Chat log from the meeting on 2019-11-26

From OpenSimulator

Jump to: navigation, search

[11:08] Andrew.Hellershanks Looking at the OS change log it was mostly changes related to mono addins this past week.
[11:09] Gavin.Hird and a lot of them too
[11:09] Gavin.Hird what happened Ubit?
[11:09] Ubit Umarov: issues with versions and signatures of the several libs it uses
[11:10] Gavin.Hird ok
[11:10] Ubit Umarov: only a subset used by opensim
[11:12] Ubit Umarov: one issue, you did hit it gavin :)
[11:14] Ubit Umarov: well part of mess was the auto build of the thing, i kept putting back older versions of the libs, from several places
[11:14] Ubit Umarov: nuget and things like that
[11:18] Andrew.Hellershanks This years OpenSim conference is now only two and a half weeks away.
[11:20] Andrew.Hellershanks I don't see a conference schedule up on the site yet. It may still be a touch early for that.
[11:20] Bill.Blight only 100 DAYS till OpenSim Fest , shameless plug
[11:20] Andrew.Hellershanks :)
[11:21] Andrew.Hellershanks Is that the second running of the event, Bill?
[11:21] Bill.Blight yes
[11:22] Andrew.Hellershanks It is a ways off but as you have mentioned it is there a URL or grid location where people can get more information on in case they want to get involved with the event?
[11:22] Gavin.Hird 2 weeks before Christmas is not a good timing if you ask me
[11:23] Bill.Blight web site, still needs to be updated but , it is
[11:23] Bill.Blight and we have a discord server
[11:23] Bill.Blight
[11:23] Andrew.Hellershanks Ok. Thanks, Bill.
[11:24] Ubit Umarov: don't discord with ppl, thats ugly
[11:24] Leighton.Marjoram The website is still being done and we will start promoting over the next week or so. So people can start signing up and setup.
[11:24] Bill.Blight site should be rocking and rolling in time for OSCC
[11:24] Andrew.Hellershanks Leighton, Are you talking about the site for the Fest or the OSCC?
[11:24] Leighton.Marjoram Discord is great Ubit there are some nice folx on there
[11:25] Leighton.Marjoram Opensim Fest Andrew
[11:25] Andrew.Hellershanks Leighton. Ok. Just wanted to make sure there was no mix up re: the events.
[11:26] Leighton.Marjoram I wasnt to sure which event you were referencing so went with the one I am involved in.
[11:27] Andrew.Hellershanks Leighton, np. In the past we have had someone attending this gathering giving updates on the OSCC. I was thinking that might have been you.
[11:27] Leighton.Marjoram np Andrew
[11:27] Bill.Blight Marcus usually starts showing up this close to OSCC
[11:27] Andrew.Hellershanks nods to Bill
[11:28] Andrew.Hellershanks Perhaps he will be here next week.
[11:29] Andrew.Hellershanks For those that don't mind using Discord should we point out the recent Open Simulator server on Discord?
[11:29] Andrew.Hellershanks :)
[11:29] Ubit Umarov: part of the mono addins changes where to fix its mautil so diva could fix wifi to oscc
[11:31] Andrew.Hellershanks Good to know. It was nice seeing Diva in IRC once again. It has been a while. I was trying to get WiFi running with git master and having some troubles with it so I have had to skip using WiFi for now. Let's hope that Diva gets WiFi updated soon.
[11:32] Ubit Umarov: ppl is beeing usig wifi with Snail for sometime now
[11:32] Ubit Umarov: Ai Austin has some instructions on his blog
[11:32] Andrew.Hellershanks I get lots of red text. The main one being that WiFi is having trouble accessing the MySQL database.
[11:33] Andrew.Hellershanks Ubit, you can guess what I'm going to ask next. :)
[11:33] Ubit Umarov: "time to end meeting?"
[11:33] Andrew.Hellershanks Hehe... not yet.
[11:34] Andrew.Hellershanks What is the URL for Ai's blog?
[11:34] Andrew.Hellershanks I don't read blogs very often.
[11:34] Ubit Umarov: and you think i remember it??
[11:34] Ubit Umarov: :)
[11:34] Andrew.Hellershanks :) np. Let me do a quick search and see if I can find it.
[11:34] Ubit Umarov: guess google remembers
[11:35] Ubit Umarov: but what he did is the normal
[11:35] Andrew.Hellershanks The URL I found is Note added: This is a mirror and does not include latest posts. Please use for latest posts.
[11:35] Ubit Umarov: fix the prebuild-xml files for .net 4.6
[11:35] Andrew.Hellershanks I'll look through it later for the information about using WiFi with 0.9.1
[11:35] Ubit Umarov: and install it the old way not using mono addins ??
[11:36] Ubit Umarov: ( well using them just not mautil etc )
[11:36] Ubit Umarov: well on code
[11:37] Ubit Umarov: im breaking kfm again
[11:37] Andrew.Hellershanks I've only applied the code changes that were documented in the place where I downloaded the source for the latest WiFi. If there other changes needed in the OS source tree I'll give that a try later.
[11:37] Andrew.Hellershanks I've seen you talking about KFM issues this past week. What issues have you been having with it?
[11:37] Ubit Umarov: but really don't understand it at SL now, specially the ping pong mode
[11:38] Ubit Umarov: our loop mode is also not as sl now
[11:38] Ubit Umarov: our does not reset position and rotation to values at start of each loop
[11:38] Ubit Umarov: they will
[11:38] Andrew.Hellershanks It has been a while since I last read up on KFM. IIRC, that allows it to run through the KFM actions from start to finish then do them in the reverse direction and then start to finish again, and so on.
[11:39] Ubit Umarov: now SL ping pong seems a mess
[11:39] Gavin.Hird what is SL ping pong?
[11:39] Andrew Hellershanks: Gavin, It is a feature of Keyframing.
[11:39] Ubit Umarov: it does reset positions to initial values at the end of each iteration, ( already no much sense )
[11:40] Gavin.Hird ok
[11:40] Ubit Umarov: but worse... it does it somewhere in the middle of a sequence
[11:40] Ubit Umarov: not at extremes
[11:40] Ubit Umarov: a total mess
[11:40] Gavin.Hird and when did this change?
[11:40] Ubit Umarov: no idea, ages since i tested it there
[11:41] Gavin.Hird ah
[11:41] Andrew Hellershanks: Resetting positions in the middle of a KFM sequence is going to seriously mess things up.
[11:41] Ubit Umarov: as is it just makes no sense
[11:41] Ubit Umarov: so im doing loop as theirs..
[11:41] Ubit Umarov: but ping pong as we had it
[11:41] Ubit Umarov: no reset postions ever
[11:42] Gavin.Hird but how do you test it?
[11:42] Ubit Umarov: make a thing and tell it to move :p
[11:42] Andrew Hellershanks: With no reset the position of items might drift over time.
[11:43] Ubit Umarov: kfm is prone to drifts anywhere
[11:44] Andrew Hellershanks nods
[11:44] Ubit Umarov: ok where did my thing went? lol
[11:45] Gavin.Hird too much helium?
[11:45] Andrew Hellershanks: You can reset an object location back to the start for one time KFM sequences. When you start looping that can become an issue over time.
[11:45] Andrew Hellershanks: I saw it going up and it went out of my view.
[11:45] Bill Blight: there is still one of your test things over on the border, still running
[11:45] Gavin.Hird I did too
[11:45] Andrew Hellershanks: Ubit, It is well up in the air and still rising.
[11:46] KFMtstUbit: leaving touch
[11:47] KFMtstUbit: leaving touch
[11:47] Ubit Umarov: ok this is our loop
[11:47] Andrew Hellershanks: What do you find happens when you use Ping Pong mode?
[11:47] KFMtstUbit: leaving touch
[11:47] KFMtstUbit: leaving touch
[11:47] Ubit Umarov: stop touching it :p
[11:47] Andrew Hellershanks: :)
[11:47] Andrew Hellershanks: You should ignore touches from anyone other than the owner.
[11:48] Ubit Umarov: adn one cares about that on test things...
[11:48] Andrew Hellershanks: Hm... it occasionally jumps position
[11:49] Ubit Umarov: ppl is not suposed to touch any thing they see :p
[11:49] Andrew Hellershanks: Its like giving people a big red button and telling them not to push it. ;)
[11:50] Ubit Umarov: wel at least from the list we can guess what it will do
[11:50] Ubit Umarov: can't say the same at sl
[11:51] Ubit Umarov: but yes
[11:51] Ubit Umarov: more opensim breaking changes on this
[11:51] Ubit Umarov: yeack
[11:51] Andrew Hellershanks: Have you noticed whether the physics engine in use affects KFM?
[11:51] Ubit Umarov: i know the code andrew :p
[11:52] Ubit Umarov: we had this fine at avination
[11:52] Ubit Umarov: got messed up on the big merge
[11:52] Andrew Hellershanks: Good. How long before you fix it?  :)
[11:52] Ubit Umarov: now im remaking it
[11:52] Ubit Umarov: well i helped making work there
[11:53] Ubit Umarov: but is has issues
[11:53] Ubit Umarov: like with PI rotations
[11:53] Ubit Umarov: have new code with that fixed
[11:53] Andrew Hellershanks: I made a working ferris wheel using KFM. I've been meaning to bring it in to 0.9.1 so I could test it in the latest version of grid code.
[11:53] Ubit Umarov: but still a over complex code
[11:54] Ubit Umarov: this is a damm simple thing.. grrr
[11:54] Ubit Umarov: well not much jitter here
[11:54] Ubit Umarov: current kfm is worse on windows
[11:54] Andrew Hellershanks: IIRC, my use of KFM involves using rotations
[11:55] Ubit Umarov: has more rubber banding than linux, typical
[11:55] Ubit Umarov: hope i did improved that also
[11:56] Andrew Hellershanks: Good luck with it, Ubit. The OSCC grid has moved to 0.9.1. I don't know if they have any KFM based objects running over there. If they do we will probably hear about it soon.
[11:57] Andrew Hellershanks: We are at just under 5 minutes to the top of the hour. No Kayaker here today to fire off some questions. Does anyone else have any issues to discuss at this meeting?
[11:57] Ubit Umarov: at SL wiki
[11:57] Ramesh.Ramloll just a simple question about 0.9.1
[11:57] Ubit Umarov: the example Continuous Spin
[11:57] Andrew Hellershanks: Go ahead, Ramesh.
[11:57] Ubit Umarov: does not do what they say
[11:57] Gavin.Hird not really. Been working on the Windows installer for the viewer so that is posted this morning
[11:58] Ramesh.Ramloll does anyone experience unusual delays with rezzing and taking objects into inventory?
[11:58] Bill Blight: I have not noticed any such delays, unless the region is very overloaded
[11:58] Ramesh.Ramloll Ok
[11:59] Andrew Hellershanks: Ubit, Looks like the start of something for a washing machine. ;)
[11:59] Ramesh.Ramloll also no issues with compiling very large LSL scripts (closed to the maximum allowed memory limit for one script)
[11:59] Ramesh.Ramloll ?
[12:00] Ramesh.Ramloll It used to be faster on 0.8
[12:00] Andrew Hellershanks: I remember you have also mentioned that compiling very large scripts results in the script being compiled twice.
[12:00] Ramesh.Ramloll We also get this
[12:00] Ubit Umarov: our current loop does has they claim it should.. ( no reset to start position)
[12:01] Ubit Umarov: well our current code doesn't even move to the limits grrr
[12:02] Bill Blight: not seeing double "compile successful" here
[12:03] Gavin.Hird I have never seen that
[12:03] Andrew Hellershanks: Ramesh, Do you have any information about the size of the scripts that result in that behaviour? Either in terms of lines or number of characters?
[12:04] Ubit Umarov flüstert: thats compiling using what?
[12:04] Ramesh.Ramloll well I will have to get back to you, what I do know is that we were very close to the max allowed, and had to refactor scripts
[12:04] Ramesh.Ramloll so that each was small enough
[12:05] Ubit Umarov flüstert: thats compiling using what?
[12:05] Ubit Umarov: well our current code doesn't even move to the limits grrr
[12:05] Ramesh.Ramloll Andrew, can you remind us of the compiler we are using?
[12:05] Gavin.Hird do you see it on every viewer?
[12:05] Gavin.Hird it could be a viewer issue
[12:05] Ramesh.Ramloll Yes
[12:06] Ubit Umarov: sounds like you have comunications issues
[12:06] Andrew Hellershanks: Ubit, If you mean which scripting engine it is XEngine. IIRC, it is OS running git master with mono 6.4
[12:06] Ubit Umarov: that's possible a viewer their stupid retries on http
[12:06] Leighton.Marjoram I do have to head off see you all again next week :)
[12:06] Leighton.Marjoram :waves:
[12:06] Andrew Hellershanks: ok, Leighton. Thanks for coming.
[12:06] Gavin.Hird TC Leighton
[12:07] Ramesh.Ramloll Bye TC
[12:07] Ubit Umarov: ie you region is taking 2 long.. viewer did retried to upload same script again
[12:07] Ubit Umarov: so you got 2 compiles
[12:07] Ramesh.Ramloll Ok
[12:07] Ubit Umarov: guess your issues are .. bad comms
[12:08] Andrew Hellershanks: I didn't realize (or remember) there is a limit to the maximum size of a script. My biggest script is 3,000 lines long. Not sure how many characters.
[12:08] Ubit Umarov: or overloaded regions
[12:08] Andrew Hellershanks: Overloaded in terms of number of rezzed prims, or script usage, or both?
[12:09] Gavin.Hird or slow db response
[12:09] Andrew Hellershanks: Gavin, That is a possibilty. The server has plenty of free RAM and the CPU load isn't that high.
[12:10] Ramesh.Ramloll After refactoring our scripts are pretty small actually
[12:10] Ramesh.Ramloll even the main one is just 417 lines
[12:10] Andrew Hellershanks: 417 lines isn't a very large script.
[12:11] Andrew Hellershanks: Although I know the person writing that code and he doesn't use much in the way of whitespace so that code is pretty dense.
[12:12] Ubit Umarov: 64k chars i think
[12:12] Ubit Umarov: viewer dependent
[12:12] Ramesh.Ramloll correct
[12:12] Andrew Hellershanks: I do see delays accessing a database now and then when I try to access it via phpMyAdmin. Not sure what causes that sort of delay.
[12:12] Andrew Hellershanks: It even happens on my own computer now and then.
[12:14] Gavin.Hird probably the db having to reach for the disk to fetch something, and could also be sequential search
[12:16] Ubit Umarov: Ramesh we don't notice those issues
[12:16] Ramesh.Ramloll Ok
[12:16] Ubit Umarov: that double compile its the first time i see it
[12:16] Andrew Hellershanks: Ramesh, Have you tried different viewers to see if it might or might not be viewer specific?
[12:16] Ramesh.Ramloll Were there major changes made to the way sleep works?
[12:16] Ubit Umarov: ( and viewers should not retry http.. thats nonsense, but they do it a lot )
[12:17] Ubit Umarov: llsleep?
[12:17] Ramesh.Ramloll yes
[12:17] Ubit Umarov: on X as suicidal as before
[12:17] Andrew Hellershanks: :)
[12:17] Ramesh.Ramloll Ok :)
[12:18] Andrew Hellershanks: llSleep is less so with YEngine.
[12:18] Ramesh.Ramloll Good to know
[12:18] Ubit Umarov: yeah we need to change to Y and retire X
[12:19] Ubit Umarov: but.. again some scripts ppl call good, will fail
[12:19] Andrew Hellershanks: I'm sure it will happen eventually. it just takes some time. It isn't as if you don't have anything else to do, Ubit. :)
[12:19] Ramesh.Ramloll May be we should try YEngine and see if things are better
[12:20] Ubit Umarov: we are on y here
[12:20] Ramesh.Ramloll Ok
[12:20] Andrew Hellershanks: Easy enough to switch over to Y. Scripts will need to be recompiled.
[12:21] Ubit Umarov: this simple ping pong at sl does..
[12:21] Ubit Umarov: moves to PI.. jumps to zero, moves to -pi
[12:21] Andrew Hellershanks: We are getting close to the half past mark. Do you want to say anything about the latest changes to Dayturn? Anything special being worked on that you might want to mention?
[12:21] Gavin.Hird have you reported it there
[12:21] Gavin.Hird or is there a jira for it?
[12:22] Ubit Umarov: and its one waypoint only
[12:22] Gavin.Hird As I said I have worked on the Windows installer which is a nightmare to test
[12:22] Ubit Umarov: with more.. as i said no idea where it jumps to start
[12:22] Selby.Evans Must go -- bye all
[12:22] Gavin.Hird plus a bunch of llsingleton shit, but it gave a small performance increase
[12:22] Andrew Hellershanks: Bye, Selby.
[12:22] Gavin.Hird so there are new builds this morning
[12:23] Gavin.Hird for the Windows version that is
[12:24] Andrew Hellershanks: Thanks, Gavin.
[12:24] Gavin.Hird yw
[12:25] Ubit Umarov: this is kinda what it does at sl
[12:25] Ubit Umarov: with the simple case
[12:26] Ubit Umarov: err or not lol
[12:26] Ubit Umarov: not :)
[12:26] Andrew Hellershanks: That viewer is currently Windows only and is being developed by the IMA community.
[12:26] Andrew Hellershanks: Thank you all for coming. See you again next week.

Personal tools
About This Wiki