Chat log from the meeting on 2019-03-26

From OpenSimulator

Jump to: navigation, search

[11:03] Andrew Hellershanks: Hello, everyone.
[11:03] Andrew Hellershanks: Hello, Selby.
[11:03] Selby.Evans hi everyone
[11:04] Gavin.Hird Hi Selby
[11:04] JayR.Cela Hey there :_)
[11:07] Andrew Hellershanks: I don't know if anyone else is on their way in. We might as well get started.
[11:07] Ubit Umarov: jay rebake :)
[11:07] Andrew Hellershanks: Just as I say that, here is Leighton. :)
[11:08] Gavin.Hird Hi Leighton
[11:08] Leighton.Marjoram hello all gtsy
[11:08] Andrew Hellershanks: Hello, Leighton
[11:09] JayR.Cela :_) hey there Leighton
[11:09] Ubit Umarov: so what opensim news do you bring today ? :)
[11:09] Leighton.Marjoram waves
[11:10] Ubit Umarov: jayyy rebake :p
[11:10] Gavin.Hird I am build the Windows version of the viewer with Animesh support this week
[11:10] Andrew Hellershanks: Only news from my end is that I am down to three OSSL functions still to be documented with sample scripts. One is slightly more complicated as it has two versions with different parameter list.
[11:10] Bill Blight: Well I like the new ViewerObjectCache
[11:10] Gavin.Hird non-RLV will most likely be done tomorrow
[11:11] Eldovar.Lamilton is online.
[11:11] Ubit Umarov: yeap VOC seems working on here and lbsa
[11:11] Ubit Umarov: working ok..
[11:11] Gavin.Hird what does it do?
[11:12] Gavin.Hird ... compared to before I mean
[11:12] Andrew Hellershanks: What is different about it?
[11:12] Ubit Umarov: finaly we do try to make use of viewers objects cache
[11:12] Bill Blight: We never actually had ViewerObjecCache before
[11:12] Bill Blight: it never worked
[11:13] Ubit Umarov: yeap part of my lludp facelifing
[11:13] Andrew Hellershanks: Ah, ok.
[11:13] Ubit Umarov: lifting?
[11:13] Ubit Umarov: ok .. my lludp code review :p
[11:14] Gavin.Hird is the new lludp in astate safe for testing?
[11:14] Ubit Umarov: you tell me, you are on last dev master here :)
[11:14] Gavin.Hird sitting here
[11:15] Ubit Umarov: but that did not kick in for you still
[11:15] Gavin.Hird I am sure I can do things that challenge the code in other ways :-)
[11:15] Ubit Umarov: you should now have files on viewer objects cache folder
[11:15] Ubit Umarov: if your viewer does suport it
[11:15] Ubit Umarov: the old singu i have does not for example
[11:16] Ubit Umarov: on next relog here, you should notice a little diference
[11:16] Ubit Umarov: ( or a crash ;) )
[11:17] Ubit Umarov: about that, there a big tradition of telling everyone to clear viewers cache
[11:17] Ubit Umarov: that is bad.. viewers cache are fundamental for performance
[11:17] Ubit Umarov: should one be clean when we do suspect it does have a issue
[11:17] Gavin.Hird there are 2 entries in the objectcache folder
[11:17] Andrew Hellershanks: I just said that to someone earlier today when they were talking about some textures appearing just white.
[11:18] Ubit Umarov: but "current opensim culture" is to keep cleaning it all the time
[11:18] Ubit Umarov: very bad..
[11:18] Bill Blight: yes one should be the object.cache the other file is the per region cache
[11:19] Ubit Umarov: yes each as region grid coords on name
[11:19] Gavin.Hird objects_11089_10001.slc
[11:19] Ubit Umarov: thats it
[11:19] Gavin.Hird but it suspiciously small
[11:19] Ubit Umarov: objects aren't that big
[11:19] Ubit Umarov: and that is compacted format
[11:19] Gavin.Hird is there a separate cap for it?
[11:20] Ubit Umarov: basicly the lludp binary format i think
[11:20] Ubit Umarov: no pure lludp
[11:20] Gavin.Hird ok
[11:20] Andrew Hellershanks: Is this going to help with people who use different grids? In the past using different grids with the same viewer caused issues with what you saw and cleaning viewer cache helped.
[11:20] Ubit Umarov: but a twisted protocol
[11:20] Ubit Umarov: that we where not doing
[11:21] Ubit Umarov: dif grids may cause cache invalidation, on regions in same location
[11:21] Ubit Umarov: but just that
[11:21] Andrew Hellershanks nods
[11:21] Ubit Umarov: shoudl "self heal "
[11:21] Andrew Hellershanks: That would be good.
[11:22] Ubit Umarov: well i finished the code when? yesterday? the day before?
[11:22] Ubit Umarov: i mean code without sistematic viewer crashes[11:22] Gavin.Hird and it has been deployed to OSG or just test regions?
[11:23] Ubit Umarov: last where because those particles there at arrival point
[11:23] Ubit Umarov: only lbsa is running it
[11:23] Gavin.Hird ok, I'll stop by there then
[11:23] Ubit Umarov: and it is disabled by default .. there is a option to activate it
[11:23] Gavin.Hird I don't want to update the code on xmir till the animesh merge id done
[11:23] Ubit Umarov: until more ppl test it
[11:23] Ubit Umarov: you should have it all
[11:23] Gavin.Hird is*
[11:23] Ubit Umarov: ll has this for years now
[11:24] Ubit Umarov: so you should have it all
[11:25] Ubit Umarov: we just made no use of it and did not had necessary protocol up
[11:25] Gavin.Hird are there any regions around with animated meshes on for testing? So far the windows build is being tested in Aditi
[11:25] Ubit Umarov: yes this, lbsa etc
[11:26] Ubit Umarov: hmm think last osg bin has it
[11:26] Gavin.Hird they really don't have any release notes :-)
[11:26] Ubit Umarov: see the 2 guards back there?
[11:26] Ubit Umarov: or the tiny there in the middle ?
[11:27] Gavin.Hird not today
[11:27] Gavin.Hird last time they showed
[11:27] Ubit Umarov: err not animesh.. its aniwhat :p
[11:27] Ubit Umarov: well i see them dancing
[11:27] Ubit Umarov: odd guards dancing
[11:28] Bill Blight: they are not guards they are greeters
[11:28] Bill Blight: LOL
[11:28] Bill Blight: :P
[11:28] Ubit Umarov: oops details
[11:29] Gavin.Hird another thing I noticed in Aditi is that the viewer shows the shadow of attached mesh as it should , but not in OpenSim
[11:29] Ubit Umarov: there where more changes on lludp encoding
[11:29] Ubit Umarov: actually most of it by now
[11:29] Gavin.Hird maybe there is something server side not being sent to the viewer
[11:29] Ubit Umarov: think we do send all there is to send
[11:29] Ubit Umarov: actually not that much
[11:30] Ubit Umarov: basicly a flag
[11:30] Ubit Umarov: and the start/stop animations on own packets
[11:31] Ubit Umarov: nothing about shadows
[11:31] Gavin.Hird any attached mesh should show a shadow of it
[11:31] Ubit Umarov: ask viewers devs :p
[11:31] Gavin.Hird there code is there and it works on Aditi
[11:32] Gavin.Hird we'll have to figure it out
[11:32] Ubit Umarov: well you can see the protocal
[11:32] Ubit Umarov: and protocol
[11:32] Bill Blight: well I see them, others see them with FS so obviously the right things are being sent
[11:32] Ubit Umarov: and you cna debug the messages going around
[11:32] Ubit Umarov: so, tell me what is missing
[11:33] Gavin.Hird you are logged in a s local OSG user?
[11:33] Ubit Umarov: yeap
[11:33] Gavin.Hird and I am HG user
[11:33] Bill Blight: yes I shall hg over
[11:33] Gavin.Hird so there is a difference
[11:34] Ubit Umarov: that may impact animations been asked to wrong grid etc
[11:34] Ubit Umarov: not shadown
[11:34] Ubit Umarov: or shadows
[11:35] Ubit Umarov: and that animation issue should be temporary a few seconds or less
[11:35] Gavin.Hird I have always seen them before and last time they were animating properly, but today they don't show at all
[11:35] Ubit Umarov: UNLESS using fs
[11:35] Ubit Umarov: well i did not activate viewerassets cap bc of fs
[11:35] Gavin.Hird there has been no changes in the viewer code that affect it since last week, so what changed
[11:35] Bill Blight: my HG avatar sees the aniwhat
[11:35] Ubit Umarov: liru did gave a fix to FS about that
[11:36] Ubit Umarov: need to wait for that to be release bf activate the viewerassets cap
[11:36] Ubit Umarov: you can test it on your regions removing the coment on the caps list in opensimdefaults
[11:37] Ubit Umarov: if you do have that cap that is
[11:37] Ubit Umarov: guess all viewers without it will be banned from sl if not already
[11:37] Gavin.Hird the viewer does not have any code for viewerassets
[11:37] Ubit Umarov: look better :p
[11:38] Ubit Umarov: but thats not aniwhat thing
[11:38] Ubit Umarov: is all assets
[11:39] Ubit Umarov: lludp assets fetch will be removed ( if not already ) at sl
[11:39] Ubit Umarov: we will keep it of course
[11:39] Gavin.Hird which is why we split the viewer code in 2016
[11:40] Gavin.Hird Kokua cannot log in to Opensim at all any more
[11:40] Ubit Umarov: well as i said for now disabled bc that fs bug
[11:40] Gavin.Hird Dayturn can only log on to SL till they turn off UDP assets
[11:41] JayR.Cela have a good day everyone :_)
[11:41] Gavin.Hird but I only need two more days for the windows version to pass all the animesh tests on Adidi :-)
[11:41] Selby.Evans bye , JR
[11:41] Gavin.Hird bye
[11:42] Andrew Hellershanks: Bye, JayR
[11:44] Ubit Umarov: well to test VOP you can just relog here
[11:44] Ubit Umarov: and see if there is anydif
[11:44] Ubit Umarov: VOC i mean
[11:44] Ubit Umarov: (objects cache )
[11:44] Gavin.Hird I'll do a direct login to OSG first to eliminate the HG element
[11:45] Ubit Umarov: should not matter
[11:45] Ubit Umarov: think udp does not mess up regions has caps does
[11:45] Ubit Umarov:
[11:46] Gavin.Hird there is other grid specific things the viewer does too, so I need to have a look at than again
[11:49] Gavin.Hird there are files in the cache of the form like 99d02b42-9119-4708-804e-4c7d699162ae.xmir.inv.gz
[11:49] Andrew Hellershanks: We are about 10 minutes before the top of the hour. Any other topics for todays meeting?

Personal tools
About This Wiki