Chat log from the meeting on 2022-10-18

From OpenSimulator

Jump to: navigation, search
[11:00] Andrew Hellershanks: Hello, everyone.
[11:00] Marcus.Llewellyn @grid.avacon.org:8002 grins.
[11:01] Marcus.Llewellyn @grid.avacon.org:8002: Hiyas. :)
[11:01] Ubit Umarov: hi
[11:01] Andrew Hellershanks: Hello, Marcus.
[11:01] Selby.Evans @grid.kitely.com:8002: hi everyone
[11:01] Ubit Umarov: hmm something tells me ocss time is arriving
[11:01] Marcus.Llewellyn @grid.avacon.org:8002: Teehee.
[11:01] Ubit Umarov: oscc even
[11:02] Marcus.Llewellyn @grid.avacon.org:8002: Yes, the wheels are in motion. And as always, we'd love to have all of you there. :)
[11:03] Ubit Umarov: :)
[11:03] Andrew Hellershanks: Yes, I've mentioned it a couple of times recently. It is being held on Dec  10 and 11.
[11:03] Andrew Hellershanks: Hello, Kayaker.
[11:03] Marcus.Llewellyn @grid.avacon.org:8002 waves to Kayaker. :)
[11:04] Kayaker Magic: Ubit! We missed you last week!
[11:04] Ubit Umarov: btw sorry about last week, i was not feeling well, so missed it
[11:04] Ubit Umarov: missed the meeting, i mean
[11:04] Kayaker Magic: Yeah, looks like you still have a burning in your throat
[11:05] Ubit Umarov: oh? dang.. :)
[11:05] Andrew Hellershanks: hehe
[11:05] Ubit Umarov: wel abotu opensim code changes...  not many
[11:05] Andrew Hellershanks: Sorry to hear you weren't well last week. I wondered what kept you away from joining us.
[11:06] Andrew Hellershanks: Most of the past weeks changes were making use of changes in C#.
[11:06] Ubit Umarov: more changes on the dotnet 6.0 branch a lot just updating some code stylem witgout any meaningfull improvements other than maybe readability
[11:07] Ubit Umarov: but a few shound have some performance impact, even if small
[11:07] Marcus.Llewellyn @grid.avacon.org:8002: Yay for dotnet6 branch! :)
[11:07] Andrew Hellershanks: Hello, Jamie.
[11:08] Ubit Umarov: since visual studio messages help find some cases, i do the less useful just to clean the messages list
[11:08] Ubit Umarov: for example  list a = new list();
[11:08] Jamie.Jordan @grid.kitely.com:8002: hi everybody
[11:08] Ubit Umarov: can now be  list a = new();
[11:08] Ubit Umarov: questionable..  and visual change only
[11:09] Ubit Umarov: but ive done several, as i said to take those messages out of vs list so easy spoting others
[11:09] Ubit Umarov: this region is on that branch, and so far we are alive :)
[11:10] Marcus.Llewellyn @grid.avacon.org:8002: Cool. :)
[11:10] Ubit Umarov: osgrid decide to move all their regions to it also
[11:10] Andrew Hellershanks: Ubit, it is waiting for us to get in to an interesting discussion before it will crash. ;)
[11:10] Ubit Umarov: so lbsa, event plaza, sanfboxes.. all on dotnet6
[11:11] Andrew Hellershanks: Wow. I would have thought that is taking a bit of a chance.
[11:11] Ubit Umarov: wright... all osgrid regions
[11:11] Ubit Umarov: seems very visible savings on resources like CPU and ram
[11:12] Kayaker Magic: What OS is this server running? Linux? MicroSquishy?
[11:12] Andrew Hellershanks: Sounds good. How significant are the savings of each?
[11:12] Ubit Umarov: like ram usage average down to 1GB from 4GB on one case
[11:13] Marcus.Llewellyn @grid.avacon.org:8002: Nice.
[11:13] Marcus.Llewellyn @grid.avacon.org:8002: I imagine that the entire build process has been overhauled on the dotnet branch? (sorry, I'm way behind on this news.)
[11:13] Andrew Hellershanks: Wow. I wasn't expecting it to be that significant a change.
[11:13] Ubit Umarov: yeah Marcus it is a breaking change
[11:13] Andrew Hellershanks: Marcus, you need to be here more often. :)
[11:13] Marcus.Llewellyn @grid.avacon.org:8002: Probably should yeah. :P
[11:14] Ubit Umarov: i did update prebuild so part of the migration is automatic
[11:14] Marcus.Llewellyn @grid.avacon.org:8002: What about modules that rely on Mono's add-ons? (mautil) Diva's Wifi, for example.
[11:14] Ubit Umarov: but then we still need to copy at least one dll by hand, for the local platform
[11:14] Ubit Umarov: no idea
[11:14] Marcus.Llewellyn @grid.avacon.org:8002: I'll take that up with Crista. :)
[11:15] Ubit Umarov: we also use mono addins, and those are still as compiled to framework
[11:15] Ubit Umarov: and so far working
[11:15] Ubit Umarov: ofc Diva makes a deeper use of them.. so.. no idea
[11:15] Marcus.Llewellyn @grid.avacon.org:8002: So just takes some extra finesse?
[11:16] Ubit Umarov: possible
[11:16] Marcus.Llewellyn @grid.avacon.org:8002 noda amiably.
[11:16] Ubit Umarov: i didn't look much to robust
[11:16] Ubit Umarov: but thing osgrid also did migrate it
[11:16] Ubit Umarov: think
[11:17] Ubit Umarov: it a big breakign change, but a lot on limited areas
[11:17] Ubit Umarov: like no appdomains...
[11:17] Ubit Umarov: on a lot of code, it just does it fine
[11:18] Marcus.Llewellyn @grid.avacon.org:8002: Sounds good to me. It's a change that has to happen. Eventually Mono will fade out.
[11:18] Ubit Umarov: if not i would be porting opensim to c++ :p
[11:18] Vincent.Sylvester @hg.zetaworlds.com:8002: We talked about this, no self-harming xD
[11:19] Ubit Umarov: ofc we lost several things
[11:19] Ubit Umarov: like XEngine that makes no sense without appdomains
[11:19] Marcus.Llewellyn @grid.avacon.org:8002 nods amiably.
[11:19] Ubit Umarov: all out automaitc tests need deep rewrite
[11:20] Marcus.Llewellyn @grid.avacon.org:8002: I've already got AvaCon grid using YEngine, and plan the same for the OSCC grid.
[11:20] Ubit Umarov: they where already on obsolet nunit 2.0
[11:20] Ubit Umarov: and many do depend on xengine, for example
[11:21] Andrew Hellershanks: I just checked and 2.0 is rather old. It is possible to download Nunit 3.13
[11:21] Ubit Umarov: ofc tests are only useful as development code validation
[11:22] Ubit Umarov: have no impact on real runtime code
[11:22] Vincent.Sylvester @hg.zetaworlds.com:8002: They have been useful a few times lately
[11:22] Vincent.Sylvester @hg.zetaworlds.com:8002: Detecting some potential edge cases and such
[11:22] Marcus.Llewellyn @grid.avacon.org:8002: That validation has played in important roles for things like permissions in the past, IIRC.
[11:22] Ubit Umarov: yeah they have limited use... but not null
[11:23] Andrew Hellershanks nods
[11:23] Ubit Umarov: so.. on TODO.. one day :P
[11:23] Ubit Umarov: well sadly they are a lot more limited
[11:23] Andrew Hellershanks: That would be another way for a new contributor to help out.
[11:23] Ubit Umarov: bc they are made as direct image of the current code
[11:23] Vincent.Sylvester @hg.zetaworlds.com:8002: I spent last week fixing up the tests we do have as best as possible in prep to have them at least all working again before attempting to port to nunit3
[11:23] Ubit Umarov: if that code is wrong, so they are
[11:24] Ubit Umarov: permissions is such a complex case
[11:24] Ubit Umarov: they are made as "current idea" of what this or that permission should be, not how is actually is :)
[11:24] Vincent.Sylvester @hg.zetaworlds.com:8002: There is still something broken somewhere with export permissions as well
[11:25] Ubit Umarov: in other words, they are made based on current code implementation so that code does pass, not based on a higher level of the funcionality
[11:25] Ubit Umarov: that makes them a lot useless
[11:25] Ubit Umarov: sadly a lot of them are like that
[11:26] Ubit Umarov: doen after the code so THAT code does pass
[11:26] Ubit Umarov: see what i mean?
[11:26] Marcus.Llewellyn @grid.avacon.org:8002: Yup
[11:26] Vincent.Sylvester @hg.zetaworlds.com:8002: A few are very basic indeed only loading relevant paths of code and testing if conversions or something are running as designed. Less testing more like "does math.ceil work"
[11:27] Marcus.Llewellyn @grid.avacon.org:8002: Pass/Fail only says it doesn't crash, not that it works as intended.
[11:27] Ubit Umarov: well reason why i killed several in past..
[11:27] Vincent.Sylvester @hg.zetaworlds.com:8002: In regards to the script tests at least I did figure out how to basically start a near full scene to check if the output of a script matches
[11:27] Ubit Umarov: they where plain wrong :)
[11:27] Andrew Hellershanks: Marcus, that would depend on the tests. I've written tests for a project that check that code works as expected.
[11:27] Ubit Umarov: now all disabled on dotnet6 branch :)
[11:28] Ubit Umarov: for now...
[11:28] Vincent.Sylvester @hg.zetaworlds.com:8002: The notecard line test that I been having trouble with is just checking if adding lines to it shows up in the assets data for example
[11:28] Vincent.Sylvester @hg.zetaworlds.com:8002: Fairly useless test cause unlikely that ever fails
[11:28] Vincent.Sylvester @hg.zetaworlds.com:8002: Annoyingly that one test keeps getting stuck for no reason as well breaking builds
[11:28] Andrew Hellershanks: Vincent, Definitely not going far enough.
[11:29] Ubit Umarov: i did little changes on both master and dotnet
[11:29] Ubit Umarov: like fix a old setting on HG and killing respective warnin
[11:30] Vincent.Sylvester @hg.zetaworlds.com:8002: I think apart from the notecard tests I have gotten all tests we have and did have working again
[11:30] Ubit Umarov: mantis 9032
[11:30] Vincent.Sylvester @hg.zetaworlds.com:8002: Even the performance tests
[11:30] Ubit Umarov: ohh fixed a bug on ubode
[11:31] Ubit Umarov: fy = m_mass * fy + m_force.Z + m_forceacc.Y;
[11:31] Ubit Umarov: was not cool :)
[11:31] Ubit Umarov: another of my c&p typos
[11:31] Marcus.Llewellyn @grid.avacon.org:8002 pretends he knows what that refers to and looks horrified. :)
[11:32] Andrew Hellershanks: That does look a little suspect
[11:32] Ubit Umarov: this was a recent bug.. so only a few got it
[11:32] Ubit Umarov: well fixed
[11:32] Vincent.Sylvester @hg.zetaworlds.com:8002: There is still another bug somewhere in there causing massive lag with mesh physics, but I am still tracking down which commit introduced that specifically. Think I mentioned that before. Annoyingly it's time consuming testing commit by commit so will be a while
[11:32] Marcus.Llewellyn @grid.avacon.org:8002: Yay!
[11:33] Andrew Hellershanks whispers: Vincent, you can't use bisect to help track it down?
[11:34] Ubit Umarov: possible that bug is on the side of the mesh creator :p
[11:34] Vincent.Sylvester @hg.zetaworlds.com:8002: With how I do builds no I gotta do one by one
[11:35] Vincent.Sylvester @hg.zetaworlds.com:8002: I think it's something with meshing actually, triggering not immediately
[11:35] Ubit Umarov: there where changes also on unmanaged ubode
[11:35] Andrew Hellershanks: Too bad. Bisect was meant to reduce number of code builds needed to track down when a bug was introduced.
[11:35] Ubit Umarov: but those impact all meshes
[11:35] Vincent.Sylvester @hg.zetaworlds.com:8002: It's just 20 or so, that's not a big deal really, had worse
[11:36] Ubit Umarov: bisec is just a binary search
[11:36] Andrew Hellershanks: Yes, I have used it on very rare occasions to find the source of a bug.
[11:37] Ubit Umarov: and region still alive! ;)
[11:37] Vincent.Sylvester @hg.zetaworlds.com:8002: It's odd, it only triggers after a certain amount of time, not immediately. I think something is causing re-meshing of physics mesh or something like that. Region crossing does make it appear pretty instantly. Also not on all regions either, some are totally fine
[11:37] Ubit Umarov: well its like 3rd week on dotnet 6 no?
[11:37] Ubit Umarov: ofc several changes..
[11:37] Vincent.Sylvester @hg.zetaworlds.com:8002: I never get the easy stuff, I get the bugs that demand hours of work to track down and fix, just my luck
[11:38] Andrew Hellershanks: Vincent, it wouldn't be half the fun if it was easy. :)
[11:38] Ubit Umarov: i made a "deep" change on ubode
[11:38] Vincent.Sylvester @hg.zetaworlds.com:8002: Makes fixing them more satisfying that's for sure
[11:39] Andrew Hellershanks: indeed.
[11:39] Ubit Umarov: moved the avatars broadband phase to managed code
[11:39] Ubit Umarov: so saving a lot of managed/unmanaged crossings..
[11:40] Ubit Umarov: wel some.. :)
[11:40] Ubit Umarov: i may go on doing that on other cases
[11:40] Ubit Umarov: since dotnet 6.0 math code is now closer to c++
[11:40] Ubit Umarov: at least on intel cpus
[11:41] Andrew Hellershanks nods
[11:41] Ubit Umarov: sadly we can't use all improvements
[11:42] Ubit Umarov: for example they added Vector math, sometime ago that does use SIMD automaticly
[11:42] Ubit Umarov: in System.Numerics
[11:42] Ubit Umarov: that is alrady old news.. there since net4.5 or so
[11:42] Ubit Umarov: and was pretty bad... a lot worse than our Vector3 on libomv
[11:43] Ubit Umarov: but they keep trying and improving.. so now it may be a bit better
[11:43] Ubit Umarov: but we just can't use
[11:43] Ubit Umarov: bc all out code does depend on our Vector3
[11:44] Ubit Umarov: bahh fun fun
[11:44] Ubit Umarov: and on things like this we cant ever make a vector3 as good as native one
[11:45] Ubit Umarov: bc MS ( like mono did ) does cheat
[11:45] Andrew Hellershanks nods
[11:45] Marcus.Llewellyn @grid.avacon.org:8002 wonders if there's a good equivalent to numpy for C#.
[11:45] Ubit Umarov: the JIT code generator knows ONLY about those native types and does optimizations, even replacing entire code, only on those
[11:46] Ubit Umarov: the opensim source c# of vector3 is just totally igoned at runtime
[11:46] Ubit Umarov: JIT sees it is a vector3 and replaces it all by c++ made code
[11:47] Andrew Hellershanks: We are just under 15 minutes to the top of the hour. Before we continue Marcus has something to say about the upcoming OSCC event.
[11:47] Andrew Hellershanks: Marcus, go ahead if you are ready.
[11:48] Marcus.Llewellyn @grid.avacon.org:8002: I'd like to invite everyone to attend! :) As Andrew mentioned earlier, it will be held on Decemeber 10-11 this year. We're still accepting proposals (one of you has already submitted one, yay!)
[11:48] Marcus.Llewellyn @grid.avacon.org:8002: We're also still planning on highlighting work by core devs, so would love to have you guys again. :)
[11:49] Marcus.Llewellyn @grid.avacon.org:8002: I plan to upgrade the OSCC grid to 0.9.2.1. I am just waiting on Diva, because we rely heavily on one of her modules as part of conference registration.
[11:50] Ubit Umarov: oh 0.9.2.1.. so old
[11:50] Ubit Umarov: :)
[11:50] Ubit Umarov: but guess there will not be a 0.9.2.2 net4.6 release soon
[11:50] Andrew Hellershanks: It will be good to have that grid updated to a new version of code. It should have been updated to a newer release a couple of years ago but circumstances prevented that from happening.
[11:50] Marcus.Llewellyn @grid.avacon.org:8002: It's the last stable release. I would love to run the latest and greatest, but only have the mental bandwidth to use a build designated as stable. Too many other tasks. :)
[11:51] Ubit Umarov: but 21 guess is ok :)
[11:51] Marcus.Llewellyn @grid.avacon.org:8002: It was upgraded for last year's conference.
[11:51] Kayaker Magic: Yeah! I recall it was running YEngine last year@!
[11:51] Andrew Hellershanks: Hm... I must have last checked what version it was running before it had been updated.
[11:51] Marcus.Llewellyn @grid.avacon.org:8002: You asked for it, and you got it. ;)
[11:51] Kayaker Magic: hehe
[11:51] Andrew Hellershanks: :)
[11:51] Marcus.Llewellyn @grid.avacon.org:8002: Well, we're always far behind master.
[11:52] Ubit Umarov: well for a reason we do have the so called "releases" :)
[11:52] Marcus.Llewellyn @grid.avacon.org:8002: Yeah.
[11:52] Andrew Hellershanks: It was at one point when 0.9 had been out for a year or two and the OSCC grid was stlil on 0.8, IIRC.
[11:52] Ubit Umarov: master can be suicidal :)
[11:52] Ubit Umarov: wel 0.9 had a lot of resistence
[11:53] Andrew Hellershanks: Yes, there was also that.
[11:53] Ubit Umarov: still many out ther on 0.8.x
[11:53] Marcus.Llewellyn @grid.avacon.org:8002: I inherited grid admin duties for OSCC, but not sudo access at that time. That's all been cleared up.
[11:54] Andrew Hellershanks: Marcus, I don't know if I should say congratulations or commiserate with you. :) Fortunately the OSCC grid isn't that large.
[11:54] Marcus.Llewellyn @grid.avacon.org:8002: Oh, and I spose I ought to refer a link to the conference. https://conference.opensimulator.org/
[11:54] Ubit Umarov: :)
[11:55] Andrew Hellershanks: Marcus, good idea. ty
[11:55] Marcus.Llewellyn @grid.avacon.org:8002: Hehe. OSCC isn't too bad, no. I reworked a of the scripts that automate things.
[11:56] Andrew Hellershanks: Are you talking scripts that automate things in-world or on the servers, or both?
[11:56] Marcus.Llewellyn @grid.avacon.org:8002: bash scripts on the server side.
[11:56] Andrew Hellershanks: ok.
[11:57] Andrew Hellershanks: It takes time to work out ways to make it easier to admin a lot of regions for a grid.
[11:58] Marcus.Llewellyn @grid.avacon.org:8002: I mostly just have a bunch of scripts that fire up regions grouped into custom .screenrc files. Some other stuff for backups, auto-kicking bad HGers, etc.
[11:58] Kayaker Magic: And every grid manager rolls their own management code....
[11:58] Andrew Hellershanks: Kayaker, so true.
[11:58] Marcus.Llewellyn @grid.avacon.org:8002: Yeah. neb and my ways aren't too dissimilar, fortunately.
[12:00] Marcus.Llewellyn @grid.avacon.org:8002: And I like using ANSI codes. ;)
[12:00] Andrew Hellershanks: I'm only dealing with a very small grid that has about a half dozen regions. I use some shell scripts to maintain it. When I was dealing with three servers and several hundred regions I had a C# program that provided a web based system to aid in the management of the regions.
[12:01] Marcus.Llewellyn @grid.avacon.org:8002: I've thought about doing a server/agent style management system in C#. Never got around to it, and I'm inclined not to fix what ain't broke. Heh.
[12:01] Andrew Hellershanks: Marcus, It probably isn't worth the effort for a grid the size of the OSCC.
[12:01] Marcus.Llewellyn @grid.avacon.org:8002 nods amiably.
[12:02] Ubit Umarov: (that did remind me on dotnet6 the performancegap btw debug and release mode is larger)
[12:02] Marcus.Llewellyn @grid.avacon.org:8002: Good to know.
[12:03] Ubit Umarov: ( well also large on .net 4.6 / mono... not as on old days where was abotu the same )
[12:03] Andrew Hellershanks: Marcus, good luck with your admin duties and on the OSCC. Was there anything else you wanted to say about the event or the dates for any events leading up to it?
[12:04] Marcus.Llewellyn @grid.avacon.org:8002: Just to please check out the website! We'd love to see people there either as presenters, volunteers (we can always use more greeters!) or as an attendee! :)
[12:05] Kayaker Magic: Wasn't the deadline for presentations YETERDAY?
[12:05] Andrew Hellershanks: I'll be at the Keynote if it is once again with the core devs. I'm just glad the times for the OSCC are west coast times. I've seen RL conferences that have their Keynotes starting at 8am. I'm not up that early. :)
[12:06] Marcus.Llewellyn @grid.avacon.org:8002: It was. It's been extended, and we've pretty much always been relaxed about late submissions anyway.
[12:06] Marcus.Llewellyn @grid.avacon.org:8002: Some people are never on time. ;)
[12:06] Andrew Hellershanks: True.
[12:06] Andrew Hellershanks: Thank you again, Marcus.
[12:07] Marcus.Llewellyn @grid.avacon.org:8002: Sure thing!
[12:07] Andrew Hellershanks: We are now just past the top of the hour. Did anyone else have any questions/comments for todays meeting before people need to leave?
[12:07] Vincent.Sylvester @hg.zetaworlds.com:8002: If anyone is interested to help out a little we have mantis issue numbers 3535, 3644 and 2357 concerning texture position and rotation behavior that evidently at one point did not work correctly. I have re-tested them and couldn't find any issue, but it would be useful to have some more feedback in order to verify that it is fixed.
[12:07] Marcus.Llewellyn @grid.avacon.org:8002: I have an annoying grid issue I can't figure out with baking, if I may pick some brains? (braaaaains)
[12:08] Ubit Umarov: mantis 3k was what opensim 0.6.x ?
[12:08] Andrew Hellershanks: Thanks, Vincent.
[12:08] Vincent.Sylvester @hg.zetaworlds.com:8002: Last update on them claimed bug still present in .7x and I had issues that said the same and were broken still
[12:09] Andrew Hellershanks: Ouch. That would make them extremely old bugs. BTW, nunit 2.0 is from 2002.
[12:09] Andrew Hellershanks: The mantis issues Vincent mentioned are from 2008 and 2009.
[12:10] Andrew Hellershanks: Marcus, you can go ahead with your issue about baking.
[12:10] Marcus.Llewellyn @grid.avacon.org:8002: Thanks. :)
[12:12] Marcus.Llewellyn @grid.avacon.org:8002: So, about two months ago AvaCon grid has a catastrophic RAID crash. Happily, I had offsite SQL backups. Everything seems to be fine except that avatar baking is really bad. It isn't like there is all that much to XBakes, so I'm been banging my head againast my desk trying to figure out what I might be missing.
[12:13] Marcus.Llewellyn @grid.avacon.org:8002: That grid is currently on 0.9.2.1
[12:13] Ubit Umarov: Xbakes is basicly a cache
[12:13] Marcus.Llewellyn @grid.avacon.org:8002: Right. And I see hits to it, and no errors. But agents don't seem to be getting the baked textures of other agents in a timely manner, if at all.
[12:13] Ubit Umarov: using last fs release?
[12:14] Marcus.Llewellyn @grid.avacon.org:8002: Yup, latest FS.
[12:14] Ubit Umarov: well dunno.. remember we had a came back of bakng issues some time ago
[12:14] Ubit Umarov: but think all viewer side
[12:15] Ubit Umarov: and mostly related to Universal wearable etc
[12:15] Andrew Hellershanks: What viewer are you using, Marcus?
[12:15] Marcus.Llewellyn @grid.avacon.org:8002: Firestorm 6.6.3.67470
[12:16] Ubit Umarov: well ofc HG tps still do a mess :)
[12:16] Ubit Umarov: but most time a manual rebake does fix
[12:16] Marcus.Llewellyn @grid.avacon.org:8002: Hehe, yeah, that's HG. :) But this is with local accounts.
[12:17] Marcus.Llewellyn @grid.avacon.org:8002: I dunno. I'm clearly missing something to the point where I'm questioning if my Mono version os bad, or maybe the version of libgdi or something.
[12:17] Ubit Umarov: shoul dbe on last mono stable
[12:17] Andrew Hellershanks: Marcus, If Ubit is right about the baking issues that came back are (mostly?) all viewer side have you tried using an older viewer to see if it makes a difference?
[12:17] Marcus.Llewellyn @grid.avacon.org:8002: I haven't. I can give that a go.
[12:18] Ubit Umarov: and no.. rebake is viewer side operation.. no libgdi used
[12:18] Marcus.Llewellyn @grid.avacon.org:8002: Okay, so that's good.
[12:18] Ubit Umarov: server only servies files
[12:18] Marcus.Llewellyn @grid.avacon.org:8002: I'm on Mono 6.12 right now.
[12:19] Ubit Umarov: dunno
[12:19] Marcus.Llewellyn @grid.avacon.org:8002: Anyway, was hoping this might be somethings like, "Oh yeah, I remember that we had that problem and we fixed it!" or something, but seems I'm not so lucky. hehe. I'll just keep banging on it.
[12:19] Marcus.Llewellyn @grid.avacon.org:8002: Appreciate the time. :)
[12:20] Ubit Umarov: missing assets or very slow service of them..  can bbe several things
[12:20] Ubit Umarov: i don't remember if 0.9.2.1 did had a issue
[12:20] Ubit Umarov: well nothing on mantis
[12:20] Marcus.Llewellyn @grid.avacon.org:8002: Everything else on the grid is served really quick! It is truly puzzling.
[12:21] Marcus.Llewellyn @grid.avacon.org:8002: I'm sure I've done something boneheaded (or forgot to do something.)
[12:21] Marcus.Llewellyn @grid.avacon.org:8002: Anyway, thank you guys. :)
[12:22] Andrew Hellershanks: Marcus, I would also check the consoles just in case there are any odd messages that might provide any additional hints.
[12:22] Marcus.Llewellyn @grid.avacon.org:8002: yeah, I do that for sure. It hasn't been volunteering any good clues. :P
[12:22] Ubit Umarov: viewer logs may help also :)
[12:23] Marcus.Llewellyn @grid.avacon.org:8002: Hmm... viewer logs. Good idea!
[12:23] Andrew Hellershanks: Or run the viewer from the console to see what messages roll by if you want to see them in real time.
[12:24] Marcus.Llewellyn @grid.avacon.org:8002 nods amiably.
[12:24] Marcus.Llewellyn @grid.avacon.org:8002: Great idea. I'll try that. :)
[12:24] Andrew Hellershanks: Marcus, let us know how you get on. There is the mailing list and IRC if you need to follow up with any information/questions before next week.
[12:24] Marcus.Llewellyn @grid.avacon.org:8002: Yes, I'm on the mailing list still, and I know where IRC is, even if I haven't been there in quite a while. :)
[12:24] Ubit Umarov: for sanity check and also try a 0.9.2.2 region
[12:24] Ubit Umarov: :)
[12:25] Marcus.Llewellyn @grid.avacon.org:8002: Er... did I miss a point release?
[12:25] Andrew Hellershanks: Good idea. No point looking for a problem that might have been fixed by a change in a more recent version of code.
[12:25] Ubit Umarov: i really don't remember if i had to make any region side code changes
[12:25] Ubit Umarov: since ..2.1 on rebakes
[12:25] Selby.Evans @grid.kitely.com:8002: bye all
[12:26] Andrew Hellershanks: Bye, Selby.
[12:26] Ubit Umarov: cya selby.Evans
[12:26] Marcus.Llewellyn @grid.avacon.org:8002: Laters! :)
[12:26] Andrew Hellershanks: We will be wrapping up this meeting shortly as it is almost half pas the hour.
[12:26] Ubit Umarov: miss release point? no
[12:26] Marcus.Llewellyn @grid.avacon.org:8002 nods affably. :)
[12:26] Ubit Umarov: master
[12:26] Marcus.Llewellyn @grid.avacon.org:8002: Ah... lol
[12:27] Ubit Umarov: it is been stable
[12:27] Andrew Hellershanks: Marcus, You will be busy for a while with all those avenues to pursue while tracking down what may be causing the baking issue.
[12:27] Marcus.Llewellyn @grid.avacon.org:8002: Yeah, I guess if all else fails, I can try master.
[12:27] Jamie.Jordan @grid.kitely.com:8002: great meeting thanks yall
[12:27] Andrew Hellershanks: yw, Jamie.
[12:27] Ubit Umarov: we we don't see any severe rebake issues
[12:28] Ubit Umarov: even most HG ones self heal
[12:28] Marcus.Llewellyn @grid.avacon.org:8002: I figured if there had been, it would have been news. So I broke something. Prolly.
[12:28] Ubit Umarov: :)
[12:29] Ubit Umarov: Xbakes kicks in doing tps btw in grid regions
[12:29] Andrew Hellershanks: You could also paste part of your ini files for review if you think it might be as simple as a config issue.
[12:29] Ubit Umarov: as i said it is a grid side cache
[12:30] Marcus.Llewellyn @grid.avacon.org:8002: Maybe. I've looked at the INIs until my eyes cross. lol! But sometimes it's one little char that your eyes just refuse to see. I'll let you know. :)
[12:30] Ubit Umarov: but just saves region having to ask viewer to remake them all again
[12:30] Ubit Umarov: something pretty damm heavy
[12:30] Marcus.Llewellyn @grid.avacon.org:8002: I'm currently thinking it might be inventory corruption.
[12:30] Ubit Umarov: well its rebake :)
[12:31] Ubit Umarov: get all ladies naked for inspection
[12:31] Andrew Hellershanks: It probably won't be the issue but you can also check CPU and memory usage on the system to make sure they have nothing to do with the issue.
[12:32] Ubit Umarov: well broken COF sure will cause issues
[12:32] Kayaker Magic: I've heard women claim that was a conspiracy of the male developers to see them all naked.
[12:32] Marcus.Llewellyn @grid.avacon.org:8002: Yeah, and while not the beefiest grid server, it doesn't seem to be a resource issue.
[12:32] Andrew Hellershanks nods
[12:32] Ubit Umarov: ohh really kayaker ?
[12:33] Ubit Umarov: how did they foud abotu that, i wonder..
[12:33] Ubit Umarov: :)
[12:33] Marcus.Llewellyn @grid.avacon.org:8002: Not interested in the ladies myself. ;)
[12:34] Andrew Hellershanks: There are a couple of ways to interpret that but let's not go there. :)
[12:34] Ubit Umarov: but inv issues would possible be localized to this or that avatar
[12:34] Ubit Umarov: not all
[12:34] Andrew Hellershanks: Right
[12:35] Marcus.Llewellyn @grid.avacon.org:8002: Sooner or later I'll have a Eureka moment. I hope. :)
[12:35] Ubit Umarov: or have everyone white during oscc :)
[12:35] Marcus.Llewellyn @grid.avacon.org:8002: Anyway, thank you for the meeting, and letting me talk about OSCC. :)
[12:36] Andrew Hellershanks: Marcus, np. Drop by any time. Let us know about your progress in solving the problem.
[12:36] Marcus.Llewellyn @grid.avacon.org:8002: Will do. Take care, guys. :)
[12:36] Andrew Hellershanks: On that note, I will call this meeting to a close as we are now past the half hour mark.
[12:37] Andrew Hellershanks: Thank you all for coming. See you again next week.
Personal tools
General
About This Wiki