User:Zero Linden/Office Hours/2007 Dec 18
Jump to navigation
Jump to search
Transcript of Zero Linden's office hours:
| Saijanai Kuhn: | was just going to issue a call for all Groupies to wear there tags. Psycholigcal intimidation | |
| Jason Swain: | Hello Zero | |
| Rex Cronon: | especially when u see the body and clothes ok, but the eyes are gray | |
| Saijanai Kuhn: | ho, Hi Zero | |
| Zha Ewry worries that one day, we will see an airplane fly in through the roof on "inbound zero" | ||
| Head Replacement: | Hi zero | |
| Neas Bade: | hey zero | |
| Zero Linden: | Intimidation? | |
| Zero Linden: | Hello all | |
| Jarod Godel: | I worry that Zero devides himself too much | |
| Zha Ewry: | Afternoon, Zero. | |
| Penumbro Eclipse: | hello | |
| Rex Cronon: | hi zero | |
| Jurin Juran: | wb gigs | |
| Zero Linden: | Indeed - I do out of necessity | |
| Gigs Taggart: | thanks | |
| Saijanai Kuhn: | intervention time for Linden Labs... | |
| Head Replacement: | Zero, why is it that sometimes you need to wait x minutes before you can login again? My poor alt Tao is waiting in that queue :-) | |
| Penumbro Eclipse: | I guess that means you are undefined! | |
| Bounced Offcourse says nothing.. on that topic.. | ||
| Penumbro Eclipse tries fervently to divide by zero, but fails! | ||
| Zero Linden: | I have people managerial responsibility, I assist the exec. staff on strategy development, I work with non-dev. parts of the company on developing products, I work inside dev. on architecture | |
| Zero Linden: | AND - I work with you! | |
| Zha Ewry grins | ||
| Head Replacement: | thisi actually might be the topic today ;-) | |
| Zha Ewry: | So.. can we get the RC and Windlight clients fixed to support login to OpenSim again :_) | |
| Saijanai Kuhn: | Show and tell time, Zero. Got a python bot almost working | |
| Zero Linden: | There is a fair bit of interlock in the system to ensure that you don't log in while another region thinks you are still in world | |
| Zero Linden: | being in twice would break many many things | |
| sculpt test: | BlŒ bŠr rššda | |
| Zero Linden: | Alas, writing out attachements is a significant burden when you log out | |
| Saijanai Kuhn: | That poor ruthed clone of me is a python script | |
| Head Replacement: | ic | |
| Saijanai Kuhn: | pure python no libsl | |
| Bounced Offcourse: | hehe, then .. when you eventually did get logged out.. it might log your new login out too :P | |
| Zero Linden: | Sometimes, the presence entry gets stuck - | |
| Head Replacement: | now at least I know why it happens, it's still sometimes annoying and Head has to go to all those office hours instead of Tao ;-) | |
| Zero Linden: | consider what happens if the region you are logging out from crashes | |
| Zero Linden: | or worse, hangs for a while | |
| Seifert Surface: | wouldnt it be nice if the timeout was related to how much attachments data there actually is, rather than a default 5 minutes or so... | |
| Head Replacement: | actually it wouldn't happen if the viewer wouldn't crash ;-) | |
| Zha Ewry: | Hangs is *much* worse | |
| Zero Linden: | while the system is waiting to see if the hung process comes back to life (of course we'd rather not kill it) | |
| Saijanai Kuhn: | Zero, right now this bot is getting 1900 packets while the server waits for the last login step | |
| Seifert Surface: | then people would realise perhaps how much data theyre carrying around | |
| Zero Linden: | if you were in the middle of being logged off... well, the system just has to wait before letting you back in | |
| Head Replacement will try to crash earlier before office hours then ;-) | ||
| Zero Linden: | One project that has been on hold for ages would enable detection of when attachement data could be just re-written, or even better yet, not written at all | |
| Stipe Gorbunov: | hay | |
| Zero Linden: | we've talked about a call that script could make that would signal - "I have no crucial state" - which would | |
| Zero Linden: | mean that that script needn't be written out on logout - we'd just drop it | |
| Zero Linden: | when you log in, you'd get the last state | |
| Zero Linden: | again | |
| Zero Linden: | for most things like HUDs and AOs, this is just fine | |
| Zero Linden: | All - welcome to my office hours | |
| Zero Linden: | as always - transcript -> wiki => speak = open | |
| Head Replacement: | I wonder how much we need to know about that when it comes to region domain etc. :-) | |
| Penumbro Eclipse: | This is very, very interesting, but RL beckons. Thanks all! | |
| Head Replacement: | cya Penumbro | |
| Bounced Offcourse: | What's the relationship between estates and regions? :D | |
| Rex Cronon: | bye penumbro | |
| Zero Linden: | Bit of scheduling: Today and Thursday are the last office hours | |
| Zero Linden: | until January | |
| Jarod Godel: | awww | |
| Head Replacement: | I nearly guessed that much ;-) | |
| Bounced Offcourse: | ooh. okay | |
| Zero Linden: | specifically January 8th | |
| Zero Linden: | I'll be out for two weeks | |
| Harleen Gretzky: | An esate i a group of region(s) | |
| Zero Linden: | Actually - I'm not going anywhere: vacation at home - the best kind --- | |
| Jarod Godel: | First Christmas with a kid? | |
| Head Replacement: | I wish I would have 2 weeks at home... | |
| Zero Linden: | if you see my alt in world - feel free to say hello - but I won't be talkin' shop! | |
| Saijanai Kuhn: | spring cleaning in the mdidle of winter | |
| Zha Ewry nods | ||
| Saijanai Kuhn: | Go Go Igo! | |
| Zha Ewry: | Grab your alt and play some Go | |
| Zero Linden: | Jarod - no - my son is 6 and we had a foster kid before him - so experienced in "it is all unwrapped by 5am" | |
| Zha Ewry grins | ||
| Bounced Offcourse: | hehe, isn't it that way everywhere? | |
| Zero Linden: | So now, estates? regions? was there a question there? | |
| Bounced Offcourse: | Well.. how would you describe an estate | |
| Zha Ewry: | You shoiuld have had a few years of "You can't come out of the bedroom until *THIS* number is bigger than 6" | |
| Head Replacement: | actually I think some of us would like to talk about collaboration and documentation | |
| Head Replacement: | and how AWG goes on | |
| Saijanai Kuhn picks his head up from restling with UDP packets | ||
| Bounced Offcourse leaves it up to the group to decide :D | ||
| Zha Ewry: | And. Yes. Collaboration, and.. in particular, documentation and stable interface points, or the lack of the same | |
| Head Replacement: | and what the process is or could be | |
| Saijanai Kuhn: | or is supposed to be but isn't | |
| Head Replacement: | main problem is I think that we all are trying to figure out a) how things work today and b) how they migth then work tomorrow. Example these days is login | |
| Saijanai Kuhn: | ...and why the sea is boiling hot, and whether pigs have wings. | |
| Saijanai Kuhn: | and estalbishing presence | |
| Zero Linden: | well, in SL pigs DO have wings, and the sea is boiling hot because it is very compute intensive to render | |
| Saijanai Kuhn: | pointing to bot that gets 1900 packets while the server waits for last message to establish presence | |
| Rex Cronon: | forget the pigs, login more interesting | |
| Zha Ewry: | And in particular, wha's the intended behairo of the region simulators on login, what actually happens, and what's the path forward to even bgin to talk coherently | |
| Head Replacement: | and at least I personally feel that being able to ask questions 2 times a week is not really sufficient to go along. At least I don't see that much progress being made that way as in fact most things seems mostly known to Lindens | |
| Zha Ewry: | about how to break out the agent part of the region code | |
| Jarod Godel: | Zero, how many Brazillians must your ocean kill? | |
| Zero Linden: | Well- | |
| Al Sonic thinks Jaron answered his own question. | ||
| Al Sonic: | (typo on Jarod... not that it matters...) | |
| Zero Linden: | So I'll be brutally honest and say that indeed I agree with the assesment of how well things have gone to date | |
| Jarod Godel: | Al, I only slaughter words, not Brazilians. ;) | |
| Zero Linden: | I'm not set on any particular process or procedure - I see this sort of open community design to be a rather novel and new approach without many antecedants for me to point at | |
| Saijanai Kuhn: | Zero we want to help, but things can get very confusing... | |
| Zero Linden: | So, here are my thoughts - please take them as thoughts, not edicts, and recognize that now I'm speaking as one Linden, not all of Linden Lab | |
| Zero Linden: | One possible direction - | |
| Head Replacement: | It's also not to blame anybody of course it's just that it's hard to move somewhere if not totally doing some stuff out of the blue. | |
| Zero Linden: | Assuming I can muster several engineering resources internally to do this, we could have say two Lindens | |
| Zero Linden: | authoring proposed protocols, accepting input from you all, and modifying them as needed | |
| Zero Linden: | while, say, both a small group of Lindens implements this on a dev grid | |
| Saijanai Kuhn: | Zero, taking a step back. We need help understanding what goes on NOW before we can help design something new | |
| Zero Linden: | while, at the same time, one or more groups here (OpenSim and LibSL, say) | |
| Zero Linden: | implement them externally | |
| Zha Ewry: | That.. becomes.. anything but an open process | |
| Zha Ewry: | At the moment, we're getting a lot of pushback from OpenSim (with good reason) | |
| Rex Cronon: | almost anybody should be able to implement the network protocols if their description is available | |
| Zha Ewry: | about th new authetnication process, because it's gone that way | |
| Zha Ewry: | And.. in fact | |
| Neas Bade: | yeh, it would be reallly good to get some details on newauth out there, as well as the loginuri (or equiv) back in | |
| Zha Ewry: | the way it went was people pushed back hard on some details | |
| Zha Ewry: | and linden, so far, has responded to most of that feedback by | |
| Zha Ewry: | saying "Trust us" | |
| Jarod Godel: | Isn't that SL's motto for developer relations? (not to be glib...) | |
| Zero Linden: | Let's be fair about some things here | |
| Head Replacement: | for a start I'd like to have some Linden on IRC sometimes to be able to ask questions here and then.. like about login | |
| Zero Linden: | Linden Lab is busy trying to do two things | |
| Zero Linden: | 1) Run an existing virtual world | |
| Zero Linden: | 2) Transition to an open metaverse approach | |
| Zero Linden: | No matter how much we are committed to working on that second thing, the first is still there and will have to proceed | |
| Saijanai Kuhn: | sure, and has to be your top priority... | |
| Zero Linden: | There are 100 people or so in engineering, and at this time 99 of them are working on the current system | |
| Saijanai Kuhn: | BUT... | |
| Zha Ewry nods "Nobody is expectign otherwise" | ||
| Zha Ewry: | Going out of business, is a bad plan for the open metaverse. | |
| Neas Bade: | Zero, no arguments there. The new auth issue ends up being quite a bit of a show stopper on the OpenSim front, as all the new clients can't be made to work with OpenSim | |
| Jarod Godel: | Agree with Zha here. | |
| Zero Linden: | AND - that current system is going to change without perfect input, or at times even any input, from the outside | |
| Rob Linden: | hi all (was double booked today) | |
| Saijanai Kuhn: | hey Rob | |
| Rex Cronon: | hi rob | |
| Zero Linden: | Not that we would like to have no input, but it is going to happen | |
| Malburns Writer: | Hi Rob | |
| Zha Ewry: | Afternoon Rob | |
| Saijanai Kuhn take a drink | ||
| Head Replacement: | I think we are more talking about the output in terms of docs at least after the change ;-) | |
| Bounced Offcourse: | Greetings Rob | |
| Zero Linden: | The current system is, as all of you know, VERY squirlley and not really a good candidate for standardization | |
| Head Replacement: | Hi Rob! | |
| Zero Linden: | that LibSL and OpenSim work with it is a MAJOR testiment to those engineers talents | |
| Neas Bade: | So any ability to get OpenSim collaboration gets flushed pretty quickly if the client is no longer common ground | |
| Saijanai Kuhn admires packets flying by his screen. Yeah, squirreling (no offense Ms Wood) | ||
| Zha Ewry nods "Tho, note, in this case, plenty of input was given, and some assurances of its being taken were given" | ||
| Jarod Godel: | Zero, is there a reason you guys even ask for input? Not to be snide (I hope I don't sound that way) but I've nver seen Flickr or anyone ask "How do we fixup our API for you?" Would SL's open-ness go faster without the community nagging the engineers? | |
| Saijanai Kuhn: | Jarod: OPensource, Open Standards | |
| Head Replacement: | I think it's a good sign they ask for input, I'd also like flickr to do this ;-) | |
| Neas Bade: | not looking for perfect standards here, just enough enough to keep OpenSim interoperable. | |
| Zero Linden: | We ask for input because we really do care - and we really do believe that what the community does with our tool is important | |
| Head Replacement: | and in fact flickr of course receives feedback regarding their API | |
| Harleen Gretzky: | Why can't OpenSim support the new auth? | |
| Neas Bade: | I'm speaking about the very specific instance of new auth, not really anything broader at this point | |
| Head Replacement: | and in fact people work together with flickr to e.g. do things with them like OAuth | |
| Head Replacement: | but back to SL :) | |
| Saijanai Kuhn: | Neas, di you see the script Tao worked up? | |
| Rex Cronon: | aren't there lots of programmers in the community, jarod? | |
| Bounced Offcourse: | It can.. however, They're asking for more documentation on it. | |
| Neas Bade: | Harleen, 1) it isn't documented. 2) the Client has secondlife.com hard coded for login | |
| Bounced Offcourse: | .. to be able to implement it. | |
| Jarod Godel: | Rex, in the SL community or Flickr? | |
| Head Replacement: | I think the main problem is not my script here. libsl people know how to login | |
| Head Replacement: | the problem is simply that clients cannot be directed to some different login page if I understand this right | |
| Zero Linden: | I belive there is infact a command line option for that | |
| Zha Ewry the other problem is one of sheer communication | ||
| Rex Cronon: | sl | |
| Zha Ewry: | There is in fact not, one yet | |
| Zero Linden: | if there isn't i know that internal group has planned for it | |
| Zha Ewry: | And. there have been repeated discussions on this | |
| Rob Linden: | we're all going to get better at collaborating over time. if you take these first awkward steps as a sign of how it's always going to be, then you'll get pretty depressed :) | |
| Saijanai Kuhn: | ||
| in case anyone wonders how the new login works | ||
| Neas Bade: | yeh, we could get around 1 as there has been some work on libsl. | |
| Head Replacement: | I think the opensim thing is one issue, the documentation issue is another one | |
| Neas Bade: | zero: it doesn't work in either of the RCs for sure | |
| Jarod Godel: | Rex, sure! I'm just throwing out a question. I love SL+community, but I was curious if it was the best option. Zero seems to say it is, so... | |
| Head Replacement: | like we might need to understand what actually happens between client and server today in order to not miss any important things | |
| Head Replacement: | so it's either docs we need or somebody to ask and we can write it down | |
| Head Replacement: | (a faster wiki would help though ;-) ) | |
| Saijanai Kuhn: | bu t reverse engineering is fun | |
| Saijanai Kuhn: | !!! | |
| Head Replacement: | yes, if you have lots of time | |
| Zha Ewry looks at Saij funny | ||
| Zha Ewry: | We all really do want to make this work | |
| Jarod Godel: | and a razor blade and the original developer | |
| Neas Bade: | so until there is a client drop that supports something equivalent to loginuri, there is some definite frustration, as there is no use in working out the bits of new auth and integrating them | |
| Rob Linden: | this may seem like a tangent, but it's not entirely: | |
| Zero Linden: | Okay - I'm going to go back to the team and talk with them about communication. I will try to get them to engague more | |
| Rex Cronon: | i wasn't comming down on u jarod, i just want to make sure what u mean:) | |
| Head Replacement: | like I said, I would like IRC best as I am not constantly in SL, I am not sure about Lindens or others here | |
| Zha Ewry: | Chuckle. | |
| Zha Ewry: | Doesn't look like a tangent to me | |
| Zero Linden: | Well - at this time LL cannot afford an engineer to sit in IRC all day | |
| Head Replacement: | some sort of sprint would of course also be nice | |
| Jarod Godel: | Rex, that's cool. I've found myself taking the opposite tack than i usually do a couple of times today...so, no worries. | |
| Zero Linden: | It is an excellent idea | |
| Head Replacement: | well, does not need to be all day | |
| Zero Linden: | And I think we should have support engineers of that sort | |
| Head Replacement: | but maybe more than 2 hours a week which are usually also filled with other stuff | |
| Jarod Godel: | Zero, what about Twitter? | |
| Zero Linden: | Well, Head - I'm pretty sure it would be a full time job - even if it is only 1/2 time in IRC - the other 1/2 time is going to be being sure they are up on the technology | |
| Head Replacement: | not sure twitter really is suited for that.. | |
| Jarod Godel: | at least have engineer chat logs post there? | |
| Head Replacement: | but once we are up with technology we can help :) | |
| Saijanai Kuhn: | and it would be a good way to provide feedback when we try to apply existing docs (such as the y are) and they don't work, or do something flakey | |
| Rob Linden: | a number of us are on irc.efnet.org #opensl quite a bit | |
| Head Replacement: | or some other channel, does not need to be IRC | |
| Head Replacement: | like we maybe could use the mailing list ;-) | |
| Zero Linden: | I think, no matter what the channel, the issue is that you are asking for a resource to be your technical liason | |
| Zero Linden: | it is a completely reasonable request | |
| Head Replacement: | and then I also wonder how we go from here.. I have my basic login sitting here but I am not sure what to do with it now. How do we write this down would be one question | |
| Zero Linden: | In fact, I think I'm going to present this as something LL MUST do in the coming months | |
| Saijanai Kuhn: | and how to test it, etc | |
| Zha Ewry: | Well, its the intersection etween being able to read code, which often | |
| Neas Bade: | Rob, fwiw, we moved the #opensim* channels to freenode over the weekend, as we found a lot of host blocking on efnet | |
| Zha Ewry: | reads as "We hacked on the client until it worked" | |
| Neas Bade: | wasn't sure if you found us again yet | |
| Daaneth Kivioq: | more communications is a good thing - especially any method for use to communicate to you lindens about pressing issues | |
| Zha Ewry: | and being able to konw what is intended | |
| Saijanai Kuhn: | I've got my login almost working and I noticed a strange pathology when teh avatar isn't responding as the server obviously expects. Who do I tell about that? | |
| Head Replacement: | so I might try #opensl and #libsl more again maybe | |
| Gigs Taggart: | kelly's on opensl a lot but he stays pretty busy | |
| Head Replacement: | but of course at some point somebody who knows what the AWG needs might be good then | |
| Wyn Galbraith: | Does anyone know how to set up IRC in Trillian? | |
| Head Replacement: | well, I talked to Lindens in there but I wasn't always sure how busy they are and if I take their precious time away | |
| Jarod Godel: | At the risk of being dense, let's try to remember that getting this engineering liason likely means losing an active engineer. | |
| Jarod Godel: | Is that a trade-off that's going to help? | |
| Head Replacement: | of course it's always a resource problem | |
| Zero Linden: | Yes, it does | |
| Zha Ewry: | Well, the trade off, is getting the OpenSim, and the openSource | |
| Zha Ewry: | community engaged | |
| Rob Linden: | Jarod: precisely....some of these things are zero sum, unfortunately | |
| Head Replacement: | but I think LL is crucial here and if there are no resources for AWG then I don't see much future in an open project | |
| Jarod Godel: | I like docs as much as anyone (and examples!), but I'd rather SL get a protocol that was working and stable first | |
| Saijanai Kuhn: | and getting AWGroupies up to speed on understanding things | |
| Saijanai Kuhn: | cart before the horse, Jarod | |
| Daaneth Kivioq: | and kill fricking voice! | |
| Head Replacement: | Jarod: but then only LL can do that protocol | |
| Head Replacement: | and then we are again at this resource problem as somebody still has to write it down | |
| Saijanai Kuhn: | and if they can't document, is it really a protocol? | |
| Rob Linden: | one thing that doesn't have direct costs is working more transparently, and that's something that we'll be trying to get peopel to do more | |
| Jarod Godel: | if they get it working, then they'll have time to document it | |
| Zero Linden: | Actualy, voice is done primarily by an outside company- it takes very little engineering resources, and only microscopic server resources on our side | |
| Head Replacement: | indeed, if e.g. more discussions would be done more openly it would maybe also help | |
| Saijanai Kuhn: | no, protocols have to be documented first or its never a protocol | |
| Neas Bade: | jarod: you obviously have never worked in teh software industry :) | |
| Daaneth Kivioq: | then why does voice seem to cause so many problems? | |
| Jarod Godel: | Neas, only the software side of education | |
| Zha Ewry: | Not clear that it gets you an open one, Jarod, nor does it let OpenSim and LibSL.. and other players actually prove that it is a prortocol, not a set of softwar hwich happens to work otgether | |
| Head Replacement: | well, once it's coded there is little chance of docs ;-) | |
| Neas Bade agrees with Head | ||
| Head Replacement: | and should they do that I think opensim and maybe AWGroupies might go another route in the meanwhile | |
| Zha Ewry: | Right now, SL doesn't really have a protocol, since they control both ends of the pipe, and can (and looking at the code do, tweak them somehwat at random | |
| Head Replacement: | well, maybe ;-) | |
| Jarod Godel: | Oh, wait. You mean code documentation! I thought you meant protocol documentation. | |
| Jarod Godel: | you're talking about /* comments */ | |
| Zha Ewry: | They are, one and th4e same Jarod | |
| Neas Bade: | protocol documentation as well. | |
| Zha Ewry: | No, we're not talking about comments | |
| Saijanai Kuhn: | code can change but the protocol has to stay the same | |
| Zha Ewry: | A protocol, acutally needs a spec | |
| Saijanai Kuhn: | or be documented as changing | |
| Head Replacement: | well, at least for now it would help I think to understand login more | |
| Daaneth Kivioq: | or the new protocol has to be backwards compatible | |
| Zha Ewry: | One of the problems with "Read the code" for protocols | |
| Zha Ewry: | is that it doesn't dfefine the prototocl, i defines one possible prtocol, at the moemnt the code was written | |
| Zha Ewry: | And. in fact, worse, "As the engineeers managed to make it work" | |
| Head Replacement: | but in the end we of course need some closer collaboration because right now it does not fell that much like it. I am not sure though how this could look like | |
| Rex Cronon: | the protocol can have an optional extension that makes it backwards comptible | |
| Jarod Godel: | I guess I was thinking of it backwards. | |
| Saijanai Kuhn: | an example is that the seedcap (from what I understand from John Hurliman) has many caps available through it but in every case there is a UDP packet that does the same thing except TP | |
| Head Replacement: | I also wonder how the state of AWG at Linden Lab internally is | |
| Head Replacement: | is there work being done? discussions? planning? | |
| Saijanai Kuhn: | I can find the docs for teh UDP on the wiki though theyre not always accurate. There's NOTHING I can find on teh CAPS version. I didn't even know there were two versions | |
| Zha Ewry: | Well, Which and Co. are banging awayon c=http, and esroew | |
| Head Replacement: | I guess this UDP thing is historically because in the beginning there only was UDP ;-) | |
| Daaneth Kivioq: | the only caveat with extending a protocol, is that it can become cumbersome, and at some point you end up wanting to cut out the old stuff and force a migration | |
| Zha Ewry: | And.. if the rest of LL could doucment and be as transparent as which is.. there wouuld be a lot less complinaing | |
| Saijanai Kuhn: | thats what CAPS and the AWG are all about, Daaneth | |
| Head Replacement: | actually what Which etc. do seems to be quite open to me but I am not that much into the topic at the moment | |
| Head Replacement: | but status updates etc. are good to have and the commit-list is also helpful | |
| Head Replacement: | but I also understand that this is maybe not working for the LL server code | |
| Head Replacement: | but as said, I think a sprint would be great.. have a 1-2 day meeting, work on some code or ideas together and after that people will better understand what can be done | |
| Jarod Godel: | in-world flash conference? | |
| Head Replacement: | so come all to the SnowSprint in January ;-) | |
| Head Replacement: | well, not sure how to do it practically, i mostly was involved in RL sprints up to now | |
| Head Replacement: | (which have the additional socializing aspect) | |
| CSI:NY Toolbar: | Script run-time error | |
| CSI:NY Toolbar: | Stack-Heap Collision | |
| Neas Bade: | well, instead of trying to solve the grander issues here, I would try to focus on a specific instance like new auth. Making the loginuri a priority on the client so that the community can start playing with new auth implementations, and an hour of someones time to sketch out what is there on a wiki would be really helpful | |
| Daaneth Kivioq: | (LOL) | |
| Jason Swain: | lol | |
| Head Replacement: | yep, I also don't think that explaining login can be that much work | |
| Zha Ewry: | That.. and.. a discussion, on what is supposed to happen, one UDP packet into login, would be excellent firsts teps | |
| Head Replacement: | and what's important in that login process etc. | |
| Rex Cronon: | whoever has the csi thing, u need to get a new copy, your is broken, due to blitz(most likely) | |
| Zero Linden: | Zha - do you mean what currently happens? | |
| Saijanai Kuhn: | yeah, and what is supposed to happen if a mistake is made | |
| Saijanai Kuhn: | I'm not joking, Zero, about the client getting 1900 packets | |
| Zero Linden: | Okay - Concerns heard by me loud and clear. I agree with vast majority of what you are all asking for. | |
| Head Replacement: | a good idea for AWG is maybe to write down a usecase with all the exceptions | |
| Head Replacement: | but for this it also might be useful to know what's happening today | |
| Daaneth Kivioq: | Ideally, it should be a black box in the end | |
| Zero Linden: | I will take this to the larger organization immediatly. I will lobby for more.... | |
| Head Replacement: | great | |
| Wyn Galbraith: | Thank you Zero. | |
| Rex Cronon: | yes. black boxes are nice | |
| Neas Bade: | awesome, thanks zero | |
| Zero Linden: | ...but I also caution you --- it isn't likely that most of LL engineering is going to suddenly start documenting | |
| Zero Linden: | like Which does | |
| Zha Ewry: | Right now,what seems to happen aftert the first packet | |
| Zha Ewry: | in login 9the first UDP packet up) | |
| Zha Ewry: | is a flood of packets down | |
| Zha Ewry: | Some of which clearly need to be acked | |
| Head Replacement: | tell them if they don't document then they need to spend their time in IRC and answer questions all day long ;) | |
| Zha Ewry: | But.. it's almost impossible to determine what's needed first, and | |
| Zha Ewry: | The trace logs are different than they were a few months back | |
| Zha Ewry: | But.. why.. isn't at all clear | |
| Daaneth Kivioq: | unless the packet gets shot down by Ack-Ack | |
| Zha Ewry: | One expects that there is some off | |
| Saijanai Kuhn: | Message # 1968 Flags: 0x40 ('LayerData', 'Trusted', 'Unencoded') sequence # (1303,) Message # 1969 Flags: 0x00 ('CoarseLocationUpdate', 'Trusted', 'Unencoded') sequence # (1304,) Message # 1970 Flags: 0x60 ('EnableSimulator', 'Trusted', 'Unencoded') sequence # (918,) | |
| Zha Ewry: | *odd | |
| Head Replacement: | maybe we can also come up with some process where we can discuss the details in chunks like small milestones | |
| Zero Linden: | I'm going to admit that the current set of protocols do not have clear documentation even internally | |
| Saijanai Kuhn: | that bot I had here when you arrived got those back | |
| Zha Ewry: | combinatoin of "what should happen" and "wha does happen" going on | |
| Daaneth Kivioq: | That way lies madness, Zero | |
| Head Replacement: | so we might have some initial meeting with Linden involvement to get details clear about one component | |
| Saijanai Kuhn: | 1970 of them | |
| Neas Bade: | zero, honestly, when we talk about documentation it's more like the kind of sketch that will save us a day from figuring it out ourselves. It doesn't have to be super robust | |
| Head Replacement: | and get clear what the next goal is.. for now the goal for Q4 was login via legacy-login caps if I remember that right | |
| Saijanai Kuhn: | obviously I goofed but how, and why the spam? | |
| Rob Linden: | speaking of madness, I've booked for meetings all day today. next..... | |
| Zero Linden: | So I have only limited hope for the existing protocols - my hope is to muster enough troops, internall and external, to get to new protocols that are well defined | |
| Head Replacement: | cya Rob | |
| Zero Linden: | Alas, I have a two oclock too | |
| Zha Ewry nods at Zero | ||
| Zero Linden: | So I'm out of here, sigh | |
| Daaneth Kivioq: | we could always port the code to FORTRAN (:-) | |
| Zha Ewry: | We're eager to help | |
| Neas Bade: | thanks zero | |
| Rex Cronon: | that looks like a big blob broken in parts, sai | |
| Zha Ewry: | But.. We do need to | |
| Jason Swain: | Thank you Zero | |
| Zha Ewry: | Be able to | |
| Jarod Godel: | later, Rob | |
| Zha Ewry: | get the door open | |
| Zha Ewry: | Thanks Zero | |
| Gigs Taggart: | me too, later | |
| Zha Ewry: | And Rob | |
| Head Replacement: | thanks Zero :-) | |
| Zero Linden: | I promise I'm going to bring this feedback to all of engineering - I'm going to be your voice in getting more involvement | |
| Rex Cronon: | bye rob | |
| Malburns Writer: | Bye Rob | |
| Malburns Writer: | Thanks Zero | |
| Saijanai Kuhn: | thanks Rob. | |
| Saijanai Kuhn: | Thanks Zero | |
| Zero Linden: | See you Thursday | |
| Daaneth Kivioq: | Thanks ero | |
| Daaneth Kivioq: | Zero | |
| Rex Cronon: | bye zero | |