User:Zero Linden/Office Hours/2008 Dec 16

From Second Life Wiki
Jump to navigation Jump to search
  • [13:00] Infinity Linden: okay
  • [13:00] Infinity Linden: it's 1PM by my clock
  • [13:00] Infinity Linden: shall we start?
  • [13:00] Infinity Linden: agenda items?
  • [13:01] Infinity Linden: i told morgaine i would try to get a little clarification for our open efforts moving forward...
  • [13:02] Zha Ewry: listens
  • [13:02] Infinity Linden: and unfortunately the most i can say right now is... "your request has been passed to the appropriate office and we are awaiting an answer."
  • [13:02] Dahlia Trimble: lol
  • [13:02] Mirt Tenk: Xugu!
  • [13:02] Saijanai Kuhn: Transcript of this morning's meeting for those who missed it: https://wiki.secondlife.com/wiki/AW_Groupies/Chat_Logs/AWGroupies-2008-12-16
  • [13:02] Lim Catteneo: "cannot comment on an ongoing investigation" ;)
  • [13:02] Xugu Madison: hi!
  • [13:03] Timeless Prototype: hi all
  • [13:03] Timeless Prototype: ah just missed it then :)
  • [13:03] Infinity Linden: in other words, there were enough meta-issues with the question of "where are we headed with the viewer?" that peeps eschelons above me want to make sure we provide as accurate an answer as possible
  • [13:03] Saijanai Kuhn: ANd transcript of today's OpenSim meeting as well: "
  • [13:04] Infinity Linden: anything else?
  • [13:04] Nepal Tairov: what kind of meta-issues?
  • [13:04] Lim Catteneo: the question was more along the lines of how does this supposed major work impact its open source status, I think
  • [13:04] Infinity Linden: the type of meta-issues I would be wise to leave to others to answer
  • [13:04] Infinity Linden: and honestly
  • [13:05] Infinity Linden: Robla (and to a lesser degree Soft) are the ones who will be taking lead on the FLOSS questions
  • [13:05] Saijanai Kuhn: well it doovetails with the discussion we had on the IMprudence plugin/.scripting thread: [1]
  • [13:05] Infinity Linden: you're going to get a much better answer out of Robla
  • [13:06] Dahlia Trimble: so are the imprudence client side scripting issues focused on user interfacing?
  • [13:06] Infinity Linden: cool link
  • [13:07] Saijanai Kuhn: I think its the same issues we've discussed for pyogp and likely for libmov
  • [13:07] Dahlia Trimble: I dont get the connection Sai
  • [13:07] Saijanai Kuhn: you've got several possible levels of interfacing/API: GUI, mid-level API (utilties) and packet level composition
  • [13:08] Infinity Linden: right. Zero, Tess and I generally focus on open protocols more than open source software
  • [13:08] Saijanai Kuhn: and how do you write plugins to handle them
  • [13:08] Infinity Linden: but it is just one point where "open" meets SL
  • [13:08] robomind Maximus: g'day grubers
  • [13:08] Nepal Tairov: so why reinvent the stack?
  • [13:09] Saijanai Kuhn: which stack?
  • [13:09] Infinity Linden: @Nepal... which stack?
  • [13:09] Dahlia Trimble: why scripting level packet handlers?
  • [13:09] Infinity Linden: jink
  • [13:09] Infinity Linden: jinx
  • [13:09] Nepal Tairov: application
  • [13:09] Saijanai Kuhn: Nepal, the GPL viwewer from Linden Lab is old, creaky, kinda sucky around the edges, and VERY hard to extend
  • [13:10] Infinity Linden: which is, of course, Sai's way of saying the code is "mature"
  • [13:10] Saijanai Kuhn: Dahlia, for testing purposes and for test bots like ENus wants to use iwth pyogp
  • [13:10] Infinity Linden: but the point that it is somewhat difficult to extend is cogent
  • [13:10] Dahlia Trimble: having spent some time developing an independent viewer, I now have more respect for the Linden client
  • [13:10] Infinity Linden: and apropos
  • [13:10] Zha Ewry: Its a non trivial bit of code
  • [13:11] Infinity Linden: yup. but i think we all recognize that the viewer code could be improved.
  • [13:11] Infinity Linden: hm
  • [13:11] Saijanai Kuhn: notes that it rivals the original Star Wars missile shield code in complexity (million lines of code)
  • [13:11] Infinity Linden: uh.. robomind
  • [13:11] Infinity Linden: i have a whip
  • [13:11] robomind Maximus: hello
  • [13:11] Videlia Yuitza: lol
  • [13:11] Infinity Linden: (which is my way of saying... you just sat on me...)
  • [13:11] Aimee Trescothick: I think that was what he was bargaining on
  • [13:12] Infinity Linden: awkward
  • [13:12] Timeless Prototype: there's a seat here sapre :)
  • [13:12] Videlia Yuitza: Robomind aka get your butt up
  • [13:12] Infinity Linden: thx
  • [13:12] Timeless Prototype: spare*
  • [13:12] Zha Ewry: Hard to type when your elbows are filled with viscera
  • [13:13] Infinity Linden: but i guess i have to take some blame... had i only used a few more pints of the patented developer blood, sweat and tears, we might not have the "two people sitting in the same space" problem
  • [13:13] Infinity Linden: but
  • [13:13] Dahlia Trimble: so have any specific goals been defined for client side scripting?
  • [13:13] Infinity Linden: back to the discussion...
  • [13:13] Infinity Linden: at Linden?
  • [13:14] Dahlia Trimble: anywhere
  • [13:14] Nepal Tairov: Dahlia Trimble: why scripting level packet handlers?
  • [13:14] Nepal Tairov: this in interesting
  • [13:15] Infinity Linden: i believe there is a general interest in making the client support arbitrary scripting languages, but i think we've all been much more focused on making mono on the simulator side an extremely good experience
  • [13:15] Infinity Linden: so
  • [13:15] Zha Ewry: Boith are good goals ;-)
  • [13:16] Infinity Linden: to answer the question... i don't think there are any official projects to support aribtrary client scripting given our finite resources
  • [13:16] Infinity Linden: what about OpenSim?
  • [13:16] Infinity Linden: looks at Zha
  • [13:16] Dahlia Trimble: I'm not aware of any client side scripting efforts around opensim
  • [13:17] Saijanai Kuhn: I've seen things flash by in OPenSim meetings about scripting and the like
  • [13:17] Saijanai Kuhn: I think it was in reference to realXtend
  • [13:17] Zha Ewry: Well, they'd not really be OpenSim related, unless one defined some hooks in the sim side
  • [13:17] Infinity Linden: ah ah
  • [13:17] Saijanai Kuhn: realXtend is sorta officially part of OpenSim?
  • [13:17] Bartholomew Kleiber: hi all
  • [13:17] Timeless Prototype: waves
  • [13:18] Dahlia Trimble: no it's more of the beginnings of a collaboration sai
  • [13:18] Lim Catteneo: sai, not really
  • [13:19] Saijanai Kuhn: KK my bad. Was scanning transcripts
  • [13:19] Infinity Linden: this awkward silence brought to you by *crickets* brand software...
  • [13:20] Zha Ewry: So. one thing which came up at AWG this morning, was would it be possible to get sort of an overall
  • [13:20] Zha Ewry: technical roadmap for the next six months?
  • [13:20] Zha Ewry: takes a deep breath
  • [13:20] Dahlia Trimble: OGP roadmap? :D
  • [13:20] Infinity Linden: i think that is doable
  • [13:20] Infinity Linden: i'll actually drag Zero back to his office hours for that, though.
  • [13:20] Zha Ewry: Well, I'm hoping for Lidnen's perspective
  • [13:20] Lim Catteneo: there is such thing? ogp roadmap? ;D
  • [13:20] Zha Ewry: sort of top to bottom
  • [13:21] Zha Ewry: overall goals, major projects, from Central services, infrastructure, through to Sim, OGP and Client...
  • [13:21] Zha Ewry: tries hard not to look *too* greedy
  • [13:21] Saijanai Kuhn: generic plugin meta-architecture...
  • [13:21] Infinity Linden: about the only thing i can do at the moment is ask the community's indulgance while we work through how our initiatives and projects map to long term strategy
  • [13:22] Zha Ewry: fails
  • [13:22] Lim Catteneo: oh well
  • [13:22] Nepal Tairov: how can we communicate unless there is a modular/open stack of generic stack of protocols... think of it like OSI
  • [13:22] Zha Ewry: nods
  • [13:22] Infinity Linden: hmm... this is beginning to sound like we need to get Joe in the loop
  • [13:22] Infinity Linden: which is vaguely difficult as his time is usually quite impacted
  • [13:23] Infinity Linden: we're hitting up upon issues that cross the boundary between open source and open protocols
  • [13:23] Zha Ewry: nods
  • [13:23] Zha Ewry: You are
  • [13:23] Infinity Linden: and are starting to ask for more high level guidance
  • [13:23] Saijanai Kuhn: Nepal OGP is supposed to be the start of that
  • [13:23] Zha Ewry: Part of this, Infi.. is that
  • [13:23] Tree Kyomoon: ahhhhhhhhhhhhhhh
  • [13:23] Zha Ewry: we're clearly somewhat off Zero's roadmap from 2007
  • [13:23] Dahlia Trimble: the cowboy era is over?
  • [13:23] Ollj Oh: i hurd trusting a server with login data is the most common issue
  • [13:24] Infinity Linden: (that being said... i don't have control of Joe's schedule... and it's entirely possible he might direct people under him to communicate some of the higher level stuff.)
  • [13:24] Zha Ewry: Which. hey, this is software, roadmaps are printed on tissue paper, and used by senior management to clean up spillerd coffee
  • [13:25] Saijanai Kuhn: for those playing catchup, anything from about april of this year (around the time Zha did the first SL OpenSim teleport) is good readign for catchup: https://wiki.secondlife.com/wiki/AW_Groupies#Chat_Logs
  • [13:25] Infinity Linden: and everyone knows that the real documents are written on napkins in bars at standards meetings
  • [13:25] Saijanai Kuhn: https://wiki.secondlife.com/wiki/Category:Grid_Interoperability_Chat_Logs
  • [13:26] Zha Ewry: ignroes the beer stained notes from WSA meetings and whistles to herself
  • [13:26] Infinity Linden: well.. since we don't have a roadmap for 2008
  • [13:26] Infinity Linden: what should be on it?
  • [13:27] Nepal Tairov: I think I'm in the wrong meeting, where do I go for the TCP Networking 101?
  • [13:27] Infinity Linden: lol
  • [13:27] Tree Kyomoon: /tree kyomoon owns "thinkonanapkin.com" so there
  • [13:27] Videlia Yuitza: lol
  • [13:27] SeanMcPherson Senior: pulls a receipt out of his wallet and is fully prepared ot take notes
  • [13:27] Infinity Linden: lol
  • [13:28] Tree Kyomoon: /I hate receipts these days...so small and crowded you cant write any ideas on them
  • [13:28] Dahlia Trimble: contemplates the stone wall bordering the nothwest sides of the parcel
  • [13:28] Lim Catteneo: so to smarize, LL on their commitment to open standards and open source "we'll get back to you" ;)
  • [13:28] Aimee Trescothick: TCP? That's sooo 1974
  • [13:28] Zha Ewry: What would we like on the roadmap?
  • [13:28] Infinity Linden: the process by which we came to agreement WRT doing teleport first was something like
  • [13:28] Zha Ewry: Lets see:
  • [13:29] Infinity Linden: "well... we kinda need to do login and teleport before we can do anything else..."
  • [13:29] Xugu Madison: Roadmap? Erm. Buy shiny things, hang on tree, celebrate end of year in preferred manner?
  • [13:29] Zha Ewry: A security roadmap. A c;oemt rpad,a[
  • [13:29] Zha Ewry: ah
  • [13:29] Zha Ewry: my
  • [13:29] Zha Ewry: Client Roadmap
  • [13:29] Zha Ewry: some story abnout where the RRR work fits in
  • [13:29] Tree Kyomoon: looks forward to being decorated
  • [13:30] Infinity Linden: @Lim... to summarize... Infinity Linden, who does not officially set open source policy for Linden Lab is, as a service to the community, spending time tracking down the people who do set this policy and expressing concerns from teh community.
  • [13:30] Infinity Linden: scribbles notes re: RRR
  • [13:30] Dahlia Trimble: the community is grateful :)
  • [13:31] Xugu Madison: For a 2009 roadmap, how's inventory between grids going? I'm not above sending cookies if it'll make development go faster...
  • [13:31] Infinity Linden: and yes... a security roadmap... like maybe something like "what the heck is this Trust Phase 0 and what does it buy me? and when will we see it deployed>"
  • [13:31] Ollj Oh: has anything done besides changing a network without client relog?
  • [13:31] Infinity Linden:  ?
  • [13:32] Tree Kyomoon: can we teleport to open life without relogging?
  • [13:32] Infinity Linden: @Ollj... you mean changing grids? or changing from OSGrid to SL?
  • [13:32] Infinity Linden: or changing from wired to wireless entworks?
  • [13:32] Infinity Linden: desperately hopes not to get into a discussion about SIP registration or MobileIP
  • [13:34] SeanMcPherson Senior: I'm not sure of the mechanics or who'd need ot hold the authenitcation tokens, as it were, but it'd nice to have an OpenID -likeprocess to tie avatars form various grids together, where if you're logged into Grid A (LL's grid) and want to go to Grid B (OpenSim) then you have to have gone thru a set of checks to confirm that you have all 3 sets of credentials, and that one each end you've confirmed that you want to allow xfer across grid namespaces. Is this what's currently on the roadmap, or something completely different?
  • [13:34] Infinity Linden: okay... security... inventory... how RRR impacts current work... a statement about the future of the client wrt Big Spaceship's work
  • [13:34] BlueWall Slade: eyes a large amt. of napkin scribbling over at #opensim-dev
  • [13:35] Tree Kyomoon: does linden labs :1. acknowledge open life 2. embrace open life 3. differentiates from open life because...?
  • [13:35] Infinity Linden: there is actually an open task to talk about OpenID
  • [13:35] Saijanai Kuhn: wants to have a dictionary definedalal AppleScript for generic client services so we can have a universal plugin/scripting model clientside
  • [13:35] Dahlia Trimble: wants belgian dark chocolate
  • [13:35] Saijanai Kuhn: for ANY client, not just realxtend/Imprudence/GPL/pyogp/Openviewer/etc
  • [13:35] Bartholomew Kleiber: lol
  • [13:35] Kyrah Abattoir: Anybody could tell me what are LL's plan to control the asset's migration from SL to foreign grids?
  • [13:36] Infinity Linden: in oher words... i'm happy to go up on stage and talk about what we perceive the specific weaknesses of the current implementation of OpenID to be
  • [13:36] Infinity Linden: though
  • [13:36] SeanMcPherson Senior: Applescript/hypercard's 'style' of extensibility and sanity would be a geat model for it, good idea :)
  • [13:36] Saijanai Kuhn: and yes, its perfect.y doable. Just define the servic3es and the syntax for sending the command from the script
  • [13:36] SeanMcPherson Senior: <- no relationship to OpenID nor any specific love for it, just using it as a generic reference people have heard of :)
  • [13:37] Lim Catteneo: i know that jhurliman is working on adding extended attributs to open id to support work on opensim
  • [13:37] Lim Catteneo: [2]
  • [13:37] Infinity Linden: cool. personally... i'm a little more impressed with SAML...
  • [13:37] Infinity Linden: and
  • [13:37] Dahlia Trimble: I would be interested in LL's perspectives of the weaknesses in openID
  • [13:37] Lim Catteneo: there is already patch to opensim asset servers which will allow to blacklist/whitelist speccific avatars/grids from using the assets
  • [13:38] Infinity Linden: if you squint, you could probably make out a way to extend the semantics (but not the syntax) of the current OGP spec to do OpenID auth
  • [13:38] Lim Catteneo: via transperent openId login
  • [13:38] Tree Kyomoon: so...will we ever be able to teleport from SL to OL?
  • [13:38] Bartholomew Kleiber: @lim: cool
  • [13:38] Videlia Yuitza: hiya
  • [13:38] Lim Catteneo: that sort of depends on LL ;)
  • [13:38] bend Bowie: sup
  • [13:38] Infinity Linden: and in a world where i had a lot more time, i would probably spend the time driving down to the googleplex to hash out how to overcome some of the difficulties i have with OpenID
  • [13:39] SeanMcPherson Senior: At some point it will be a policy question as opposed to a technology question, I'd say.
  • [13:39] Nepal Tairov: so with open ID is this another authentication method we're developing?
  • [13:39] Lim Catteneo: no no, this is not AWG or LL project, its libomv/opensim effort atm
  • [13:39] Infinity Linden: @Tree.. that is specifically a bizdev kinda question... transporting from SL to a foreign grid
  • [13:40] Infinity Linden: but... from a technology perspective we have demonstrated taht the SL codebase can support teleports between SL and assorted OpenSim instances
  • [13:40] Dahlia Trimble: there's currently a patch for opensim under consideration to allow the user servers to be openID providers
  • [13:41] Bartholomew Kleiber: oh?
  • [13:41] Tree Kyomoon: I hope the do allow it...and then just take the high road and differenciate themselves with more stable sims, more features etc. rather than just trying to pretend the open sims dont exist from a business perspective
  • [13:41] Infinity Linden: part of the security task that Zha mentioned earlier includes the technology behind making the decision as to allow an AV to tp into SL from an external source
  • [13:41] Infinity Linden: or vice versa
  • [13:41] Infinity Linden: so
  • [13:41] Infinity Linden: i think what we're going for here is "Mechanism, but not policy"
  • [13:41] Bartholomew Kleiber: should follow mailing lists more often.
  • [13:42] Dahlia Trimble: I dont think the intent of the patch is for grid access authentication though
  • [13:42] Infinity Linden: meaning... we'll define extensions to the OGP/Teleport protocol that very likely reference existing standards (like X.509 and PKIX) to provide not only SL with the ability to make such access decisions
  • [13:42] BlueWall Slade: the OpenSim OpenID is for a server in the (AD) part for some of teh comming subsystems to use
  • [13:42] Tree Kyomoon: Mech and poli are so intertwined...forming a monster that is beyond comprehension.. I call him PoliMech! And it is our job to snatch the hapless client from its deadly clutches!
  • [13:42] Infinity Linden: but anyone who's operating a grid
  • [13:43] Nepal Tairov: forest and trees
  • [13:44] Infinity Linden: okay... BlueWall... i'm getting some motivation to finish my security pieces in hope that we don't wind up using OpenID in a way we shouldn't
  • [13:44] Mirt Tenk: ty, bye
  • [13:44] Zha Ewry: Its got to be generic, and X.509 is our best hoap there
  • [13:44] Infinity Linden: but the idea is we want to define a mechanism that works well for both LL and OpenSim users
  • [13:44] Dahlia Trimble: now would be a good time to express one's concerns about openID :)
  • [13:44] Infinity Linden: and yeah... there have been a number of discussions
  • [13:44] Nepal Tairov: yes used this along time ago... 1974 I think,
  • [13:45] BlueWall Slade: yes, at the very least, the mech should be pluggable
  • [13:46] Infinity Linden: and X.509/PKIX is the technology taht seems to have the right bits to support what we want to do
  • [13:46] Zha Ewry: Note OpenID
  • [13:46] Zha Ewry: is about end user auth
  • [13:46] SeanMcPherson Senior: My biggest 'concerns' as a user is that I want ot be sure if I have to use a 3rd party auth provider that in a way sllowing for that provider to go poof without me losing everything, and for me to require that someone have all sets of keys displayed initially and at some recurrring basis to 'continue' the ability to get in. If I have to demonstrate I have credentials on both grids that I want to port from/to once every <X> interval, (kind of like Pop before SMTP back in the day) it's okay with me.
  • [13:46] Bartholomew Kleiber: people say it's crap but it's most likly considered a standard? :-)
  • [13:46] Zha Ewry: the X.509 leg is about the component level auth
  • [13:46] Zha Ewry: Then the thirfd leg is plicy
  • [13:46] Infinity Linden: well... i leave that as an exercise to the developer... to make this kind of stuff pluggable
  • [13:46] Zha Ewry: Policy
  • [13:46] Zha Ewry: So we can know the sim is somewhat trustable
  • [13:46] Zha Ewry: the user really is in fact Morgaine
  • [13:46] Zha Ewry: and then decide to give her non blingy shoes
  • [13:46] Infinity Linden: ack
  • [13:47] Infinity Linden: was hoping to not have the discussion on security policy as i have only 16 hours before i MUST get some sleep
  • [13:47] Zha Ewry: Three totally seperable, but entangled issues
  • [13:47] bend Bowie: 16 is way plenty
  • [13:47] Videlia Yuitza: lol
  • [13:47] Zha Ewry: 16 is enough to argue our way through the first few inches of the mile of stuff in there
  • [13:47] Infinity Linden: you've never written a distributed security policy language
  • [13:48] Zha Ewry: leases five minutes of a lock ont he key to the right to modify the ACL of the Infinity's whip
  • [13:48] Infinity Linden: <- wrote the distributed security policy language for SecureTreo and ALP
  • [13:48] Infinity Linden: and let me tell you
  • [13:48] Infinity Linden: it's like nailing jello to a tree
  • [13:48] Nepal Tairov: @infidlity and expensive, this still seems a authentication and directory problem
  • [13:48] SeanMcPherson Senior: ponders writing security policy and starts to salt the muzzle of a pistol. :)
  • [13:48] Infinity Linden: right.. auth is core to the discussion
  • [13:49] bend Bowie: pulls the trigger for him
  • [13:49] SeanMcPherson Senior: (If I'm gonna jam this thing in my mouth, it's at least gonna remind me a marguerita till the last possible second *grin*
  • [13:49] Infinity Linden: a policy description language is useless if you don't have a high degree of confidence that the person you're talking to will honor the assertions in the policy language
  • [13:49] Infinity Linden: and there are a couple aspects to auth here....
  • [13:49] [[User:emoteur [scriptemoteur]|emoteur [scriptemoteur]]]: Script run-time error
  • [13:49] [[User:emoteur [scriptemoteur]|emoteur [scriptemoteur]]]: Stack-Heap Collision
  • [13:49] Nepal Tairov: kerberos
  • [13:49] Nepal Tairov: kerberos
  • [13:50] Nepal Tairov: kerberos
  • [13:50] SeanMcPherson Senior: And Trust but Verify requires the mechanism for verify to mean something.
  • [13:50] Ollj Oh: but we should know better than thet about sl clients
  • [13:50] Infinity Linden: there's the auth that happens when a user logs in to the agent domain
  • [13:50] Infinity Linden: and the auth that a simulator (or agent host) does when talking to another agent domain or simulator
  • [13:50] Zha Ewry: user auth and component level auth
  • [13:50] Nepal Tairov: there should be only one auth
  • [13:51] Zha Ewry: Nepal, I can't decide to trust the sim, based on the auth of the people logged onto it
  • [13:51] Zha Ewry: I need it to prove to me, that it is who it says it is, and has a policy which I want to liev with
  • [13:51] Infinity Linden: and... at the end of the day, we have a business requirement that we be able to pass something like a PCI or SAS70 audit... so the likelihood we'll be using a TGS instead of a RA or CA is pretty low
  • [13:51] Zha Ewry: That's not a user level issue
  • [13:51] bend Bowie: shivvers at the word audit
  • [13:52] Nepal Tairov: no that's PKI and kerberos covers that
  • [13:52] Infinity Linden: kerberos... the TGS.. is neuberg ryuppel
  • [13:52] Infinity Linden: not PKI
  • [13:52] Infinity Linden: later versions of the kerberos software fromMIT did include support for PKI
  • [13:52] Infinity Linden: but
  • [13:53] Infinity Linden: i dont think we're going to specify a specific implementation of software in the standard
  • [13:53] Zha Ewry: In the web, X.509 and friends, are pretty much de-jure for PKI
  • [13:54] Infinity Linden: sighs but accepts the inevitability
  • [13:54] Nepal Tairov: X.509 is sooo over kill
  • [13:54] Infinity Linden: btu yes... X.509 for all it's flaws... does do what it does fairly well
  • [13:54] Infinity Linden: Nepal... what do you recommend?
  • [13:55] Zha Ewry: It is a bit of overkill, but, since you're likely to establish a http session, to move the data, it's pretty nicely webish
  • [13:55] Nepal Tairov: who is going to be the CA? revocations... etc. nasty
  • [13:55] Zha Ewry: grimaces and nods
  • [13:55] Zha Ewry: Oh, its all ugl, but whether is X.509, or any other PKI
  • [13:55] Zha Ewry: you have the exact same revocation issue
  • [13:55] Infinity Linden: right. i know of this company in San Francisco that already manages something like 15 million user accounts...
  • [13:55] Zha Ewry: (or you can do positive validation of public keys, if someone gets that actualy built)
  • [13:56] Zha Ewry: if you just do "shared secrets" you need a way to get the https sessions established, which is in theory, easy
  • [13:56] Nepal Tairov: like a passport?
  • [13:56] Infinity Linden: and for CA... it could be linden for SL or linden could just be the RA, delegating CA functions to *shiver* VeriSign
  • [13:56] Zha Ewry: but..99% of the web libraries, what they want is an https:// urI
  • [13:57] Zha Ewry: and a X.509 cert at the other end
  • [13:57] Infinity Linden: we're getting close to the end of the hour
  • [13:57] Infinity Linden: maybe the thing to do is for me to run off and write a bit more about the proposal
  • [13:57] Nepal Tairov: we use x.509 on firewalls and it goes to pot as soon as you're not in control of the CA
  • [13:57] Trident Numbers: hehe
  • [13:57] Zha Ewry: chuckles and recalls owing infinity part of that
  • [13:58] Infinity Linden: @Nepal.. how?
  • [13:58] Tree Kyomoon: is there a person we should ask about the bizdev plan of teleporting to openlife?
  • [13:58] Infinity Linden: ginsu
  • [13:58] Infinity Linden: or M
  • [13:58] Nepal Tairov: because you don't own the Issuing authority
  • [13:58] Infinity Linden: so?
  • [13:59] Dahlia Trimble: thanks :)
  • [13:59] Nepal Tairov: and in a highly meshed enviroment you hae to
  • [13:59] Zha Ewry: sighs and watches this head to stuff drink territory, where you get to argue over root certifates and signing
  • [13:59] Infinity Linden: i should ask if you're talking about an IPSec VPN or a PPP over TLS style VPN
  • [13:59] Infinity Linden: and exactly what you're using the cert for
  • [13:59] Nepal Tairov: IPSEC
  • [13:59] Infinity Linden: yeah... i'm waiting for that too
  • [14:00] Infinity Linden: that's when i mention that i was one of the three women who wrote hayden for VeriSign
  • [14:00] Saijanai Kuhn: calls for a recap of meeting. Where we go from here? Etc.
  • [14:00] SeanMcPherson Senior: notes his receipt is almost covered in notes but at this rate I'll have another bar receipt available shortly as I'll feel like I need a drink :)
  • [14:00] Nepal Tairov: Check Point communities or Cisco DMVPN the same applies
  • [14:00] Infinity Linden: oaky..
  • [14:00] Infinity Linden: so
  • [14:00] Infinity Linden: we have a request for a roadmap
  • [14:00] Tree Kyomoon: ok I sent my question to ginsu thanks infinity
  • [14:00] Infinity Linden: and it's going to be a technical roadmap
  • [14:01] Infinity Linden: not a bizdev type thing
  • [14:01] Zha Ewry: right
  • [14:01] Infinity Linden: and it's gonna have some security bits on it
  • [14:01] Zha Ewry: ideally with some hints of how to play with the various communities
  • [14:01] Infinity Linden: and there's a request to add inventory to it as well
  • [14:02] Saijanai Kuhn: groupo IM...
  • [14:02] Saijanai Kuhn: or at least a starting point for experimentation
  • [14:02] Infinity Linden: careful Sai... i have a whip
  • [14:02] Saijanai Kuhn: promises promises
  • [14:02] Zha Ewry: I have to run off to a Sl voice meeting
  • [14:02] Infinity Linden: and i'm not afraid to use it against people who use four letter words (like IM)
  • [14:03] Infinity Linden: but seriously...
  • [14:03] Infinity Linden: right
  • [14:03] Tree Kyomoon: thanks for the meeting folks. I have to run to
  • [14:03] Infinity Linden: we'll put IM on it
  • [14:03] Infinity Linden: and i think we have a bit of stuff to talk about specifically related to that
  • [14:03] Zha Ewry: nods
  • [14:03] Nepal Tairov: bye
  • [14:03] SeanMcPherson Senior: Just put in an IRC plugin and be done with it ;P
  • [14:03] Zha Ewry: Much of a muchness in all this
  • [14:03] Dahlia Trimble: IRC!!! :D
  • [14:03] Saijanai Kuhn: wants to mention the client plugin model. Is there a way to describe it so that any client could use?
  • [14:03] Infinity Linden: yup
  • [14:04] Infinity Linden: mmm... that's a tough one sai
  • [14:04] Infinity Linden: it's not impossible... consider NSAPI
  • [14:04] Infinity Linden: it was originally only supposed to be used for Netscape web servers
  • [14:04] Dahlia Trimble: is quite interested in plugin model use cases :)
  • [14:04] Saijanai Kuhn: well, like I said, AppleSCript dictionary could define generic utiltiies and higher-level functions that any regular client could expect to support
  • [14:04] Multi Gadget: v1.57.1 by Timeless Prototype
  • [14:04] Infinity Linden: but the IIS folk picked it up as well
  • [14:05] Saijanai Kuhn: login/logout/TP/inventory management, etc
  • [14:05] Infinity Linden: and personally.... i really like plugin architectures... making applications nothing but a message bus for a collection of plugins with well defined semantics
  • [14:05] Infinity Linden: the problem is.... defining those semantics can take a lot of effort
  • [14:06] SeanMcPherson Senior: the issue with using a plugin architecture for something that inherent is knowing when you can trust the black box the info comes out of and then you're back to layering security on security thru semantics and policy
  • [14:06] Infinity Linden: especially when you have a diverse developer community
  • [14:06] Saijanai Kuhn: right, though some can be trivial (like current OGP protocols)
  • [14:06] Infinity Linden: i hear you... and i really wish i had another 12 hours each day
  • [14:06] Saijanai Kuhn: a CORE" dictoinary of services plus extensions
  • [14:07] Infinity Linden: yup. in a sense, apache is almost like that
  • [14:07] Infinity Linden: they have a small core that knows how to deal with generic HTTP request flow
  • [14:08] Infinity Linden: and just about everything else is handled with a mod_somethingorother
  • [14:08] Infinity Linden: and i think one could argue that a) apache is a pretty darn'd stable system and b) fairly widely used
  • [14:08] Saijanai Kuhn: well, Enus wants to work on a pyogp-specific messaging thing. THinking we can use that as an example ofr something generic
  • [14:09] Infinity Linden: i'll have to link up with Enus again. i think the both of us have waaay too many project and waaay to little time
  • [14:09] Saijanai Kuhn: along the lines of Morgaines external scripting interface
  • [14:09] Morgaine Dinova: Back.
  • [14:09] Saijanai Kuhn: [3]
  • [14:09] Morgaine Dinova: Wow, just in time to see my name mentioned, lol
  • [14:10] Infinity Linden: okay... i gots to be on my way
  • [14:10] Infinity Linden: cheers all
  • [14:10] Infinity Linden: thanks for coming
  • [14:10] Saijanai Kuhn: take care Infinity. Thanks for hosting this
  • [14:10] Morgaine Dinova: waves bye
  • [14:10] Videlia Yuitza: yw
  • [14:10] Videlia Yuitza: nice to see you
  • [14:10] Dahlia Trimble: thanks :)
  • [14:10] PouletFritesMayO Boucher: bye
  • [14:10] Trident Numbers:  :)
  • [14:10] SeanMcPherson Senior: Thanks Infinity
  • [14:11] SeanMcPherson Senior: bamfs so folks who need to work, can.
  • [14:11] Dahlia Trimble: heads out, bye all :)
  • [14:11] PouletFritesMayO Boucher: bye all
  • [14:13] Saijanai Kuhn: OK, so lots covered or at least mentioned. I'd like to try to get the client-writers thinking about the semantics of a client dictionary. Some things are obvious, like login/logout/TP and some are not.
  • [14:13] Morgaine Dinova: Are you posting the transcript Sai?
  • [14:13] Saijanai Kuhn: and moving towarsd specalized plugins will be even harder I suspect
  • [14:13] Saijanai Kuhn: yep
  • [14:13] Morgaine Dinova: Coolness