User:Zero Linden/Office Hours/2007 Dec 06

• [8:33] Mircea Lobo: It won't? So there will never be a connection between something like opensim with the linden grid?
• [8:33] Morgaine Dinova: SS: that's what LL's storage backend should be, a distributed GFS filestore or similar, over iSCSI, with local disks doing only caching.
• [8:33] Melanie Milland: no
• [8:33] Teravus Ousley: Well, it won't happen until permissions are checked before committing by the asset server.. or some other 'overseer'
• [8:33] Melanie Milland: an "evil" sim can copy all content
• [8:34] Melanie Milland: make a new object, upload it
• [8:34] Melanie Milland: no way to guard
• [8:34] Melanie Milland: so it won't happen
• [8:34] Dr Scofield: evil viewers can as well
• [8:34] Mircea Lobo: Thats sad. i hope permissions will get fixed soon
• [8:34] Melanie Milland: no, they cannot copy scripts
• [8:34] Rex Cronon: hello everybody
• [8:34] Dr Scofield: ah, zero incoming
• [8:34] Melanie Milland: a sim can copy scripts
• [8:34] Melanie Milland: source and bytecode
• [8:34] Morgaine Dinova: It won't be guarded in an interop'd distriuted supergrid. It can't. Permissions will be advisary only.
• [8:34] Mircea Lobo: anyway, how about a way to export your inventory on this account to your computer? So, i could download my contents here to use in OS or something else
• [8:35] Morgaine Dinova: Morning, Zero
• [8:35] Mircea Lobo: many want to take their content from her ein opengrid
• [8:35] Neas Bade: morning zero
• [8:35] Melanie Milland: don't talk about that too loud
• [8:35] Zero Linden: Good morning
• [8:35] Zante Zapedzki: yay : )
• [8:35] Mircea Lobo: hello
• [8:35] Squirrel Wood: Morning ^^
• [8:35] Husserl Dreamscape: hi
• [8:35] Rex Cronon: hello zero
• [8:35] Dr Scofield: hi zero
• [8:35] Jason Swain: Hello Zero ^^
• [8:35] Zha Ewry: Welcome back oh hopefully less sickly Zero
• [8:35] Kerry Giha: Good Morning
• [8:35] Zero Linden: Well, still working from home
• [8:35] Pixel Gausman: hopes Zero isn't infectious
• [8:35] Harleen Gretzky: Hi Zero :)
• [8:35] Morgaine Dinova: Benefit of VW's ... can't pick up Zero's bug :-)
• [8:35] Zero Linden: One of the worst colds I've had in over a decade
• [8:35] Kovu Lian: greetings
• [8:35] Dr Scofield: hates colds
• [8:36] Rex Cronon: the bugs are evolving fast. humans have to catch up
• [8:36] Tsukhoi Culdesac: greetings Zero
• [8:36] Morgaine Dinova: A cold in early December is *good* --- you're immunized for Xmas.
• [8:36] Teravus Ousley: :D One of the benefits of using rubber gloves sticking out of a glass box for hazardous materials :D
• [8:37] Zero Linden: I wonder at what point VW viruses will evolve....
• [8:37] Zero Linden: There was "tinies"
• [8:37] Melanie Milland: to make avies sneeze? LOL
• [8:37] Morgaine Dinova: Heh, that's actually a good topic for interop
• [8:38] Dr Scofield: snow crash
• [8:38] Tsukhoi Culdesac: may i ask a question before u get too technical for me ?
• [8:38] Neas Bade: the "no pants" virus
• [8:38] Pixel Gausman: Neas: nice.
• [8:38] Dr Scofield: in just a few moments zero is going to pull out that piece of paper with static on it
• [8:38] Tsukhoi Culdesac: setting affinity to a double processor MAC ? Any way to do that ?
• [8:38] Dr Scofield: just...it's not really static ;-)
• [8:39] Zero Linden: Well - welcome to my office hours
• [8:39] Morgaine Dinova: Anyone want to take a guess how long before someone claims copyright on a picture of static?
• [8:39] Zero Linden: the transcripts are posted to the public wiki
• [8:39] Zero Linden: so speak openly
• [8:39] Mircea Lobo: i may be a bit semi-afk, so if you want to ask me something and i dont notice send me a im here
• [8:40] Zero Linden: I have no set agenda for today - I've been totally out-of-it for over a week now
• [8:40] Tsukhoi Culdesac: may i ask a question before u get too technical ?
• [8:40] Tsukhoi Culdesac: setting affinity to a double processor MAC ? Any way to do that ?
• [8:40] Morgaine Dinova: Zero: here's a question then. As Engineering Director, are you now overseeing the IM devs too?
• [8:41] Neas Bade: Zero, I think there is a lot of interest in the 1.8.6 RC viewer, and the new auth method. Especially as it seems to remove -loginuri, which means it is a no go with current OpenSim
• [8:41] Morgaine Dinova: We were wondering what your new scope was :-)
• [8:41] SS Sakai:  :(
• [8:41] Zha Ewry: Neas beats Zha to that
• [8:41] Teravus Ousley: seconds Neas's question
• [8:41] Zha Ewry: I was typing that question
• [8:41] Dr Scofield: thirds it
• [8:41] SS Sakai: lol everyone wants to know it!
• [8:41] Dr Scofield: also, is there doc on how it works on the wiki?
• [8:41] Morgaine Dinova: Fourthed! :-)
• [8:41] So Kuu: Hehe, indeed.
• [8:42] Rex Cronon: the documentation is the code:)
• [8:42] Zero Linden: whew
• [8:42] Zero Linden: well.....
• [8:42] PierreJoseph Proudhon: +1
• [8:43] Zero Linden: First, Tsuk. - this isn't the right forum for that sort of Q.
• [8:43] Zero Linden: Second, About my role
• [8:43] Zero Linden: True enough, I've been 'promoted' to "Engineering Director"
• [8:43] Zero Linden: and in LL form, "we'll figure out what that role is when we get there"
• [8:44] HUD Listener: v2.00.64 ALPHA: Script run-time error
• [8:44] HUD Listener: v2.00.64 ALPHA: Stack-Heap Collision
• [8:44] Zero Linden: Which is to say, I'm still trying to define it
• [8:44] Morgaine Dinova: Any initial ideas?
• [8:44] SS Sakai: well i hope it at least got u a bigger office?
• [8:44] Zante Zapedzki:  :]
• [8:44] aliceinwire Bleac: :D
• [8:44] Zero Linden: but there is a second part of to M's question that I need to discuss: "...are you now overseeing the IM devs too"
• [8:44] Harleen Gretzky: lol
• [8:44] aliceinwire Bleac: that'ssa good idea
• [8:44] Zero Linden: by "IM" do you mena Instant Messaging?
• [8:44] Morgaine Dinova: Yes
• [8:45] Dr Scofield: ss sakai: in-world?
• [8:45] Zero Linden: because one aspect of Linden engineering organization has been that we don't have
• [8:45] Neas Bade: congrats on the new position zero
• [8:45] Zero Linden: engineers that are long term assoiciated with a feature
• [8:45] Morgaine Dinova: Because, of course, IM issues are giving AWGroupies a huge headache.
• [8:45] aliceinwire Bleac: Halllooooo
• [8:45] aliceinwire Bleac: kitty
• [8:45] Kitty Tandino: Halllooooo
• [8:45] Kitty Tandino: alice
• [8:45] Zero Linden: Which is a bit different than most places I've worked
• [8:46] Tsukhoi Culdesac: Hello Kitty
• [8:46] Zero Linden: The good: There are no artificial boundaries of where functinality must start and stop, no pre-concieved limits on features, no one is kept out of code they need to fix to get things going
• [8:46] Kitty Tandino: hey hun
• [8:47] Kitty Tandino:  :)
• [8:47] Morgaine Dinova: So does LL make up for that by someone assigning work priorities, and someone grabs from the top?
• [8:47] Teravus Ousley: wonders what would be the better forum for the 1st question.. which was Tsk-ed at.
• [8:47] Zero Linden: The bad: there are no artificial boundaries between functionality (otherwise known as "Interfaces")
• [8:47] Zha Ewry: winces at that image
• [8:47] Kitty Tandino: (congrats zero)
• [8:47] Dr Scofield: ouch
• [8:48] Zero Linden: Yes -
• [8:48] Zero Linden: I perceive an issue
• [8:48] Neas Bade: heh
• [8:48] Zero Linden: So, M., yes, there is a large chunk of work, it is roughly prioritized by people
• [8:48] Zero Linden: The studio directors pick up high-pri. projects
• [8:49] Zero Linden: engineers choose to work with studios
• [8:49] Zero Linden: and code gets done
• [8:49] Mircea Lobo: Ok, there is something I wanted to ask Zero Linden about a certain part. Is it ok if i do? Sorry if i missed something couldnt follow this chat well
• [8:49] Zero Linden: And that benefit is that it is very flusid, quick to respond
• [8:49] Mircea Lobo: about opensim
• [8:49] Zero Linden: But, as it gets bigger, we need those Interfaces to be better, well, nailed down
• [8:49] Melanie Milland: it sounds just like an environment i could work in
• [8:50] Zero Linden: and not so fluid
• [8:50] Neas Bade: sounds like pretty classic growing pains. I know that one well :)
• [8:50] Melanie Milland: especially since i am likely to pull out the less popular things - data handling is my specialty :)
• [8:50] Morgaine Dinova: That should work fine, lots of places do something similar, with a dispatcher assigning priorities. Although I haven't seen studios before. But has IM somehow fallen between studio stools?
• [8:50] Zero Linden: I thinkthat, to tie this back to me, that I the essence of my role as Engineering Director (one of three), is to
• [8:50] Zero Linden: focus on the architecture and it's embodyiment
• [8:51] Mircea Lobo: Tell me when i can ask about this, just a little something i wanted to ask about that would help very very much
• [8:51] Zero Linden: and to help lead the engineering group to formailize it, whilst keeping their very fluid style
• [8:52] Zha Ewry: nods and then, looking a little pained, says "Can we drag this cool discussion over to -loginuri and openSim and similar for a moment?"
• [8:52] Zero Linden: The rpoblems with IM are at this point structural -
• [8:52] Zero Linden: they need some re-design and that is going on now, but will take some time
• [8:52] Zero Linden: NO
• [8:52] Zero Linden: NOW
• [8:53] Zero Linden: (sorry, forgot the W)
• [8:53] Zero Linden: So - -loginuri and OpenSim
• [8:53] aliceinwire Bleac: ao off kitty :P
• [8:53] SS Sakai: lol almost thought u were getting angry hehe
• [8:53] Mircea Lobo: Yes id like to talk about the -loginuri thing
• [8:53] Kitty Tandino: :P no
• [8:53] Kitty Tandino: lol
• [8:53] Kitty Tandino: ...L...
• [8:53] Kitty Tandino: ...O...
• [8:53] Kitty Tandino: ...L...
• [8:53] Kitty Tandino: ...Giggles...
• [8:53] aliceinwire Bleac: :D
• [8:53] aliceinwire Bleac: ypu levitate
• [8:53] Kitty Tandino:  :)
• [8:54] Zero Linden: So -
• [8:54] aliceinwire Bleac: great pose
• [8:54] Mircea Lobo: I'd like to know, Linden does intend to keep us compatible by releasing every client with support for -loginuri, correct?
• [8:55] Zero Linden: As in you want the command line option and code to be there, even though at some point
• [8:55] Zero Linden: Linden's own servers will not accept it
• [8:55] Zero Linden:  ?
• [8:55] SS Sakai: yes
• [8:55] Zha Ewry: Well, for now.. and a way to match what Linden does, eventually, I expect, in time
• [8:55] Rex Cronon: so if that will become absolete, than what is the new protocol?
• [8:55] Dr Scofield: right
• [8:55] Zha Ewry: Its mostly a mater of making sure we don't break things for the majority of people, while doing the right thing in the long term
• [8:55] Melanie Milland: and a command line option to redirect the viewer to that opensim login server, when that is done
• [8:55] Dr Scofield: any docs on that?
• [8:56] Zero Linden: I thought the details were communicated in the wiki and on SLDev
• [8:56] Teravus Ousley: :D Even if the 'login screen' says, (warning, this is not Linden Labs SeconfLife, this is something else)
• [8:56] Neas Bade: zero, right, but the new client can only connect to the linden infrastructure
• [8:56] Saijanai Kuhn: I've missed those details Zero
• [8:56] Dr Scofield: me too
• [8:56] Mircea Lobo: at the login screent
• [8:56] Mircea Lobo: *screen
• [8:56] Dr Scofield: didn't notice them on sldev
• [8:56] Zero Linden: Okay - then I'll get on that team to fix it
• [8:56] Mircea Lobo: It would be nice if a manual IP could be specified from the client
• [8:56] Neas Bade: where the old client could connect to any server that talked the right handshake
• [8:56] Squirrel Wood: so login is now based on html
• [8:56] Rex Cronon: where exacltly are those "details" documented on the the wiki?
• [8:57] Dr Scofield: and you get a CAP back?
• [8:57] Saijanai Kuhn: has been *creating* the login details lately
• [8:57] Dr Scofield: and send it in how?
• [8:57] Morgaine Dinova: Zero: the existence of the commandline option isn't related to LL servers accepting anything. It's required in the client so that people can point to other servers. The LL servers don't need to know anything about the option.
• [8:57] Rex Cronon: for the new protocol, saijanai
• [8:58] Saijanai Kuhn: and properly in a meta-grid world, itshouldn't be a commandline either, but part of the normal login GUI
• [8:58] Rob Linden: I suspect this was an oversight. we'll look into it
• [8:58] Saijanai Kuhn: REx: right, but like as not, we'll have to rework them to make them useable
• [8:58] Morgaine Dinova: Saijanai: agreed. The commandline should really just override the internal default, no more.
• [8:59] Zero Linden: Okay - so wait
• [8:59] Zero Linden: listen up
• [8:59] Mircea Lobo: Ok, so what will exactly be the fate of the -loginuri command, or how will logging in to other servers remain compatible with the sl client?
• [8:59] Zero Linden: Here is the non-technical details new flow
• [8:59] Mircea Lobo: in the fture
• [8:59] Neas Bade: listens intently to zero
• [8:59] Teravus Ousley: listens :D
• [9:00] Dr Scofield: listens as well
• [9:01] Zero Linden: a) User opens viewer, which displays web page X1, fills out log in form, which then POSTs to X2 and gets cap L back
• [9:01] Zero Linden: b) User opens webbrowser, surfs to web page X0, fills out log in form, which then POSTs to X2 and gets cap L back, and launches viewer
• [9:02] Dr Scofield: X2 is REST?
• [9:02] Zero Linden: THEN, c) Viewer POSTS to cap L to complete login, gets info that old LOGINURL used to provide
• [9:02] Zero Linden: Yes, X2 is REST
• [9:02] Zero Linden: SO
• [9:03] Zero Linden: The key is that the only configurable part that points the viewer to LL's grid, vs. OpenSim's is X1 or X0
• [9:03] Zero Linden: everyone see that?
• [9:03] Dr Scofield: so, libsl can post to X2 and get cap L and post to it?
• [9:03] Dr Scofield: yes
• [9:03] Mircea Lobo: so... the login form of the furute clients will be a web page?
• [9:03] Mircea Lobo: *future
• [9:03] Melanie Milland: it alsready is
• [9:03] Melanie Milland: in 1.18.6
• [9:03] Saijanai Kuhn: But needs to get X2 from X1 or X0
• [9:04] Melanie Milland: what will happen to the "remamber password" thing?
• [9:04] Dr Scofield: is X2 fixed?
• [9:04] Mircea Lobo: so how will connecting to external addresses like opengrid actually work?
• [9:04] Squirrel Wood: remember password => cookie ?
• [9:04] Zero Linden: Well, I imagine that either OpenSim would tell it's users to use method b and surf to Y0
• [9:04] Dr Scofield: opengrid has it's own X2'
• [9:04] Mircea Lobo: I see...
• [9:04] Zero Linden: or there could be a viewer option to start with an initial URL Y1, rather than the default X1
• [9:04] Melanie Milland: some people, me included are violently opposed to atrting a browser to log into sl
• [9:05] Melanie Milland: also, in linux, the browser cannto launch SL
• [9:05] Tree Kyomoon: violently? eek.
• [9:05] Zero Linden: SW = exactly - just like it already is on the SL website
• [9:05] Mircea Lobo: so you would neet to open the web browser separately with the client at each logon?
• [9:05] Morgaine Dinova: Don't think of it as a "browser", just because the login page is HTML.
• [9:05] Mircea Lobo: and go on a web page each time?
• [9:05] Zero Linden: MM = then use the browser that is within the client - nothing new to start
• [9:05] Saijanai Kuhn: thats what I meant about the command line should be in the GUI. LL vs opensims vs whatever *0 page should be a drop down list
• [9:05] Teravus Ousley: thinks the client has a browser in it.
• [9:06] Zero Linden: it is running already when you see the log in screen - that pretty image is an HTML page
• [9:06] Dr Scofield: zero, X2 is a fixed URI?
• [9:06] Mircea Lobo: it does
• [9:06] Melanie Milland: yes. i meant using a separate one
• [9:06] Zero Linden: Ds - no, of course not - it is ecoded in the HTML returned by X0
• [9:06] Zero Linden: or X1
• [9:06] Melanie Milland: no need for firefox's memory footprint, things are slow enough already
• [9:06] Tree Kyomoon: whats the URL of that page?
• [9:06] Dr Scofield: ouch
• [9:06] Zero Linden: Why is that ouch?
• [9:07] Dr Scofield: well, then you have to parse X0's/X1's HTML
• [9:07] Zero Linden: This enables OpenSim to have a totally differentauthentication mechanism -
• [9:07] Saijanai Kuhn: https we assume?
• [9:07] Dr Scofield: to get at X2
• [9:07] Melanie Milland: personally, it hink, phishing just got easier
• [9:07] Mircea Lobo: What i wanted to knw is, when i log in, can i still just open the SL client, wriithe the usernames and password ald log in? without anything else needed?
• [9:07] Zero Linden: Well, LL's will be HTTPS, I'm sure
• [9:07] Morgaine Dinova: I like Sai's idea on that. And alternative LCC clients can ignore the initial visuals entirely and just head straight for the URL in the drop-down list, or by commandline override.
• [9:07] Mircea Lobo: *know
• [9:07] Rex Cronon: u don't need a browser to get hmtl pages, just a library that does that
• [9:07] Zero Linden: ML - yes
• [9:07] Zero Linden: it looks identical
• [9:07] Mircea Lobo: in that case i even find that a great idea
• [9:07] Mircea Lobo: if i understood everything right
• [9:07] Groove Market: Zero, Im not sure if this is an appropriate question but we seem to have a great team of volunteers here working very hard to reverse engineer code and I appreciate that portions of your code may contain parts which are copyright by other companies. Is the plan to release some parts as open source or is the intention to get the community to reverse engineer the entire codebase? Is this something you can discuss?. It just seems a wasted effort to me when their time could be spent by these bright people on enhancing the architecture.
• [9:08] Zero Linden: well, not bit-for-bit identical, but functionally identical
• [9:08] Melanie Milland: the concept is alright - just rough around the edges
• [9:08] aliceinwire Bleac: wb kitty
• [9:08] Zero Linden: hold on, GM
• [9:08] Groove Market: ok
• [9:08] Kitty Tandino: ty love
• [9:08] Saijanai Kuhn: Groove, I'm not trying to reverse engineer anything. LIbsl may be going that route still for some legal issues
• [9:08] Zero Linden: Now, lastly, I have to admit, for awhile, "cap L" was just a UUID token -
• [9:08] Neas Bade: yeh, I think the new approach has a lto more flexibility. We'll just need to sort out the X0, X1 for OpenSim
• [9:09] Zero Linden: and that isn't general enough to work for OpenSim
• [9:09] Zero Linden: I don't know of the state of the current deploy code
• [9:09] Zero Linden: but before -loginuri ever goes away, then we'll have to have that part of the protocol be a real cap
• [9:09] Saijanai Kuhn: eventually, you may have login presence to the entire meta-grid from any vendor's login session, right?
• [9:09] Dr Scofield: so, a libsl client will have to go to X0 or X1, get the HTML, parse it to extract X2?
• [9:09] Zero Linden: that way X2 (or some other grid's Y2) can return a URL cap right into their login servers
• [9:10] Dr Scofield: or will there be a way to invoke a REST call to get X2?
• [9:10] Mircea Lobo: But if loginuri goes away, will we still be able to use opensim with the same client? How exactly... sorry missed a bit
• [9:10] Zero Linden: I suppose libsl should be supplied with X2 or Y2 directly, yes?
• [9:11] Dr Scofield: if X2 is fixed, yes
• [9:11] khires Blackburn: y
• [9:11] khires Blackburn: eeeeeeeeeeee
• [9:11] Mircea Lobo: I mean, the SL client normally connects to Linden when opened. What will give it that information to look for another address when started?
• [9:11] khires Blackburn: eeee
• [9:11] Zero Linden: ML - it'll need an option, say -loginpage
• [9:11] Dr Scofield: "screen scraping" HTML to get at X2 is not really that reliable
• [9:12] Morgaine Dinova: Jeez, no screen scraping pls ;-)
• [9:12] Saijanai Kuhn: L-X2 would eventually be non-fixed I would expect.
• [9:12] Neas Bade: yeh, as long as there is an option like that, it will be fine
• [9:12] Melanie Milland: it's not precisely screen scraping - it will be the action of the form on the page
• [9:12] Zero Linden: eh, I call that screen scraping
• [9:12] Neas Bade: it would also be ideal if the search url wasn't hard coded, and was returned to the client somewhere in the X2 handoff
• [9:12] Saijanai Kuhn: and again, should have the option to have a drop down to list possible pages
• [9:12] Zha Ewry: Just needs to be easy enogh to do
• [9:13] Neas Bade: but that's for another day :)
• [9:13] Zha Ewry: And yes, any of the URLs which tie in, should be passable like neas mentioned
• [9:13] Mircea Lobo: Will we be able to specify what page it will connect to at startup?
• [9:13] Zero Linden: especially as we do things like occasionally put survyes on that page
• [9:13] Dr Scofield: just fix the X2 URL and do load balancing via DNS or so
• [9:13] Neas Bade: zha, I wouldn't want searchurl to be passible
• [9:13] Melanie Milland: can't mandate that
• [9:13] Dr Scofield: in which case there will be multiple forms?
• [9:13] Neas Bade: I'd want it to get that from the agent domain
• [9:13] Teravus Ousley: well, I suppose, you could write that part in XHMTL :D
• [9:13] Melanie Milland: not if it should remain scalable
• [9:13] Zha Ewry: That's what I mean
• [9:13] Zha Ewry: Passable
• [9:14] Zha Ewry: Not on the login page, but.. not fixed,, but asusmption
• [9:14] Rex Cronon: zero, on which page do u put surverys, on X0 or x1?
• [9:14] Zero Linden: DR - either that or do one step, LLSD based call to X0 that gets X2
• [9:14] Melanie Milland: search url must come from X2
• [9:14] Neas Bade: the 1.8.5 client has the fun of search being on the main grid when you are in an opensim
• [9:14] Zero Linden: RC - both!
• [9:14] Melanie Milland: don't want opensim users to get SL results
• [9:14] Dr Scofield: that would be fine as well
• [9:14] Saijanai Kuhn: Zero, so its webpage/browserpage => X2cap => xmlrpc cap ?
• [9:14] Morgaine Dinova: Well let's not forget that eventually there will be zillions of these login points if it goes as planned. So, there definitely will be search engines supplying the entry points.
• [9:15] Zero Linden: no, we eliminate the xmlprc cap I believe, and just return the same data, but now as a nice tidy LLSD
• [9:15] Zero Linden: BUT
• [9:15] Zero Linden: talk all I say with a grain of salt
• [9:16] Zero Linden: we need Tess here to tell us where they actually are in all this
• [9:16] Squirrel Wood: Here's your salt :p
• [9:16] Zha Ewry: Lets try and get that to happen
• [9:16] Zero Linden: LOL
• [9:16] Neas Bade: heh
• [9:16] Kitty Tandino: lol
• [9:16] Kitty Tandino: ...L...
• [9:16] Kitty Tandino: ...O...
• [9:16] Kitty Tandino: ...L...
• [9:16] Kitty Tandino: ...Giggles...
• [9:16] Mircea Lobo: would appreciate it if -loginuri would still remain there just as a backup, as it doesnt harm anyone to leave that little command line supported
• [9:16] Zha Ewry: We need to not churn people up over this, for no good reason
• [9:16] aliceinwire Bleac: :D
• [9:16] aliceinwire Bleac: is a big salt
• [9:17] Zero Linden: Okay - Someone else earlier own had an OpenSim question for me (!)
• [9:17] Kitty Tandino: hehhe
• [9:17] Zante Zapedzki: is talking about html on a prim taboo? : p
• [9:17] Mircea Lobo: yes i wanted to ask something erlier
• [9:17] Melanie Milland: well, loginuri actually is a big deal.
• [9:17] Mircea Lobo: *eariler
• [9:17] Mircea Lobo: if thas ok
• [9:17] Kitty Tandino: WOOOT WEB ON PRIM!
• [9:17] Kitty Tandino:  :)
• [9:18] Melanie Milland: since the login part is a browser window now, loginuri means to also have the old panel in there
• [9:18] Zero Linden: go for it, ML
• [9:18] Mircea Lobo: Thank you. Ok, what i wanted to ask is;
• [9:18] Zero Linden: MM - yes, that would be true
• [9:18] Saijanai Kuhn: well, we need a way to reroute the viewer to sniffers fro testing/debugging purposes, and eventually people will want to have their own private sims for logging into if they don't do that already
• [9:18] Teravus Ousley: notes that his login implementation question was answered.. if it was him that Zero thinks stil had the question
• [9:18] Melanie Milland: also...
• [9:18] Melanie Milland: some open source patches made modifications to that login panel
• [9:18] Mircea Lobo: Opensim is currently developing, between many things, scripting and LL functions, which we have to intrioduce manually for scripting to work there fully like here
• [9:18] Melanie Milland: most notably one, to remove the ability to log in anywhere but last location
• [9:19] Mircea Lobo: But it will take a lot. I have thought that even if lindens softwate is c++ (if im right) and OS is c#,
• [9:19] Mircea Lobo: Linden could be so kind as to send us their complete LSL database
• [9:19] Mircea Lobo: WIth the right recognitions for every ll function in scripots
• [9:19] SS Sakai: smiles...
• [9:19] Zero Linden: As in the list of calls?
• [9:19] Ash Qin: Language shouldn't matter too much, it's the actual implementation that is hard.
• [9:19] Zero Linden: Or as in their implementation
• [9:19] Mircea Lobo: Yes, the LSL database. That tells what each ll command does
• [9:19] Mircea Lobo: implementation
• [9:20] Rex Cronon: why would u want to force people to login to the same sim?
• [9:20] Ash Qin: ...What?
• [9:20] Mircea Lobo: like tha tcode that says llSay makes it say That in the console
• [9:20] Ash Qin: Just use the LSL functions wiki?
• [9:20] Mircea Lobo: no,
• [9:20] Saijanai Kuhn: libsl is C#. The modular client, as little as we've done, is language neutral. We've got login scripts in Python and LSL and also in Java for example
• [9:20] Melanie Milland: RestrainedLife, the bdsm viewer. it has an otpion to prohibit teleporting. being able to relog to anoither sim subverst that
• [9:20] Ash Qin: It is unlikely the code from LL would work with opensim
• [9:20] Mircea Lobo: i meanthe database that implements scripts to work
• [9:20] Mircea Lobo: the code that says what llSomething means and does
• [9:20] Zero Linden: ML - I'm pretty sure we don't have a database of that sort
• [9:20] Zero Linden: there is a listing of the functions and arguments and short descriptions
• [9:21] Teravus Ousley: :D The LSL Wiki is pretty complete
• [9:21] Zero Linden: and that listing file is part of the common parts of the sima nd viewer
• [9:21] Mircea Lobo: I mean scripts work here with their ll function. I wonderred if Linden could send us that database
• [9:21] Saijanai Kuhn: there's the compiler data in the source...
• [9:21] Zero Linden: so it is in the open source viewer
• [9:21] Ash Qin: Mircea, the LSL wiki has all the functions listed and their parameters.
• [9:21] Saijanai Kuhn: AND its in the source code
• [9:21] Saijanai Kuhn: the tool tips info
• [9:22] SS Sakai: sauce?? intersting...
• [9:22] Mircea Lobo: just a sec
• [9:22] Dr Scofield: (but viewer source is GPL and libsl is BSD...)
• [9:22] Morgaine Dinova: Sai's point is important. For development, we need the ability to route the official viewer through our tools. With the sparcity of documentation, a lot depends on our ability to see what's going on through various filters amd proxies etc.
• [9:22] Squirrel Wood: open sauce, add code.
• [9:22] Saijanai Kuhn: a list of function names and parameters is for a language. I thought those were open to anyone to use, regardless of licensing
• [9:23] Saijanai Kuhn: wait, those are more than just the language NM
• [9:23] Mircea Lobo: here is what i am talking about : [1]
• [9:23] Mircea Lobo: this is the implementation code i meant for a function. one example of course
• [9:23] Groove Market: Mircea, so you mean the server code (c++) for running LSL scripts?
• [9:23] Ash Qin: The code LL uses likely would not work in opensim
• [9:24] Mircea Lobo: I asked is Linden could send us theirs for each llSomething
• [9:24] Zero Linden: We are not ready to release the implementation of the server source
• [9:24] Zero Linden: (this ties into GM's question)
• [9:24] Mircea Lobo: yes
• [9:24] aliceinwire Bleac: [2] is down
• [9:24] Mircea Lobo: im sorry if its not compatible and it was silly to ask, i just tohught it could work
• [9:24] Saijanai Kuhn: besides right now, you'd also get the bugs too.
• [9:24] aliceinwire Bleac: database problem
• [9:24] Mircea Lobo: I mean for each ll function to work, there is a database for each ll function with such code.
• [9:24] Dr Scofield: lol
• [9:24] aliceinwire Bleac: toomany coonection
• [9:24] Saijanai Kuhn: though perhaps you need the bugs for compatibility
• [9:25] Zero Linden: no - no database, just a giant C++ file or two
• [9:25] Mircea Lobo: so, the Linden one wouldnt be competible with Opensim?
• [9:25] Melanie Milland: yes, so much in sl worsk on sls side effects
• [9:25] Teravus Ousley: :D probably not.
• [9:25] Ash Qin: imagines the HTTP and xml-rpc stuff scale across many files and servers
• [9:25] Zero Linden: I don't think our code would help OpenSim much
• [9:25] Melanie Milland: the side effects (=bugs) need to be ported
• [9:25] Zero Linden: for things like string handling, perhaps the implementation would help you get the odd corner cases
• [9:25] Mircea Lobo: ah ok now i nuderstand
• [9:26] Teravus Ousley: Particularly, some things that may help though.. is with regards to things like llMovetoTarget
• [9:26] Mircea Lobo: *understand
• [9:26] Zero Linden: right, but for things like parcel functions, the implemetnation isn't going to help a wit
• [9:26] Teravus Ousley: I think Andrew discussed how that function works though in a previous meeting.
• [9:27] Saijanai Kuhn: maybe a list of functions could be worked up ?
• [9:27] Neas Bade: yeh, I don't think direct code would help much
• [9:27] Teravus Ousley: we're talking procedural psudo code :D
• [9:27] Saijanai Kuhn: we need to know details of this, this and this...
• [9:27] Zha Ewry: it may be worth having a discussion on things like how LSL is gong to handle low lvel strings in mono and such.. if you're mostlygoing to use standard libraries of stuff.. that could be a good way to keep things consistent on the low level
• [9:27] Melanie Milland: use the source, luke
• [9:27] Ash Qin: A list of functions is pn the LSL wiki
• [9:27] Neas Bade: the wiki reference is actually very good at behavior specification for most of LSL
• [9:27] Zero Linden: The language designer in me would of course much prefer to see the wiki become a formal spec for LSL
• [9:27] Saijanai Kuhn: they need implementation details, especially for the oddbal ones with sideeffects
• [9:27] Zero Linden: rather than take our particualr implementaiton and turn that into an operational semantic
• [9:28] aliceinwire Bleac: that's can be great
• [9:28] Melanie Milland: yes but it is just as important to watch for what it _doesn't_ say
• [9:28] Zero Linden: Yikes - look at the time
• [9:28] Mircea Lobo: Also another thing id like to know, is there a irc channel where lindens and sl users hang together, where you could talk with someone each time?
• [9:28] Zero Linden: Alas, one side effect of doing this at 8:30 is that I have a 9:30am meeting
• [9:28] Zero Linden: SO - I can't linger at all
• [9:28] Morgaine Dinova: Zero: that would be excellent, because then we'd just add O(1) arrays into the wiki, and it would be in the spec. ;-))) /me giggles
• [9:28] Saijanai Kuhn: carry on
• [9:28] Ash Qin: This reminds me of what wine does -- they implement how Windows behaves, not what the MSDN (Microsoft's official API specs), because the purpose of the Wine engine is to run Windows PE programs, not to implement a 'correct' API that windows does not implement.
• [9:28] Rex Cronon: bye zero
• [9:28] Zero Linden: Thank you all for coming
• [9:29] Dr Scofield: thx for your time, zero
• [9:29] Kitty Tandino: thanks for the info zero :)
• [9:29] Neas Bade: thanks zero
• [9:29] SS Sakai: ciao Zero
• [9:29] aliceinwire Bleac: thank you zero
• [9:29] Morgaine Dinova: Cheers Zero
• [9:29] Zante Zapedzki: thanks Zero : )
• [9:29] Mircea Lobo: Wait! i remembered something else that is important. i would like to talk about this here
• [9:29] Ash Qin: Good bye, Zero.
• [9:29] Teravus Ousley: Thanks Zero :D
• [9:29] Zero Linden: See you all on Tuesday