Chat log from the meeting on 2018-08-21

From OpenSimulator

Jump to: navigation, search

Summary:
- Mantis #5844 - NULL_KEY not being treated as FALSE
- What's keeping grids from transitioning from 0.8 to 0.9?
(jepp, that's one heck of a topic^^)

[11:10] Andrew.Hellershanks @hg.osgrid.org: I don't have any thing much to report regarding code changes this week. I made one small patch to fix a long standing problem. Other than that, its been quiet on the development front.
[11:11] Ubit Umarov: hmm what code changes this week ?
[11:11] Arielle.Popstar @hg.osgrid.org: what was the long standing prob Andrew?
[11:12] Andrew.Hellershanks @hg.osgrid.org: There has been a long standing bug in how a key variable is treated in an if statement when used as a boolean.
[11:12] LaNani Sundara: a UUID?
[11:12] Andrew.Hellershanks @hg.osgrid.org: Mantis #5844.
[11:13] Sheera Khan: the NULLKEY not being treated as FALSE?
[11:13] LaNani Sundara: ah k
[11:13] Andrew.Hellershanks @hg.osgrid.org: Yes. If key_var has the value of NULL_KEY and you do "if (key_var)" it was always treated as true.
[11:13] Bill Blight: yeah but easy enough to fix with different if statement rather than changing the code
[11:13] Bill Blight: but not really a big deal either way
[11:13] Andrew.Hellershanks @hg.osgrid.org: Bill, you can go ahead and tell everyone who has done that to fix their code. :)
[11:13] Bill Blight: has been that way FOREVER
[11:13] Bill Blight: lol
[11:14] Andrew.Hellershanks @hg.osgrid.org: Yes, it has been that way a long time.
[11:14] LaNani Sundara: and that popped up where for example?
[11:14] LaNani Sundara: which module
[11:14] Andrew.Hellershanks @hg.osgrid.org: Someone did that recently and pointed it out to me. I thought the problem sounded familiar and confirmed that thought when i found the mantis.
[11:14] LaNani Sundara: ok
[11:15] LaNani Sundara: ah its a scripting thing i see
[11:15] LaNani Sundara: *nods
[11:15] Andrew.Hellershanks @hg.osgrid.org: Correct.
[11:16] Andrew.Hellershanks @hg.osgrid.org: It was a simple fix.
[11:16] LaNani Sundara: i happen to be in VS on my other screen :)
[11:16] Andrew.Hellershanks @hg.osgrid.org: :)
[11:16] Bill Blight: if (!key_var) is not treated as a null key, if a key is present at all it would return true , rather than seeing a NULL_KEY as false ... Long standing fix was to do if(key_var == NULL_KEY) , to get a false result ..
[11:17] Bill Blight: or if(key_var != NULL_KEY) for true
[11:17] LaNani Sundara: so its a bad scripting .habit?
[11:17] Bill Blight: would not call it a bad habit, would call it a an adaptation to the way the code has been forever
[11:18] Ubit Umarov: its Xengine feature
[11:18] Ubit Umarov: :p
[11:18] LaNani Sundara: cos "null" and false are not the same in many programming languages
[11:18] Bill Blight: right
[11:18] LaNani Sundara: so why assume they are
[11:18] Ubit Umarov: ( that also cames from c#
[11:18] Andrew.Hellershanks @hg.osgrid.org: It might be something documented in SL that some people expect to work in OpenSim. I haven't checked to see what SL defines as the expected behaviour.
[11:18] Ubit Umarov: there is no if(athing)
[11:19] LaNani Sundara: *nods
[11:19] Bill Blight: LaNani , in this case you are testing for the presence of a valid key, so if NULL_KEY which is a key of all zeros , NULL KEY is not a NULL
[11:19] Andrew.Hellershanks @hg.osgrid.org: Ubit, what do you mean by that?
[11:20] LaNani Sundara: if statements should evaluate boolean expressions, not "anything"
[11:20] LaNani Sundara: i think thats what he means :)
[11:20] Arielle.Popstar @hg.osgrid.org: in light of the name of this region Dev Outreach, i present this link and point out how many grids are stilll on .8.2 and ask what is it thats needed to get these grids on .9? What are they balking about? and what can the core devs do to answer their concerns for switching?
[11:20] Arielle.Popstar @hg.osgrid.org: https://docs.google.com/spreadsheets/d/1flswhynCWaav2-fLk8XJWq7-aBL9CqO57YazzBPcQd4/edit#gid=0
[11:20] Bill Blight: Constant: string NULL_KEY = "00000000-0000-0000-0000-000000000000";
[11:20] Bill Blight: so we are evaluating that not an actual NULL
[11:20] Ubit Umarov: yes.. there are some auto casts to bool, but not good practice on c#
[11:20] Bill Blight: I know it is confusing to the pure coders
[11:20] Andrew.Hellershanks @hg.osgrid.org: LaNani, to say that means you shouldn't do something like "if (flag)" where flag is a boolean.
[11:20] LaNani Sundara: but it is also not false or true
[11:21] Bill Blight: sends me for a loop sometimes too
[11:21] Ubit Umarov: but Xengine issue is a bit deeper
[11:21] LaNani Sundara: a boolean is a boolean expression
[11:21] Sheera Khan: Arielle: in one case I can firmly say: it's the (wo)man power needed - we don't have the time as we are very busy iRL :-/
[11:21] Ubit Umarov: well we do have Yengine now... it should that correctly
[11:22] Andrew.Hellershanks @hg.osgrid.org: Treating a key or string as if they have a boolean value when used on their own isn't very good practice.
[11:22] Bill Blight: well the test is bascially so see if the key is a NULL_KEY if it not, then an avatar or object key is present, then it would be true, if it is NULL_KEY it is false, so pure coding wise it is a bit screwed up,. but how it is used in LSL is , just the way it is ..
[11:22] Arielle.Popstar @hg.osgrid.org: i was in metro this morning and wasnt impressed
[11:22] LaNani Sundara: exactly Andrew
[11:22] Ubit Umarov: me and others before me did tried to make key a proper type.. bad got smoke
[11:22] Arielle.Popstar @hg.osgrid.org: clouds and stuttering
[11:22] Arielle.Popstar @hg.osgrid.org: its been over 3 years now
[11:22] LaNani Sundara: who needs clouds anyway ;p
[11:23] Bill Blight flüstert: even from the LSL SL wiki
[11:23] Arielle.Popstar @hg.osgrid.org: what are all these grids having an issue with?
[11:23] Bill Blight: NULL_KEY is a string. However it is only really useful as a key.
[11:23] LaNani Sundara: haha first thing i do in my viewer graphics settings is slide Sky settings to the left
[11:23] Bill Blight: Despite fitting the syntax criteria to be a valid key, when fed to a Conditional as a key it executes as FALSE.
[11:24] Leighton.Marjoram @grid.opensim.life:8002: @Arielle the problem was vocal community members slating 0.9 and its kinda stuck
[11:24] Ubit Umarov: lsl key is internally a string, that is why we have that issue
[11:24] Andrew.Hellershanks @hg.osgrid.org: Ubit, I tried that too. If you make the key type a key and not a string you have problem just trying to do something as simple as -> key_var = "<uuid>".
[11:24] Bill Blight: evidently it is how it works in sl too
[11:24] Bill Blight: http://wiki.secondlife.com/wiki/NULL_KEY
[11:24] LaNani Sundara: ok
[11:24] Ubit Umarov: i did tried changing from the parser also... yeack.. got smoke.. gave up
[11:24] Arielle.Popstar @hg.osgrid.org: so you think there is no specific valid objection?
[11:24] Ubit Umarov flüstert: IT IS XENGINE FEATURE
[11:24] Ubit Umarov: lol
[11:25] Ubit Umarov: like the llSleep
[11:25] Bill Blight: People preaching the bad about .9x and not the improvements and people not wanting to update, cause more issues than actual code issues
[11:25] Ubit Umarov: the vectors expressions needing more ()
[11:25] Ubit Umarov: etc
[11:25] LaNani Sundara: transfering to 0.9... would loading a 0.8 made OAR be an issue?
[11:25] Bill Blight: nope
[11:25] LaNani Sundara: ok
[11:25] Bill Blight: only if you switch physics engines and have really bad mesh
[11:25] Bill Blight: there could be an issue
[11:25] Ubit Umarov: and some evaluation order different from sl
[11:25] Arielle.Popstar @hg.osgrid.org: only if they load to a UBode physics
[11:25] Leighton.Marjoram @grid.opensim.life:8002: should be ok with a caveat some mesh might go wonky if its not well made
[11:26] Arielle.Popstar @hg.osgrid.org: it will break their meshes
[11:26] Bill Blight: no
[11:26] Andrew.Hellershanks @hg.osgrid.org: I've been wanting to dig in to the vector math bug and see if it can be fixed without a lot of work.
[11:26] Bill Blight: false statement
[11:26] Ubit Umarov: wo iil let that mantis stay on ice for now :)
[11:26] Leighton.Marjoram @grid.opensim.life:8002: no thats part of the problem Arielle some mesh will break
[11:26] Bill Blight: it will break SOME mesh, that was actually incorrect from the start
[11:26] Ubit Umarov: so...
[11:26] LaNani Sundara: so stick to BulletSim and it should be ok?
[11:26] Arielle.Popstar @hg.osgrid.org: it was correct for the time
[11:26] Leighton.Marjoram @grid.opensim.life:8002: Nods @ LaNani
[11:27] Ubit Umarov: grr i refuse to go back into that old mesh useless discution
[11:27] Bill Blight: ditto
[11:27] Arielle.Popstar @hg.osgrid.org: or if the use Meshmerizer instead of UbodeMeshmerizer
[11:27] LaNani Sundara: so thats the answer to your question Arielle?
[11:27] Bill Blight: BAD IDEA
[11:27] Ubit Umarov: documentation is avaiable on our site
[11:27] Bill Blight: but have fun
[11:27] Arielle.Popstar @hg.osgrid.org: works for me
[11:27] Ubit Umarov: that is plain Stupid arielle
[11:27] LaNani Sundara: lol
[11:27] Arielle.Popstar @hg.osgrid.org: but in any case the point is about what is needed to get grids to switch
[11:28] LaNani Sundara: now you've done it Arie ;p
[11:28] Arielle.Popstar @hg.osgrid.org: as i have many times in past
[11:28] Ubit Umarov: but it is your region.. do whatever you want.. just don't complain about issues
[11:28] Arielle.Popstar @hg.osgrid.org: there isnt any Ubit
[11:28] Bill Blight: People to quit telling people to use invalid configurations that will ultimately make stuff break in the future , would be a good start to get people to convert .
[11:28] LaNani Sundara: so sims and OAR no problem... for grid the Robust side is of course important
[11:29] Ubit Umarov: yeah guess you can't see the issues.. well its your problem
[11:29] Ubit Umarov: follow instructions, or ignore them.. it is your game.. just keep it yours then
[11:29] Arielle.Popstar @hg.osgrid.org: i am not talking about mesh issues except as a by and by
[11:29] LaNani Sundara: lets just assume they use BulletSim
[11:29] Sheera Khan: Arielle: and the database migration ...
[11:29] Bill Blight: there were minor changes to bulletsim in master, but nothing really huge
[11:30] Arielle.Popstar @hg.osgrid.org: my question is whether the devs should outreach to address those grid objections
[11:31] Bill Blight: Well it is obvious that most people want to hear what they want, and don't want to listen to the actual facts, case in point, " use Meshmerizer instead of UbodeMeshmerizer"
[11:31] LaNani Sundara: the question was, what is needed for a grid to migrate from .8 to .9.... sims seem no issue? so yes what about the databases and Robusts
[11:31] Leighton.Marjoram @grid.opensim.life:8002: *looks around* we are at a dev meeting they could come along to this.
[11:31] Andrew.Hellershanks @hg.osgrid.org: A wiki page, or just a paragraph or two, talking about the things to watch for when migrating would be useful.
[11:31] Arielle.Popstar @hg.osgrid.org: yes good point Andrew
[11:31] LaNani Sundara: *nods agreed Andrew
[11:31] Bill Blight: That too Leighton
[11:31] Ubit Umarov: instructions are clear, to use ubOde you need to use ubOdeMeshmerizer
[11:32] Ubit Umarov: if you ignore those, you are on your own.. have fun
[11:32] LaNani Sundara: not talking about that now Ubit :)
[11:32] Arielle.Popstar @hg.osgrid.org: you said you werent going to get into that |Ubit
[11:32] Ubit Umarov: just making it clear
[11:32] LaNani Sundara: ok :)
[11:32] Ubit Umarov: i never said that
[11:32] Bill Blight: Case in point, we are inside a totally mesh building here, ubODE with ubODEmesher, and it is good mesh and no issues
[11:32] Arielle.Popstar @hg.osgrid.org: so what can YOU do to help alay people's fears of moving over?
[11:33] Ubit Umarov: ubodeMeshmerizer does more things than normal one
[11:33] Bill Blight: if your mesh is busted to begin with it does not work
[11:33] Leighton.Marjoram @grid.opensim.life:8002: I dont think its the Devs job to convince people to upgrade.
[11:33] Ubit Umarov: for ubOde use
[11:33] LaNani Sundara: hmm
[11:33] Arielle.Popstar @hg.osgrid.org: i disagree Leighton
[11:33] Bill Blight: That is a valid point Leighton
[11:33] LaNani Sundara: then who's is it
[11:33] Arielle.Popstar @hg.osgrid.org: it is beneficial to have most on the same page
[11:34] Bill Blight: that is really odd coming from you Arielle
[11:34] Andrew.Hellershanks @hg.osgrid.org: Arielle, point out what issues (if any) they will encounter. We want to provide facts and dispel all the rumours and false impressions some people may have about the code.
[11:34] Bill Blight: considering you just stated you don't do what the devs recommend
[11:34] Leighton.Marjoram @grid.opensim.life:8002: I think the devs are already doing enough, there is plenty of information about upgrading, there are past dev meeting notes and good comments in the code.
[11:34] Arielle.Popstar @hg.osgrid.org: opensim is often about workarounds
[11:34] Leighton.Marjoram @grid.opensim.life:8002: It has been available for about 2-3 years now
[11:34] Bill Blight: yep
[11:35] Bill Blight: not like .9 just poppped in to existence 3 months ago
[11:35] Andrew.Hellershanks @hg.osgrid.org: Leighton, It would be worth having the information in a single place so people don't have to search meeting notes.
[11:35] Arielle.Popstar @hg.osgrid.org: yes@ Andrew
[11:35] Andrew.Hellershanks @hg.osgrid.org: It would also allow us to just point to that place if someone says they are worried about switching.
[11:35] Leighton.Marjoram @grid.opensim.life:8002: Shall I go and upgrade for them, we live in an information age if they cant find, sort and analyse information I am not sure I can do that for them
[11:36] Arielle.Popstar @hg.osgrid.org: decreasing complexity is always good
[11:36] Leighton.Marjoram @grid.opensim.life:8002: but I do agree that making it easier might help
[11:36] LaNani Sundara: hmm... just having it on opensimulator.org does not mean people will upgrade, does this group encourage grids to upgrade? is it a goal?
[11:36] Leighton.Marjoram @grid.opensim.life:8002: its their choice I assume
[11:36] Arielle.Popstar @hg.osgrid.org: as long as a region stays on bulletsim, everything should be fine
[11:37] Bill Blight: .9x is NOT any harder to setup than previous versions of opensim, IF you read and use the new config files adjust for your setup as the wiki already states, I'm not sure where the issues are
[11:37] LaNani Sundara: of course its a choice
[11:37] Bill Blight: bullet is still the default physics engine
[11:37] Andrew.Hellershanks @hg.osgrid.org: It is up to grid owners whether they upgrade or not. If they don't they won't get the benefits of new features and/or bug fixes in the latest code.
[11:37] Bill Blight: so that would not cause issues
[11:37] Arielle.Popstar @hg.osgrid.org: maybe a page that states ther benefits of an upgrade?
[11:38] Andrew.Hellershanks @hg.osgrid.org: For everyone that have HG enabled grids, if there was ever a change in the code that affect HG TP's they could find themselves having problems with people going to, or coming from, other grids.
[11:38] LaNani Sundara: Bill thats not the issue i think, i do not know what the objectives are here, of the "board" of opensimulator.org ... is the new version promoted, or not?
[11:38] Sheera Khan: Andrew, that reminds me of that project: "ChangeLog" ...
[11:38] Andrew.Hellershanks @hg.osgrid.org: :)
[11:38] Bill Blight: same as this one pretty much http://opensimulator.org/wiki/0.9.0.0_Release
[11:38] LaNani Sundara: if this is just the dev meeting then, fine
[11:39] Bill Blight: http://opensimulator.org/wiki/0.9.0.1_Release
[11:39] Gavin.Hird @grid.xmir.org:8002: I suspect it is often a combination of lack of resources both manpower and kit to do a successful migration, in combination with own code that must be ported, and the fear of losing face and users if something goes wrong.
[11:39] Arielle.Popstar @hg.osgrid.org: nods @ Gavin
[11:40] Andrew.Hellershanks @hg.osgrid.org: The release notes were some of the basic information about the changes. There is a lot of items not mentioned in the 0.9 release note.
[11:40] Sheera Khan aggrees with Gavin
[11:40] Gavin.Hird @grid.xmir.org:8002: If you have paying customers, you don't just upgrade
[11:40] Gavin.Hird @grid.xmir.org:8002: like that I mean
[11:40] Bill Blight: This is not really any different than the "Sky is falling" moments from .6.9 up .. each major release has had it's share of fearmongering
[11:40] Arielle.Popstar @hg.osgrid.org: is there any regressions left between .8.2 and .9?
[11:40] LaNani Sundara: so everyone here is fine with people running their 0.8 grids and never updating, i am
[11:40] Arielle.Popstar @hg.osgrid.org: easy for you Bill having a smaller grid
[11:41] Ubit Umarov: 0.9x has no regressions has evolution
[11:41] Ubit Umarov: :p
[11:41] Bill Blight: pffft
[11:41] Leighton.Marjoram @grid.opensim.life:8002: Easier for Bill because he understands the code base
[11:41] Arielle.Popstar @hg.osgrid.org: bugs are now features you mean?
[11:41] Bill Blight: Well not scared of it anyway
[11:41] Arielle.Popstar @hg.osgrid.org: that too Leighton
[11:41] LaNani Sundara: i understand it too thats why i am sticking with my own ;p
[11:41] Bill Blight: LOL Arielle
[11:42] Bill Blight: Bugs have been features since Opensimulator was made
[11:42] Arielle.Popstar @hg.osgrid.org: yes but specifically what are potential regressions
[11:42] Arielle.Popstar @hg.osgrid.org: what is expected behavious that is not currently avalable in .9
[11:42] Bill Blight: Sit target offset, the ENTIRE BULLETSIM Mesh rendering using visual mesh as physics shape is technically a bug, because NOBODY in the 3d game industry does that ..
[11:43] Ubit Umarov: yes Arielle your settings for ubOde show how good you are doing debug... whatever
[11:43] Arielle.Popstar @hg.osgrid.org: it works for me....i know you hate that but thats the way it is
[11:44] Arielle.Popstar @hg.osgrid.org: if there was an issue, i am sure Hyacinth would have told me
[11:44] Gavin.Hird @grid.xmir.org:8002: the classic message: "An Error occurred becaue an Error occured!"
[11:44] Ubit Umarov: ooh now you blame Hyacinth also?
[11:45] Arielle.Popstar @hg.osgrid.org: no, i credit her as being knowledgable about the debugging process
[11:45] Arielle.Popstar @hg.osgrid.org: i blame you for not being able to present to me a valid instance where my config would NOT work
[11:46] Aiden Frimon: I feel that some grid owners are fairly happy with the way .8 works. They don't want to upgrade as they see issues with .9 but that if often the case with upgrades until everything is smoothed out.
[11:46] Ubit Umarov: sorry i gave up explaining things to you
[11:46] Arielle.Popstar @hg.osgrid.org: true too Aiden
[11:46] Ubit Umarov: why should i? you just ignore it
[11:46] Arielle.Popstar @hg.osgrid.org: you never started Ubit
[11:46] Arielle.Popstar @hg.osgrid.org: its always JUst because i said so
[11:47] Ubit Umarov: yeap bad memory is nasty.. whatever
[11:47] Arielle.Popstar @hg.osgrid.org: i have logs of or conversations
[11:47] Ubit Umarov: a tip.. test CONVEX on prims...
[11:47] Bill Blight: you don't put the wrong parts on the wrong car, even though they might work for a while , they are still the wrong parts, why should someone have to tell you it is wrong when you already know it ..
[11:48] Bill Blight: Has she told you that is a good idea, or just patted you on the head and said move along? Because doing it that way is really dumb, because regard less if your broken mesh is working , it is not meshing good mesh correctly it is just letting you keep using your broken mesh ..
[11:48] Ubit Umarov: if you do know what CONVEX on prims means, of course
[11:48] Arielle.Popstar @hg.osgrid.org: Bill, i am not a graphical artist and dont know how to fix a mesh building that doesnt have a physics map
[11:48] Ubit Umarov: then USE BULLET
[11:48] Arielle.Popstar @hg.osgrid.org: on the other hand, i prefer using Ubode physics
[11:49] Ubit Umarov: or replace the object
[11:49] Arielle.Popstar @hg.osgrid.org: so my config allows both
[11:49] Ubit Umarov: those are the specs.. stay on them
[11:49] Bill Blight: Not really
[11:49] LaNani Sundara: the convex issue will older "bad"mesh is why people stick with BulletSim, the known "can not enter a door" .
[11:49] Arielle.Popstar @hg.osgrid.org: yes Nani
[11:49] LaNani Sundara: for example
[11:49] Arielle.Popstar @hg.osgrid.org: 4 lines of code
[11:49] Arielle.Popstar @hg.osgrid.org: as per Ubit
[11:49] LaNani Sundara: 4 lines of code?
[11:49] Arielle.Popstar @hg.osgrid.org: mhmm
[11:49] LaNani Sundara: well .. add them?
[11:50] LaNani Sundara: lol
[11:50] Arielle.Popstar @hg.osgrid.org: he wont because h e demands perfection damn the cost
[11:50] Ubit Umarov: well i said i would not go back to that issue and already did 2 much
[11:50] Bill Blight: The cost is busted ass mesh
[11:50] Bill Blight: not working
[11:50] Arielle.Popstar @hg.osgrid.org: anyways it isnt my point
[11:50] Ubit Umarov: do whatever you want on yr regions.. just do not complain...
[11:50] Bill Blight: If it's broke, don't want it anyway
[11:50] Arielle.Popstar @hg.osgrid.org: i am not
[11:50] Arielle.Popstar @hg.osgrid.org: \
[11:51] Arielle.Popstar @hg.osgrid.org: you are
[11:51] LaNani Sundara: i think the meeting has devolved into arguments :)
[11:51] Arielle.Popstar @hg.osgrid.org: it has
[11:51] Andrew.Hellershanks @hg.osgrid.org: Time for a new topic before I have to send those two in to different corners of the bar for a time out.
[11:51] Arielle.Popstar @hg.osgrid.org: i posit an outreach of devs to help bring people on board with .9
[11:51] Bill Blight: yes but it is an argument on how NOT to implement .9x and how some people are publicly telling people how to do it WRONG
[11:52] Arielle.Popstar @hg.osgrid.org: imagine what would have ensued had i let him drink
[11:52] Andrew.Hellershanks @hg.osgrid.org: That goes back to my earlier suggestion. Have a page that talks about migrating from 0.8 to 0.9. Tell people what they need to do, what problems they may encounter, etc.
[11:53] Gavin.Hird @grid.xmir.org:8002: teleport one to height 10000 m at one corner of the sim and the other to height -200 at the opposite corner
[11:53] Arielle.Popstar @hg.osgrid.org: :)
[11:53] Andrew.Hellershanks @hg.osgrid.org: Gavin, :)
[11:54] Andrew.Hellershanks @hg.osgrid.org: Dealing with sit targets and bad mesh are two of the topics for such a page.
[11:54] LaNani Sundara: good suggestion Andrew
[11:54] Arielle.Popstar @hg.osgrid.org: ahh
[11:54] Arielle Popstar: good thing for me to look at
[11:55] Andrew Hellershanks: If there is incorrect information or rumours out there about 0.9 we need to dispel them with facts.
[11:55] Arielle Popstar: first potential negative use case scenerio i have heard
[11:55] Bill Blight: Sounds like a lot of work for something that 90% will not read or ignore anyway, as the evidence of record indicates ... but worth a try I guess ..
[11:55] Arielle Popstar: its whether the grid owners/managers read it
[11:56] Arielle Popstar: 20 owners would can swing the 60% on .8.2 over
[11:56] Andrew Hellershanks: Bill, I know we can never make them read something but if someone says they don't want to migrate to 0.9 we can ask why or just point to the page so they know what to expect and possibly correct some bit of misinformation about the code they may have gotten from someone.
[11:56] Bill Blight: This is no different than people who don't want to upgrade to Windows 10 or whatever, not gonna make them regardless of the info that is available , until they want to or HAVE to ..
[11:56] LaNani Sundara: hmm... with that attitude one could say 0.9 was a lot work when 90% of old grids will stick to 0.8
[11:57] Andrew Hellershanks: Bill, it shouldn't be a lot of work unless there is a lot of issues someone may have to deal with in making the switch. I don't think there are that many issues.
[11:57] Bill Blight: Exactly La Nani
[11:57] Arielle Popstar: there was some bad feeling about how master was forked over to avination code Lanani
[11:57] Bill Blight: but the fact is they did not do the work, so what are they crying about .. HAHA
[11:58] Arielle Popstar: but after 3+ years its time to move on
[11:58] Arielle Popstar: or continue on with developing .8.2
[11:58] Andrew Hellershanks: Arielle, it wasn't forked over to Avination code. It picked up a lot of changes to code that grid had made.
[11:58] LaNani Sundara: its open source :) so let them all do what they want ;P
[11:58] Bill Blight: Exactly LaNani
[11:59] Bill Blight: they can use whatever version they want until it is no longer compatible or secure
[11:59] Arielle Popstar: how many grids on current master?
[11:59] Bill Blight: .9x with newer .NET already has security improvements
[11:59] Gavin.Hird @grid.xmir.org:8002: hopefully not many
[11:59] Gavin.Hird @grid.xmir.org:8002: grids are not supposed to produce on dev code
[12:00] Bill Blight: I"d agree with that Gavin, we need an official release , just to quite the roar
[12:00] Arielle Popstar: will it make the upgade path more extreme?
[12:05] Bill Blight: I fully understand the pain of the larger grids, I really do, but .9x has been around for over 2 years now, not testing before now, and getting a plan together not in the eleventh hour is not a problem for the core devs ..
[12:05] Andrew Hellershanks: I was about to make note of a third issue re: migrating but I forgot what it was before I could write it down. :P
[12:06] Arielle Popstar: i suspect Bill that as aiden said, a lot are sort of ok with it
[12:06] Bill Blight: if these grids are so LARGE and so well run, that they don't have dev servers and testing servers, and a test grid setup for code changes, then I question how well they are actually run ..
[12:07] Gavin.Hird @grid.xmir.org:8002: many are probably run on a shoestring
[12:07] Arielle Popstar: maybe benefits of .9 vs .8.2 have not been stated well enough
[12:07] Gavin.Hird @grid.xmir.org:8002: even securing a good set of database backups before migration could be a real issue for most
[12:08] Arielle Popstar: especially with asset db sizes
[12:08] Bill Blight: They have been stated but certain Loud voices in the community over shadow the benefits , as you know a bad word goes twice as far as a good one ..
[12:08] Arielle Popstar: that was a while ago now Bill
[12:08] Bill Blight: It still goes on every day
[12:09] Arielle Popstar: but tbh, i dont know what the real benefits are except parity with .8.2 and less clouuds
[12:09] Arielle Popstar: less stuttering too
[12:09] Andrew Hellershanks: I thought of something else which is item #4 (apparent memory leak in Robust) as it wasn't what I had been about to write down.
[12:10] Bill Blight: that memory leak I don't think is opensimulator specific
[12:10] Bill Blight: as it only seems to hit certain config setups
[12:10] Andrew Hellershanks: Bill, it is of some concern when we don't know what causes the issue or how to fix it.
[12:11] Bill Blight: I see it on my emby server which is a mono app on my debian server
[12:11] Bill Blight: but after updating mono and kernel on the server it stopped
[12:12] Andrew Hellershanks: Bill, which mono are you using?
[12:12] Arielle Popstar: whats this new viewer i heard a rumout about?
[12:12] Andrew Hellershanks: and which kernel?
[12:12] Bill Blight: between 5.12 and 5.14
[12:12] Bill Blight: I'd have to check my debian server
[12:12] Andrew Hellershanks: ok. I'm using mono 5.12
[12:12] Bill Blight: but I can check my robust
[12:12] Andrew Hellershanks: Arielle, it is just an ugly rumour.  ;)
[12:13] Aiden Frimon: New viewer is still a rumor, I suppose
[12:13] Andrew Hellershanks: hm... No idea, Arielle. I haven't heard anything recently about a new viewer.
[12:13] Bill Blight: 4.15.0-13-generic
[12:13] Bill Blight: on my robust
[12:13] LaNani Sundara: i has some people who complained about unstable sims, suggested updating to newer mono and it fixed it
[12:13] Andrew Hellershanks: Bill, ty
[12:13] Bill Blight: Time now is 8/21/2018 12:13:46 PM
Server has been running since Sunday, 8/5/2018 3:33:07 PM
That is an elapsed time of 15.20:40:39.2088240
[12:14] Bill Blight: using barely 4 gig of ram
[12:16] Andrew Hellershanks: Time to call the meeting as we just lost about half those who were here.
[12:17] Andrew Hellershanks: Any last comments before we close todays meeting?

Personal tools
General
About This Wiki