Chat log from the meeting on 2017-09-12

[11:11] Andrew.Hellershanks @hg.osgrid.org: I was just checking if there was anything new in the way of development this past week but apparently not on the master branch.
[11:11] Andrew.Hellershanks @hg.osgrid.org: There has been some discussion about releasing a post fixes version of 0.9.
[11:12] Andrew.Hellershanks @hg.osgrid.org: Part of the issue with making a post fixes releases about the amount of feedback of 0.9, or rather the lack of feedback about it.
[11:12] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: how has 0.9 been doing out in the wild do we know?
[11:12] Andrew.Hellershanks @hg.osgrid.org: That's just it. AFAIK, there has been little feedback.
[11:13] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: Oh, I have heard very little on social media? Have been seeing grids doing upgrades but not sure if thats to the 0.9 version.
[11:13] Sheera Khan: By the way, the numbers switched from 0.9.0 to 0.9.1 quite a bit ago. What is the state of 0.9-Release?
[11:13] Andrew.Hellershanks @hg.osgrid.org: People may still be avoiding it, or they have tried it and are happy with it (or just not found any issues), or they haven't gotten around to providing any feedback.
[11:14] Kayaker Magic: The server in my barn went down just before I left on a 6 week trip, so I have been unable to test 0.9 for two months... :(
[11:14] Sheera Khan: is it a rather current 0.9.1? or is it some final 0.9.0?
[11:14] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: I have the release version (from the website) running with no problems
[11:15] Selby.Evans @grid.kitely.com:8002: Kitely is feeling pressure for upgrade to 0.9, but the have other stuff keeping them busy
[11:16] JayR.Cela @grid.3rdrockgrid.com:8002: We are running 0.9.Final on the KazGrid / the only issue we have noticed it a slight stutter with vehicle's crossing Sim borders / standard 256x256
[11:16] Sheera Khan: It looks like 0.9.1 runs pretty well in our 0.8.3 grid
[11:16] Kayaker Magic: Ilan at Kitely believes 0.9 needs more testing
[11:17] Andrew.Hellershanks @hg.osgrid.org: Sheera, the 0.9.1 designation would be the master branch. I just checked the code and i see there is a post fix branch of 0.9
[11:18] Selby.Evans @grid.kitely.com:8002: Kitely needs to add and test proprietary things before switch.
[11:18] Andrew.Hellershanks @hg.osgrid.org: JayR, which physics engine are you using in KazGrid?
[11:18] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: Just so I understand there is an official 0.9 release version, then 0.9 post fixes and master is at 0.9.1
[11:19] Ubit Umarov: 0.9 post fixes == release at this time
[11:19] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: ok thanks Ubit
[11:19] Ubit Umarov: master has a few fixes now
[11:19] JayR.Cela @grid.3rdrockgrid.com:8002: @Andrew / we are using Bullet Sim / we tried ubODE but none of the Power Boats we tried would cross standard Sim borders at all / so that was a real surprise and we switched back to Bullet
[11:20] Ubit Umarov: that will be the postfixes release
[11:20] JayR.Cela @grid.3rdrockgrid.com:8002: @Ubit / OK thanks for the heads up
[11:20] Sheera Khan: thank you for clarifying that Ubit :-)
[11:21] Ubit Umarov: and about releases / master
[11:21] Andrew.Hellershanks @hg.osgrid.org: JayR, what problems did you have with the power boats when crossing a sim boundary when using ubODE?
[11:21] Ubit Umarov: grids that do have coders, should keep a eye on master
[11:21] Ubit Umarov: and get the fixes from it
[11:22] Ubit Umarov: releases are more for folks without the skills to do that
[11:22] Andrew.Hellershanks @hg.osgrid.org: good advice, Ubit. I know I watch master and have cherry picked fixes in the past for a grid.
[11:22] JayR.Cela @grid.3rdrockgrid.com:8002: @Andrew / the boats worked fine within a 256x256 Region but when we came to a border the boats would all just freeze up
[11:22] Ubit Umarov: of course this is valid now.. that master is mainly on a bugs fix mode
[11:22] Ubit Umarov: after some point master will get back into full dev mode
[11:23] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: Will look into how to do that, have done it a couple of times. Gulp.
[11:23] Ubit Umarov: ( i did some dev changes on a branch for example )
[11:23] Sheera Khan: Would it be much of a hassle to make an announcement when that change happens?
[11:23] Andrew.Hellershanks @hg.osgrid.org: Leighton, git cherry-pick --help
[11:24] Ubit Umarov: well in my case guess it will a merge of the branch httptests into master :)
[11:24] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: ah ok is it similar to applying a patch
[11:24] Sheera Khan: it would help us to see, when to to take our "snapshot" for our next edition ...
[11:25] Ubit Umarov: we will take some time still before releasing a postfixes
[11:25] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: just googled it @ Andrew doesnt look to complicated thank you
[11:26] Ubit Umarov: we are waiting for people to really starting using 0.9 and report issues
[11:26] Ubit Umarov: about running 0.9 on 0.8x grid it should work
[11:26] Andrew.Hellershanks @hg.osgrid.org: yw, Leighton
[11:27] Ubit Umarov: was a initial transition goal
[11:27] Ubit Umarov: but in the time it took to actually release 0.9 some mb lost ;)
[11:27] Ubit Umarov: but guess minor issues.. if any
[11:27] Andrew.Hellershanks @hg.osgrid.org: I have someone using 0.8 that wanted to try out ubODE. The region is throwing lots of mesh related messages. I'll be getting them to try out 0.9 soon.
[11:28] Ubit Umarov: 0.8 does not have ubOde
[11:28] Ubit Umarov: and you can't just "merge" code
[11:28] Andrew.Hellershanks @hg.osgrid.org: oh. :P
[11:28] Ubit Umarov: that merge is called 0.9
[11:28] Ubit Umarov: :p
[11:28] Andrew.Hellershanks @hg.osgrid.org: I've been dealing with 0.9 so long I'm use to ubODE "just being there". :)
[11:28] Andrew.Hellershanks @hg.osgrid.org nods
[11:30] JayR.Cela @grid.3rdrockgrid.com:8002: on the better side of things ubODE exposed all of the improperly uploaded MESH stuff I have been bringing over to THE KazGrid / LOL / that was trip running around putting up invisiprims all over the place / :_))))
[11:30] JayR.Cela @grid.3rdrockgrid.com:8002: and then removing them all after the switch back to Bullet :_)
[11:31] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: that doesnt sound much fun
[11:32] Ubit Umarov: not all is bad...
[11:32] JayR.Cela @grid.3rdrockgrid.com:8002: @ Leighton / haaa / your right it was not fun at all :_)
[11:32] Ubit Umarov: some are just to be used with type NONE or CONVEX
[11:32] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: maybe saving and loading oars would have made it a little easier @ JayR
[11:32] JayR.Cela @grid.3rdrockgrid.com:8002: @ubit / well we know that know
[11:33] Ubit Umarov: but lots of houses around yes :(
[11:33] Ubit Umarov: i did fix 2
[11:33] Ubit Umarov: where very simple.. just added a dozen prims
[11:33] JayR.Cela @grid.3rdrockgrid.com:8002: I dont own the Grid / I am an advisor / so It is really something I have little control over
[11:33] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: out of 14 regions I was lucky I only had to replace two pieces of mesh, phew.
[11:33] Ubit Umarov: and recover full physics collisions easy
[11:34] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: that makes sense JayR
[11:34] JayR.Cela @grid.3rdrockgrid.com:8002: :_)
[11:35] Ubit Umarov: btw don't forget that changing physics engines means 2 settings on opensim.ini !!!!
[11:35] Ubit Umarov: you need to select the Meshmerizer and the physics engine
[11:35] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: nods @ Ubit
[11:35] Ubit Umarov: or meshs will smoke :p
[11:36] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: Fun fact I really like the word Meshmerizer :)
[11:36] Ubit Umarov: don't know who named it.. Dahlia possible
[11:37] Ubit Umarov: +90% of both versions are Dahlia code
[11:38] Andrew.Hellershanks @hg.osgrid.org: The other setting that needs to be changed when changing physics engine is the one called "meshing" that can be found just above the area in OpenSim.ini where you have the "physics" setting.
[11:38] Ubit Umarov: yes andrew i said Meshmerizer and engine
[11:39] Ubit Umarov: good they are very close on the file
[11:39] Ubit Umarov: but don't forget to change both!!
[11:39] Ubit Umarov: ( yes.. i do forget also... )
[11:39] Andrew Hellershanks: Ubit, yes. I was pointing out the actual name of the setting and not just one of the values for it.
[11:40] JayR.Cela @grid.3rdrockgrid.com:8002: @ Andrew & Ubit / thanks for the tips I will pass this information along to Kazuko Yoshida the Grid owner :_)
[11:40] Ubit Umarov: we will remove that need sometime in future
[11:40] Ubit Umarov: either making the one for ubOde part of the engine, or merging both
[11:40] Ubit Umarov: most likely the last option
[11:41] Ubit Umarov: that dilema is why we still have 2 :p
[11:41] Andrew Hellershanks: Hm... I just had a thought. One easy option is to have the code spit out red text if there is a mismatch between the physics engine and the mesher. My other thought is if someone went in to the code to do that they could just as easily override the mesher option when ubODE is selected.
[11:42] Ubit Umarov: well you will notice it pretty fast :p
[11:42] Andrew Hellershanks nods
[11:42] JayR.Cela @grid.3rdrockgrid.com:8002: ???? @Ubit do you mean merge ubODE and Bullet ??
[11:42] Ubit Umarov: NONO
[11:42] James atLLOUD: lol
[11:42] Andrew Hellershanks: lol
[11:42] OtakuMegane Desu: That is a horrifying thought.
[11:42] Ubit Umarov: ubOdeMeshmerizer and MeshMerizer
[11:42] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: so what happens to the mesh when the wrong meshmerizer is used?
[11:43] Ubit Umarov: physics engines can't be merged
[11:43] James atLLOUD: to be honest @ jay, I had that thought too, but dismissed it fast
[11:43] JayR.Cela @grid.3rdrockgrid.com:8002: Ahhh / OK that now sounds realistic / :_)))
[11:43] Ubit Umarov: you will see a few mesh errors
[11:43] Ubit Umarov: and strange things inworld :p
[11:44] Ubit Umarov: ( collisions mainly )
[11:44] Leighton.Marjoram @grid.lighthousepoint.co.uk:8002: ah ok thanks Ubit
[11:44] Ubit Umarov: there should be no damaged expect possible on phsical objects
[11:44] Andrew Hellershanks: Ubit, People will see a problem quickly. I still find it nice if the code explicitly comes out and tells you what you messed up instead of having to figure out from the symptoms.
[11:45] Ubit Umarov: those can get into odd places/state
[11:45] Ubit Umarov: andrew you are core.. mb you can work that out ?
[11:45] Ubit Umarov: :p
[11:46] Andrew Hellershanks: Ubit, I'm sure I could. First step would be to have a mantis I can watch to remind me to look in to the issue.
[11:47] Ubit Umarov: you did the suggestion.. so..
[11:47] Ubit Umarov: :p
[11:47] Andrew Hellershanks: I've been working on the software for a personal project and setting up a backup server for a grid in the last while so I've been a bit busy.
[11:47] Andrew Hellershanks: There is still additional information that needs to go in to the 0.9 release notes.
[11:48] Andrew Hellershanks: Some of the points I suggested for it haven't been added yet.
[11:49] Ubit Umarov: so crossing issues with 0.9 ?
[11:49] Ubit Umarov: crossings are very sensitive to network "weather"
[11:50] Ubit Umarov: they depend on fast comms btw the 2 regions and grid
[11:50] Ubit Umarov: 0.9 does reduce that a bit.. but still there
[11:51] Ubit Umarov: ahh and viewers..
[11:51] Ubit Umarov: hmm meeting nap time?
[11:51] Andrew Hellershanks: Kayaker, have you done any tests with 0.9 since the recent raycasting changes were made?
[11:52] Kayaker Magic: Unfortunately my server is down, so I don't have a place to install new 0.9 versions until I get that working again.
[11:52] Andrew Hellershanks: ah, yes. You did mention that and the changes were quite recent.
[11:53] Ubit Umarov: and andrew means Bullet changes
[11:53] Kayaker Magic: Bullet has an integrated castray now?
[11:54] Andrew Hellershanks: yes those were changes for raycasting made by Robert for BulletSim physics.
[11:54] Kayaker Magic: Cool.
[11:54] Kayaker Magic: I'm replacing the server this weekend, I will try tests with that as soon as I can.
[11:54] Andrew Hellershanks: Kayaker, I know you have (or had) a mantis about raycasting issues so I thought you would want to try it out when you can.
[11:55] Andrew Hellershanks: Almost the top of the hour. Any other topics for today?