[14:03] Geo Meek: and may we share this metting?
[14:03] Trinity Coulter: Zero, what do you mean?
[14:03] Rex Cronon: what exactly is a composite object?
[14:03] Infinity Linden: as well as the semantics of "i want this item to live on the main grid and grid X, but not grid Y"
[14:03] Zero Linden: some points:
[14:03] Morgaine Dinova: No it does not. Each asset should be allowed or restricted separately.
[14:03] Zero Linden: 1) This meeting is public - proclaim the transcript far and wide
[14:03] Dahlia Trimble: it could be checked recursively and only allowed to be set to true if all subcomponents are also true
[14:03] Geo Meek: in voice people my give a ear
[14:03] Morgaine Dinova: A composite object can be composed of assets from multiple worlds. Therefore your objection is wrong.
[14:04] Saijanai Kuhn shudders at the thought of recursive checking based on a client action
[14:04] Zero Linden: 2) Composite objects have linked prims, obejct inventory, textured faces --- and notecards have embedded things, and clothing and gestures....
[14:04] Rex Cronon: in voice there is no transcript
[14:04] Zero Linden: it is a complex topic!
[14:04] Trinity Coulter: Morgaine, I'm sure Zero has a valid, point, it might help more to hear the specifics tho
[14:04] Zero Linden: and - really - please - go work out the implementation of that - even on OpenSim --
[14:04] Infinity Linden: @rex. torley uses mechanical turk to turn an MP3 recording of his meetings into a typed transcript
[14:04] Latif Khalifa: Zero, have you tested how current permission work? :P It already uses least permissive bits as the components
[14:04] Infinity Linden: we're told it's very affordable
[14:04] Geo Meek: Rex make your own
[14:05] Zero Linden: Latif - believe me, I know all about the current permissions :-P
[14:05] Rex Cronon: u don't know to read eo:)
[14:05] Rex Cronon: geo*
[14:05] Latif Khalifa: then you objection about composite objects makes no sense
[14:05] Geo Meek: true Rex
[14:05] Trinity Coulter: Latif, from what i see, Zero's comment makes perfect sense
[14:05] Trinity Coulter: its just not a detailed comment
[14:05] Latif Khalifa: the object would be grid tranferable only and only if all of its parts had that permission
[14:06] Morgaine Dinova: Objects are transferred one asset at a time, so there is not recursive pre-processing required. It is trivial to check that the interop bit is set at transfer time with no addtional overhead, so the objection is false on overhead grounds too.
[14:06] Geo Meek: feel better Rex?
[14:06] Saijanai Kuhn: I'm not sure that's how it works Morgaine.
[14:06] Infinity Linden: morgaine. go prototype that solution on an OpenSIm instance and get back to me
[14:06] Rex Cronon: i feel very good, an u?
[14:06] Zero Linden: would you transfer objects Morgain that contained objects in inventory that couldn't be transfered?
[14:06] Trinity Coulter: Morgaine... you might help your case if you stopped using the term "object" and said "asset" instead
[14:06] Zero Linden: wouldn't that render then broken?
[14:07] Trinity Coulter: because it sounds more like you mean asset
[14:07] Zero Linden: Trinity - it wouldn't -- there are objects that aren't assets
[14:07] Zero Linden: And there are asset referencs without permissions bits
[14:07] Latif Khalifa: zero, if one part of the object is not transferable the whole object would not be... same way it works now with transfer permission
[14:07] xstorm Radek: true
[14:07] Zero Linden: so, in the current scheme of SL - this thing is VERY complicated to design correctly
[14:07] Trinity Coulter: you are talking about Inventory items, right? not objects only
[14:07] Morgaine Dinova: Zero: you only transfer one asset at a time, there is no other way. Transferring a composite object just breaks down into transferring lots of individual ones.
[14:07] Zero Linden: Latif - that amounts to a recursive evaluation
[14:08] Trinity Coulter: and a generic term besides "Inventory item" would be "asset" wouldn't it?
[14:08] Latif Khalifa: you do that already for in-world transfers
[14:08] Morgaine Dinova: It is *NOT* recursive. Assets are transferred one at a time.
[14:08] Zero Linden: I don't think you understand enough about how objects are transfered to say that ----
[14:08] Infinity Linden: but the good news is... we can take what we've learned in OpenSim and in implementing SL and propose an interface and pdu format that lets us upgrade our implementation some time in the future
[14:08] Talin Sands: if the sum of the interop bits == number of assetts then trans allowed is TRUE? or have i missed a step?
[14:08] Zero Linden: You can choose to transfer them one at a time, but the objects won't work unless the recursive tree is transferable
[14:09] Zero Linden: if you transfer after checking only one level -- you are potentially transfering an objec tthat will be totally broken on the receiving side
[14:09] Infinity Linden: and let's not forget the UI / UX implications
[14:09] Latif Khalifa: ok, so we are back again at nothing can be implemented on the sl grid...
[14:09] Morgaine Dinova: Even if your transfer code were written in a recursive manner to handle arbitrary sized hierarchical objects in the future, *STILL* your validation would not be recursive but only O(1), because your actually transfer the assets one at a time.
[14:09] Infinity Linden: one, small non-transferrable linked prim in a build would halt the transfer
[14:09] Saijanai Kuhn ignores the grapes just out of reach. They're probably rotten anyway
[14:10] Trinity Coulter: yes, that's ok, isn't it, Infinity?
[14:10] Dahlia Trimble: Infinity, if what I'm hearing is that LL will be desiging new protocols a piece at a time and publishing and standardizing them as they are implemented, then that makes me happy :)
[14:10] Trinity Coulter: that's the point of permissions :)
[14:10] Infinity Linden: you would want to haev a UI tool that allows you to hilight the non-transferrable part so you can decide how to replace it
[14:10] Zero Linden: So look
[14:10] xstorm Radek: it seems to me as the content is in inventory there needs to be a way for people to backup this content on there own computer or server
[14:10] Trinity Coulter: why would you want to break content just in order to move it?
[14:11] Infinity Linden: right. that's what's supposed to happen
[14:11] Zero Linden: It is VERY easy to imagine simple solutions here - it is harder to engineer them to work practically
[14:11] Trinity Coulter: if a designer puts in a non-transfer script ina shoe, they expect that the permissions will apply that way
[14:11] Trinity Coulter: so you're talking about breaking an object to force it to move over
[14:11] Infinity Linden: because if you have a coalesced object composed of multiple assets, one of which can't be transferred
[14:11] Infinity Linden: you'll probably want to know why your object can't be transferred
[14:11] Trinity Coulter: when that wouldn't be the logical implied desire of its creator
[14:11] Geo Meek: good point Zero
[14:12] xstorm Radek: only the content made by the owner needs to be backup
[14:12] Saijanai Kuhn: Darn, where's that gate to the castle...
[14:12] Zero Linden: This area is non-trival -
[14:12] Niaht Nakamichi: Trinity, I have no doubt someone would find a reason, unrelated to content theft issues. The creator would simply not add that bit to any part of their shoe, and thusly, no part would be transferable.
[14:12] Infinity Linden: ohoh trinity... i'm thinking of when you're building something out of parts you got from other people
[14:12] xstorm Radek: textures can be saved on a computer now as it is
[14:12] Trinity Coulter: you should have a tool that tells the creator at the time they make something if they have permissions to permit it to go to another grid
[14:12] Saijanai Kuhn: eventually, from other grids...
[14:13] Infinity Linden: i think you're talking about the situation where you're the creator of an object and you've maybe accidentally left the xfer bit off of one component
[14:13] Trinity Coulter: if they are told that it can't go to another grid, then the box is grayed out for them
[14:13] Morgaine Dinova: Zero: of course you can be transferring an item that is totally broken. You need to stop thiking in a centralist fashion --- an object may be composed of assets from countless worlds, and absolutely no way do you have control over all of them. It may contain a texture that does not have access rights in SL, so you may have to present a default texture for example. These things are not under your control in an interop world.
[14:13] Trinity Coulter: you shouldn't replace parts of the item just to make it transfer over
[14:13] Geo Meek: do let me know when you guy's goto voice
[14:13] Trinity Coulter: its not a word doc
[14:13] Trinity Coulter: and you're not just missing a font
[14:14] Morgaine Dinova: Zero: you've coupled asset service and region in your mind, that's the trouble here.
[14:14] xstorm Radek: voice ? NNNOOOO
[14:14] Geo Meek: Thank you for this metting
[14:14] Geo Meek: Voice YESSSSSSSSSSS
[14:14] xstorm Radek: i keep logs
[14:14] Saijanai Kuhn poitns out that there's plenty of low hanging fruit we're ignoring...
[14:14] Zero Linden: Morgaine - perhaps you have to start thinking in a practical fashion, where the expereience of the common user and the common operations are essential to optimize for
[14:14] Saijanai Kuhn: and it might not even be rotten
[14:14] Zero Linden: AND ON THAT NOT
[14:14] Zero Linden: AND ON THAT NOTE
[14:14] Trinity Coulter: nite?
[14:14] Trinity Coulter: :)
[14:14] Zero Linden: I'm calling the patient
[14:14] Zero Linden: thank you all for coming
[14:14] Geo Meek: voice well save on your tiny hard drive
[14:15] Latif Khalifa: lol
[14:15] xstorm Radek: bye Zero :-) and thank you
[14:15] Twisted Laws: ty and bye
[14:15] Rex Cronon: nice. i crash and when i come back teh show is over
[14:15] Zero Linden: later all
[14:15] Zha Ewry: TY Zero
[14:15] Saijanai Kuhn: laters 0
[14:15] Rex Cronon: bye zero
[14:15] Infinity Linden: yup. i gotta run too. there's some stuff i want to push out to the ogpx list before i go home
[14:15] Infinity Linden: cheers folks
[14:15] Trinity Coulter: nite :)
[14:15] Rex Cronon: tc infinity
[14:15] Whump Linden: If the meetings are in voice, then people will realize I'm not really a robot.
[14:15] Saijanai Kuhn: laters ∞
[14:15] xstorm Radek: bye Infi
[14:15] Twisted Laws: ^^
[14:15] Dahlia Trimble: bye zero, I :)

User:Zero Linden/Office Hours/2009 jun 02

Navigation menu

Search