Chat log from the meeting on 2019-11-05

From OpenSimulator

Jump to: navigation, search

[10:45] Ubit Umarov: so adding BoM to dayturn gavin?
[10:45] Ubit Umarov hides
[10:46] Gavin.Hird It will hit you with a big BOM one of these days!
[10:46] Ubit Umarov: :)
[10:46] Gavin.Hird apart from supporting copybotters I frankly don't see much urgency for it right now
[10:46] Gavin.Hird but I might be wrong
[10:47] Melanie Milland: no idea what you're talking about
[10:47] Bill.Blight Why not just use Lanani's code people seem to think she has BoM without viewer support
[10:47] Ubit Umarov: bet sacrarium will start to fill with BoM things
[10:47] Ubit Umarov: gavin hit Bill with your BOM thing
[10:47] Gavin.Hird has anyone looked at her changes?
[10:47] Bill.Blight yeah, yeah
[10:48] Melanie Milland: whatever BoM is
[10:48] Bill.Blight Bake On Mesh
[10:48] Melanie Milland: ahja
[10:48] Ubit Umarov: i stopped looking whe she removed lludp throttles
[10:48] Melanie Milland: see i wasn't in here for so long
[10:48] Melanie Milland: i'm missing stuff
[10:48] Gavin.Hird it is a spillover from SL
[10:48] Melanie Milland: yes
[10:48] Gavin.Hird where it is the latest fad
[10:49] Melanie Milland: i had a version of my own that's still in the codebase
[10:49] Bill.Blight Her code Applies base textures from the system level, it does not bake them as BoM does so still individual texture calls, twice as many actually
[10:49] Bill.Blight Lanani that is
[10:49] Melanie Milland: the advantage of the code i have is that it works pretty much transparently
[10:49] Ubit Umarov: yeah ll BoM is a bit more complex
[10:49] Melanie Milland: so you just wear stuff normally
[10:49] Gavin.Hird that would be ideal Melanie
[10:50] Bill.Blight BoM takes 11 texture calls and makes it one ..
[10:50] Bill.Blight (I think it is 11)
[10:50] Ubit Umarov: No
[10:50] Ubit Umarov: it makes up to 11 :p
[10:50] Ubit Umarov: from up to 65 i think
[10:50] Bill.Blight yeah had that backwards
[10:51] Bill.Blight needs more coffee
[10:51] Ubit Umarov: it just went back to classic wearables
[10:51] Ubit Umarov: each can define a texture for a avatar part
[10:52] Ubit Umarov: baking code cooks all them per each bake part
[10:52] Bill.Blight so ends up one per layer, right, I think that is where my brain got confused , I think anyway
[10:52] Ubit Umarov: and only one is distributed
[10:52] Ubit Umarov: upper body, head.. etc
[10:52] Ubit Umarov: up to 11 of those parts
[10:52] Gavin.Hird if anyone can point me to a video explaining BoM
[10:52] Ubit Umarov: or bakes
[10:52] Ubit Umarov: use google :p
[10:53] Gavin.Hird google does not get anywhere near my machines
[10:53] Ubit Umarov: well with normal wearables it is still at most 5 or 6 bakes
[10:53] Bill.Blight
[10:53] Gavin.Hird except for YT
[10:53] Ubit Umarov: that works fine even on older versions of opensim
[10:54] Gavin.Hird Apr 9, 2018
[10:54] Gavin.Hird this video is terrible
[10:54] Ubit Umarov: but they added a new wearable called Universal..
[10:54] Ubit Umarov: that can define more baked textures
[10:54] Ubit Umarov: up to 11 now
[10:54] Bill.Blight Then find one yourself , geeze
[10:55] Gavin.Hird a video which is made with the released version and not being a commercial for all kinds of products
[10:55] Ubit Umarov: and viewers do send them on a pack with 45 texture entires now
[10:55] Ubit Umarov: old opensim only supports 32
[10:56] Ubit Umarov: ok made all fall asleep again?
[10:56] Melanie Milland: zzzzzZZZZZ
[10:56] Sheera Khan: nope ;-)
[10:56] Ubit Umarov: :)
[10:57] Ubit Umarov: SL goal is to reduce the need for those crazy onion mesh avatars
[10:57] Ubit Umarov: etc
[10:57] Gavin.Hird I get that and agree
[10:58] Ubit Umarov: also no need for alpha layers
[10:58] Ubit Umarov: viewers do them automaticly
[10:58] Ubit Umarov: well some
[10:59] Gavin.Hird What's new at Kitely?
[11:00] Ubit Umarov: im wearing a BoM mesh
[11:00] Ubit Umarov: im so pretty no?
[11:00] Selby.Evans They hope to be able to publicize efucational uses at Kitely soo
[11:01] Selby.Evans soon
[11:01] Ubit Umarov: thats the thing on ground
[11:01] Gavin.Hird What does that mean Selby?
[11:01] Andrew.Hellershanks blinks
[11:01] Leighton.Marjoram Big waves to everyone
[11:02] Andrew.Hellershanks Hello, Melanie. Quite a surprise to see you here. Welcome.
[11:02] Selby.Evans re SL: I can now report this morning that Linden Lab has laid off staff as part of what appears to be a significant reorganization and reallocation of staff resources. My source (who shall remain anonymous, but whom I consider a good source) tells me that 30 people were let go. The rumour mill this weekend has been saying that 20 to 30 people were laid off, so the figure would appear to be accurate.
[11:02] Kayaker Magic: Waves and Beaches!
[11:02] Gavin.Hird Ubit looks like he's got a major sunburn in his face
[11:02] Ubit Umarov: oh?
[11:02] Gavin.Hird Sansar staff?
[11:02] Andrew.Hellershanks You need a newer viewer program.
[11:03] Ubit Umarov: we knew they did hired rendering devs for sl
[11:03] Selby.Evans yes - sansar but more:
[11:03] Ubit Umarov: one senior
[11:03] Selby.Evans I heard a rumour Linden Lab got bought, about two/three weeks ago; I don’t fully trust the source (it was definitely a rumour), but that’d explain it.
[11:04] Gavin.Hird I wonder who would pick up that wreck
[11:04] Sheera Khan:
[11:04] Gavin.Hird thanks for the link
[11:04] Selby.Evans It could have been for sale cheap.
[11:04] Sheera Khan: And Nyx Linden seems to have vanished...
[11:07] Andrew.Hellershanks The LL people that got laid off worked on SL or their other new offering?
[11:07] Selby.Evans I hope Firestorm people read the handwriting on the wall
[11:08] Selby.Evans The LL peopl worked on Sansar -- That might be cosidered the new offering
[11:08] Kayaker Magic: OK, let's talk about OpenSim! What do you all think about my patch to add osPermissionToCall to the OSSL functions? (mantis 8618)
[11:08] Ubit Umarov: ( seems gavin is using Dayturn ..)
[11:08] Ubit Umarov hides
[11:08] Gavin.Hird he does
[11:08] Ubit Umarov: what made it crash? :p
[11:08] Gavin.Hird there is some strange interaction between safari 13 and the viewer which kills the connection
[11:09] Ubit Umarov: ok
[11:09] Gavin.Hird actually it is not strange at all because the SL plugin is a particularly hairy piece of software
[11:09] Ubit Umarov: well about LL my concerns are will show up on viewers
[11:10] Ubit Umarov: making things even harder
[11:10] Gavin.Hird how so?
[11:10] Ubit Umarov: or things deleted from viewers
[11:10] Ubit Umarov: the usual problem gavin.Hird
[11:10] Ubit Umarov: like the EEP things
[11:11] Gavin.Hird you know my biggest concern is they write a new renderer and close source it
[11:11] Ubit Umarov: that may happen also
[11:12] Ubit Umarov: and if owner of LL changed, all policies may change
[11:12] Ubit Umarov: lets see
[11:12] Melanie Milland: the time is long past to divorce from the sl viewers
[11:12] Andrew.Hellershanks Kayaker, the patch is simple enough. One question I have is whether that is the best name to use for the function.
[11:12] Bill.Blight I like osPerm2Use
[11:12] Bill.Blight much shorter
[11:12] Melanie Milland: i'm not convinced of that name
[11:13] Melanie Milland: although it fits in with lllist2List etc
[11:13] Ubit Umarov hits Bill with a large BOM
[11:13] Andrew.Hellershanks Bill, or something like IsAllowed
[11:13] Gavin.Hird someone should make a particle BOM thingy
[11:14] Andrew.Hellershanks Bill Of Materials? (sorry, been doing too much engineering work lately)  :)
[11:14] Melanie Milland: yeah that was the first thing that came to my mind
[11:14] Melanie Milland: but the talk didn't fit in twith that meaning
[11:14] Bill.Blight pats his stack of materials that he is of ...
[11:14] Bill.Blight :P
[11:15] Arielle.Popstar BoM has a bom
[11:15] Gavin.Hird it is sort a bill or materials if you think about it
[11:15] Ubit Umarov: i just used the BOM gavin tried to throw at me :)
[11:15] Andrew.Hellershanks Gavin, true
[11:15] Melanie Milland: BOM, EBOM, MBOM.....
[11:15] Andrew.Hellershanks hm... I don't know those last two.
[11:16] Melanie Milland: Engineering BOM and Manufacturing BOM
[11:17] Melanie Milland: the EBOM comes from CAD data
[11:17] Kayaker Magic: Well, pick another name for llPermissionToCall, I'm agnostic about that, I'd like to see the functionality by any name.
[11:17] Melanie Milland: whereas the MBOM comes from apps like ProcessSimulate
[11:17] Andrew.Hellershanks Melanie, ok. ty
[11:18] Gavin.Hird but you also have BOMB
[11:18] Gavin.Hird which goes BOM
[11:18] Ubit Umarov: but ok... so Andrew what's new on opensim?
[11:19] Arielle.Popstar where do we get your new BoM viewer Bill?
[11:19] Andrew.Hellershanks Ubit, It would be better to ask the person who is doing most of the coding.  ;)
[11:19] Kayaker Magic: OK, moving on: I wanted to have Gavin and Ubit talk about why YEngine does not work on ARM processors.
[11:19] Bill.Blight click the dancing BoM guy out front
[11:19] Gavin.Hird I have no idea kayaker because I have not tried it
[11:20] Kayaker Magic: I tried running it again and saving the log, YEngine just stops responging and OpenSim never finishes coming up.
[11:20] Bill.Blight the one with the text over it's head that says to download the viewer
[11:20] Arielle.Popstar ok ty
[11:20] Ubit Umarov: and i will break my droid phone :p
[11:20] Kayaker Magic: I see nothing in the log except "setting YEngine to default"
[11:20] Gavin.Hird I was acutally planning to try make it work
[11:20] Kayaker Magic: What needs to be done?
[11:21] Ubit Umarov: well not much i can do with that info :(
[11:21] Arielle.Popstar hgluv been having issues and wondering if its with the viewer cache code or people visiting with the BoM viewer
[11:21] Gavin.Hird I have no idea right now
[11:21] Andrew.Hellershanks There have been a number of miscellaneous bug fixes to OpenSim in the last week and several fixes related to YEngine.
[11:22] Ubit Umarov: guess that ARM issue will take some deep debuging
[11:22] Ubit Umarov: sure it is Y ?
[11:22] Andrew.Hellershanks Is there a downloadable version of mono for ARM or did you have to compile mono for the ARM based system?
[11:22] Gavin.Hird flüstert: did you just take the 0.9.1 release and substitute the libs and then run it?
[11:22] Gavin.Hird apt install mono
[11:23] Gavin.Hird installs mono 6.4.x
[11:23] Gavin.Hird at least on rasbian which is Debian
[11:23] Andrew.Hellershanks ok, good. Saves having to compile mono for ARM. I could give it a try on my Pi 4 if I can find some time.
[11:23] Kayaker Magic: I apt-get'ed the mono that was supposed to be for my version of Linux.
[11:23] Kayaker Magic: mono was able to compile a recent version of OpenSim and run it, with libraries from Gavin.
[11:24] Ubit Umarov: many distros use old mono versions.. or used...
[11:24] Gavin.Hird good
[11:24] Ubit Umarov: opensim needs at least 5x as you know
[11:24] Kayaker Magic: Oh, my mono is 5.18
[11:24] Ubit Umarov: should work then
[11:24] Andrew.Hellershanks I'm ok. I have 6.4.0
[11:24] Gavin.Hird I intend to test run the 0.9.1 release on the Pi, but till I have done I'm blank
[11:25] Kayaker Magic: Yeah, I have 6.4 on my x86 system
[11:25] Ubit Umarov: about code state
[11:26] Ubit Umarov: yes a made a few bug fixes about issues ppl reported after release ( of course )
[11:26] Ubit Umarov: expect more :p
[11:26] Ubit Umarov: have change that will make me get killed
[11:26] Andrew.Hellershanks It wouldn't be the first time.
[11:27] Ubit Umarov: ie to limit the max number of a attachments to 38, as viewer limit
[11:27] Arielle.Popstar This Yengine is not hypergrid compatible?
[11:27] Ubit Umarov: but allow all on same point
[11:27] Ubit Umarov: as viewers expect
[11:27] Kayaker Magic: I have no problems with YEngine and HG on x86 processors.
[11:28] Melanie Milland: script engines and HG are not related
[11:28] Andrew.Hellershanks Here is a question for you about YEngine. What result do you get if you do "if (some_key_var)" where some_key_var is set to NULL_KEY?
[11:28] Bill.Blight Yenging is perfectly HG compatible ... Just if you HG from Y to X or X to Y , you should expect strange behavior ..
[11:28] Arielle.Popstar as long as the destination region also defaulting to Yengine?
[11:28] Melanie Milland: you get false
[11:28] Ubit Umarov: yeah it does that as SL does ( not as X does )
[11:28] Andrew.Hellershanks Melanie, ok. That wasn't handled properly in XEngine.
[11:29] Bill.Blight flüstert: Y to Y works perfectly fine local or HG
[11:29] Ubit Umarov: on X you need to compare explicity
[11:29] Arielle.Popstar can it mess with the destination grid if mismatched then?
[11:29] Melanie Milland: the script engine and HG relate only in the way that you have to program your script in a way that sidesteps the differences
[11:29] Andrew.Hellershanks I have a patch for that for XEngine but never pushed that to master because it was considered controversial at the time.
[11:29] Melanie Milland: else the destination grid will simply not be able to compile the script
[11:30] Arielle.Popstar but won't break anything on the region itself like BoM?
[11:30] Melanie Milland: no
[11:30] Arielle.Popstar ok
[11:30] Melanie Milland: the script is recompiled on TP
[11:31] Melanie Milland: so it can only do what the destination engine allows
[11:31] Ubit Umarov: good X code should compile ok on Y
[11:31] Kayaker Magic: It is possible for a bad programmer to write FORTRAN in any language....
[11:31] Andrew.Hellershanks True but there is a lot of bad X code out there.
[11:31] Bill.Blight Only issue is script states, X to Y, mostly ok, Y to X , state lost
[11:31] Andrew.Hellershanks Most people wouldn't know if the X code is good or bad if it appears to be working.
[11:32] Bill.Blight yeap
[11:32] Bill.Blight LOL
[11:32] Ubit Umarov: well will learn the hard way :p
[11:32] Melanie Milland: most X scripts expect state loss
[11:32] Andrew.Hellershanks nods
[11:32] Arielle.Popstar is there a way where something like an ao can restart after hg jumping to another grid without having to rewear the ao?
[11:33] Melanie Milland: on arrival, the script gets recompiled and reset
[11:33] Bill.Blight (that's what they made viewer AO's for, to eliminate that problem)
[11:33] Gavin.Hird use the viewer ao
[11:33] Melanie Milland: it should init and run then
[11:33] Arielle.Popstar i dont like the viewer ao's because often the walks dont work right
[11:33] Gavin.Hird when FS release their next version there will be a number of improvements to viewer ao behavior
[11:34] Arielle.Popstar especially in FS
[11:34] Kayaker Magic: Other HUDs besides AOs also often fail on a region crossing, TP or HG jump.
[11:34] Arielle.Popstar that too Kay
[11:34] Melanie Milland: i don't know why everyone loves FS so much
[11:34] Bill.Blight Most of those fixes are already in my test viewer
[11:34] Gavin.Hird yeah
[11:34] Kayaker Magic: It is a race condition: Sometimes the script is not compiled in the destination region.
[11:34] Gavin.Hird but you can't release till they have Bill
[11:35] Ubit Umarov: well and about HG and TPs we can only talk about both regions on 0.91
[11:35] Bill Blight: lol, LGPL license does not prohibit that
[11:35] Bill Blight: they can say that all they want
[11:35] Melanie Milland: and i wish the viewer codebase that I donated and diva rewrote to c# would finally see use
[11:35] Ubit Umarov: to/from older versions.. well..
[11:35] Bill Blight: but per the license they can't stop you
[11:35] Arielle.Popstar i thought it couldnt be used?
[11:35] Arielle.Popstar @Mel
[11:35] Gavin.Hird true Bill, but be nice
[11:35] Andrew Hellershanks: I thought viewer AOs use animations available in ones inventory so use of viewer AO shouldn't not mess up a walk.
[11:35] Bill Blight: but my viewer is only for testing, currently
[11:35] Melanie Milland: diva rewrote it to c#/xenko
[11:36] Gavin.Hird right
[11:36] Melanie Milland: no idea how far it got
[11:36] Gavin.Hird I said release
[11:36] Gavin.Hird there is a difference ;-)
[11:36] Arielle.Popstar has Diva been around?
[11:36] Melanie Milland: i've skyped with her
[11:36] Melanie Milland: but she's been scarce in irc
[11:36] Andrew Hellershanks: I haven't seen her in IRC in quite a while.
[11:37] Arielle.Popstar
[11:38] Gavin.Hird not touching it with a pitch fork
[11:39] Arielle.Popstar nothing new on it in 11 months
[11:39] Arielle.Popstar 10
[11:39] Arielle.Popstar is there a core branch with the additional modules?
[11:41] Melanie Milland: why are you saying you won't touch it?
[11:41] Gavin.Hird it does not run on macOS or iOS ever
[11:41] Melanie Milland: not true
[11:41] Melanie Milland: xenki is multiplatform
[11:41] Gavin.Hird no
[11:41] Arielle.Popstar does it work on android?
[11:42] Gavin.Hird last time I checked no macOS
[11:42] Melanie Milland: it doesn't target devices
[11:42] Melanie Milland: but runs on all three PC platforms
[11:42] Andrew Hellershanks: It uses xenko now? I knew there was talk of basing it on something else than Unity(IIRC).
[11:42] Melanie Milland: at least i think I saw mac support in the source
[11:42] Ubit Umarov: ( Xenko )
[11:43] Gavin.Hird the reaosn why LL has to lay off people is because they hace completely failed to address the mobile market
[11:43] Gavin.Hird which is where all the growth is
[11:43] Bill Blight: from their faq, "Xenko currently supports Windows, Android, iOS, Xbox One and Universal Windows Platforms (UWP). "
[11:43] Melanie Milland: mobile devices are still not powerful enough for a full experience
[11:43] Gavin.Hird bull
[11:43] Melanie Milland: i know it runs on linux. i use it there
[11:44] Gavin.Hird how powerful was the device in 2010 that provided the full experience?
[11:44] Melanie Milland: you have to cross-compile on windows, though
[11:44] Arielle.Popstar well Lumiya is broken for core too
[11:44] Bill Blight: "Xenko also supports most popular VR headsets and has experimental support for Linux."
[11:44] Bill Blight: forgot to post te second half
[11:44] Melanie Milland: ok, so no mac
[11:44] Arielle.Popstar so opensim doesnt support mobile any longer either
[11:44] Andrew Hellershanks: experimental? hm... Probably leaves me out as it did before.
[11:44] Melanie Milland: i don't use it on mac i just thought i saw something in the code there
[11:45] Bill Blight: I disagree for an "opensim" experience, considering te Oculus Quest is just a mobile android device, the horsepower is there
[11:45] Ubit Umarov: not much we can do about lumiya when its dev stopped working on it
[11:45] Bill Blight: just a matter of getting proper access to it.
[11:45] Arielle.Popstar well think the prob is server side
[11:45] Melanie Milland: the oculus is much more powerful than a mobile phone
[11:45] Arielle.Popstar so could be made to work again
[11:45] Gavin.Hird server side clearly does not scale for a mobile audience
[11:46] Andrew Hellershanks: The Xenko site doesn't make it easy to find the system requirements needed to be able use it.
[11:46] Bill Blight: Display panel: OLED.
Display resolution: 1440 x 1600 per eye (Oculus Rift had 1080×1200 per eye)
72Hz refresh rate.
Qualcomm Snapdragon 835 processor.
Lithium-ion battery with 2-3 hours playtime, depending on what you are playing.
[11:46] Bill Blight: pretty much same as my LG phone
[11:46] Bill Blight: LOL
[11:46] Ubit Umarov: we could work with lumiya dev if there was one..
[11:46] Melanie Milland: then i guess you have a modern phone
[11:46] Melanie Milland: not everyone does
[11:47] Gavin.Hird there is plenty of power on handsets and tablets to provide a - maybe not stellar, but good experience
[11:47] Melanie Milland: a 2 year old device just coming out of contract would be too slow i think
[11:47] Gavin.Hird by the time you have it developed you have another 2 generations of handset in the market
[11:47] Arielle.Popstar i've used the Lumiya VR mode and found it worked ok
[11:47] Ubit Umarov: no going to do blind server side changes with no idea what the issue is
[11:48] Kayaker Magic: Lumiya seems to be abandoned, I cannot get a copy for my new cellphone.
[11:48] Arielle.Popstar well i get the crash issue to send to the dev but not figured out how to divert it to you
[11:48] Melanie Milland: i fail to see the point on a phone
[11:48] Melanie Milland: tablets, sure
[11:49] Melanie Milland: but phone screens are a bit small
[11:49] Gavin.Hird phone act as a gaming device sending output to a TV
[11:49] Arielle.Popstar
[11:49] Melanie Milland: at least at my age, tiny stuff becomes indescernible
[11:50] Melanie Milland: anyway, i'm ducking out early
[11:50] Leighton.Marjoram Phone and tablets are just connection points between people to interact with. Big complex worlds with lots of detail arent that important. Think communication and interaction.
[11:50] Andrew Hellershanks: ok, Melanie. Thanks for coming. See you again another time.
[11:50] Melanie Milland: so maybe it's not the final choice
[11:50] Arielle.Popstar use the phone headsets, then it becomes big
[11:50] Melanie Milland: but now it's DINNERTIME!
[11:50] Gavin.Hird Agreed Leighton
[11:50] Arielle.Popstar $10 nowadays
[11:50] Gavin.Hird bon apetit
[11:51] Andrew Hellershanks: Enjoy
[11:51] Ubit Umarov: :)
[11:52] Arielle.Popstar newer version Kay
[11:52] Arielle.Popstar 3.4.2 is the newest
[11:52] Ubit Umarov: well seems a viewer declared to safely suport opensim BoM
[11:53] Ubit Umarov:
[11:54] Kayaker Magic: Do I have time to bring up a 0.9.1 issue?
[11:54] Andrew Hellershanks: We are almost at the top of the hour. Starting a countdown. It is now just over 5 weeks remaining until this years running of the OSCC.
[11:54] Arielle.Popstar yes
[11:54] Ubit Umarov: "The Cool VL Viewer is now officially the first viewer to fully support bake-on-mesh in OpenSim."
[11:54] Ubit Umarov: they say
[11:54] Ai Austin: worked second time
[11:54] Kayaker Magic: I am running the 0.9.1 release on OSGrid and having a problem with the INI files.
[11:54] Arielle.Popstar says it will be able to use the proper baking depending on the region release
[11:55] Ai Austin: I hung for two minutes on first TP attempt
[11:55] Arielle.Popstar would have been nice if henri had been here himself
[11:55] Kayaker Magic: The new GridCommon.ini defines services, like FriendsServerURI that I never had in my old ini files.
[11:56] Ubit Umarov: 0.91 Snail does provide necessary information for viewers to identify it suports full BoM
[11:56] Kayaker Magic: If I leave them in, they get defined wrong for OSGrid (which uses sub-domains) and then I get timeouts and errors starting Opensim.
[11:56] Ubit Umarov: well i didn't tested it.. just passing the information :)
[11:56] Kayaker Magic: If I comment out the services I don't have a URL for, then it starts OK.
[11:57] Ubit Umarov: for osgrid you need to use their settings
[11:57] Arielle.Popstar why wasnt there a pre realease anyway? Thjats the usual flow
[11:57] Ubit Umarov: better just use their distro
[11:57] Bill Blight: was going to say, osgrids INI files define them correctly
[11:57] Ubit Umarov: they still organize ini files in a diferent way
[11:57] Bill Blight: [FriendsService]
 ; Change this to your grid-wide friends server
FriendsServerURI = ""
[11:58] Ubit Umarov: so a mix can cause issues
[11:58] Kayaker Magic: OSGrid distro has a 5 year old set of INI files that is nothing like the new ones, I'm trying to upgrade.
[11:58] Kayaker Magic: Where did you learn that Bill?
[11:58] Bill Blight:
[11:58] Gavin.Hird make 2 installations - one for OSG and one for the rest
[11:58] Ubit Umarov: the code is ours
[11:58] Gavin.Hird then you don't have to mix ini files
[11:59] Ubit Umarov: but ini files are theirs, and as bill said, old formats
[11:59] Ubit Umarov: ( with new things added )
[11:59] Arielle.Popstar people look at the osg one or use it to convert to a more recent release for other than osgrid
[12:00] Ubit Umarov: well we did a release now :)
[12:00] Ubit Umarov: possible a not far
[12:00] Arielle.Popstar yes but at the rate of putting them out, it will be another year? :)
[12:00] Ubit Umarov: depending on issues reports
[12:00] Kayaker Magic: I'm trying to make an INI file I load after the example ones that overrides all the wrong URLs with the right ones form OSGrid.
[12:00] Arielle.Popstar will there be a post fixes ??
[12:01] Bill Blight: why not just use the INI's that come with the osgrid download?
[12:01] Bill Blight: they appear to be correct
[12:01] Ubit Umarov: im not using "postfixes" anymore
[12:01] Arielle.Popstar oh
[12:01] Andrew Hellershanks: Kayaker, make your ini file load all the ones from osgrid then you can put in the settings you want to change after the includes.
[12:01] Arielle.Popstar executive decision?
[12:01] Ubit Umarov: next release will be and may be mostly fixes
[12:01] Ai Austin: I mentioned to Ubit that we would usually label the dev master and postfixes would be ... But ...;-)
[12:01] Gavin.Hird sounds good to me
[12:02] Ai Austin: It was Diva who added the fourth digit for that remember
[12:03] Ubit Umarov: master is working thing
[12:03] Ubit Umarov: so its work 0 :p
[12:03] Arielle.Popstar whats the reasoning behind changing how Opensim has done things for the past 12 years?
[12:03] Arielle.Popstar was it broken?
[12:03] Andrew Hellershanks: Ubit, I didn't like "post-fixes" because it often came out before the actual release.
[12:04] Gavin.Hird so a pre-fix
[12:04] Arielle.Popstar but andrew, at least people knew what to expect
[12:04] Ai Austin: postfixes was just a label while development occurred was it noty. then it got releaseed as the next fourth digit +1 on the last release
[12:05] Andrew Hellershanks: Arielle, I didn't as it was counter-intuitive to how I would do things and to how a lot of other projects I'm using do things.
[12:06] Kayaker Magic: I agree with Arielle, there is a procedure ( and it does not seem to have been followed. Where was the discussion? Where was the release canidate?
[12:06] Ai Austin: Main thing is to make sure it got a new number... but I think all the recent zeros and ones look too similar !!
[12:06] Gavin.Hird I like the idea of a pre-fix; Fixing bugs before they have been released.
[12:06] Ai Austin: sound good Gavin. saves a lot of work :-)
[12:06] Arielle.Popstar yes the release canditate was an important naming convention\\
[12:06] Ai Austin: unless someone actually has to fix then ;-)
[12:07] Arielle.Popstar people knew it was for testing
[12:07] Andrew Hellershanks: We are now past the hour mark. Any last minute items for today?
[12:07] Gavin.Hird it is sort of like losing weight while you eat
[12:07] Andrew Hellershanks: Gavin, that is what release candidates are used for.
[12:07] Ai Austin: I have a quick one for Arielle...
[12:07] Andrew Hellershanks: Go ahead, Ai
[12:08] Ai Austin: HGLub grid uses the Nani version of OpenSim
[12:08] Ai Austin: is theer anything seriously different in there we ought to consider for core?
[12:09] Arielle.Popstar she is looking into an issue that seems to be getting triggered by the way the new code goes for the viewer cache
[12:09] Arielle.Popstar it is messing with inventory when visitors get back to hgluv
[12:09] Ai Austin: that would be good, I have a few issues I have been mentioning to Ubit.. not pinned them down yet as I am travelling so my setup is not my usual kit
[12:10] Arielle.Popstar she is still trying to pin down the exact flow for the problem
[12:10] Arielle.Popstar her and Nani are both looking into it
[12:10] Ai Austin: good.
[12:11] Gavin.Hird it would be immensely more helpful if she actually published her changes and not just a file dump
[12:11] Arielle.Popstar but Hy just informed me it isnt because of Bill's BoM viewer so thats good
[12:11] Ai Austin: when I get back to Edinburgh I will do more etsyign on what I am seeing too. I had some wierd ones where textures never show unless I clear cache and a second viewer on one desktop did not show a second avatar when logged in
[12:12] Ai Austin: I am using a bog standard FS 6,.3.2.58056 just compiled as is with OpenSim fflag as a comparison as well as Bill's OpenStorm
[12:12] Bill Blight: if you use the same viewer to go to SL and to OS , they have been doing something with the cache over there that seems to screw with some people
[12:12] Ai Austin: ah, good to know Bill, thanks
[12:12] Arielle.Popstar ok yes, i have used the viewer on both grids
[12:13] Bill Blight: It is not something nailed down, but many users on my grid have expressed the issue
[12:13] Kayaker Magic: SL, I don't go there any more. Next I will abandon RL, which takes up entirely too much of my time!
[12:13] Bill Blight: they clear cache, before logging in to OS or use viewers with different caches , and the problem goes away
[12:13] Ai Austin: much too overrated that RL
[12:13] Arielle.Popstar :)
[12:14] Arielle.Popstar still have to figure out how to upload ourselves here
[12:14] Ai Austin: I resisted using BoM to come along when Ubit invited me :-)
[12:14] Ai Austin: but I am VERY keen on BoM as we develop the next Release candidates for Ruth 2.0 and Roth 2.0 mesh bodies. That is underway
[12:14] Ubit Umarov: there was no BoM here before that :p
[12:15] Arielle.Popstar oh and a poster in HGB asked for some links to more info on YEngine
[12:15] Arielle.Popstar is there a wiki anywhere/
[12:15] Ai Austin: I found 5 or so typical types of script errors I had to change when I switched to YEngne on my Ailand experimental grid
[12:15] Ubit Umarov: no
[12:16] Ai Austin: nothing serious, but they did need changing
[12:16] Ubit Umarov: and possible there will only be one, when its options are defined
[12:16] Ai Austin: Not so good for folks who don't have mod permissions
[12:16] Andrew Hellershanks: Ai, care to document them for the OS wiki? :)
[12:16] Ubit Umarov: for now it is a X replacement
[12:16] Ai Austin: sure. I put them on a blog post alreday
[12:17] Andrew Hellershanks: Where is the blog post?
[12:17] Kayaker Magic: Yes! I want to see this!
[12:17] Kayaker Magic: I'm finding very few issues with YEngine.
[12:17] Ai Austin:
[12:17] Andrew Hellershanks: Ai, ty
[12:18] Ai Austin: YW
[12:18] Kayaker Magic: Although I wrote most of my scripts to run on SL, IW and OpenSim, so found many minor syntax issues that way.
[12:18] Arielle.Popstar and Bill, didnt you have some Yengine info somewhere?
[12:18] Ai Austin: if anyone can think of best place to put them I will add the notes on OS wiki
[12:18] Ai Austin: should we have a conversion guide Ubit?
[12:19] Andrew Hellershanks: It could be added to the end of the release notes for 0.9.1
[12:19] Ai Austin: happy to do that if Ubit is happy. he was annoyed at a note I added on the osslEnabled.ini change :-)
[12:19] Andrew Hellershanks: :)
[12:20] Arielle.Popstar somebody nudge Bill
[12:20] Andrew Hellershanks: For now there is your blog post. Others can find that now that we have the URL in todays meeting log.
[12:20] Ai Austin: Good
[12:21] Ubit Umarov: Ai
[12:21] Arielle.Popstar I'm sure there was some info on a page of his too
[12:21] Bill Blight: I"ll see if I can find it again
[12:21] Ubit Umarov: break is not a LSL keyword
[12:21] Ubit Umarov: is it?
[12:21] Ai Austin: I know... but it clearly worked in XEngine
[12:21] Ubit Umarov: did nothing
[12:21] Gavin.Hird For the transcript there is a new macOS version fo the viewer with an updated vivox client that works much better on the Mac
[12:21] Gavin.Hird the Win version will follow shortly
[12:21] Ubit Umarov: X could accept odd things
[12:21] Andrew Hellershanks: break shouldn't be an LSL keyword as there is no programming structure in stock LSL that would use it.
[12:21] Ubit Umarov: without a error
[12:22] Ai Austin: I know.. it was in an XEngione script worked in XEngine - it was in a spacecraft flight controller (from lani grid)
[12:22] Andrew Hellershanks: or is there. My LSL Fu is getting rusty.
[12:22] Ubit Umarov: and break is a Yengine keyword, using some options :p
[12:22] Ai Austin: skydancer shuttle
[12:23] Ai Austin: Precendence brackets were needed in a couple of places too - and some other things
[12:23] Ubit Umarov: better fix that blog :)
[12:23] Ubit Umarov: seems you are telling break is a legal LSL
[12:23] Ai Austin: I will add "not valid LSL but worked in XEngine"
[12:24] Ai Austin: if one script had it from someone and it worked I am sure it will turn up for others
[12:24] Andrew Hellershanks: Replacing break with return could have serious impact on a script as they are typically used in very different ways.
[12:24] Ubit Umarov: break is just a typo, bad script
[12:24] Ai Austin: okay, I added a note in the blog on that
[12:24] Ubit Umarov: proper fix requires reading it
[12:25] Andrew Hellershanks: Can't use break to exit a loop prematurely?
[12:25] Ai Austin: every spacecraft I had from lani region had the issue
[12:26] Ubit Umarov: well i don't remember it being a valid keyword
[12:26] Ubit Umarov: it will be on Y
[12:26] Ubit Umarov: X had odd things also
[12:26] Andrew Hellershanks: LSL wiki doesn't mention continue or break.
[12:27] Ai Austin: I know. don't shoot the messenger :-)
[12:27] Ubit Umarov: some things just sneaked from LSL source to C# intermediate code
[12:27] Ai Austin: It was not my code!
[12:27] Ubit Umarov: so possible the break; did that
[12:27] Ubit Umarov: and on C# it is legal :)
[12:27] Ai Austin: Easily replaced. as I had mod rights
[12:28] Andrew Hellershanks: Thank you all for coming. See you again next week.
[12:28] Kayaker Magic: EOM, End Of Meeting
[12:28] Ubit Umarov: i almost made official the keyword Yoptions
[12:28] Arielle.Popstar is at S/L meeting too
[12:28] Ai Austin: I managed to track down all errors I was seeing on YEngine. One was tricky until Tampa gave me a way to get some hook on the item.
[12:28] Ubit Umarov: enabling new flow control
[12:28] Andrew Hellershanks: Ubit, You can do that in a later release. :)
[12:28] Ubit Umarov: and try cactch
[12:29] Ubit Umarov: but try catch has issues yeackk
[12:29] Ubit Umarov: actually finally() has issues
[12:29] Andrew Hellershanks: new flow control would be useful. Not sure about adding in try/catch/finally.
[12:29] Ai Austin: Its good to find issues before we have hundreds of users hit them
[12:29] Arielle.Popstar nods
[12:29] Ubit Umarov: yeah break is nice to have
[12:30] Andrew Hellershanks: yes, I'd vote for break and continue.
[12:30] Ubit Umarov: to have officially !!!
[12:30] Ai Austin: the diagnostics on YEngine were pretty good at giving pointers to object:script and x/y.z involved in most cases.. just a couple were not giving that data for the errors I hit
[12:30] Ubit Umarov: :)
[12:30] Ai Austin: There is a Mantis about that
[12:31] Ubit Umarov: as i said on that, some errors are hard to read :)
[12:32] Ubit Umarov: well the new keywords when public valid, will require a wiki for Y
[12:32] Ubit Umarov: of course
[12:32] Ubit Umarov: ( they are there.. ppl can play with them )
[12:32] Ai Austin: I was not arguing for different script engine constructs
[12:32] Ubit Umarov: of course more problems for scripts compatibility
[12:32] Ai Austin: just good to have advice for anythig that differs in case people try Engine
[12:33] Ubit Umarov: tps etc
[12:33] Ai Austin: Good to be able to go BACK to XEngine as well, so ideally we want script source compatibility
[12:33] Ubit Umarov: parsing order is also more sl like
[12:33] Ubit Umarov: some complex statements may give different results
[12:33] Ai Austin: Folks have asked for a few notes on what YEngine offers
[12:34] Ubit Umarov: but X did required extra () as those.. so..
[12:34] Ai Austin: I tend to add prioritizing parentheses in my code, but a lot of preexistent stuff assumes a left to right parse
[12:34] Ubit Umarov: always good to put them
[12:35] Ubit Umarov: no not be dependent on engine decisions
[12:35] Ubit Umarov: evaluation order etc
[12:35] Ubit Umarov: your example on binary & and == is such a case
[12:36] Ai Austin: yep

Personal tools
About This Wiki