[5:54 PM PT] Object: Script running [19:02 UTC] *** Logging started. *** [12:02 PM PT] Gavin.Hird @grid.xmir.org:8002: Thganks Ubit [12:02 PM PT] Gavin.Hird @grid.xmir.org:8002: Thanks* [12:02 PM PT] Ubit Umarov: anyway [12:03 PM PT] Ubit Umarov: welcome to another meeting :) [12:03 PM PT] Ubit Umarov: abot last week code changes... [12:03 PM PT] Andrew Hellershanks: :) [12:03 PM PT] Ubit Umarov: Vicednt did noticed and reported a few issues on oar save.. so we addressed those [12:04 PM PT] Ubit Umarov: vicent also helped... [12:04 PM PT] Ubit Umarov: ty [12:04 PM PT] Ubit Umarov: ohh and vincent.Sylvester also!! ( jezz ) [12:04 PM PT] Vincent.Sylvester @hg.zetaworlds.com: lol [12:05 PM PT] Lyr.Lobo @cc.opensimulator.org:8002 grins [12:05 PM PT] Ubit Umarov: there was a bug that would break a oar load if parcels have more than one entry on access list [12:05 PM PT] Ubit Umarov: and some missing parameters on parcel media store and loar [12:05 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I still have not worked out what even prompted that. Only discovered having to roll a region back due to missing objects, no idea why those vanished, so there is more lurking out there [12:06 PM PT] Vincent.Sylvester @hg.zetaworlds.com: More voodoo :) [12:06 PM PT] Ubit Umarov: well another change is kinda breaking one from SL at least [12:06 PM PT] Ubit Umarov: it is abotu the linknumber of root prims, and as seen by LSL functions [12:07 PM PT] Ubit Umarov: as you know on a single prim, the prim has link number 0, but on a linkset with more than 1 prim, it changes to 1 [12:07 PM PT] Joe Magarac: Oh, is that changing? [12:07 PM PT] Ubit Umarov: and we do that.. [12:07 PM PT] Ubit Umarov: now at SL a setting avatar also counts.. [12:08 PM PT] Ubit Umarov: so on a single prim with a sitting avatar the linknumber is also 1 [12:08 PM PT] Cuga.Rajal @rajal.org:9000: so sitting on a single prim, the prim changes from 0 to 1 and the avatar "prim" number is 2? [12:08 PM PT] Ubit Umarov: we never did that... and will never will.. so this was already a change [12:09 PM PT] Ubit Umarov: my hange now was to remove all the broken code we had to try to reproduce ll [12:09 PM PT] Ubit Umarov: now on LSL functions both numbers 0 or 1 will address root prim [12:09 PM PT] Kayaker Magic: Link numbers have always been FUBAR [12:09 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Prompting this I did check a bunch of scripts that I wrote along with a few multi sit scripts and such stuff, so far no reports of them breaking so they were written in ways that isn't impacted by this [12:10 PM PT] Ubit Umarov: see at sl we cannot make a generic script... [12:10 PM PT] Ubit Umarov: in case of a single prim LINK_ROOT fails becasue that just means 1 and the prim has 0 [12:11 PM PT] Andrew Hellershanks: Kayaker, they used to be really FUBAR when combining two different link sets but that got sorted out a while back. [12:11 PM PT] Ubit Umarov: if a vatar sits.. then it will work.. but 0 will fail.. [12:11 PM PT] Ubit Umarov: so only chance on that to work there in both cases is to use LINK_THIS [12:11 PM PT] Kayaker Magic: The viewer has had a bug forever to display the link numbers wrong. [12:12 PM PT] Ubit Umarov: i case of a link set [12:12 PM PT] Andrew Hellershanks: Kayaker, I haven't see that issue in a long time. I haven't tested it in a while though. [12:12 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Yeah there is a mantis on that if I recall [12:12 PM PT] Joe Magarac: The viewer has no idea what the link numbers are. It guesses based on the order of object updates. [12:12 PM PT] Ubit Umarov: a for a script on a child prim, both 0 and 1 will work.. (ll considers 0 working as a legacy bug they keep) [12:12 PM PT] Cuga.Rajal @rajal.org:9000: I thought they fixed that in recent months [12:12 PM PT] Ubit Umarov: but on root prim, 0 will fail... [12:13 PM PT] Ubit Umarov: so.. since we where already not that compatible, not considering sitting avatars, i just changed to bot 0 and 1 mean root prim [12:14 PM PT] Ubit Umarov: ( did not check all code paths.. but that is the rule now :) ) [12:14 PM PT] Cuga.Rajal @rajal.org:9000: does that fix LINK_ROOT for single prim? [12:14 PM PT] Ubit Umarov: yes [12:14 PM PT] Cuga.Rajal @rajal.org:9000: yay! [12:14 PM PT] Ubit Umarov: link_root measn 1 [12:15 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Otherwise doesn't impact the counting of links so it doesn't shift the sitting avatars around [12:15 PM PT] Vincent.Sylvester @hg.zetaworlds.com: So most of those scripts should be fine [12:15 PM PT] Ubit Umarov: i also added some little options to oar load and iar save... [12:17 PM PT] Ubit Umarov: on oar load i added a option to make region to replace entries in cache, even if the asset was already there.. the region will also try to reupload it ( that will fail ) [12:17 PM PT] Ubit Umarov: this is not a nice thing.. just to help on some data recovery.. like current osgrid mess [12:18 PM PT] Ubit Umarov: on iar save i added a option to not save inventory items that point to a missing or empty asset.. [12:18 PM PT] Ubit Umarov: again not a nice option.. just for some recovery operations like current osgrid mess [12:19 PM PT] Ubit Umarov: adn about current osgrid mess.. [12:20 PM PT] Ubit Umarov: like you seen last week, and on recent anoucenment ()https://www.osgrid.online/news/asset-reset/) osgrid decide to delete all assets [12:21 PM PT] Ubit Umarov: i must say i consider this a very stupid decision.. even a violation of the spirit of all those that on last disc crash raised about 5000USD to pay for the data recover.. [12:22 PM PT] Ubit Umarov: this is a decision not based on what is right.. but on what is easy [12:22 PM PT] Ubit Umarov: well and that was it :) [12:22 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Amen to that [12:22 PM PT] Andrew Hellershanks: Thank you for that update, Ubit. [12:23 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I noticed something yesterday that has me confused [12:23 PM PT] Vincent.Sylvester @hg.zetaworlds.com: asset with wrong type: 00000000-0000-2222-3333-100000001035 0 != 57 [12:23 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I see a lot of this in logs lately [12:23 PM PT] Joe Magarac: Wow. When they come back, they will have lost half the users. [12:24 PM PT] Vincent.Sylvester @hg.zetaworlds.com: 0 implies image, 57 is material [12:24 PM PT] Gavin.Hird @grid.xmir.org:8002: and 90% of registered regions [12:24 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Is this just viewers requesting images as materials? [12:25 PM PT] Ubit Umarov: yeah seems viewers store some materials with type 0 [12:26 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Figured as much since a lot of these messages are from default library assets as well [12:26 PM PT] Ubit Umarov: ohh no.. they ask for materials telling type 0 [12:26 PM PT] Ubit Umarov: yeah we do have that one stored with type 57 [12:27 PM PT] Ubit Umarov: so sorry.. not store.. ut on request [12:27 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Request is fine, that's a viewer problem, unless it starts writing back wrong types to assets it'll be fine [12:28 PM PT] Gavin.Hird @grid.xmir.org:8002: are we talkiing about the old materials vs PBR materials? [12:28 PM PT] Ubit Umarov: 57 is a pbr material [12:28 PM PT] Gavin.Hird @grid.xmir.org:8002: old materials have diffuse, spec and normal [12:28 PM PT] Gavin.Hird @grid.xmir.org:8002: yes [12:29 PM PT] Ubit Umarov: on fetching via viewerasset, viewer sends the type of the asset it wants [12:29 PM PT] Gavin.Hird @grid.xmir.org:8002: so the viewer can both write and request old materials [12:29 PM PT] Ubit Umarov: and they send 0 in several cases [12:29 PM PT] Gavin.Hird @grid.xmir.org:8002: for anything that isn't converted to PBR [12:29 PM PT] Ubit Umarov: ol materials are fectched on another cap [12:30 PM PT] Gavin.Hird @grid.xmir.org:8002: but asset type is deferent [12:30 PM PT] Ubit Umarov: well we made them assets on opensim [12:30 PM PT] Ubit Umarov: they are not true assets.. and should had never been... [12:31 PM PT] Ubit Umarov: dahlia initial code to support them made them just a prim data extention.. [12:31 PM PT] Ubit Umarov: then others change that to asset.. [12:32 PM PT] Ubit Umarov: a move that blocked any LSL for them for a few years.. [12:32 PM PT] Ubit Umarov: to avoid filling up the assets servers [12:32 PM PT] Ubit Umarov: with jsut the mousewell over some parameters [12:32 PM PT] Ubit Umarov: whell [12:32 PM PT] Ubit Umarov: i worked aroudn that adding a dead time.. [12:33 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I'll just add a condition to silence this. Seems fcache clears out those assets and then a viewer requests them anew each time spamming the logs [12:33 PM PT] Ubit Umarov: but those are other materials :) [12:33 PM PT] Ubit Umarov: the materials are fine [12:33 PM PT] Ubit Umarov: it is just a warning... ignore :p [12:33 PM PT] Ubit Umarov: ( the 0 != 57 ) [12:34 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Yeah I can ignore it, but my disk usage can't xD [12:34 PM PT] Ubit Umarov: in fact also happens with terrain textures/pbr [12:34 PM PT] Ubit Umarov: has nothing to so with disk use.. the disk is in use.. [12:34 PM PT] Ubit Umarov: as i said is on FECTH not store [12:34 PM PT] Vincent.Sylvester @hg.zetaworlds.com: The messages add up in the logs is what I mean [12:34 PM PT] Ubit Umarov: in fact on terrain they to that mess [12:35 PM PT] Ubit Umarov: viewer has no idea if terrain UUID is a texture or a material [12:35 PM PT] Vincent.Sylvester @hg.zetaworlds.com: fcache clears some of these out constantly and then viewer requests them again creating a new message every couple minutes, that over hundreds of regions adds up in log spam :) [12:35 PM PT] Ubit Umarov: so it asks for a texture, then a material .. same uuid [12:36 PM PT] Ubit Umarov: fcahce never cleans out our own asset... [12:36 PM PT] Ubit Umarov: stop that :p [12:36 PM PT] Vincent.Sylvester @hg.zetaworlds.com: 00000000-0000-2222-3333-100000001035 is a default asset [12:36 PM PT] Ubit Umarov: that warining is on Fectch.. and is a warning [12:36 PM PT] Ubit Umarov: ignore it [12:38 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Accounts for 5873 of 84127 lines of log file message "asset with wrong type" so it's pretty spammy [12:38 PM PT] Ubit Umarov: fien i will coment it out :p [12:38 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Nah it's fine, just saying [12:39 PM PT] Gavin.Hird @grid.xmir.org:8002: can't you just change the log type to debug? [12:39 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I'll add some cleaning to log file rotate and it'll be fine for me, just maybe worth watching log file size :) [12:40 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Some warnings are actually useful to have, especially operating with master dev version [12:40 PM PT] Ubit Umarov: [12:40 PM PT] Gavin.Hird @grid.xmir.org:8002: log level, info, war, debut [12:40 PM PT] Gavin.Hird @grid.xmir.org:8002: debug [12:40 PM PT] Gavin.Hird @grid.xmir.org:8002: set it to debug so you can see it when needed [12:40 PM PT] Ubit Umarov: so guess part of terrain.. in this case viewer asked for a material and is is a texture [12:41 PM PT] Ubit Umarov: part of the mess ll did on terrain... [12:41 PM PT] Ubit Umarov: to add pbr support [12:41 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Funny part of that is the region doesn't have the texture on terrain or really anywhere [12:42 PM PT] Ubit Umarov: vodooo ofc [12:42 PM PT] Ubit Umarov: :) [12:42 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Would explain why it comes up every other hour or so if not in use gets cleaned out [12:43 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Gonna dig through the texture entries maybe it is in there somewhere [12:43 PM PT] Vincent.Sylvester @hg.zetaworlds.com: asset archeology lol [12:43 PM PT] Ubit Umarov: :) [12:44 PM PT] Kayaker Magic: So I ran off to make a couple OARs here on OSGrid. I'm getting lots of ARCHIVER errors about references to missing or damaged items. Does that mean my OARs are too little too late? [12:44 PM PT] Gavin.Hird @grid.xmir.org:8002: I would say yes [12:44 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Possibly, but also normal for it to not find everything in some cases, especially on old things the assets can contain references that never worked to begin with [12:44 PM PT] Ubit Umarov: soem are false fails [12:44 PM PT] Cuga.Rajal @rajal.org:9000: Making OARs always throws some errors even when ok [12:45 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Remember the uuid gatherer tries to search things for everything relevant so it even searches script code [12:45 PM PT] Cuga.Rajal @rajal.org:9000: at least the errors gives you the asset names so you can search for them [12:45 PM PT] Ubit Umarov: on oar saves the code sees for a exmple a UUID on a script and has no way to know what it is [12:45 PM PT] Cuga.Rajal @rajal.org:9000: or decide how important [12:46 PM PT] Ubit Umarov: so it tries to get a asset... and if it fails it reports it as possible fail [12:46 PM PT] Kayaker Magic: Ah, I used to write code that had UUIDs for several different systems in the comments and that caused no end of problems. [12:46 PM PT] Ubit Umarov: in last weeks.. well yeah possilbe is a missing asset [12:46 PM PT] Vincent.Sylvester @hg.zetaworlds.com: It somewhat depends on the ratio. A region with only a handful of objects throwing hundreds of errors then yeah those assets might be screwed, but getting a few on a region with a thousand prims is somewhat normal [12:47 PM PT] Ubit Umarov: fun thing... possilble osgrid did not lost any asset [12:47 PM PT] Cuga.Rajal @rajal.org:9000: there are cases with the error says "are missing assets or is not a problem" [12:47 PM PT] Ubit Umarov: some are just out of visibility temporary stored elsewhere [12:47 PM PT] Ubit Umarov: or maybe... [12:47 PM PT] Gavin.Hird @grid.xmir.org:8002: I bet the bits are still there [12:48 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Put a avatar uuid into a script and it might complain about that not being an asset even if that is a debatable stance xD [12:48 PM PT] Ubit Umarov: :p [12:48 PM PT] Cuga.Rajal @rajal.org:9000: I searched hard on the assets with errors and couldn't find any issues [12:48 PM PT] Andrew Hellershanks: I have had issues before where I see a report of a missing asset but no idea what object or script referenced the asset. [12:48 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Because the uuid gatherer also doesn't know [12:49 PM PT] Vincent.Sylvester @hg.zetaworlds.com: The code is somewhat recursive so it digs losing the hierarchy that got it there [12:49 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If it finds a referenced script it grabs that and might dig into that to find more stuff it needs to fetch [12:49 PM PT] Cuga.Rajal @rajal.org:9000: so an object in contents could throw the error? Contents of an object n contents? [12:50 PM PT] Andrew Hellershanks: I figured that was the case. Probably not an easy thing to fix. [12:50 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Can't say how far it would go, that code is not so easy to grasp. A bit abstract, but then it has to be to be able to dig like this [12:51 PM PT] Cuga.Rajal @rajal.org:9000: how do you have a UUID inside a script? does it actually look inside the code? [12:51 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Yes [12:51 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If you reference a particle texture uuid, probably good idea to have that as part of an OAR [12:51 PM PT] Ubit Umarov: scans letter by letter... pff [12:52 PM PT] Ubit Umarov: well almost... [12:52 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I wrote a uuid gatherer in python a while back, it's quite the pain to catch all the things to have an analog copy of what it fetches [12:52 PM PT] Vincent.Sylvester @hg.zetaworlds.com: There is a lot to parse [12:53 PM PT] Cuga.Rajal @rajal.org:9000: What if the script references the asset name and not the UUID? [12:53 PM PT] Cuga.Rajal @rajal.org:9000: same? [12:53 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If you reference by name then the only real option is that the asset is on inventory of the prim [12:53 PM PT] Ubit Umarov: asset name means a asset on the prim... [12:53 PM PT] Ubit Umarov: found elsewhere [12:54 PM PT] Cuga.Rajal @rajal.org:9000: would that throw the missing asset error? [12:54 PM PT] Ubit Umarov: igonred [12:54 PM PT] Cuga.Rajal @rajal.org:9000: kk [12:54 PM PT] Ubit Umarov: it sees the prim stored there [12:54 PM PT] Ubit Umarov: of thing [12:54 PM PT] Ubit Umarov: or.. [12:54 PM PT] Joe Magarac: Viewers tend not to expose full UUIDs to deter copybotting. If you run Sharpview and look at the log, all the UUIDs fetched are in the log, but part of the UUID says "CENSORED". [12:55 PM PT] Vincent.Sylvester @hg.zetaworlds.com: In my tests it finds 99,7% of all the references that could be there, but the parsing is different depending oh what ends up doing it. In python it did not find some things, but found others the C# version didn't catch [12:55 PM PT] Ubit Umarov: this is on internal things joe [12:55 PM PT] Ubit Umarov: to same OARs for example [12:55 PM PT] Ubit Umarov: err to save.. [12:55 PM PT] Joe Magarac: Oh, I though you were concerned about assets. [12:55 PM PT] Ubit Umarov: yes [12:55 PM PT] Vincent.Sylvester @hg.zetaworlds.com: IARs and OARs use the same system for this as well [12:56 PM PT] Ubit Umarov: you you save a object you need to save the assets it needs [12:56 PM PT] Ubit Umarov: so a hard scan is done... [12:56 PM PT] Vincent.Sylvester @hg.zetaworlds.com: That reminds me [12:56 PM PT] Ubit Umarov: on each scrip.. notecard... etc [12:57 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If now storing the media texture uuid in the oar does that grab the asset for that texture? [12:57 PM PT] Ubit Umarov: we have a monster called uuidgather to do that :) [12:58 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I suppose it likely is somewhere on the region if you have it set to show video on it or something so likely to catch it otherwise [12:59 PM PT] Ubit Umarov: if it is there on the code.. [12:59 PM PT] Ubit Umarov: land goes not have gather [12:59 PM PT] Joe Magarac: Gotta go, next meeting. Thanks/ [1:00 PM PT] Andrew Hellershanks: ok, Joe. See you next week. [1:00 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I mean that's one texture, that's not that tragic in the end [1:01 PM PT] Ubit Umarov: pff i also need to try to save a oar of this [1:01 PM PT] Andrew Hellershanks: Does anyone have a different topic for todays meeting they want to talk about before others here need to leave? [1:01 PM PT] Gavin.Hird @grid.xmir.org:8002: nope [1:02 PM PT] Andrew Hellershanks: ok, np. Continue. :) [1:02 PM PT] Cuga.Rajal @rajal.org:9000: anyone get webrtc to work? [1:02 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Have yet to try it [1:02 PM PT] Andrew Hellershanks: I haven't tried iet. [1:02 PM PT] Gavin.Hird @grid.xmir.org:8002: haven't tried [1:02 PM PT] Andrew Hellershanks: s/iet/it/ [1:02 PM PT] Cuga.Rajal @rajal.org:9000: k thx :) [1:02 PM PT] Lyr.Lobo @cc.opensimulator.org:8002: sorry, Cuga, me either [1:03 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I'll help you debug if you want [1:03 PM PT] Ubit Umarov: not me [1:03 PM PT] Cuga.Rajal @rajal.org:9000: \need to talk to someone with it avtually working, will ping MrBlue [1:03 PM PT] Ubit Umarov: it is robert own project on his github [1:04 PM PT] Ubit Umarov: (for now at least) [1:04 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Looking forward to it eventually working, but still worry about feasibility, time will tell [1:05 PM PT] Ubit Umarov: maybe a linux only solution [1:05 PM PT] Lyr.Lobo @cc.opensimulator.org:8002: Thanks! very true! Pardon me, stepping away to work and lurking -- have a great day [1:05 PM PT] Cuga.Rajal @rajal.org:9000: and Mac :p [1:05 PM PT] Motoko.Karu @login.digiworldz.com:8002: Bye lyr.... [1:05 PM PT] Andrew Hellershanks: You too Lyr. See you next time. [1:05 PM PT] Cuga.Rajal @rajal.org:9000: tc Lyr [1:05 PM PT] Gavin.Hird @grid.xmir.org:8002: tc Lyr [1:06 PM PT] Vincent.Sylvester @hg.zetaworlds.com: We'll just compile janus with cosmopolitan, I'm sure that'll work first try /s [1:06 PM PT] Cuga.Rajal @rajal.org:9000: janus compiles fine:) I think its the config and.or plugin [1:07 PM PT] Lyr.Lobo @cc.opensimulator.org:8002: You too, Gaving *smiles* [1:07 PM PT] Lyr.Lobo @cc.opensimulator.org:8002: Many thanks to all of you [1:07 PM PT] Ubit Umarov: yeha janus is the thing [1:07 PM PT] Ubit Umarov: does not compile on win [1:08 PM PT] Lyr.Lobo @cc.opensimulator.org:8002: Gavin, Cuga, Andrew, Vincent, Ubit, Joe, Motoko, Jagga, kayaker [1:08 PM PT] Motoko.Karu @login.digiworldz.com:8002: *waves* [1:08 PM PT] Ubit Umarov: :) [1:08 PM PT] Kayaker Magic: Waves and beaches! [1:08 PM PT] Ubit Umarov: el calls [1:08 PM PT] Andrew Hellershanks: :) [1:08 PM PT] Ubit Umarov: and RL [1:08 PM PT] Motoko.Karu @login.digiworldz.com:8002: =) [1:09 PM PT] Ubit Umarov: see you all next week.. here or there.. [1:09 PM PT] Andrew Hellershanks: ok, Ubit. Sounds like it is time to wrap up todays meeting. [1:09 PM PT] Gavin.Hird @grid.xmir.org:8002: have a great week all [1:09 PM PT] Andrew Hellershanks: Thank you all for coming. See you again next week. [1:09 PM PT] Cuga.Rajal @rajal.org:9000: take care everybody :) [20:09 UTC] *** Logging stopped. ***