Chat log from the meeting on 2016-02-16

From OpenSimulator

Jump to: navigation, search
[11:00] Nebs Metal Bar Stool v1.5 (w/sit & launch): Hello Nebadon Izumi, enjoy your sit..
[11:04] Richardus.Raymaker andrew ? if you ever have a demo train rolling or some other one. like to see it
[11:05] Andrew Hellershanks: Richardus, I have gotten it rolling. It just keep stopping after a while.
[11:06] Nebadon Izumi: I was fixing the sit height i was floating :)
[11:06] Richardus.Raymaker Not sure what system you use ? everything with llPos i describe as bad this days
[11:06] Andrew Hellershanks: nebadon, Someone broke the sit targets again?  :)
[11:06] Kayaker Magic: Nebanon, I used to have 6 regions in OSGrid, but they have not been up for a while, since before the Great Crash. Can you check and see if they were deleted in that cleanup phase?
[11:06] Nebadon Izumi: these chairs were probably like this for some time
[11:06] Kayaker Magic: named Panthalassa0 through Panthalassa5
[11:06] Nebadon Izumi: Kayaker if you are the owner just log into the website
[11:06] Richardus.Raymaker But llpos is the most easy way. but very update intensive
[11:06] Andrew Hellershanks: yea, that's what it uses. Its a little jerky but not too bad. At least it does move.
[11:07] Nebadon Izumi: and look at your profile
[11:07] Nebadon Izumi: you can erase them if needed
[11:07] Kayaker Magic: logging...
[11:07] Nebadon Izumi: you can also search on the map
[11:07] Nebadon Izumi:
[11:07] Andrew Hellershanks: You can tell someone is a programmer if they start numbering regions with a 0. :)
[11:07] Richardus.Raymaker Andrew, then you hit the LSL timeout
[11:07] Nebadon Izumi: :)
[11:08] Kayaker Magic: LOL
[11:08] Nebadon Izumi: or a spreadsheet enthusiast
[11:08] Andrew Hellershanks: Richardus, what LSL timeout?
[11:08] Robert Adams: I wish I had the time to redesign simulation physics. solutions like keyframe movement are more useful than generalized physics
[11:09] Andrew Hellershanks: nebadon, Spreadsheets start row number on 1
[11:09] Nebadon Izumi: well i meant for custom columns
[11:09] Nebadon Izumi: if you want to do sorting
[11:09] Nebadon Izumi: it helps
[11:09] Andrew Hellershanks: ok
[11:09] Robert Adams: not to say that the current keyframe solution isvideal
[11:09] Nebadon Izumi: that is one thing I have not ever used is keyframes
[11:09] Nebadon Izumi: i would not even know where to start
[11:09] Andrew Hellershanks: Robert, what sort of redesign would you like to do if you had the time?
[11:09] Robert Adams: just that creators need specific actions and not general physics
[11:10] Richardus.Raymaker i use it much for movements. when i use it
[11:10] Andrew Hellershanks flüstert: I haven't used keyframing either but I someone gave me a script that uses it so I could have a look.
[11:10] Nebadon Izumi: the number one thing I use physics for, is to drop objects onto terrain so i dont have to spend a lot of time aligning them
[11:10] Richardus.Raymaker only sad that LSL is sometimes MeH in speed and event order , especially in SL
[11:10] Nebadon Izumi: mostly static vehicles lol
[11:10] Andrew Hellershanks: Dang fingeritis turned that last comment in to a whisper.
[11:11] Andrew Hellershanks: Robert, how is BS in handling impulses?
[11:11] Nebadon Izumi: as much physics testing I do thats literally the only thing I actually use it for
[11:11] Kayaker Magic: I have been thinking that if I was king of the Metaverse, I would abandon physics engines and just have a collision engine instead.
[11:12] Richardus.Raymaker and now i got intressted in llCastRay. work ok in SL but i hear bad things with OS
[11:12] Nebadon Izumi: yes I would avoid using llCastRay for anything serious
[11:12] Nebadon Izumi: it will fail you
[11:12] Robert Adams: vehicles need articulation, for instance. Trying to do that with physical constraints is nearly impossible. An articulation tool box would provide the visible action in a more correct fashion
[11:12] Nebadon Izumi: and cause high cpu
[11:12] Nebadon Izumi: the C# implementation is not efficient at all or reliable
[11:12] Kayaker Magic: I have submitted a bunch of Mantisi for llCastRay recently, is anyone looking at those?
[11:13] Nebadon Izumi: hard to say, I doubt anyone can fix the C# versions
[11:13] Nebadon Izumi: it needs to be redone using physics engines
[11:13] Nebadon Izumi: which is highly opimized for raycasting
[11:13] Richardus.Raymaker Andrew, it can be that you hit this limit, with llPos in a for loop it's easy " EventLimit = 30"
[11:14] Nebadon Izumi: I think that is something on the todo list
[11:14] Kayaker Magic: The original version of llCastRay has a problem with var regions, the new version detects phantom prims it should not.
[11:14] Kayaker Magic: There isn't a castray that cn be used seriously....
[11:14] Nebadon Izumi: ya and both versions
[11:14] Nebadon Izumi: if you use more than 4-5 raycasters
[11:14] Nebadon Izumi: they all start failing
[11:15] Nebadon Izumi: even with 1 raycaster that i tested was using on_click event
[11:15] Nebadon Izumi: if you rapidly click
[11:15] Kayaker Magic: Who wrote the new V3 castray?
[11:15] Nebadon Izumi: maybe 3 out of 10 register
[11:15] Robert Adams: sadly, there don't seem to be many core developers with time to do deep development
[11:16] Nebadon Izumi: Magnuz
[11:16] Nebadon Izumi: who probably isnt around anymore
[11:16] Nebadon Izumi: ya unfortunately development seems to have come to screeching halt at the moment
[11:16] Nebadon Izumi: Avination is working on new viewer tech
[11:17] Nebadon Izumi: so they are a bit preocupied trying to get that going
[11:17] Nebadon Izumi: but beyond that no one else is submitting patches or anything
[11:17] Nebadon Izumi: so far there have been zero commits in February
[11:17] Nebadon Izumi: and only a total of 18 or something this year
[11:17] Kayaker Magic: You are scaring me.
[11:17] Nebadon Izumi: and my time is extremely limited at the moment
[11:17] Nebadon Izumi: need to pay rent and eat still
[11:17] Nebadon Izumi: heh
[11:18] Nebadon Izumi: well this happens once or twice a year
[11:18] Nebadon Izumi: we go a few weeks with nothing
[11:18] Nebadon Izumi: not trying to worry anyone, but we do need more developers
[11:19] Nebadon Izumi: what I think we really need though is a new viewer
[11:19] Richardus.Raymaker That sounds a bit scary. it's already running sofar behind with many things
[11:19] Nebadon Izumi: my personal opinion, but the LL viewer is at an end, its being developed even less than opensim is
[11:19] Nebadon Izumi: and honestly its going to be hard for opensim to get any better if we keep insisting on using the LL viewers
[11:19] OtakuMegane Desu: While not the only factor the viewer really is a stumbling block at this point I think.
[11:20] Richardus.Raymaker I know.
[11:20] Richardus.Raymaker and scripting is a few notches behind too. etc etc
[11:20] Nebadon Izumi: so what Avination is doing is really important
[11:20] Robert Adams: was hoping someone like hifi could be leveraged but they seem to be behind also
[11:20] OtakuMegane Desu: It may even discourage a bit of development because it's a pain to get things going viewer-side to match OS changes.
[11:20] Nebadon Izumi: I think it is going to be key to opensim moving forward
[11:20] Nebadon Izumi: ya HiFi is terrible
[11:20] Nebadon Izumi: i tried it again yesterday actually
[11:20] Nebadon Izumi: it hasnt really improved at all
[11:20] Nebadon Izumi: no one is really using it
[11:20] Nebadon Izumi: kind of sad
[11:20] OtakuMegane Desu: Hmm
[11:21] Nebadon Izumi: problem when you hype something before you have anything
[11:21] Richardus.Raymaker Hifi, am right now a bit done with there "oculusgasm" mind
[11:21] Robert Adams: the OS simulator itself is getting long in the tooth also... hobbled by its SL roots
[11:21] Nebadon Izumi: ya it makes it quite difficult for new people to get involved
[11:22] Nebadon Izumi: unless you have deep roots in Sl
[11:22] Nebadon Izumi: its really hard to get a grasp on opensim
[11:22] Nebadon Izumi: makes it very hard to attract serious devleopers and money
[11:22] Richardus.Raymaker Sad to say, but SanSar sounds betetr then hifi
[11:22] OtakuMegane Desu: OS has its own path to be followed. Breaking from SL has been inevitable for some time, just a matter of when we reach that stage.
[11:22] Nebadon Izumi: no one has any clue about SanSar
[11:22] Nebadon Izumi: again its all hype
[11:22] Nebadon Izumi: I have zero faith in SanSar being good at this point
[11:22] Nebadon Izumi: not seen or heard enough to make an informed decision
[11:23] Nebadon Izumi: its 100% speculation
[11:23] Andrew Hellershanks: What is SanSar?
[11:23] Nebadon Izumi: even people very close to SanSar have no idea what it will be
[11:23] Nebadon Izumi: its LL's new Virtual World
[11:23] Nebadon Izumi: coming soon
[11:23] Andrew Hellershanks: oh, its got a name now.
[11:23] Richardus.Raymaker I think it's not much betetr then hifi. but the seems to understand some basic needs that . hifi ignores or throw bakc to the users
[11:23] Andrew Hellershanks: or new name.
[11:23] Nebadon Izumi: i spoke to Max Graf from SL a few weeks ago
[11:24] Richardus.Raymaker Andrew, no it's still project sansar
[11:24] Nebadon Izumi: he is probably one of the top content creators in SL, he is always one of the first to try new SL stuff
[11:24] Simulator Version v0.5 ruft: OpenSim Dev        367b7d7: 2016-01-31 03:08:24 +0100 (Unix/Mono)
[11:24] Nebadon Izumi: even he has no idea at all about SanSar yet
[11:24] Nebadon Izumi: its all very secret at this point
[11:24] Andrew Hellershanks: Richardus, I thought it had a different name when it was first being mentioned.
[11:24] Richardus.Raymaker Neb, if he have. he cannot tell
[11:24] Nebadon Izumi: no he doesnt even know
[11:24] Nebadon Izumi: he would have told me he was under NDA
[11:24] Richardus.Raymaker ok
[11:24] Nebadon Izumi: but no one outiside the lab is even on NDA yet
[11:25] Nebadon Izumi: its completely inside confidential at this point
[11:25] Richardus.Raymaker Opensim would be good. or possible betetr. but neb is right. the viewer is the problem
[11:25] Nebadon Izumi: I think he is on the list to be a tester though
[11:25] Nebadon Izumi: yes we are at a crucial point
[11:25] Nebadon Izumi: where if we don't get our own viewer soon, opensim just cant improve much
[11:25] Nebadon Izumi: it will always be a second rate SL
[11:26] Richardus.Raymaker But other viewer is not helping to improve and speedup lsl :O
[11:26] Nebadon Izumi: I have a lot of hope in what Avination is doing though
[11:26] Nebadon Izumi: but its going to take a good bit of money to get to the finish line
[11:26] Nebadon Izumi: well not even the finish line, more like to even get to the starting blocks
[11:27] Richardus.Raymaker so opensim is in the whacky races...
[11:27] Nebadon Izumi: I am sort of working on that a bit too
[11:27] Nebadon Izumi: unfortunately I can't gaurantee anything at this point
[11:27] OtakuMegane Desu: Avination came through well for the merge and new stuff for OS so if viewer stuff is in the works now I'd probably have the best hope for that over any others right now.
[11:27] Nebadon Izumi: but im optimistic
[11:27] Nebadon Izumi: yes we are far from dead
[11:27] Robert Adams: I keep buying lottery tickets but that hasn't worked yet ;-)
[11:27] Nebadon Izumi: but we have to make some hard decisions I think moving forward
[11:27] Richardus.Raymaker REally not know what to choice. except that hifi is not useable
[11:28] Nebadon Izumi: it doesnt mean abandoning SL stuff
[11:28] Nebadon Izumi: I dont want to give the wrong impression
[11:28] Nebadon Izumi: but at some point things have to change
[11:28] Nebadon Izumi: if we want to improve
[11:28] Nebadon Izumi: new protocols/ new methods
[11:28] Nebadon Izumi: can't fit a square peg in a round hole
[11:28] Richardus.Raymaker you cannot dump lsl. but with a new viewer you could add a betetr language
[11:29] Richardus.Raymaker improve framerate etc
[11:29] Nebadon Izumi: no we really dont want to dump any of it
[11:29] Nebadon Izumi: that is not the goal at all
[11:29] Nebadon Izumi: but right now the SL viewer is holding us back significantly
[11:29] OtakuMegane Desu: I see very little of what is present being removed. A lot is improving the existing functions and adding new things.
[11:29] Nebadon Izumi: it will be impossible to sell opensim as modern software
[11:29] Nebadon Izumi: if we continue to use the completely outdated SL vieweer
[11:30] Nebadon Izumi: its like lets Broadcast HD TV to Black and White TV Tube sets
[11:30] OtakuMegane Desu: Improvements will eventually break compatibility with SL. It's just going to have to happen no matter what viewer path is chosen.
[11:30] Nebadon Izumi: perhaps but that doesnt mean we have to totally abandon that content either
[11:30] Nebadon Izumi: we just need to find new ways to make it work
[11:30] Nebadon Izumi: that is all
[11:30] OtakuMegane Desu: Yeah, I meant that. :)
[11:31] Nebadon Izumi: its like moving from prims to mesh
[11:31] OtakuMegane Desu: But changing things like protocols is a pretty big change, though badly needed.
[11:31] Nebadon Izumi: its painful but ultimately worth it
[11:31] Richardus.Raymaker betetr mesh and mesh upload would be very nice
[11:31] Nebadon Izumi: yea
[11:31] Nebadon Izumi: that is something i am really struggeling with
[11:31] Nebadon Izumi: I work on very very veeeery large models
[11:31] Nebadon Izumi: doing a model right now that is 35 768x768 vars
[11:31] Nebadon Izumi: equal to 315 SL regions
[11:31] Richardus.Raymaker i have the feeling that SL mesh is not really optimized compared to FBX in virtual world
[11:31] OtakuMegane Desu: The ultimate boon I think would be if mesh could be worked into the viewer itself. Even if relatively simple compared to what can be done externally.
[11:32] Nebadon Izumi: working with that many borders
[11:32] Andrew Hellershanks: 35 3x3's? wow.
[11:32] Nebadon Izumi: and being forced to work on corneer areas
[11:32] Nebadon Izumi: its a nightmare
[11:32] OtakuMegane Desu: No doubt lol
[11:32] Nebadon Izumi: and thats just one of the dozen large models im currently working on
[11:33] Nebadon Izumi: and unfortuantely i spend all the time i could be working on and testing opensim
[11:33] Nebadon Izumi: dealing with all the limitations it currently has
[11:33] Nebadon Izumi: sucks
[11:33] Nebadon Izumi: but have to pay the bills
[11:34] Richardus.Raymaker the rare cases i tried mesh in opensim. the viewer up[loader gives me already problems or a hangup (singularity)
[11:34] Nebadon Izumi: but things are looking up I am excited about the new viewer potential
[11:34] Nebadon Izumi: and not only the potential this new viewer can offer
[11:34] Andrew Hellershanks: I also run in to texturing issues with mesh.
[11:34] Nebadon Izumi: but also that the new protocol it will use
[11:34] Nebadon Izumi: will allow for many viewers
[11:34] Nebadon Izumi: it will open the door for WebGL once that blossoms into something more useable
[11:34] Nebadon Izumi: possibly unity interfaces
[11:35] Nebadon Izumi: and then lots of new potential features that go along with that as well
[11:35] Robert Adams: if you are talking to them, some of those details getting out would be good
[11:35] Nebadon Izumi: in hardware inputs and displays etc..
[11:35] Nebadon Izumi: yes, I think it is still a bit to early for that but I agree
[11:35] Nebadon Izumi: if I am going to attract investors
[11:36] Nebadon Izumi: i need that info as well
[11:36] Nebadon Izumi: because i want to be clear, in no way can this get done for free
[11:36] Nebadon Izumi: its just too much work
[11:36] Nebadon Izumi: and not enough people to do this on a volunteer basis
[11:36] Robert Adams: it would cause a lot of discussion but it would get people excited
[11:36] Nebadon Izumi: while eventually it will be supported and maintained by volunteers and be open, that is the ultimate goal
[11:37] Nebadon Izumi: unforutnately that wont get us out of the starting gates
[11:37] Andrew Hellershanks: The size of the OS code base and lack of docs is one of the big problems keeping some new developers from jumping in.
[11:38] Andrew Hellershanks: If I didn't have a lot of projects already on my plate I'd be interested in looking at the scripting engine.
[11:38] Nebadon Izumi: yea that doesnt help, but you also don't really need to know the entire code base to fix bugs
[11:38] OtakuMegane Desu: That is an issue. Some may also be hesitant because getting major changes or fixing stuff in core has a tendency to create drama.
[11:38] OtakuMegane Desu: Or get blocked entirely at times.
[11:39] Nebadon Izumi: well generally the drama arises from monolithic patches that are hard to review
[11:39] OtakuMegane Desu: Sometimes for good reason but others it seems to come down to "Because I like it this way."
[11:39] Andrew Hellershanks: nebadon, depends on the nature of the bugs.
[11:39] Nebadon Izumi: people who start out with small fixes and get to know the community tend to do well here
[11:39] Nebadon Izumi: those who take the sledge hammer approach will not do well
[11:40] Nebadon Izumi: begginers anyway
[11:40] Nebadon Izumi: the fact that people are so passionate about it though is also a positive
[11:40] Nebadon Izumi: it means there is a lot of interest in making it work
[11:41] Nebadon Izumi: but you will find on any large scale opensource project
[11:41] Nebadon Izumi: that tempers flair
[11:41] OtakuMegane Desu: Of course lol
[11:41] Andrew Hellershanks: One bug I want to fix is how ini files are read. I haven't had the to really dig in to it to even figure out what block of code (or library) is the main one handling ini files.
[11:41] Nebadon Izumi: I have probably had some kind of argument with all the devs at one point
[11:41] Robert Adams: OS is "modular" but not in a useful way... mods should be easy to add so people can do it "their own way" of they wish
[11:41] Nebadon Izumi: threating to quit or fork will ownly get you shown the door
[11:42] Nebadon Izumi: ive wanted to quit a 1000 times
[11:42] Billy.Bradshaw ++Robert
[11:42] Nebadon Izumi: i have had melanie tell me no 1000 times
[11:42] Nebadon Izumi: we have had screaming matches
[11:42] Nebadon Izumi: but im still here
[11:42] Nebadon Izumi: she is still here
[11:42] Nebadon Izumi: we make it work
[11:44] Billy.Bradshaw We believe there is still a lot of improvements to be gained from the opensim  
As said a big issue for other developers with the necessary dotnet skills is becoming familiar with opensim architecture, that bar requires lowering
[11:44] Nebadon Izumi: yes I think we need to expand our capabilities too
[11:44] Nebadon Izumi: that will attract more people
[11:44] Nebadon Izumi: modern game engine
[11:45] Nebadon Izumi: ability to create what looks and feel like modern games
[11:45] Andrew Hellershanks: Harder still for those not familiar with .NET either.
[11:45] Nebadon Izumi: I think once people realize you are still stuck in the SL bubble
[11:45] Nebadon Izumi: its a big turn off for most
[11:45] Nebadon Izumi: it makes it impossible to attract real business
[11:45] OtakuMegane Desu: Getting a new viewer lined up will help a lot I think. Most people still see Opensim as a SL clone of sorts.
[11:45] Nebadon Izumi: right now its mostly being proped up by folks trying to emulate SL
[11:45] Nebadon Izumi: which is bad
[11:46] OtakuMegane Desu: Even I usually end up describing it as some variation os "It's like SL but better"
[11:46] OtakuMegane Desu: of*
[11:46] Nebadon Izumi: yea its hard to sell SL but better
[11:46] Nebadon Izumi: when hard core people from SL log in and the first 10 things they try dont work
[11:46] Nebadon Izumi: seriously go to SL and drag over the most hard core person you can find
[11:46] Billy.Bradshaw Sounds strange to some but we have a lot of fun profiling Opensim,
[11:46] Nebadon Izumi: let them try opensim for 10 minutes
[11:46] Nebadon Izumi: and see if they think its better
[11:47] Nebadon Izumi: you will probably never see them here ever again
[11:47] OtakuMegane Desu: It does of course depend on what they're after.
[11:47] Nebadon Izumi: yea
[11:47] Andrew Hellershanks: Its easier to add the features to make it more SL like or to add features people are used to having in SL. To radically improve things would require some person(s) to do a complete redesign of everything and how it should look and feel. That's a big effort, after which, someone has to code it.
[11:47] Nebadon Izumi: the problem is SL is now improving
[11:47] Nebadon Izumi: and we can't follow them
[11:47] Richardus.Raymaker Depends what the person is doing. but.. No simborder verhicle crossing. no good v erhicles. no exp[erience tools. slow scripting. think the run quick out
[11:48] Nebadon Izumi: becuase the TPVs have practicly disbanded
[11:48] OtakuMegane Desu: Almost everything I do in OS is far better than SL. Just for lack of all the rules and limitations imposed for financial reasons.
[11:48] Nebadon Izumi: its unlikely OpenSim is going to see any of these SL improvements
[11:48] Nebadon Izumi: unless something seriously drastic happens on the TPVas
[11:48] Richardus.Raymaker The only weak point in SL is the tier. that is the only really strong point in opensim
[11:48] Nebadon Izumi: there was definitely a point where I would have said opensim was better
[11:48] Robert Adams: we now have sim border crossing after the merge
[11:48] Nebadon Izumi: but I dont think that argument still holds water
[11:49] Nebadon Izumi: opensim is better at some things, and sl is better at others
[11:49] Richardus.Raymaker Ok. but cannot say OS is better then SL
[11:49] OtakuMegane Desu: Tier and too many rules was my breaking point honestly. I'll take some broken or buggy stuff in exchange for being able to make what I want when I want. I'm not the majority though lol
[11:50] Billy.Bradshaw I saw a TVP in world a few days ago, looks like there are active, cannot remember where
[11:50] Robert Adams: so what is OS? a totally new platform for grids?
[11:50] OtakuMegane Desu: What is SL exactly? :)
[11:50] Nebadon Izumi: SL #1 is a social platform
[11:50] Nebadon Izumi: opensim is not
[11:51] Richardus.Raymaker Not sure how it is this days, but something simple like saving a script did end many times in drama. old script showingg. new script not svaed etc.
[11:51] Nebadon Izumi: SL's primary function is social platform, second is economy
[11:51] Billy.Bradshaw not sure I agree with that, opensim is whatever you want it to be
[11:51] Nebadon Izumi: right but its primary function is not to be a social platform
[11:51] Nebadon Izumi: its a VR Framework
[11:51] OtakuMegane Desu: Opensim has largely been for the creative and nerdy among us in my observations. When SL took a turn towards social function a bunch of people like myself jumped ship to OS or sometimes other things.
[11:52] Nebadon Izumi: right now I dont use OpenSim for Social stuff at all hardly
[11:52] Billy.Bradshaw Too nerdy can be a problem
[11:52] Nebadon Izumi: im using it for City Planning
[11:53] Nebadon Izumi: I personally have very little interest in the social side
[11:53] OtakuMegane Desu: A VR framework sounds good, actually. Hadn't thought quite like that before.
[11:53] Nebadon Izumi: yes this is why its important we have our own viewer
[11:53] Nebadon Izumi: right now we come off as a cheap SL knock off
[11:53] Nebadon Izumi: if we are to be taken seriously on a global stage
[11:53] Nebadon Izumi: we need our own viewer tech
[11:54] Nebadon Izumi: one that doesnt require jumping through hoops
[11:54] Andrew Hellershanks: I wonder if a new viewer is still going to be done in C++ or not.
[11:54] Nebadon Izumi: if you have ever tried to use OpenSim in university or corporate network, or government network
[11:54] Nebadon Izumi: let me just tell you there are not enough hoops in the world you can jump through
[11:54] Nebadon Izumi: its like sonic the hedghog
[11:55] Andrew Hellershanks never played that game
[11:55] Nebadon Izumi: lol
[11:55] OtakuMegane Desu: It nwould depend what exactly we used as the base. We don't have the resources and may not for many years to make a new viewer. It's gonna have to be modifying an existing one to what we need.
[11:55] Nebadon Izumi: its a never ending trail of hops
[11:55] Nebadon Izumi: hoops*
[11:55] Billy.Bradshaw yes, and opensim can be expensive to such an extent SL is less expressive
[11:55] Nebadon Izumi: you dont even want to know how much i spend a year on hardware for opensim
[11:55] Billy.Bradshaw corporate firewalls and the like, expensive
[11:55] Nebadon Izumi: its absurd
[11:56] Billy.Bradshaw training staff
[11:56] vegaslon plutonian: atleast hardware can be repurposed, that is always a good selling point
[11:56] Nebadon Izumi: yes for the average person opensim is low cost
[11:56] Nebadon Izumi: for someone using it on the scale i use it on
[11:57] Nebadon Izumi: its quite expensive
[11:57] Nebadon Izumi: right now i have a dozen or so 8 core xeon servers with 48 to 64gb ram
[11:57] Nebadon Izumi: and thats not counting OSCC and Avacon and OSGrid hardware
[11:58] Nebadon Izumi: the OSCC server while i dont actually pay any hosting fees, the hardware cost around 6500$
[11:58] Nebadon Izumi: paid for by a grant thankfully :)
[11:58] Nebadon Izumi: but still it wasnt cheap
[11:58] Billy.Bradshaw question, the unknown user problem, do people within a grid see the problem or is it just HG travelers?
[11:59] Nebadon Izumi: its generally a HG problem
[11:59] Nebadon Izumi: however
[11:59] Billy.Bradshaw exclusive, that is important
[11:59] Nebadon Izumi: people inside the grid can see if if people ar HG'ing in
[11:59] Nebadon Izumi: the problem is the region can cache this bad data
[11:59] Nebadon Izumi: then spread it around
[11:59] Richardus.Raymaker i mean 1 server or al
[11:59] Richardus.Raymaker what does it cost every mont to run it ?
[11:59] Nebadon Izumi: in a completely walled garden it will never happen
[12:00] Nebadon Izumi: not ever
[12:00] Billy.Bradshaw I am sure last friday party someone from osg said they saw unknown, but could not contact them before they left
[12:00] Andrew Hellershanks: Is UU more a viewer issue, back-end code, or a combo of both?
[12:00] Nebadon Izumi: yes local grid users will see it
[12:01] Nebadon Izumi: if Hypergrid users are coming into their grid
[12:01] Billy.Bradshaw I known what happens for HG users, its figuring out how it can happen for single grid users
[12:01] Nebadon Izumi: you dont have to acutally initiate hypergrid to see it happen
[12:01] Nebadon Izumi: just have to have HG visitors come into a region at some point
[12:01] Nebadon Izumi: and cache that bad data
[12:01] Nebadon Izumi: then only a region restart will clear it
[12:01] Billy.Bradshaw the cache is a result of the issue, not the cause
[12:01] Nebadon Izumi: correct
[12:01] Nebadon Izumi: there are probably a 1/2 dozen causes
[12:01] Nebadon Izumi: or more
[12:02] Nebadon Izumi: there wont be one singular fix
[12:02] Billy.Bradshaw why do you say that?
[12:02] Andrew Hellershanks: Sounds like a design flaw that needs fixing.
[12:02] Nebadon Izumi: there are various ways it can be triggered
[12:02] Billy.Bradshaw it is a design flaw
[12:02] Nebadon Izumi: you will notice the UMMA sometimes appears differentl;y
[12:02] Nebadon Izumi: Justin had it show different name at the various points of failure
[12:02] Nebadon Izumi: unfortuntaely the viewer is the problem
[12:02] Nebadon Izumi: not opensim
[12:02] Nebadon Izumi: and fixing it is quite difficult
[12:03] Andrew Hellershanks: ah, that is what I was wondering.
[12:03] Robert Adams: many developer hours have been spent looking for the unknown user problem
[12:03] Nebadon Izumi: yes we spent months chasing it
[12:03] Nebadon Izumi: for OSCC
[12:03] Billy.Bradshaw interesting speculation
[12:03] Nebadon Izumi: with very little success
[12:03] Nebadon Izumi: we improved it slightly
[12:03] Andrew Hellershanks: well, the 4096 TP limit was considered a problem that would be difficult to impossible to fix and it did get fixed.
[12:03] Nebadon Izumi: yes nothing is truly impossible but it can be quite painful
[12:04] Richardus.Raymaker well, got to run. bye all
[12:04] Nebadon Izumi: kk Richardus, thanks for coming :)
[12:04] Andrew Hellershanks: tc, Richardus
[12:04] Billy.Bradshaw problem with frameworks sometimes is the isolation from what actually happens on the wire
[12:05] Nebadon Izumi: well its hard to make a framework when no one truly understands the viewer
[12:05] Nebadon Izumi: there are parts of the viewer code that literally no one understands
[12:05] Nebadon Izumi: even LL employees do not
[12:05] Billy.Bradshaw I mean dotnet framework
[12:05] Nebadon Izumi: ah
[12:05] OtakuMegane Desu: Hahaha
[12:06] Nebadon Izumi: this is really why its important we have our own viewer
[12:06] Nebadon Izumi: because then we have a much better understanding of the full dynamics
[12:06] Nebadon Izumi: using an engine that is well documented
[12:06] Nebadon Izumi: like Unreal or Unity or something
[12:06] Billy.Bradshaw The viewer sources are availble
[12:06] Nebadon Izumi: the LL viewer is a major black hole
[12:06] Nebadon Izumi: and the info is basically lost forever
[12:06] Nebadon Izumi: unless some super genius prodigy dude comes along and spells it all out
[12:07] Nebadon Izumi: which im not holding my breath for
[12:07] OtakuMegane Desu: Super genius prodigy dude would probably just rewrite it instead. :P
[12:07] Nebadon Izumi: right
[12:07] Nebadon Izumi: haha
[12:07] Andrew Hellershanks: yeah
[12:07] Nebadon Izumi: the LL viewer is 12 years of layered duct tape and spray paint
[12:08] Andrew Hellershanks: and the paint is cracking and peeling off
[12:08] Billy.Bradshaw that can be said for a lot of applications
[12:08] Billy.Bradshaw else why patch things every 5 mins
[12:08] Nebadon Izumi: haha well lots of patches isnt a bad thing
[12:09] Nebadon Izumi: however in the case of LL they kept trying to add on functionality that didnt make sense
[12:09] Nebadon Izumi: 1/2 the time abandoning the intiiative but leaving loads of useless code intact
[12:09] Nebadon Izumi: like Windlight
[12:09] Nebadon Izumi: my god
[12:09] Nebadon Izumi: what a butcher job
[12:09] OtakuMegane Desu: Not to mention they've probably rotated out devs a number of times
[12:09] Nebadon Izumi: Windlight was amazing technology
[12:09] Nebadon Izumi: that they butchered and destroyed
[12:10] Nebadon Izumi: windlight had an amazing weather system
[12:10] Nebadon Izumi: for rain and snow
[12:10] Nebadon Izumi: they gutted it
[12:10] OtakuMegane Desu: Any real documentation is likely sitting in binders at LL, never to be released. All we tend to get are some code comments and whatever people manage to puzzle out on their own.
[12:10] Nebadon Izumi: haha ya
[12:11] Andrew Hellershanks: Was Windlight originally something from outside of LL?
[12:11] vegaslon plutonian: sacrifices had to be made to keep letting it view the crazy amount of vertices in one world
[12:11] Nebadon Izumi: thanks everyone for coming
[12:11] Nebadon Izumi: i'll try to find some more info on viewer and when i can share I will
[12:11] Nebadon Izumi: and yes Andrew
[12:11] Robert Adams: thanks neb. any info would be good
[12:11] Nebadon Izumi: LL bought windlight
[12:11] Nebadon Izumi: and butchered the hell out of it
[12:12] Nebadon Izumi: if you google im sure you can probably find some info on the original windlight system
[12:12] Andrew Hellershanks: ok. I was wondering if there was any chance to put back the parts that were ripped out.
[12:12] OtakuMegane Desu: I remember reading a lot about it when they were first picking it up
[12:12] Nebadon Izumi: no they bought the company
[12:12] Nebadon Izumi: i doubt they ever shared the original source
[12:12] Andrew Hellershanks: that's too bad.
[12:12] OtakuMegane Desu: The potential was awesome. The, pretty sky and that's about it.
[12:13] Nebadon Izumi: yea
[12:13] Nebadon Izumi: ok well thanks guys, be on IRC if anyone needs me :)
[12:16] vegaslon plutonian: still a few opensim compatible viewers that are updated alot, one of them in has the new project bento skeleton in it. Skeleton works well in opensim, just lacking support for storing the new attachment points
Personal tools
About This Wiki