AW Groupies/Chat Logs/AWGroupies-2010-04-20

From Second Life Wiki
Jump to: navigation, search
[09:32] Morgaine Dinova: I only hear Zha
[09:33] Morgaine Dinova: Yep, who was that?
[09:33] Latif Khalifa: yeah
[09:33] Joshua Linden: me
[09:33] Rex Cronon: hello everybody
[09:33] Meadhbh Oh: good grief... i cn't convince my client that i really, really, really want sound to come over the headphones
[09:33] Morgaine Dinova: Coolness. Oh, dratted positional audio.
[09:33] Zha Ewry: Brookstone, Kopilo, Krise, Rex, XLR*, Joshua is goign to be pitchign slides in voice, taking Q&A in voice and chat
[09:33] Joshua Linden: Lemme know when to start, Zha
[09:33] Morgaine Dinova: I'll relocate for this, to not have some people behind me.
[09:33] Latha Serevi: Joshua, your slide display is overlapping the cat-tail vegetation just a bit.
[09:33] XLR8RRICK Hudson: ratz voice is fail
[09:34] Zha Ewry: Much better location
[09:34] Joshua Linden: thx latha, just fixed
[09:34] Joshua Linden: lemme move to the middle of the area so voice is centralized
[09:34] Meadhbh Oh: also.. is there going to be a NOTE WELL notification? is this an official IETF thingie?
[09:34] Zha Ewry: Quorum call
[09:35] Meadhbh Oh: and ugh. voice fail
[09:35] Fleep Tuque: (Try relogging)
[09:35] Joshua Linden: vwrap@ietf.org
[09:36] Meadhbh Oh: yay! voice!
[09:36] Krise Shepherd: i will relog i hear nothing
[09:36] Zha Ewry: And with that, I give you Joshua, and.. warn you that he CAN in fact, flick a whip properly (I have RL evidience of this)
[09:36] Morgaine Dinova: What format for questions Josh? At end, or interrupt?
[09:37] Zha Ewry: /
[09:37] Zha Ewry: I think we can catch Josh with a questoin if its urgent, but ideally, hold it to the ned if possible
[09:37] Free Radar HUD v1.1 by Crystal Gadgets
[09:37] Zha Ewry: I'll also flag them in voce for anone who asks in Chat
[09:38] Morgaine Dinova: And do you have untill 11am as usual here, or need to leave earlier?
[09:38] Meadhbh Oh: +1
[09:39] Meadhbh Oh: blame the asset server
[09:39] Joshua Linden: https://docs.google.com/a/lindenlab.com/present/edit?id=0AfAytYlNCPECZGZ2M24yMl8yN2N0OGJueGNy&hl=en
[09:39] Zha Ewry: Blame Infinity Linden.
[09:40] Saijanai Kuhn: meeting is in voice BTW
[09:42] Zha Ewry: Moundsa
[09:42] Mystical Demina: was a question about http transactions
[09:42] Moundsa Mayo: Electronic noise from the machine.
[09:43] Morgaine Dinova: Switch your in-world slide
[09:43] Meadhbh Oh: blargh
[09:52] Latha Serevi: scenegraph = high latency? (did you mean low?)
[09:52] Latif Khalifa: you mean low-latency?
[09:54] Latha Serevi: How bold and tongue-in-cheek is the URL proposal?
[09:55] Latha Serevi: Human readable, bah.
[09:55] Zha Ewry: "Sophont parsable?"
[09:55] Dzonatas Sol: I look at the slide and i think some assumptions were made that the server would have a message queue (on note chat 'wouldn't be good for http')... the message queue can be a peer to the server... like: server<->mq<->client instead of direct server<->client
[09:55] Morgaine Dinova: Joshua: when you say "A Web Page", could you please clarify that you're not advocating page scraping as a standard approach? I'm hoping you mean "web page" containing nothing but LLSD for example, and not pages decorated for human access.
[09:56] Meadhbh Oh: we could add it later if this decision turns out to be incorrect
[09:56] Kopilo Hallard: xD
[09:56] Latha Serevi: MOrgaine, I think he exactly DOES mean that we have to screen-scrape.
[09:56] Mystical Demina: i guess we are assuming these services are being done in 2D
[09:56] Morgaine Dinova: Latha: hence want clarification
[09:57] Free Radar HUD v1.1 by Crystal Gadgets
[09:57] Morgaine Dinova: I suggest an interrupt question, because Joshua writes "a web page" repeatedly.
[09:57] Mojito Sorbet: Parsing a web page, in multuiple languages, makes it kind of difficult for intelligent text-to-speech processing
[09:58] Zha Ewry: I'll type
[09:58] Zha Ewry: XML, Parsing, Web page? http?
[09:58] Morgaine Dinova: Jeez
[09:58] Morgaine Dinova: Joshua - -
[09:58] Latif Khalifa: Joshua, how do you standardize interface betweeen a web page and the client, linden use secondlife:// namespace, should that be standadized
[09:58] Mojito Sorbet: That is what PEOPLE interact with
[09:58] Mystical Demina: i do think it makes since to seperate presentation of the services for the implementation of the service, seems we need navigation standars, but we may not need to dfine the user interface to that
[09:58] Mojito Sorbet: People who can read
[09:58] Newfie Pendragon: people are 10% of the consumers of data
[09:59] Morgaine Dinova: We don't need human pages standardized. Interop protocols need to be automatable
[09:59] Boroondas Gupte: Wouldn't protocol negotiation solve this better? The result could still be a website in some cases when that's the lowest common denominator.
[09:59] Mojito Sorbet: HTML would be ok as long as the metadata is predictable so we can parse it
[09:59] Meadhbh Oh: i think the quote was "there is NOTHING in search / update that requires it to be integrated into the protocol."
[09:59] Newfie Pendragon: separate presentation from model
[09:59] Meadhbh Oh: s/update/profile update/
[09:59] Mojito Sorbet: Ah, that is a separate question
[10:00] Mojito Sorbet: Yes, leave search out of VWRAP, but do not at the same time say how it should be done
[10:00] Latif Khalifa: we need standardized way to perform client actions, like mime-type
[10:00] Mystical Demina: if we define the virtal world objects in a way we can access like SQL does data, then would be a good standard, not sure that standard is part of that
[10:00] Latif Khalifa: what LL uses as secondlif:// uri space
[10:00] Morgaine Dinova: Search is a separate issue. But you mention MANY things as to use a "web page", and many of them are infrastructure
[10:00] Mojito Sorbet: Just leave SEARCH out. I am fine with that. But do not suggest solutions
[10:00] Morgaine Dinova: I think Zha is answering, but can't hear
[10:00] Meadhbh Oh: the secondlife:/// URI should go away
[10:01] Latif Khalifa: Meadhbh, but we need a standard way to perform actions
[10:01] Zha Ewry: I'ms uggesting that the distcintion may be that there is a "SearcH tool
[10:01] Techwolf Lupindo: That is a good question. ADA stuff needs to be able to search objects in order to work.
[10:01] Latif Khalifa: search without teleport is useless
[10:01] Zha Ewry: and therne there is "how do I mark up prims, and regoins ansd such"
[10:01] Zha Ewry: so that we expose search items
[10:01] Fleep Tuque: nod Techwolf
[10:01] Zha Ewry: Plowing onwards
[10:01] Latha Serevi: What about having the "opaque URL to a human-readable page" as one lowest common denominator option, but not the end-all? Doesn't it prevent automated _anything_?
[10:01] Meadhbh Oh: but here's the question... if VWRAP defines the mechanism for delivering these named "services on the web page," does VWRAP need to define the names of the services?
[10:02] Mojito Sorbet: Coding a query as a standardized URI format, but with additional options to selct response formats.
[10:02] Morgaine Dinova: Your answer is only "Yes" to browser interface because you've given yourself the false requirement of "human web page".
[10:03] Mojito Sorbet: Making the SL viewer act like a bad web browser was Linden's decision. It does not have to be that way
[10:03] Morgaine Dinova: Mojito++
[10:03] Latha Serevi: I liked these slides more when I was certain that Joshua didn't really mean it.
[10:03] Latif Khalifa: that doesn't work, "search is a web page", you need to define how do you tp from a web page... web page -> client actions need defining
[10:03] Meadhbh Oh: maybe define the names as UIRs, then?
[10:03] Mojito Sorbet: There are other paradigms
[10:03] Meadhbh Oh: s/UIRs/URIs/
[10:04] Meadhbh Oh: we have a standard for that
[10:04] Meadhbh Oh: client application launch message
[10:04] Zha Ewry: Client Applicatoin Launch
[10:04] Latif Khalifa: those triggers need stardadizing
[10:04] Meadhbh Oh: what about the client application launch message?
[10:04] Meadhbh Oh: could we just use that?
[10:04] Morgaine Dinova: TP from a web page is useful as a way in from the Web, but it's not VW-VW interop.
[10:04] Meadhbh Oh: since, it's defined and implemented?
[10:04] Mojito Sorbet: If the viewer itself is doing this query, it can just do the TP itself.
[10:04] Mystical Demina: if search is local to a virtual world, i see less need for a stadard, but if search crosses across virtual worlds then i see a need for a standard
[10:05] Boroondas Gupte: vwrap:// ?
[10:05] Dzonatas Sol: <object ...
[10:05] Latif Khalifa: Mead, it's not only TP, "show on map", etc etc
[10:05] Mojito Sorbet: Consider a response to a serach that is formatted as a map..
[10:05] Meadhbh Oh: @latif: "here is an entity representing a location in the virtual world, please show it on a web page for me."
[10:05] Zha Ewry: I think theres things beyond "Teleport to" that youmay want to do with the map api
[10:05] Zha Ewry: +1 mea
[10:06] Rex Cronon: this is not FB
[10:06] Dzonatas Sol: <object> tags probably more reliable means than url mangles for search/tp
[10:06] Latif Khalifa: i can think of dozens search things that go beyond map and tp
[10:06] Meadhbh Oh: or even... "here is an entity representing an agent, please get the location then show the location on the web page"
[10:06] Mojito Sorbet: "Show me on a map where the things matching this query are located"
[10:06] Dzonatas Sol: yet urls should be optional
[10:06] Zha Ewry: URI
[10:06] Zha Ewry: URI
[10:06] Latif Khalifa: you find a person, how do you IM them, "open profile" action or some such
[10:06] Zha Ewry: URI
[10:06] Meadhbh Oh: +1 zha
[10:06] Meadhbh Oh: URIs not URLs
[10:07] Mojito Sorbet: A URI to identify an avatar.
[10:07] Meadhbh Oh: also. it seems to tick off zero
[10:07] Dahlia Trimble: are these slides available online?
[10:07] Latif Khalifa: https://docs.google.com/a/lindenlab.com/present/view?id=0AfAytYlNCPECZGZ2M24yMl8yN2N0OGJueGNy&hl=en
[10:07] Dzonatas Sol: url is subtset of uri
[10:07] Dahlia Trimble: ty :)
[10:08] Morgaine Dinova: Inventory has no business being world-side anyway. It should be in the client, and only assets should live outside in asset stores/
[10:08] Mojito Sorbet: Morg++
[10:08] Latha Serevi: The lowest common denominator says, "This URI deal with land opaquely". But MANY users of the protocol will want to add some structure to that. "This URI manipulates the following set of XML semantics". What if we define, in VWRAP, how to say WHETHER there are semantics and WHERE to find them? w/o trying to pre-define the semantics.
[10:08] Mojito Sorbet: Or in the AD maybe
[10:09] Latif Khalifa: lol
[10:09] Meadhbh Oh: or if we wanted to be fancy, we could say XML + XSLT
[10:09] Latha Serevi: How can we avoid ad hoc screen-scraping? We need a way to add structure as needed.
[10:09] Mojito Sorbet: How about we just eliminate the viewer and the server synthesize video to stream it to us? I think maybe you are taking this "everything is a web page" too far.
[10:09] Rex Cronon: r these "web pages" going to be in xlm?
[10:10] Latif Khalifa: anyone who has worked with html "tree" controls, knows how big a fail that is lol
[10:10] Meadhbh Oh: @rex. i think the answer is "maybe"
[10:10] Morgaine Dinova: Mojito: it's the webbies, they have only one tool in their toolkit, so everything has to be a nail.
[10:10] Ceawlin Steamweaver: If everything is going to be a web page, wouldn't it make more sense to make SL a subset of a web browser (firefox plugin) instead of vice-versa? <_<
[10:10] Zha Ewry waves at Alexandria Linden
[10:10] Meadhbh Oh: LOL
[10:11] Techwolf Lupindo: How will the client search though inventory when all it has is a human readable web page?
[10:11] Newfie Pendragon: How the data is presented/managed to the user has to be separated from the data, otherwise it makes automated tools subject to any non-functional (look/feel) change of the page
[10:11] Mojito Sorbet: "find my lost socks"
[10:11] Rex Cronon: it could be possible to make a java applet to explore sl:)
[10:11] Zha Ewry: WebDav?
[10:11] Saijanai Kuhn: eToys inventory
[10:11] Meadhbh Oh: @rex. i'm hoping to have a JavaScript + Web3G client
[10:12] Mojito Sorbet: "Here is my store catalog - I present it to you as an Inventory tree"
[10:12] Rex Cronon: i don't think that js is powerful enough for 3d rendering
[10:12] Latif Khalifa: no he's saying just send html view of the inventory
[10:12] Morgaine Dinova: Web technology == good, but that does NOT need to mean using human-style web pages.
[10:12] Dzonatas Sol: snow-375 lets me make an deatch inventory window... drag and drop
[10:13] Latif Khalifa: we need data, not UI
[10:13] Mojito Sorbet: Yes, use XML or something easy to parse
[10:13] Morgaine Dinova: Zha++
[10:13] Mojito Sorbet: I don't want to see " " in my inventory
[10:13] Rex Cronon: xml is not that easy/fast to parse:)
[10:13] Dzonatas Sol: snow-375 is data
[10:13] Zha Ewry: Not in the prgramtaic pipe, Mojtio
[10:13] Mojito Sorbet: It is better than unconstrained HTML
[10:13] Kopilo Hallard: xhtml >_>
[10:13] Meadhbh Oh: and w000t! you get Web3d view of your items!
[10:14] Latif Khalifa: Zha, that is not what Joshua is saying
[10:14] Saijanai Kuhn: http://www.squeakland.org/showcase/everyone/
[10:14] Newfie Pendragon: the webpage should be providing the business logic to convert the human-UI interaction to the API interface
[10:14] Mystical Demina: architecturely as long as we decouple the source and destination with some kind of indirection mechanism we can use any source and destination as long as we provide the proper adapter
[10:14] Latif Khalifa: he's saying "no standard way to transmit inventory" except for UI as a web page
[10:15] Morgaine Dinova: Zha: it comes down to one basic question: are we supporting webpage scraping as an "interop technique"? I'm going to fight that tooth and nail.
[10:15] Morgaine Dinova: KK Zha, that's different.
[10:15] Latha Serevi: I don't think it's VWRAP's job to to decide exactly how MUCH of the to standardize, right now, for all time. VWRAP just needs an extensible way to describe AT WHAT LEVEL pairs of entities are interacting, one if which might be "opaque URI for topic X".
[10:16] Meadhbh Oh: so the idea is "do both programatic interface AND HTML interface. protocol comes in when you do something 'interesting' like rez something in world or give it to someone else."
[10:16] Dzonatas Sol: WebDav may be RESTful sidekick if it is passive
[10:16] Mojito Sorbet: Perhspas a small numbers of conformance levels defined?
[10:16] Latha Serevi: Disagree that we must focus on "minimum." We must focus on "extensible".
[10:16] Latif Khalifa: Mead, inventory is interesting thing
[10:16] Lor Gynoid: Pity VRML didn't fulfill its promise. :)
[10:16] Morgaine Dinova: Latha++
[10:16] Mojito Sorbet: Collada?
[10:16] Zha Ewry: +1 :Latha
[10:17] Mystical Demina: there are assets that the server cares about, like objects with physics, but there may be objects that are redenered that the only the client sees, like video stream
[10:17] Meadhbh Oh: collada is a little too heaviweight
[10:17] Latif Khalifa: html is not a data format, it's UI standard
[10:17] Mojito Sorbet: We have HTML 1.0, HTML 2.0, etc. They build
[10:17] Morgaine Dinova: Latha += 1000. Not "minimal", but "minmal extensible"
[10:17] Meadhbh Oh: but it's great for an upload service
[10:17] Kopilo Hallard: html is not a ui standard... >_>
[10:17] Morgaine Dinova: Zha++
[10:17] Mojito Sorbet: HTML is extensible
[10:18] Newfie Pendragon: XHTML is extensible
[10:18] Mojito Sorbet: The web succeeded becuase the fundamental mechanism was extensible
[10:18] Dzonatas Sol: eXtensiable Markup Language
[10:18] Techwolf Lupindo: The minumoiun http is "GET /". Everything else is an extention of that.
[10:18] Latif Khalifa: if the only standard way of getting inventory data is html, that's a big fail
[10:18] Meadhbh Oh: @latif. no. that's not the proposal
[10:18] Meadhbh Oh: the standard way of presenting the inventory is HTML
[10:19] Kopilo Hallard: actually it is a tree...
[10:19] Dzonatas Sol: Those inventory objects should be self-contained
[10:19] Rex Cronon: the protocol doesn't say hownotecards work, it only says how their dat is transmitted
[10:19] Dzonatas Sol: yet we not there yet to implement self-contained
[10:19] Latif Khalifa: mead, that's still a fail
[10:19] Meadhbh Oh: so in other words... you don't ask the server to "give me a list of all of my inventory" but rather "give me an URL to a web page that will display my inventory"
[10:19] Dahlia Trimble: a file?
[10:19] Rex Cronon: thier data*
[10:19] Rex Cronon: thier8
[10:19] Dzonatas Sol: a notcard that carried it's own script on how to read/write it's object, for example
[10:19] Rex Cronon: i can't type*
[10:19] Latif Khalifa: Mead, that assumes client is a human reading some html page
[10:20] Mojito Sorbet: A human who can read
[10:20] Mystical Demina: shoiuldn't interop provide a way to descover if these services are available and what protocol is being used for them
[10:20] Latif Khalifa: yes
[10:20] Free Portal: i think html5 can.
[10:20] Zha Ewry: So.. Taking this as a provocative discussion.. The goal is to take questoins you have out to the VWRAP mailing list and raising those questoins explicitly
[10:20] Rex Cronon: if the viewer has to parse human readable html pages doesn't that add additionals delays?
[10:20] Saijanai Kuhn: In some cases probably yes (combat game worlds E.G.)
[10:20] Lor Gynoid: Does your Friends list need to be a tree? :)
[10:21] Mojito Sorbet: Parsing HTML is very fast, compared to 3D rendering
[10:21] Kopilo Hallard: lor: yes but is it no :p
[10:21] Free Portal: look this http://googlecode.blogspot.com/2010/04/html5-quake-ii.html
[10:21] Dzonatas Sol: Does VWRAP really need to be extensible, as changes could destablize interop
[10:21] Newfie Pendragon: html pages are subject to regular changes as the PR team wants the logo/theme/etc changed....that stuff causes unnecessary parsing burden on automated consumers
[10:21] Mojito Sorbet: It MUST be extensible
[10:21] Morgaine Dinova: "Group" has many different meanings. I think we should split them apart for VWRAP. For example, a communicating "group" should use a discrete term for its members than a term denoting some other grouping.
[10:22] Mojito Sorbet: But lets try to avoid the Microsoft-vs-everybodyelse situation
[10:22] Dzonatas Sol: another protocol can be extensible... it doesn't have to be part of vwrap
[10:22] Kopilo Hallard: brb
[10:22] Lor Gynoid: You need an open-ended way to attach attributes to Friend entries?
[10:22] Meadhbh Oh: yuo mean someone thought it would be a be-all end-all?
[10:22] Mojito Sorbet: yes yes!
[10:23] Mojito Sorbet: Groups insersect with VW only as regards permissions
[10:23] Fleep Tuque scratches her head.
[10:23] Lor Gynoid: You need an open-ended way to attach (na,ed) attributes to Groups, and FRiends?
[10:23] Lor Gynoid: named*
[10:23] Fleep Tuque: You'd have to decouple a lot of things in groups to not make it integral - land permissions, obj permissions, all sorts of thigns.
[10:23] Kopilo Hallard: pluse fail breb
[10:24] Zha Ewry: And, in many ways, Fleep, that m ight make all of that easier.
[10:24] Techwolf Lupindo: A good thing would be to somewhat standerized so that one cleint that is currently connected to one grid can IM someone on another grid.
[10:24] Dzonatas Sol: shoulds collada be part of vwrap... nah... it's it own protocol... be vwrap could refer to an dae uri
[10:24] Mojito Sorbet: Permissions can refer to "entities", whether those are Groups, Accounts, or something else. Another URI?
[10:24] Dzonatas Sol: but*
[10:24] Fleep Tuque: nod sure, current implementation of groups is like a hydra, way too much going on there.
[10:24] Saijanai Kuhn thought he invented "tourist" o well
[10:24] Mojito Sorbet: This rock here seems to be owned by http://ibm.com.
[10:25] Joshua Linden: Oops, skipped forward a slide
[10:25] Morgaine Dinova: Joshua: it's incorrect to infer that the tourist scenario implies no direct access to land. Rights may be as comprehensive in remote lands as in local lands.
[10:25] Meadhbh Oh: return objects?
[10:26] Meadhbh Oh: we'll talk about it offline
[10:26] Mojito Sorbet: I am a "tourist" in my friend's house, but he might let me use the bathroom.
[10:26] Zha Ewry: The sim can hold them for later
[10:26] Meadhbh Oh: it's not as bad as you might think within a trust domain
[10:26] Pixel Gausman: +1 on removing up/download for assets/textures
[10:26] Zha Ewry: No, no, it's when I'm on YOUR sim, with my AD and asset server that it gets messier
[10:27] Zha Ewry: +200
[10:27] Lor Gynoid: You might get kitchen rights in a friend's house, too...
[10:27] Morgaine Dinova: There is no single trust domain with max rights. You may have as many rights to local land as to remote land.
[10:27] Mojito Sorbet: But stranger walking up to the door gets no bathroom rights
[10:27] Latha Serevi: Joshua, I love a lot of this talk -- the emphasis on boldly chopping things out of the required set; and specifics about what might be in or out. But then you seem to be saying that everything in the "out" pile should be opaque URIs to VWRAP, now and for all time. This would really suck, since we want programmatic ways of doing interop. If VWRAP can't encapsulate and name these, we will work around VWRAP or smuggle binary blobs through it.
[10:27] Zha Ewry: For these, what's really hard, is what parts of this need to expose to the users
[10:27] Lor Gynoid: Region tree?
[10:28] Mojito Sorbet: Just probe a coordinate set, and get properties back
[10:28] Zha Ewry: Volumes, not 2d parcels, btw ;)
[10:28] Mojito Sorbet: coordinate ST
[10:28] Newfie Pendragon: resource allocation
[10:28] Mojito Sorbet: SET
[10:28] Rex Cronon: i am sorry. but every vw will have at least one parcel
[10:28] Mojito Sorbet: Yes, volumes
[10:28] Morgaine Dinova: Joshua: well phrased there --- "allows but does not require"
[10:29] Mojito Sorbet: An upper bound on my house box prevents my red walls going up 2000m
[10:29] Lor Gynoid: If a parcel is time-related it is 4-D? :)
[10:29] Mojito Sorbet: Number of dimenions is a per-world decision
[10:29] Newfie Pendragon: only if you can travel that 4th dimension
[10:29] Meadhbh Oh: blarg. voice fail
[10:29] Zha Ewry looks at Lor and squinches her eyes with an insta-migraine
[10:29] Mojito Sorbet: I am travelling thru time just sitting here... at 1 sec per sec
[10:29] Saijanai Kuhn: http://www.squeakland.org/showcase/everyone/
[10:30] Meadhbh Oh: i've been looking for a place to insert a tensor in the protocol
[10:30] Zha Ewry gloswers at Mea
[10:30] Lor Gynoid: Arguably, having access based on a parcel rental period...
[10:30] Saijanai Kuhn: etyoys has a vector behavior added
[10:30] Mojito Sorbet: To his disadvantage, of course. And charge him VAT
[10:30] Kopilo Hallard: lol
[10:30] Kopilo Hallard: points to whoever said volcanos
[10:30] Techwolf Lupindo: tp failures...how to grafully fail. :-)
[10:31] Lor Gynoid: Lootable inventories, with access methods? :)
[10:31] Dzonatas Sol: client-side scriptability... http://twitpic.com/1gl71s
[10:31] Kopilo Hallard: data types >_>
[10:31] Meadhbh Oh: blargh. my voice is hosed. brb.
[10:31] Joshua Linden: https://www.ietf.org/mailman/listinfo/vwrap
[10:31] Newfie Pendragon: data->readable = easy, readable->data = hard
[10:31] WordzKicker Payer v1.03: $10 WordzKicker - Latha Serevi, you were KICKED from slot 6 by Molly Eberdene! You won $9!
[10:32] Morgaine Dinova: Please note that LLSD is serializable using bianry formats too. Therefore, please don't only think "web page" as a delivery mechanism for LLSD.
[10:32] Rex Cronon: we do discuss things, but not in voice:)
[10:32] Zha Ewry: And yes, not in voice, in general
[10:32] Dzonatas Sol: The issue I see with VWRAP, at hand, is if there are multiple peers... like what I do with SNOW-375... two programs on the same machine, one handles the inventory the other handles the render
[10:33] Techwolf Lupindo: web pages are not easelly scrapable for parsing. Like text to speech devices.
[10:33] Lor Gynoid: Rendering some stuff in Arabic with glyps is hopefully not a troubling issue. :)
[10:33] Morgaine Dinova: Tech++
[10:33] Meadhbh Oh: yay! running code!
[10:33] Zero Linden: VWPP = Virtual World Politics Protocol -- still not chartered
[10:33] Kopilo Hallard: Techwolf - xhtml-strict can be but it is definatly not as easy to parse as say cleartext
[10:33] Zero Linden: heh
[10:34] Dahlia Trimble: does that mean that open source implementations are the true standard?
[10:34] Lor Gynoid: Meta politics...
[10:34] Zha Ewry: Code, Open Source or Proporiatary
[10:34] Zha Ewry: IETF doesn't care which
[10:34] Dzonatas Sol: Does LL have a beta grid with VWRAP implemented?
[10:34] Dahlia Trimble: how can it be proprietary?
[10:35] Rex Cronon: talking about implementations, will ll ever provide a sim for testing vwrap?
[10:35] Zha Ewry: Well, the IETF model is "Here is the RFC, and you use code to validate it)
[10:35] Newfie Pendragon: protocol != code, protocol may be open, code might be closed
[10:35] Dahlia Trimble: it would need to be open
[10:35] Morgaine Dinova: Joshua: while Hurli's work was great, it was NOT done on the VWRAP list. It was done behind closed doors in negotiations between LL, IBM and Hurli. Please stomp on that.
[10:35] Eddy Stryker: yeah, put a stop to that joshua
[10:36] Joshua Linden: the spec is the standard. implementations are informational only
[10:36] Joshua Linden will got on that, Eddy
[10:36] Dahlia Trimble: how could you tell if the protocol portion is closed?
[10:36] Meadhbh Oh: fwiw... this is one of the differences between the VWRAP effort and Hypergrid and MPEG-V... different focuses... different ideas about how to build a standard
[10:36] Morgaine Dinova: Eddy == Hurli. I'm not sure if John actually meant he was advocating negotiations behind closed doors.
[10:36] Joshua Linden: ;-)
[10:36] Meadhbh Oh: and OpenCobalt
[10:36] Latha Serevi: Josh, just so we don't spend all of our time rebutting something you're NOT saying ... could you say a few words about extensibility? What do you think should be done by people who have a different idea of "minimal set" than you do? Screen scraping? Non VWRAP protocol? I hope neither of those, but then what?
[10:36] Meadhbh Oh: (i didn't foget about you, sai)
[10:36] Eddy Stryker: we need to break through the closed doors of #vwrap on Freenode and committing the code to openmetaverse.googlecode.com :-)
[10:37] Newfie Pendragon: J2EE is open, but websphere isn't.
[10:37] Newfie Pendragon: tomcat is open
[10:37] Lor Gynoid: Need a standard for 'closed doors'? :)
[10:37] Meadhbh Oh: @latif. let's gab about your requirements
[10:37] Saijanai Kuhn: I firmly believe that there's no One True Way. OpenCobalt will have a place in VWRAP, I am positive
[10:37] Meadhbh Oh: J2EE is a crime against humanity
[10:37] Kopilo Hallard: lol
[10:37] Zha Ewry: Prepare to sudder learning obscure XML formats ;) if you want to submit a draft, but we'll help
[10:37] Latha Serevi: Josh - so VWRAP must stop at opaque URL, and not even give a _name_ to sub-protocols?
[10:37] Newfie Pendragon: lol
[10:37] Zha Ewry: *suffer
[10:37] Latif Khalifa: i think there is a disconnect here. about what v
[10:37] Meadhbh Oh had to implement CorbaSEC Level 2 a couple times
[10:37] Latif Khalifa: what VWRAP is
[10:38] Morgaine Dinova: VWRAP is what the working group says it is.
[10:38] Meadhbh Oh: isn't there a blog post about it somw=ewhere?
[10:38] Dzonatas Sol: we know it is not a UI
[10:38] Joshua Linden: Remember: I want to be told "I'm wrong" - but with details! Ideas! Implementations!
[10:38] Pixel Gausman: carving into stone tablet == IETF tools
[10:38] Meadhbh Oh: lol
[10:38] Joshua Linden: +1 Morgaine
[10:38] Meadhbh Oh: yes
[10:38] Rex Cronon: voice failling(
[10:38] Morgaine Dinova: VWRAP is NOT what a closed subgroup decides it is behind closed doors.
[10:38] Meadhbh Oh: i'm actually getting pretty good at it
[10:38] Latha Serevi: So Josh's proposal for extensibility is "Go prepare a draft somewhere else"?
[10:38] Latif Khalifa: it is clear to me now that VWRAP does not aim to create 3D web with a universal client. it looks like we will still have to have "lidnen viewer", "osgrid viewre" "xxx world viewer"
[10:39] Dahlia Trimble suspects the open source implementation wins
[10:39] Latha Serevi: What's your proposal, Josh?
[10:39] Meadhbh Oh: should we have another meeting to talk about what ewe've got?
[10:39] Zha Ewry: I suspect the combo of code and spec wins
[10:39] Meadhbh Oh: LLIDL
[10:39] Meadhbh Oh: LLSD
[10:39] Meadhbh Oh: deployment models
[10:39] Pixel Gausman: thx, Josh. especially the analysis on the existing UDP protocol
[10:40] Dahlia Trimble: what people use wins, the spec should reflect it
[10:40] Meadhbh Oh: but there are SOME ideas about extensibility right now
[10:40] Meadhbh Oh: i'll do it
[10:40] Meadhbh Oh: i have some spare time this week
[10:40] Latha Serevi: When we're responding to this talk, what are we responding to? I think it's "vague unspoken stuff that Josh half said".
[10:40] Eddy Stryker: Dahlia: whatever people implement and document (and people use) will win, true
[10:40] Newfie Pendragon: what people use = MS Office (shudders)
[10:40] Meadhbh Oh: we REALLY should document what we've got now
[10:40] Meadhbh Oh: or what we think there's general agreement on
[10:41] Latha Serevi: But there is no analysis in the vrap list or drafts, right? So we're responding to a void?
[10:42] Meadhbh Oh: yup. also
[10:42] Meadhbh Oh: the charter listss what we think we're going to do
[10:42] Newfie Pendragon: "when will X be standardized?" instead "will X be part of VWRAP, and if so, when?"
[10:42] Morgaine Dinova: Joshua: I like your general call for a minimalist set of interfaces standardized as an extensible protocol. But big proviso on the "web page" issue, because LLSD is binary serializable and that binary serialization is not necessarily sent over HTTP (even though it may be kicked off through a cap).
[10:43] Latif Khalifa: the key message is each world will have their own client, no universal client, no 3d web
[10:43] Fleep Tuque: Hm I didn't understand that, Latha
[10:43] Pixel Gausman: Latif: hope you are wrong
[10:43] Zha Ewry: Odddly I took jjust the oppotise Latif
[10:44] Latif Khalifa: Pixel, "inventory is just a web page" as a standard
[10:44] Latif Khalifa: no data standard
[10:44] Mojito Sorbet: So far it looks like a common 3D viewer, sitting alongside FireFox...
[10:44] Pixel Gausman: rez out of inventory does
[10:44] Meadhbh Oh: also. i put together a list of docs we agree'd to during the chartering process here
[10:44] Meadhbh Oh: http://www.meadhbh.org/vwrap/documents
[10:44] Pixel Gausman: but inventory viewing/renaming, etc is web based
[10:44] Dzonatas Sol: each world will have their own client... nah
[10:44] Dzonatas Sol: each avatar may have their own client
[10:44] Morgaine Dinova: Zha++
[10:44] Joshua Linden: just like each web site has its own client... oh wait
[10:45] Fleep Tuque: nod that was my take, if you take out the specific junk the better?
[10:45] Mojito Sorbet: It is true that decluttering the 3D protocol, the better
[10:45] Lor Gynoid: A universal client might have plug-ins to handle some world's extra services?
[10:45] Mojito Sorbet: IE?
[10:45] Mojito Sorbet: bleh
[10:45] Fleep Tuque: Thank you Joshua
[10:45] Zero Linden: thanks Josh for putting that all together
[10:45] Meadhbh Oh: w00t! +1 josh / zha for organizing this
[10:45] Zero Linden: it was clearly a lot fo analysis behind those ideas
[10:45] Dzonatas Sol: thank you =)
[10:45] Techwolf Lupindo: Requiring to have IE to view stuff is a bad idea.
[10:45] Dahlia Trimble: TY Josh :)
[10:46] Rex Cronon: what all world will have in common will be the ability to copybot thing from one to the other:)
[10:46] Saijanai Kuhn: http://waveplace.com/resources/tutorials/movie.jsp?id=106
[10:46] Lor Gynoid: Thanks for presentation and discussion.
[10:46] Meadhbh Oh is getting back to work
[10:46] Rex Cronon: using a browser:)
[10:46] Meadhbh Oh: cheers folks
[10:46] Zha Ewry: So, one last reminder "the mailing list is the place for much of this discussion"
[10:46] Saijanai Kuhn: I can see services that are references to downloadable plugins or scripts that provide new capabilities on the fly
[10:46] Eddy Stryker: ah, i had a meeting and was late to the discussion. will there be any more VWRAP gatherings in SL before april 30th?
[10:47] Zha Ewry: and.. casual informal discussoin here over the next few weeks is excelent
[10:47] Latif Khalifa: Eddy, the policy has changed
[10:47] Latif Khalifa: they killed the worst stuff
[10:47] Zha Ewry ducks out
[10:47] Dzonatas Sol: Sai +1 exactly .... 20+ year old discussion
[10:47] Saijanai Kuhn: e.g. the etoys showcase page
[10:47] Fleep Tuque: wait TPV policy you mean
[10:47] Latif Khalifa: yeah
[10:47] Fleep Tuque: oh right ok sorry got confused there
[10:47] Mojito Sorbet: I thikn Eddy meant the TPV policy
[10:48] Fleep Tuque: that meeting is at noon right?
[10:48] Latha Serevi: Hi Eddy, not that I know of.
[10:48] Zha Ewry: tpv policy gota tweak, and Joe Miller's got another discussion brrownbag on that at when, noon SL?
[10:48] Fleep Tuque: with Joe
[10:48] Fleep Tuque: nod
[10:48] Morgaine Dinova: I see the protocol as trafficking in "links to data provided by services" ... it tells you where the service is, and you use whatever protocol is required to obtain the data.
[10:48] Mojito Sorbet: But they have already responded to last week's beating up
[10:48] Eddy Stryker: morgaine: +1
[10:48] Fleep Tuque: Meeting with Joe is at Linden Estate Services (157,134,37)
[10:48] Zha Ewry: There's some tangle that's not entirely clear in my head as extactly when you want to say 'Herey ou get an HTML renderable service" and "here you get a programtic API"
[10:48] Fleep Tuque: at noon SLT
[10:48] Latha Serevi: In preparation for the Joe brown-bag, is there anything online I can read ahead of time?
[10:48] Saijanai Kuhn: Morgaine right. It might be an HTML web page only, or it might be a list of options or it might be a pure CAP type thingie
[10:48] Dzonatas Sol: I look at Collada to see a working-model for the bounds of extensibility
[10:49] Saijanai Kuhn: Dzonantas. Hardly
[10:49] Mojito Sorbet: Latha, looks at onlline TPV
[10:49] Saijanai Kuhn: Imagine etoys + OpenGL for example
[10:49] Fleep Tuque: See the changes at http://secondlife.com/corporate/tpv.php
[10:49] Latif Khalifa: http://pastebin.com/raw.php?i=Yd1j1EdE
[10:49] Mojito Sorbet: WHo has the new TPV URL?
[10:49] Latha Serevi: Anybody got a diff?
[10:49] Dzonatas Sol: Collada has do its refinery, Sai...
[10:49] Mojito Sorbet: there we go
[10:49] Mojito Sorbet: Latif has a diff
[10:49] Kopilo Hallard: viewerdirectory.secondlife.com or something
[10:49] Saijanai Kuhn: NOt at all the same thing, Dzonatas
[10:49] Zha Ewry: http://pastebin.com/Yd1j1EdE
[10:49] Rex Cronon: i g2g to do something really fast. tc people
[10:49] Morgaine Dinova: Plugins are unavoidable, because otherwise VW clients would have to contain *everything* within them.
[10:49] Eddy Stryker: wow!
[10:49] Kopilo Hallard: http://viewerdirectory.secondlife.com/
[10:50] Zha Ewry: That sould be the diff
[10:50] Eddy Stryker: they actually fixed it
[10:50] Latha Serevi: Thanks, looking at the pastebin diff.
[10:50] Latif Khalifa: Eddy, yeah
[10:50] Fleep Tuque: Yes I think they made somesignificant changes
[10:50] Latif Khalifa: lljoe needs a cookie :P
[10:50] Fleep Tuque: that should satisfy a lot of people's concerns
[10:50] Mojito Sorbet: The bad stuff is generally removed
[10:50] Eddy Stryker: satisfies mine
[10:50] Latif Khalifa: mine too
[10:50] Zha Ewry: Joe really wants to be a good guy on this, I think
[10:50] Fleep Tuque: 7a in particular
[10:50] Eddy Stryker: what am i going to complain about now?
[10:50] Fleep Tuque: hehe
[10:50] Latif Khalifa: lol
[10:50] Zha Ewry: Ummm
[10:50] Zha Ewry: Heh
[10:51] Eddy Stryker: how about stalled vwrap progress
[10:51] Mojito Sorbet: What surprised me was that Jow was able to get TPTB to yield so quickly
[10:51] Zha Ewry: We'llfin you something Eddy, don't want you to be lazy
[10:51] Zha Ewry: AHHH
[10:51] Zha Ewry: yes
[10:51] Zha Ewry: Code Faster Zha
[10:51] Eddy Stryker: i haven't seen the draft changes to the LLSD docs
[10:51] Dzonatas Sol: Collada has much input for what is 'minimal' for various formats and combined them into a single protocol... it extensibilities means that what custom XYZ data added to a collada file means it needs custom refiniery XYZ to compile it for platofmr ABC
[10:51] Mojito Sorbet: TPTB The Powers That Be
[10:51] Dzonatas Sol: platform*
[10:51] Saijanai Kuhn: mind you I'm not advocating etoys per se, only pointing out the capabilities and how they might fit into a 3D shared experience. Not ust collaborative tools and games, but more pervaisive stuff too
[10:51] Eddy Stryker: i defined a mock LLSD serialization for avatar appearance the other day. was 1.5MB of JSON or 500KB of LLSD. hooray for terrible JSON binary serialization
[10:51] Morgaine Dinova: 1c is not GPL compliant. You may not impose "further restrictions" other than those stated in GPLv2 associated with your license.
[10:52] Mojito Sorbet: Then later on they say "It wasnot our intent to violate the GPL". Thats something...
[10:52] Mojito Sorbet: INTENT
[10:52] Saijanai Kuhn: at one end, 2 or more people might agree to DL an etoy-like collaboration tool or game
[10:52] Fleep Tuque: Hm Morgaine
[10:53] Saijanai Kuhn: at the otehr, entire physics and other 3D world behaviors might end up as scripted extensions to your client
[10:53] Eddy Stryker: zha: don't code too fast you'll make me look bad. i still don't have rez_avatar/request in opensim
[10:53] Morgaine Dinova: GPLv2 is very specific about the "no further restriction". It even gives an example in the FAQ to highlight it. TPV clause 1c is in total conflict, since it places a restriction on distribution beyond the restrictions inherent in GPLv2.
[10:53] Zha Ewry: 6. Nothing in this Policy is intended to modify the terms of the GPL.
[10:53] Fleep Tuque: Morgaine: If they addeded ".. that connect to Second Life"? i.e. If you are a Developer of Third-Party Viewers that connect to Second Life
[10:53] Mojito Sorbet: Yes, they need to resolve that
[10:54] Morgaine Dinova: Zha: if nothing is intended, why does it state that "must disclose"
[10:54] Dzonatas Sol: Saijanai, of course... that's what we mean by self-contained inventory obejcts
[10:54] Zha Ewry: Well, I need to sort out an update to the 0.7 tree that includes the current OGXP code too
[10:54] Saijanai Kuhn: Dozantas, OK
[10:54] Mojito Sorbet: GPL is for software distribution. That is separate for Terms of Service in connecting
[10:54] Latha Serevi: That's probably the topic for today -- rephrasing 1c to refer to "if you're a dev, and you want your TPV to be allowed to connect..." or some such.
[10:54] Fleep Tuque: nod
[10:55] Mojito Sorbet: That is a good way to phrase it
[10:55] Zha Ewry: I *think* that the 0.7 code is almost real enough that I can imagine gettign back into the tree and looking at how to wedge the OGPX/VWRAPX code in
[10:55] Kopilo Hallard: sleepy time for me
[10:55] Kopilo Hallard: enjoy
[10:55] Eddy Stryker: zha: we almost have it done, just no commits yet
[10:55] Fleep Tuque: They can't regular TPVs that don't connect to their services, but they can regulate TPVs that DO connect to then. Or have a default "Connect to SL grid" option or osmething.
[10:55] Fleep Tuque: er regulate
[10:55] Zha Ewry: Oh.
[10:55] Zha Ewry: Eddy, we should talk
[10:55] Morgaine Dinova: Mojito: not good enough. GPL developers for SL are not 2nd class GPL citizens compared to GPL devs who do not develop for SL.
[10:55] Chaley May: where can i see the slideshow that joshua was showing? i came late
[10:56] Zha Ewry: (I looked in September/October and it changed so much as I was coding it I finally threw up my hands and wiated)
[10:56] Fleep Tuque: http://www.ietf.org/proceedings/10mar/slides/vwrap-6.pdf
[10:56] Fleep Tuque: Chaley
[10:56] Saijanai Kuhn: slideshow: http://www.ietf.org/proceedings/10mar/slides/vwrap-6.pdf
[10:56] Chaley May: ty fleep :)
[10:56] PulseBurst Flow: https://docs.google.com/a/lindenlab.com/present/view?id=0AfAytYlNCPECZGZ2M24yMl8yN2N0OGJueGNy&hl=en
[10:56] Chaley May: tnx
[10:56] Fleep Tuque: :)
[10:56] PulseBurst Flow: is more up to date
[10:56] Dzonatas Sol: lunch break....
[10:56] Chaley May: ok ty
[10:56] Fleep Tuque: nod must do some work too
[10:56] Chaley May: going to go to Andrew Lindens now
[10:56] Morgaine Dinova: Zha: can you adequately represent the tourist deployment in your talks with Eddy?
[10:56] Fleep Tuque: thanks all
[10:57] Chaley May: bye
[10:57] Morgaine Dinova: I'll be heading to OSgrid meeting
[10:57] Morgaine Dinova: Oh, Zha crashed?
[10:57] Morgaine Dinova: Zha: can you adequately represent the tourist deployment in your talks with Eddy?
[10:58] Saijanai Kuhn: She's totally inadequate to that task, Morgaine...
[10:58] Zha Ewry: I didn't crash
[10:58] Morgaine Dinova: Hardly, Zha is multi-talented :-)
[10:59] Dahlia Trimble waves and poofs... bye all :)
[10:59] Zha Ewry: And, anything that impinges on spec ends p on vwrap
[10:59] Morgaine Dinova: Zha: yes/no on Q?
[10:59] Morgaine Dinova: Cyu Dahlia
[10:59] Zha Ewry: if Eddy and I chat it's about details like code checkins
[10:59] Eddy Stryker: zha never calls anymore... :-(
[10:59] Eddy Stryker: hehe
[11:00] Zha Ewry: But, yes, I'll make sure tourist is in our discussion
[11:00] Free Portal: adios every one
[11:00] Zha Ewry: Heh. I'll be back calling soon
[11:00] Zha Ewry: I'm not sure how to cope with Linden making Eddy happy tho ;)
[11:00] Morgaine Dinova: Well if there is only one asset service provided in the protocol, then there is no tourism, period. So I don't want to see a blocking of tourism by de facto implementation.
[11:01] Zha Ewry: I think every place wherey ou can get an X
[11:01] Eddy Stryker: i'm not sure how to cope with it either. that was an unexpected turn of events
[11:01] Zha Ewry: is a map, so I'm not too worried
[11:01] Zha Ewry: I really don't like singltones
[11:01] Dzonatas Sol: so the state of vwrap is that it is still mutable due to capabilities
[11:01] Dzonatas Sol: ?
[11:01] Zha Ewry: More than that, Dzon, it's also a prety extensible design
[11:02] Zha Ewry: When we mess that up, I'm sure people will slap us hard
[11:02] Morgaine Dinova: Zha: it's only extensible if there is a LIST of asset services. If there is only one, it's non-extensible.
[11:02] Dzonatas Sol: i just worry about the long poll requirement, it takes up a resource
[11:02] Zha Ewry: Well, that's a "no singleton" issue
[11:02] Latif Khalifa: i saw someone tweet the other day "designing a protocol is sort of like getting a tattoo" :P
[11:03] Dzonatas Sol: or i should say it holds a resource
[11:03] Zha Ewry: And I'm against singletons, so we'll push on that hard
[11:03] Morgaine Dinova: Zha: yeah, but we have a singleton in the old CB
[11:03] Morgaine Dinova: Cool, Zha
[11:03] Zha Ewry didn't code it ;)
[11:03] Eddy Stryker: morgaine: i can promise you that the initial prototype vwrap implementation will make almost everyone unhappy. it's a fairly complex beast to get working (if you are building a complete virtual world stack), and i don't have time to add every policy decision and service choice that people want
[11:04] Zha Ewry: But that's code, not spec
[11:04] Eddy Stryker: but by the same token, i don't have time to write the code in a way that prevents specific usages (nor do i want to), so... feel free to improve on alpha v1
[11:04] Zha Ewry: And in terms of long poll, at the spec level, I think its going to be 90% abstracted out to "Two way pipe here" if I get my way.
[11:04] Morgaine Dinova: Eddy: yeah, I know, but I'm talking about specs, not code. As long as spec says "1+ asset services", I'm happy with an initial implementation having only one.
[11:05] Zha Ewry: Spec shoudl acyually say 0+
[11:05] Morgaine Dinova: Um
[11:05] Morgaine Dinova: OK, 0+
[11:05] Zha Ewry: Seriously
[11:05] Morgaine Dinova: Eg. chat requires no asset service
[11:05] Zha Ewry: I can do perfectly nice things with that model
[11:05] Zha Ewry: right
[11:05] Eddy Stryker: zha: speaking of long poll, i now have a C# server and client for web sockets :-)
[11:05] Morgaine Dinova: Yes, agreed Zha
[11:05] Zha Ewry: Which if we get a matching one on the client side would be loverly
[11:06] Zha Ewry: I for one will be much happer when the long poll goes the way of fax machines
[11:06] Eddy Stryker: seedCaps["event_queue_websocket"] = OSD.FromUri("ws://..."); // :-)
[11:06] Morgaine Dinova: Eddy: coo :-))) You probably have more than the websocket WG then :P
[11:06] Eddy Stryker: haha
[11:06] Zha Ewry: Oh, now, now, Hybi keeps my mail box nice and warm
[11:06] Zha Ewry: Dear god those peeps can talk and argue
[11:07] Zha Ewry: Does anyone have a nice C++ websocket impl that matches?
[11:08] Eddy Stryker: actually, i remember posting the code for my unity3d client that went through a vwrap login and got a websocket capability at the end
[11:08] Eddy Stryker: hmm good question
[11:08] Zha Ewry: Yes, yes you did
[11:08] Zha Ewry: I would love to see someone slip a websocket impl into snowglobe 2
[11:08] Eddy Stryker: zha: yes! the sirikata guys
[11:08] Morgaine Dinova: Zha: I'd like to play with that too.
[11:08] Zha Ewry: Oh! Of course
[11:08] Latif Khalifa: the day snoglobe2 compiles is a good day
[11:09] Zha Ewry: And ergo, a C++ websocket impl that BSD licenses
[11:09] Eddy Stryker: they had an implementation before anyone even knew what websockets were, heh. daniel was quick to the punch
[11:09] Morgaine Dinova: Could someone give us a 1-line update on where Sirikata are nowadays?
[11:09] Zha Ewry smiels
[11:09] Zha Ewry: *smiles
[11:09] Zha Ewry: They've been pretty quiet this month. Probably coding up a storm ;)
[11:10] Latif Khalifa: doesn't chrome/chromium implemetn web sockets?
[11:10] Eddy Stryker: morgaine: they're working on building a web-based client for their platform using one of the in-browser 3d solutions
[11:10] Zha Ewry: Unity, yes?
[11:10] Morgaine Dinova: Eddy: cool. Wish they'd appear on MMOX
[11:10] Eddy Stryker: lkalif: yeah but i think the sirikata implementation is a lot more isolated and easier to drop into something like the SL viewer
[11:10] Eddy Stryker: zha: no, web3d or ogl or one of those
[11:10] Zha Ewry: Oh fark
[11:11] Zha Ewry: I keep hoping mroe people line up with unity
[11:11] Eddy Stryker: oh, no worries about that
[11:11] Latif Khalifa: linux people hate unity :P
[11:11] Morgaine Dinova: Unity people reject Linux
[11:11] Zha Ewry ponders the continual religion and sighs
[11:11] Eddy Stryker: unity is becoming huge. and once version 3 rolls out it will be even bigger. the game industry is fawning over it these days
[11:12] Latif Khalifa: yeah
[11:12] Latif Khalifa: i wonder what mono version is unity3
[11:12] Zha Ewry: It does some things I don't love, but it seems like a sweet spot
[11:12] Eddy Stryker: 2.6
[11:12] Morgaine Dinova: Big mistake, putting even more into web browsers. Fricking webbies.
[11:12] Latif Khalifa: ah, and 3.5 compiler profile?
[11:12] Mystical Demina: all my projects are on unity now
[11:13] Eddy Stryker: yeah. it will be a recent mono because they've been working with miguel and crew at novell to get better soft debugging support into mono
[11:13] Latif Khalifa: cool
[11:13] Eddy Stryker: so the hope is unity 3 will have real debugging support
[11:13] Dzonatas Sol: i got two-way going on with SNOW-375... partially optional as GETs do a one-way... and if the cap doesn;t exist for the route back then it stays one-way
[11:13] Latif Khalifa: the libomv could go almost unmodified in unity
[11:13] Eddy Stryker: right
[11:13] Latif Khalifa: i wonder if there is a way to decode jpeg2000
[11:14] Mystical Demina: the are working on XNA/.net support
[11:14] Eddy Stryker: lkalif: yeah, csj2k
[11:14] Latif Khalifa: right, csj2k is fully managed
[11:14] Eddy Stryker: of course it will need to be hacked up depending on whether or not we get System.Drawing
[11:14] Latif Khalifa: well looks like making a SL/OpenSim client would be doable in unity
[11:14] Zha Ewry: I'm not thrilled with pushign rich content (3d shader stuff) into browsers, but.. I am happier doing it once than three times, and unity seems a good way to do it as any
[11:14] Zha Ewry: *as good
[11:15] Eddy Stryker: lkalif: if you start working on it let me know
[11:15] Latif Khalifa: embedding mono has many advantages, that calling unmanaged from mono
[11:15] Latif Khalifa: they have api that totally bypasses p/invoke
[11:16] Latif Khalifa: so you could go native speed on c# -> display updates
[11:16] Dzonatas Sol: more like a way to provide a built-in DLL if you embed mono
[11:16] Latif Khalifa: Eddy, will do
[11:16] Latif Khalifa: waiting for 3.0 to materialize
[11:17] Latif Khalifa: and then see what needs cutting from libomv to make it run
[11:17] Zha Ewry: And on that happy note, code to write prior to Joe's noon brownbag