User:Enus Linden/Office Hours/2008 December 05

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

[9:32] Saijanai Kuhn: g'morning
[9:32] Enus Linden: howdy
[9:32] Enus Linden: how are ya sai?
[9:33] Saijanai Kuhn: stil asleep but awake enough for nozzzzzzz
[9:33] Saijanai Kuhn: ;-)
[9:35] Morgaine Dinova: 'Morning
[9:35] Enus Linden: hey morgaine
[9:35] Enus Linden: how are you today?
[9:35] Morgaine Dinova: How goes Enus?
[9:35] Morgaine Dinova: Fine thanks :-)
[9:35] Enus Linden: goes well here
[9:35] Enus Linden: late night
[9:35] Saijanai Kuhn: you get all those tests workign?
[9:36] Enus Linden: listening to a meeting in the ears, hosting one on the chat
[9:36] Enus Linden: sai: i updated most of them, but need to test from inside the firewall
[9:36] Enus Linden: so made changes blindly
[9:36] Morgaine Dinova: I no longer know if I do late nights, early mornings, or Martian time. Sync'd to CA time completely destroys me. (UK here)
[9:36] Enus Linden: i'll but them on a host inside today and see how i did
[9:36] Enus Linden: time is subjective
[9:37] Morgaine Dinova: Lunch time doubly so
[9:37] Saijanai Kuhn: Zha may show up
[9:37] Tammy Nowotny: hello
[9:38] Enus Linden: morgaine, i hear you maybe took a look at pyogp?
[9:38] Enus Linden: heya tammy
[9:38] Morgaine Dinova: I installed PyOGP at last. Just need Gridnauts membership to try it.
[9:38] Tammy Nowotny: still rezzing, didn't emands to seem rude
[9:38] Enus Linden: i can help with that morgaine
[9:38] Saijanai Kuhn: emand?
[9:38] Saijanai Kuhn: mean* ah
[9:38] Morgaine Dinova: I've got to delete a group slot to make room for Gridnauts, one sec
[9:38] Enus Linden: morgaine
[9:38] Tammy Nowotny: *mean
[9:39] Enus Linden: i can just do it on aditi
[9:39] Tammy Nowotny: yes, we ned more than 25 group slots on the main SL grid
[9:39] Patnad Babii: Hi everyone :)
[9:39] Enus Linden knows how to cheat
[9:39] Morgaine Dinova: lol
[9:39] Enus Linden: Hi everybody!!! :D :D :D
[9:39] Tammy Nowotny: there is always the worklaround fo creating alts
[9:40] Morgaine Dinova: OK, got rid of a group, have a slot spare now.
[9:40] Morgaine Dinova: Oh yeah, keep forgetting about alts, lol
[9:40] Saijanai Kuhn: I've ony got one alt registered for gridnauts. Guess once we really get started, could use more, but for now, one to use the regular client and one to test pyogp with, and I'm fine
[9:41] Saijanai Kuhn: fun to watch the pyogp client rez
[9:41] Saijanai Kuhn: OMGITWORKS!
[9:41] Morgaine Dinova: Thanks Enus
[9:41] Enus Linden: i'll have to manually set up aditi
[9:41] Enus Linden: will let you know when i'm done
[9:41] Enus Linden: (later my morning)
[9:41] Enus Linden: sai: it was the same fun getting the tests working last night
[9:42] Saijanai Kuhn: yeah, but at 3 am its OMGITWORKzzzzzzz
[9:42] Morgaine Dinova: Hehe
[9:42] Enus Linden: so i don't have any agenda items in mind today
[9:43] Asterion Coen: hello folks
[9:43] Enus Linden: i've started re-engaging on pyogp agani some lately
[9:43] Enus Linden: hi asterion
[9:43] Enus Linden: i've reviewed OH transcripts from earlier this week
[9:43] Saijanai Kuhn: You 've been chatting with fword and bel about changing the object model in some way...
[9:43] Morgaine Dinova: If no agenda items, do you see any vague roadmap for PyOGP?
[9:43] Asterion Coen: (rezzing)
[9:43] Enus Linden: i do see a vague roadmap morgaine, and i ought to send it to the mailing list but
[9:43] Enus Linden: lemme type what i know
[9:44] Enus Linden: 1. get tests back in line with the refactored base libs
[9:44] Enus Linden: 2. propose an object model
[9:44] Enus Linden: 3. refactor wrt to the object model and stub out future classes
[9:44] Enus Linden: 4. further remove zca from the packages
[9:45] Enus Linden: 5. improve caps handling throughout
[9:45] Enus Linden: 6. establish base packet handlers
[9:45] Enus Linden: and 7. um, plenty more
[9:45] Saijanai Kuhn: hoped this thing would work but it doesn't oh well
[9:45] Enus Linden: hi zha
[9:45] Enus Linden: emosad on emoteur
[9:46] Morgaine Dinova waves at Zha
[9:46] Zha Ewry pulls out a tissue, blows her nose and waves
[9:46] Morgaine Dinova: Bah, so restrained. /me blows a kiss at Zha
[9:46] Saijanai Kuhn: ginger, crystalized ginger for allergies
[9:46] Enus Linden hands zha some hot tea and honey
[9:46] Zha Ewry needs RL antivitus software for the atomic affiliate
[9:46] Saijanai Kuhn: put some of the ginger in it too
[9:46] Morgaine Dinova: Mmmmm, tasty
[9:47] Enus Linden: morgaine, are you on the mailing list?
[9:47] Asterion Coen: oh, a cat (that change from teddy bears)
[9:47] Tammy Nowotny: Ginger. yummy!
[9:48] Enus Linden: i'll send the tentative roadmap to the mailing list soon, early next week at the latest
[9:48] Morgaine Dinova: I'm off all mailing lists, lol, ever since my mailserver broke. But I poke around on archives a bit, though not PyOGP's yet. Will do.
[9:48] Tammy Nowotny hands Zha some mega doses of Vitamin C
[9:48] Enus Linden: of course if i say that, i'll fail to do so :)
[9:49] Saijanai Kuhn: Amrit Kalash. More powerful than C (and more expensive)
[9:49] Asterion Coen: that's the prob to have his mail server into some 386sx puters
[9:49] Morgaine Dinova: Zha: get some Echinacea, works wonders.
[9:49] Enus Linden: the object model bit is crucial to avoid big refactors down the road
[9:49] Saijanai Kuhn: oddness. Won't draw here. COuld it be a non lltouch sim?
[9:50] Tammy Nowotny: no I see some lines on yr board, Sai
[9:50] Asterion Coen: (a nice draw for sure) :)
[9:50] Enus Linden: morgaine: one thing i could use is potential use cases for what one might want to see in the short term in pyogp
[9:50] Morgaine Dinova: I think the whiteboard has a concurrency problem
[9:51] Enus Linden: it enables testing things, can be a good bot client in time
[9:51] Saijanai Kuhn: ah OK, yeah need to set official chalk-holder
[9:51] Enus Linden: but, working toward a goal or goals can help shape what gets built out when
[9:51] Saijanai Kuhn adds to do list along with mifty million onther things
[9:51] Enus Linden: mifty million is a lot!
[9:51] Saijanai Kuhn: mifty fillion*
[9:52] Saijanai Kuhn: definitely needs work before it can become a visual aid
[9:53] Asterion Coen: nice tool though :)
[9:53] Asterion Coen: usefull when u r lost in a sim with a mark "you are here"
[9:53] Saijanai Kuhn: can make floating diagrams out of it, so yeah, floating signs, etc
[9:54] Saijanai Kuhn: wasteful of prims though
[9:54] Morgaine Dinova: Enus: I don't have any specific views on PyOGP. I tend to view it through the eyes of my earlier proposal, which didn't take off -- https://wiki.secondlife.com/wiki/Multi-Process_Client_VAG_--_draft
[9:54] Asterion Coen: temp ones ?
[9:54] Saijanai Kuhn: anyway, back to pyogp
[9:54] Asterion Coen: but yes, not a good tool to draw a paint !
[9:54] Morgaine Dinova: So I'm just hanging on here at this stage, seeing where it goes.
[9:54] Enus Linden: k morgaine
[9:55] Saijanai Kuhn: well, I've been playing (sorta recently) with adding a packet display to it. Had the bare beginnings working then distracted as usual
[9:55] Enus Linden: oooo, one biig thing we need to solve is how to handle processes in the client
[9:55] Asterion Coen: hey morgaine ;)
[9:55] Morgaine Dinova: Enus: but it seems to me that PyOGP is very much inline with what I wanted, just not as refactored.
[9:55] Enus Linden: native python threads are teh suck w/o excessive care
[9:56] Morgaine Dinova: Forget threads. Trust me, you don't want to go there.
[9:56] Enus Linden: and i'm not willing to jump to python 3 as the min version
[9:56] Saijanai Kuhn nods. We need to ask Donovan for some advaice on how to use eventlets with it I think.
[9:56] Enus Linden: so eventlet comes to mind
[9:56] Saijanai Kuhn: don't think that Python 3 solves thread issues either
[9:56] Enus Linden: i have kartic at my disposal, and whish
[9:56] Enus Linden: which
[9:56] Enus Linden: so ought to be able to prototype things in eventlet soonish
[9:57] Saijanai Kuhn: its an interpreter thing. locks the entire interpreter with each instruction
[9:58] Enus Linden: so other than pyogp: what's up out there?
[9:58] Morgaine Dinova waves at Asterion :-)
[9:58] Enus Linden: ogp talk?
[9:58] Enus Linden: sl talk?
[9:58] Enus Linden: qa?
[9:58] Enus Linden loves qa... and process... and and and
[9:58] Morgaine Dinova: Enus, talking about QA ...
[9:59] Asterion Coen: ... and bears :√†
[10:00] Tammy Nowotny: like Linden Bears, you mean Asterion?
[10:00] Asterion Coen: do u know some other bears ?:)
[10:01] Morgaine Dinova: One of the things that Jacek mentions in the Imprudence viewer background is that there seem to be huge bottlenecks to LL adding things to the viewer, and she put it down to QA bottleneck. I don't really understand that. QA doesn't take long, as long as the devs are responsible for writing unit tests.
[10:01] Morgaine Dinova: Any comments on that?
[10:01] Enus Linden: asterion: bear is available on this parcel should you care to search (take a copy :))
[10:01] Asterion Coen: (i should bet it)
[10:02] Asterion Coen: (was sure the cat was to hide it)
[10:02] Enus Linden: morgaine: i have lots of visibility here, having recently taken on a roll in qa where i keep things moving
[10:02] Enus Linden: so
[10:02] Morgaine Dinova: Coolness
[10:02] Enus Linden: qa is not a gate in general
[10:03] Enus Linden: justification? well,rarely are we so busy that things queue up
[10:03] Enus Linden: granted, this past month we have been hosed
[10:03] Enus Linden: but i have no backlog today :D
[10:03] Skee jetpack (X-Flight): All Go
[10:03] Enus Linden: things DO move slowly sometimes
[10:04] Morgaine Dinova: That's what I thought. I see dev as a natural bottleneck, but not really QA
[10:04] Saijanai Kuhn: Hey Dahlia
[10:04] Dahlia Trimble: Hi :)
[10:04] Tammy Nowotny: I shd think there is also a reluctance to add things to the viewer which might not add value for the users, too.
[10:04] Morgaine Dinova waves at Dahlia
[10:04] Enus Linden: our process is improving, and should speed things up
[10:04] Saijanai Kuhn: There's our resident 3D viewer guru
[10:04] Asterion Coen: Dahlia :)
[10:04] Dahlia Trimble: guru? lol hardly
[10:05] Saijanai Kuhn: I can draw a... cube
[10:05] Asterion Coen: :)
[10:05] Dahlia Trimble: thats a start ;)
[10:05] Enus Linden: one t hing that slows us down is lake of good test plans, solved in part by getting qa involved with projects upstream and helping craft the test plans in advnace of code being ready to test
[10:05] Morgaine Dinova: Sorry Dahlia, if we say guru, it's guru. This isn't a democracy ;-)
[10:05] Enus Linden: hi dahlia. welcome <#
[10:05] Enus Linden: <3
[10:05] Dahlia Trimble: :)
[10:05] Enus Linden: lack not lake
[10:05] Enus Linden: :P
[10:06] Enus Linden: another thing that slows us down is *(^&(& bug counts tbh
[10:06] Tegg Bode: And you're up for a Guru performance review next week too „ã°
[10:06] Morgaine Dinova: Gotta refactor, or bugs become exponential.
[10:06] Enus Linden: solved in part by having specs *gasp* and better testing quality and and and
[10:06] Dahlia Trimble: lol just don't cut my pay
[10:07] Enus Linden: i'm pushing a lot of process lately, we all are i guess. i hope to see things go faster, be higher quality, and be more fun period
[10:07] Saijanai Kuhn: biggest issue with client is the original GUI + event model. Very creaky and gets messier as more stuff was added on top.
[10:07] Morgaine Dinova: Fun is good. Firefighting sucks
[10:07] Enus Linden: yeah, and some of our methods don't let us catch what becomes a firefight down the road
[10:08] Saijanai Kuhn: GUI including the original opengl drawing pane/whateer
[10:08] marek Dembo: Boo!
[10:08] Enus Linden: there are more catastrophic server side bugs than viewer, so i'm helping focus process improvemnts there
[10:08] Saijanai Kuhn: that's scary tbh
[10:08] marek Dembo: Get lost!
[10:08] Saijanai Kuhn: though I've heard it before
[10:09] Tammy Nowotny: I already am lost, Marek
[10:09] Morgaine Dinova: The worst firefights come from non-scalability, because once you're among those flames, nothing you do helps, it just moves the congestion elsewhere.
[10:09] Enus Linden: yep.
[10:09] Saijanai Kuhn: Enus, what can AW G and Gridnauts do to help the testing/QA?
[10:10] Enus Linden: 1.25.2 introduced to production a memory leak from a new c module for llsd parsing
[10:10] Enus Linden: we didnt catch it in testing
[10:10] Saijanai Kuhn: ack. There's a pervasive bug
[10:10] Patnad Babii: i heard this morning from CG its been found
[10:10] Enus Linden: i can repro it now, it's 'fixed' in the branch now
[10:10] Patnad Babii: good :)
[10:10] Enus Linden: and we'll be using the test method in divertse ways on every server build in the future, automated, click and evaluate a report
[10:11] Saijanai Kuhn: great
[10:11] Enus Linden: in this regard: help on the beta grid is useful for server side testing
[10:11] Morgaine Dinova: Neat
[10:11] Enus Linden: thinking about the best ways to test functionality and performance there is helpful
[10:12] Enus Linden: vektor is currently engaged on the beta grid, and he or I are good resources there
[10:12] Tammy Nowotny: how many sims oare on the current vewrsion of the Beta grid?
[10:12] Enus Linden: lemme go count, 1 sec
[10:13] Enus Linden: 127 regions, maybe 40 hosts or so?
[10:13] Tammy Nowotny: Thanks
[10:13] Tammy Nowotny: more than in the past, I think
[10:14] Enus Linden: i'm working on a libomv based ec2 distributed agent load test framework (pyogp will hopefully replace)
[10:14] Morgaine Dinova: So a real test environment, not a toy
[10:14] Enus Linden: basically, ~4000 agents sent through one util host = average concurrency
[10:14] Enus Linden: start there, and expand the numbers out
[10:15] Dahlia Trimble: "util host"?
[10:15] Enus Linden: and then fire up webservice based load tests targeted at specific functionality to stress things
[10:15] Enus Linden: util runs central backbones
[10:16] Dahlia Trimble nods
[10:16] Enus Linden: there are 16, evenly ditributed agent ids run through speific hosts
[10:16] Enus Linden: so agent presence, now inventory, group chat etc all run through a one of those 16
[10:16] Tammy Nowotny: so that means each one is handling about 4000-5000 avis on the live grid?
[10:17] Enus Linden: yeah
[10:17] Saijanai Kuhn: that's different than the agent IDs we see in the pcakets
[10:17] Saijanai Kuhn: ?
[10:17] Enus Linden: nope sai: same ids
[10:17] Saijanai Kuhn: ah, so each handles 1/16 the IDs?
[10:17] Tammy Nowotny: is there any way to tell which one we are on at a given moment?
[10:17] Enus Linden: correct
[10:17] Enus Linden: tammy: nope. the sim proxies it all
[10:17] Tammy Nowotny: and does that effect why some avis rezz faster than otehrs?
[10:18] Tammy Nowotny: interesting
[10:18] Enus Linden: it *could* tammy, but that is dependent on many things
[10:18] Tammy Nowotny nods
[10:20] Enus Linden: what else might we be curious about?
[10:20] Morgaine Dinova digs out the curious diary
[10:21] Morgaine Dinova: IM doesn't fall under curious, it falls under distressing
[10:21] Enus Linden: :)
[10:21] Tammy Nowotny: that util host thing was interesting. I did not know that before now.
[10:21] Saijanai Kuhn: Just wondering if we can get something into pyogp as an official baby AD. Bel's got ambitoius plans. Zha's contemplating a C# version for OpenSim. I THINK I can do a python one, but the innards, like tacking avatars and so on, are a bit more difficult than handling a few http POSTs
[10:22] Saijanai Kuhn: tracking* avatars
[10:22] Enus Linden: tammy: take a look at the architecture model over at cg's. it is very similar to what's real
[10:22] Enus Linden: sai: what's the sue case for that?
[10:23] Morgaine Dinova: That would be good. If we can get trial ADs authorizing access to free resources, there should be fast progress. Everyone wants to see object interop.
[10:23] Tammy Nowotny: Thanks... oh and what is "cg's"
[10:23] Dahlia Trimble: "sue" case? lol
[10:23] Enus Linden: CG Linden
[10:23] Tammy Nowotny: Oh OK. Thanks :-)
[10:23] Enus Linden: southeast corner of this region = his parcel
[10:23] Morgaine Dinova: "Sue" is the american version of "use". ;-)
[10:23] Enus Linden: and the model is tehre
[10:24] Enus Linden: well, Tao had a working AD in pyogp
[10:25] Saijanai Kuhn: eyah. Use case is for extending the AD in various ways: inventory, gorup IM between ADs, stuff we haven't thought of yet
[10:25] Enus Linden: we can ressurect that fairly easieven wihtout ZCA :Ply
[10:25] Enus Linden: o
[10:25] Enus Linden: m
[10:25] You decline ! Science Friday (205, 249, 22) from A group member named TJ Asp.
[10:25] Enus Linden: g
[10:25] Enus Linden: i type for crap
[10:25] Enus Linden: pyogp can produce an AD fairly easily, even without zca
[10:26] Saijanai Kuhn: yeah. The issue is the db handling. Though, I guess for a handful of avatarrs., we can fake a DB
[10:26] Enus Linden: but development would fork from LL should things wander off into new territory
[10:26] Dahlia Trimble: I have a curiosity, not sure if it's related, but are there any references to how "planar" texture mapping works available?
[10:26] Saijanai Kuhn: right. Was thinking of protocol testing of "custom" ADs
[10:27] Saijanai Kuhn: ADs will eventually offer custom services. To do it right, we need to figure out how to let other ADs and clients know what those services are
[10:27] Enus Linden: dahlia, that is so over my head that i cant even answer
[10:27] Dahlia Trimble: k ty :)
[10:27] Enus Linden: but i can help find out offline :)
[10:27] Dahlia Trimble: :D
[10:28] Enus Linden: sai: so a working AD that catches up to changes is totally feasible. we would need to identify what the base libs need to support further AD work as well
[10:28] Saijanai Kuhn: K
[10:28] Enus Linden: for example, kill threads in the current mock implementations of a client and replace with a real sustainable solution
[10:28] Enus Linden: that would enable proper caps handling
[10:29] Enus Linden: which then supports AD functioanlity prototyping
[10:29] Saijanai Kuhn: I had fun with LSL. Showed, sorta that you could do a nano AD with the new http calls. Won't work with Lnden aditi due to the secondlife.com issue though
[10:29] Saijanai Kuhn: enus, right. Though, the library appears to be able to handle things other than threads
[10:30] Saijanai Kuhn: so it shouldn't be too hard to get it towork with eventlet or whatever
[10:30] Enus Linden: agreed
[10:31] Saijanai Kuhn: my ibbest problem is figuring out how to make it work with new parts, like a wx threading model. I know twisted has a thing that works with wx, but that's way beyond me right now
[10:31] Saijanai Kuhn: my biggest* problem
[10:31] Enus Linden: just need the framework in which multiple processes can run concurrently
[10:32] Tammy Nowotny: /(when I saw the acronym "wx" I first thoiught Sai meant "weather" which is something which SL has had trouble fully implementing)
[10:32] Saijanai Kuhn: darn whiteout textures. Really got to open a jira on this. Everybody is white.
[10:32] Tammy Nowotny: speaking of weather, nice sunset :-)
[10:32] Enus Linden: i just read that there was once rain!
[10:33] Morgaine Dinova: Just use eventlet to make a noddy I/O server, and keep your concurrency outside in additional scripts.
[10:33] Saijanai Kuhn: stood up and it went away.
[10:33] Tammy Nowotny: there is... they cdn't solve the problem of figuring out of a given point is under a roof or not, so they deprecated it
[10:33] Dahlia Trimble: my neighbor's particle rain would always rain in my skybox :(
[10:33] Tammy Nowotny: I think estate owners can still make it rain
[10:34] Enus Linden: that should be fixed now then tammy, since shadows are underway and such
[10:34] Morgaine Dinova: Well wind still blows through walls, doesn't it?
[10:34] Tammy Nowotny: and also, water is not like RL water, so that wd be an iussue if we had rain here too
[10:34] Enus Linden: true
[10:34] Saijanai Kuhn: seems you'd have to have a new object parameter "roof" to make sure you could make that happen.
[10:34] Tammy Nowotny: that shd be fun. Enus
[10:34] Enus Linden: would be lovely to see a calm pond being hit with rain drops
[10:34] Morgaine Dinova: Yeah
[10:35] Tammy Nowotny: I used to be a climate researcher in a previous first life
[10:35] Enus Linden: jealous i am
[10:35] Morgaine Dinova: Oh, on a similar vien, what happened to Windlight's water trails? I used to love swimming with the wake behind me.
[10:35] Enus Linden: i alays wanted into NOAA
[10:35] Tammy Nowotny: modelling the effects of rainfall is a massively complicated problem
[10:35] Asterion Coen: :)
[10:35] Saijanai Kuhn never had a comp good enough to see those I think
[10:35] Enus Linden: morgaine: i do not know!
[10:36] Asterion Coen: at least i built the okeanos explorer for noaa on sl :) im sure hackshaven will be happy to rezz it next to your island, hehe
[10:36] Morgaine Dinova: Well it wasn't crucial, but sure was pretty :-)
[10:36] Dahlia Trimble: I did some CFD work years ago
[10:36] Enus Linden: nice asterion! should show it to me sometime
[10:36] Saijanai Kuhn: Dahlia, BTW< Qarl's office hours are at 11 AM. He'd probably be able to answer a lot of graphics questions for you
[10:37] Dahlia Trimble: oh kk ty Saij :)
[10:37] Enus Linden: nice sai, saves me a search for the right person :)
[10:38] Morgaine Dinova: Well I read a lot of NOAA papers, at the time that I was running GCM models to figure out what all the global warming stuff was about.
[10:38] Morgaine Dinova: Interesting area.
[10:38] Dahlia Trimble: heh I have a question for Qarl about how he closes the ends of sphere sculpties
[10:39] Morgaine Dinova: Dahlia: a singularity ;P
[10:40] Enus Linden: XD
[10:41] Saijanai Kuhn: son is taking a look at sculties now. He doesn't do SL, but wants to write a script to algorithmically make them without requiring a 3D interface
[10:42] Tammy Nowotny: does anything aside from SL & its derivatives use sculpties?
[10:42] Dahlia Trimble: not that I've seen
[10:42] Saijanai Kuhn: just about all pro 3D packages have their own version. Qarl based his directly on Maya's
[10:42] Enus Linden: (alrighty, time is kinda up. i'm gonna stick around for a spell to listen, but have not much to push now. will wander off in 15 i tihnk)
[10:42] Tammy Nowotny: TY Enus
[10:43] Saijanai Kuhn: they're called various things
[10:43] Dahlia Trimble: ty Enus :)
[10:43] Saijanai Kuhn: thanks enus
[10:43] Morgaine Dinova: Cheers Enus, have fun
[10:44] Dahlia Trimble poofs for a rl break.. bye all :)
[10:44] Enus Linden: see ya dahlia!
[10:44] Tammy Nowotny: cyas Dahlia
[10:44] Saijanai Kuhn: They're just a simple way f providing an easily mapped UV thing. something like a bump map but for UVs
[10:44] Morgaine Dinova: OK, so, in practical terms, what's happening with the AD thing? Waiting for Bel and co to provide?
[10:44] Tammy Nowotny nods
[10:44] Tammy Nowotny: Thanks Sai
[10:45] Enus Linden: morgaine: we can move the pyogp implementation along independently
[10:45] Zha Ewry looks up an listens
[10:45] Enus Linden: you missed the earlier bits i think zha :)
[10:45] Enus Linden: to recap: Tao had a working pyogp ad package
[10:45] Enus Linden: which i broke in removing zca
[10:46] Saijanai Kuhn: I passed her your 7 point list
[10:46] Morgaine Dinova: Is there some kind of spec for the AD on the wiki?
[10:46] Enus Linden: for the base libs
[10:46] Enus Linden: ye smorgaine
[10:46] Morgaine Dinova: Coolness, looking
[10:46] Enus Linden: Open_Grid_Protocol
[10:46] Morgaine Dinova: Heh
[10:47] Saijanai Kuhn: is there? I've seen the TP/Login sequence diagras, but OGP itself seems to be mostly about client <=> AD with only a hint about the AD <=> sim/grid
[10:47] Enus Linden: so zha: i'd be happy to bring the AD back inline in pyogp , but it requires some lib work to enable a pyogp based client to be a test framework still
[10:47] Enus Linden: sai: all the rez_avatar/place, request, rez, derev serves as a spec
[10:48] Saijanai Kuhn: right. http://wiki.secondlife.com/wiki/OGP_Explained#Login
[10:49] Saijanai Kuhn: that much I could do its teh error tracking and whatnot and the tracking of the avatars in a AD DB that gets fuzzy for me
[10:49] Patnad Babii: bye everyone im out and have a nice weekend :)
[10:49] Enus Linden: that is all server side implementation detail, not really AD specific per se
[10:49] Enus Linden: laters patnad
[10:49] Morgaine Dinova: Cya Patnad
[10:50] Asterion Coen: have fun pat :)
[10:50] Tammy Nowotny: I have to get going. have fun everyone
[10:50] Enus Linden: laters tammy
[10:50] Asterion Coen: have fun tam :)
[10:50] Morgaine Dinova: Cya Tammy
[10:50] Tammy Nowotny: :D
[10:50] Tammy Nowotny: *poofs*
[10:50] Enus Linden: zha: how activelty are you thinnking of pursueing (sp) a C# AD?
[10:52] Saijanai Kuhn pokes the IBM princess
[10:53] Morgaine Dinova hands Sai a feather
[10:53] Saijanai Kuhn: seems like I'm not the only personzzzzzzzing
[10:54] You decline Science Friday Members, SCIENCE FRIDAY (219, 211, 28) from A group member named Bjorlyn Loon.
[10:54] Saijanai Kuhn: ah, well, I post this in your office hours ENus. HEading to Q's to ask a question or two for my son
[10:55] Saijanai Kuhn reminds self to reset Tree's flash thingie. hand-regex is a bitch
[10:56] Morgaine Dinova: Is Qarl's office in this zone?
[10:56] Saijanai Kuhn: laters all heading to Q
[10:56] Enus Linden: ty sai!
[10:56] Morgaine Dinova accepted your inventory offer.
[10:56] Morgaine Dinova: Ta Sai
[10:56] Enus Linden: thanks all for coming
[10:56] Enus Linden: enjoyed it
[10:56] Saijanai Kuhn: thanks Enus talk to you later
[10:56] Morgaine Dinova: Tnx Enus

User:Enus Linden/Office Hours/2008 December 05

Navigation menu

Search