Réunion du 31-05-2022
Source : http://opensimulator.org/wiki/Chat_log_from_the_meeting_on_2022-05-31
Introduction
[10:59] Andrew Hellershanks : Bonjour, tout le monde.
[10:59] NevenNate Howell : Bonjour
[10:59] Andrew Hellershanks : oooh... chat lag
[11:00] Andrew Hellershanks : Bonjour, Selby.
[11:00] Ubit Umarov : Bonjour.
[11:00] Selby.Evans @grid.kitely.com:8002 : Bonjour à tous.
[11:00] Ubit Umarov : Bienvenue.
[11:00] NevenNate Howell : bonjour
[11:02] Vincent.Sylvester @hg.zetaworlds.com:8002 : Il n'y a pas eu beaucoup d'activité cette semaine, une modification mineure de movetotarget a été effectuée.
[11:02] Vincent.Sylvester @hg.zetaworlds.com:8002 : J'ai quelques points à suivre, mais rien de concret pour l'instant.
[11:02] Andrew Hellershanks : Bonjour, NevenNate. Bienvenue à la réunion des développeurs d'OpenSimulator. Malgré le nom, vous n'avez pas besoin d'être un développeur pour être ici.
Fork Gloebit
[11:03] Vincent.Sylvester @hg.zetaworlds.com:8002 : J'ai effectué quelques modifications mineures sur le fork Gloebit pour corriger certains crashs qu'il provoque, mais je n'ai pas encore terminé.
[11:03] Vincent.Sylvester @hg.zetaworlds.com:8002 : Dans l'idéal, il faudrait que ce module soit en mode failsafe pour ne pas faire planter tout le simulateur.
[11:03] NevenNate Howell : J'ai lu office hours avant de venir.
[11:03] Ubit Umarov : vincent, ce n'est pas un changement du code principal.
[11:03] Ubit Umarov : continuons à parler des changements du code principal d'opensim, ok ?
[11:04] Ubit Umarov : à moins que tu aies soumis un patch, je n'ai pas remarqué :).
[11:04] Vincent.Sylvester @hg.zetaworlds.com:8002 : Nan, je creusais surtout parce que ça montrait le pilote mysql comme étant la cause, mais finalement c'était juste dans le module. Je ne sais pas si tu partages cette opinion, mais je pense que les modules ne devraient jamais causer de crashs par eux-mêmes.
Bogues de mouvement des avatars : llMoveToTarget et mousewalk
[11:04] Ubit Umarov : Et oui, la semaine dernière, nous avons reçu des rapports de 2 problèmes sur le traitement de la récente mise à jour du mouvement des avatars.
[11:06] Ubit Umarov : llmovetotarget sur les pièces jointes ne fonctionnait pas bien... corrigé maintenant.
[11:06] Ubit Umarov : l'autre rapport concernait la fonction "mousewalk" (marcher à l'aide de la souris) d'un viewer.
[11:06] Ubit Umarov : en utilisant cela, l'avatar se déplace maintenant plus lentement.
[11:07] Ubit Umarov : le problème est que le vierwer envoie l'inpulsion pour la marche (?)
[11:07] Ubit Umarov : au moins trois variations.
[11:07] Ubit Umarov : en fait, il pourrait y avoir 9 combinaisons.
[11:08] Ubit Umarov : en utilisant le clavier, il envoie la commande 'nudge mov' (lance le mouvement) pour 250ms de pression sur la touche.
[11:08] Ubit Umarov : puis change pour dire 'move'(bouge) et 'move fast'(bouge vite), 2 flags en même temps.
[11:08] Ubit Umarov : donc il y a un mouvement à basse vitesse pendant 250ms, puis une vitesse normale qui est en fait normale + rapide, actuellement.
[11:09] Ubit Umarov : en plus de cela il y a run (cours) ... mais run est sur un autre message du protocole.
[11:09] Ubit Umarov : maintenant le mousewalk est une option (bidouillage) de viewers qui déclenche les clés.
[11:10] Ubit Umarov : MAIS... ils n'ont jamais réparé le timer..., la souris n'a pas de timer de pression de touche,
[11:10] Ubit Umarov : le code utilise la même pression sur la touche, mais comme il n'y a pas eu de pression sur la touche, le temps de pression est de zéro.
[11:11] Ubit Umarov : donc... le mousewalk indique seulement à la région de déplacer l'avatar lentement...
[11:12] Ubit Umarov : il semble que SL déclasse la qualité des régions, et ils n'ont qu'un seul "speed/strength" donc on ne le remarque pas ici.
[11:12] Ubit Umarov : une telle rétrogradation signifie que le pas minimal de l'avatar est de 1 à 1.5m.
[11:12] Ubit Umarov : plutôt mauvais
[11:13] Ubit Umarov : actuellement nous l'avons divisé par deux.
[11:13] Ubit Umarov : et je ne vais pas casser ça, juste parce que le code de la fonctionnalité du viewer est incomplet.
[11:14] Vincent.Sylvester @hg.zetaworlds.com:8002 : FS a maintenant un incident ouvert pour cela, peut-être qu'ils peuvent envoyer les données correctes, ce serait la meilleure solution.
[11:14] Vincent.Sylvester @hg.zetaworlds.com:8002 : Plus de code ajouté sur le mouvement ne fait que le rendre plus lent et plus lourd, pas besoin d'en rajouter.
[11:16] Ubit Umarov : ils doivent avoir un timer pour ce clic gauche sur l'avatar + clic droit.
[11:16] Ubit Umarov : j'ai aussi remarqué que sur FS, un clic gauche sur une prim avec touch_start, ouvre son éditeur.
[11:17] Ubit Umarov: alchemy ne semble pas faire cela.
[11:17] Ubit Umarov : et il a aussi la fonction "mousewalk".
[11:17] Ubit Umarov : c'est en effet très bien.
[11:18] Ubit Umarov : la vitesse lente est seulement un problème si on veut marcher une longue distance.
[11:18] Ubit Umarov : tout le monde appelle un taxi de nos jours... qui marche ?
[11:19] Ubit Umarov : ;)
[11:19] Vincent.Sylvester @hg.zetaworlds.com:8002 : Je double-clique sur téléportation, c'était cassé dans FS pendant un moment, ça se déclenchait à chaque fois, c'était... pas drôle lol.
[11:19] Ubit Umarov : :)
[11:20] Andrew Hellershanks : Si je dois faire un long trajet dans opensim, je vole (quand c'est autorisé) ou je cours.
[11:20] Ubit Umarov : il vaut mieux que ces actions de la souris soient désactivées.
[11:20] Ubit Umarov : ce n'est pas si facile de ne pas cliquer au mauvais endroit.
[11:21] Ubit Umarov : comme cliquer sur des choses animées.
[11:21] Ubit Umarov : oui, c'est très difficile à faire :)
[11:21] Vincent.Sylvester @hg.zetaworlds.com:8002 : Ce serait bien de l'avoir en combinaison avec une touche modificatrice, mais au moins je n'ai pas besoin d'utiliser la carte pour me téléporter, ce serait pénible.
[11:22] Ubit Umarov : c'est le mousewalk.
[11:22] Ubit Umarov : pas si lent que ça, je suppose.
[11:22] Vincent.Sylvester @hg.zetaworlds.com:8002 : On s'en aperçoit assez vite quand on manipule les diviseurs de la vitesse de marche.
[11:23] Andrew Hellershanks : Je dirais que c'est une vitesse de marche raisonnable.
[11:23] Vincent.Sylvester @hg.zetaworlds.com:8002 : Certaines personnes veulent avoir des boulons partout heh
[11:23] Ubit Umarov : bien sûr, cette région est par défaut.
[11:23] Vincent.Sylvester @hg.zetaworlds.com:8002 : J'ai rendu la marche encore plus lente pour correspondre à l'animation via la constante, donc je ne l'ai pas du tout remarqué.
[11:23] Ubit Umarov : oui, notre marche via le clavier ressemble déjà à une course.
[11:23] Ubit Umarov : voyez
[11:24] Vincent.Sylvester @hg.zetaworlds.com:8002 : marche lunaire en avant.
[11:24] Vincent.Sylvester @hg.zetaworlds.com:8002 : Patinage en longueur
[11:24] Ubit Umarov: hmm?~
[11:24] Andrew Hellershanks : Il y a un peu de ça.
[11:24] Ubit Umarov : Je ne suis pas Mickel Jackson.
[11:24] Ubit Umarov : tu ne fais pas de moonwalk :p
[11:24] Ubit Umarov : err je ne le fais pas
[11:25] Vincent.Sylvester @hg.zetaworlds.com:8002: Avatar key run is faster than I can ride my bike most roads [11:25] Ubit Umarov: seen on a viewer a moonwalk option [11:25] Andrew Hellershanks: The walking animation and forward motion have never been 100% synchronized. [11:26] Ubit Umarov: but didn't test it, when went back looking, did not found it :P [11:26] Ubit Umarov: No [11:26] NevenNate Howell: could there be a set speed, with a is_running bool true/false saved on sim like in most games then x2 speed [11:26] Ubit Umarov: viewers do change animation rate, based on speed [11:26] Ubit Umarov: at least should.. [11:27] Ubit Umarov: there is the always running [11:28] Ubit Umarov: tap tap hold does work on with the mouse [11:28] Andrew Hellershanks: Whether walking or running there is a slight slippage of the foot on the ground that can be seen if one looks closely. [11:29] Ubit Umarov: none of thise things was made to close look [11:29] Andrew Hellershanks: :) [11:29] Ubit Umarov: everything fails on close look [11:30] Ubit Umarov: it still takes a lot of imagination to feel the gaps [11:30] Ubit Umarov: "immersion" [11:31] Ubit Umarov: avatars looking more like cartoons may improve that actually [11:31] Ubit Umarov: brain is ness prone to compare with rl [11:31] Ubit Umarov: but just guessing [11:31] Andrew Hellershanks nods [11:32] Ubit Umarov: well on the other hand brain also does fixe those isues [11:33] Ubit Umarov: bc brain does not see what is there, but what it expects to be there.. [11:33] Ubit Umarov: ( like the mars face.. horses rinding in clouds.. etc etc ) [11:33] Selby.Evans @grid.kitely.com:8002: In RL we have quantum mechanics that is more weird than what I see with close inspection in OS. [11:33] Ubit Umarov: well yeah [11:34] Ubit Umarov: but luckly quatum effects at human size scale are rare things in rl [11:34] Ubit Umarov: tunneling would be disturbing thing [11:35] Ubit Umarov: ohh tanks here have a extra armour thing [11:36] Ubit Umarov: is they stop a X meters from target, bullets tunnel not touching them :P [11:37] Ubit Umarov: ( consequence of simulation only in timesteps of 20ms ))no whisper echo) [11:37] Vincent.Sylvester @hg.zetaworlds.com:8002: Btw the meeting logs for the 24th are still missing from the wiki [11:38] Ubit Umarov: so fun seeing in tv car wheels moving back [11:38] Ubit Umarov: or heli blades, etc [11:38] Kayaker Magic: Have I missed much? [11:38] Ubit Umarov: i did add a limit to angular motion to avoid that here [11:39] Ubit Umarov: not that nice [11:39] Andrew Hellershanks: Vincent, That's odd. I was sure I posted the log to the wiki last week. [11:39] Ubit Umarov: but does irritate me seening in games also, wheels moving back [11:39] NevenNate Howell: talk on mouse movement Kayaker [11:39] Andrew Hellershanks: Hello, Kayaker. [11:39] Andrew Hellershanks: Not that much. [11:39] Ubit Umarov: or WoT tanks tracks [11:40] Ubit Umarov: ( thats nyquist sampling theorem / sampling aliasing ) [11:41] Ubit Umarov: i still didnt look to ODE ray/terrain code :) [11:41] Ubit Umarov: still on TODO [11:41] Selby.Evans @grid.kitely.com:8002: avatar tunneling: Pass through walls You need double-click move (above) turned on. You just need to cam to the inside of a place and use the double-click to there. To cam through a wall, zoom up close to it. Then rotate your view (side-pointing arrow) till you see the other side. This won't work in a place where the TP landing point is set. [11:42] Andrew Hellershanks: Vincent, log has been posted. [11:42] Vincent.Sylvester @hg.zetaworlds.com:8002: Merci [11:43] Ubit Umarov: think i made physics sit on prims a bit tp also :) [11:43] Ubit Umarov: should realy put restrictions on them [11:44] Ubit Umarov: in fact limit those sits to avatars on same parcel [11:44] Andrew Hellershanks: That would be a good idea. [11:45] Ubit Umarov: don't remember not, but at thing they had 10m range [11:45] Ubit Umarov: think [11:45] Ubit Umarov: err and not? now ! [11:46] Ubit Umarov: but sit on a roof or something is a fast way to get someplace :) [11:47] Andrew Hellershanks: That is true but I do find that some objects don't let you sit so that doesn't always work. [11:48] Ubit Umarov: A) you need ubode [11:48] Ubit Umarov: b) the prim must have nice physics mesh [11:49] Andrew Hellershanks: As it doesn't always seem to work with standard prims it could be the regions weren't using ubode. [11:50] Ubit Umarov: it works with standard non phantom prims [11:51] Andrew Hellershanks: We are at 10 minutes left in the hour. Any other topics for discussion today? [11:51] Ubit Umarov: see? [11:52] Andrew Hellershanks nods to Ubit [11:52] Ubit Umarov: if you hit a phantom, it does like bullet.. puts you on a silly place [11:53] Vincent.Sylvester @hg.zetaworlds.com:8002: After the fourth time being catapulted into space I made a habit now to add sit and stand targets to everything [11:53] Ubit Umarov: mesh makers also forget this [11:54] Ubit Umarov: the mesh for physics should also be ok for sits, and also ok for raycasts [11:54] Ubit Umarov: not only script but the camera ones also [11:54] Andrew Hellershanks: I suspect a lot of mesh makers may not think about the raycasting part. [11:55] Ubit Umarov: many don't even think physics :) [11:55] Andrew Hellershanks: :) [11:55] Ubit Umarov: well seems ll plans to change meshs and materials [11:55] Ubit Umarov: so, we may lose viewers, if thy decide to change to that losing "legacy" [11:55] Andrew Hellershanks: Is there any note(s) in the wiki about mesh and what one needs to take in to account re: sits and raycasts? [11:56] Ubit Umarov: possible they can't bc SL is filled with "legacy" [11:58] Ubit Umarov: have a viewer dev giving me such news, as if that will be the end of opensim [11:59] Ubit Umarov: when it will, at most, be the end of that viewer on opensim, unless it stays complatible [11:59] Ubit Umarov: and compatible [11:59] Andrew Hellershanks nods [11:59] Ubit Umarov: but well people said things.. [11:59] Ubit Umarov: lets wait and see [12:00] Andrew Hellershanks: Right. If it is just something that LL is "planning on doing" there is little point in worrying about it until they say they are doing it. [12:01] Ubit Umarov: yeah [12:01] Ubit Umarov: actuall more than that [12:01] Ubit Umarov: until a TPV does add that code is a usable way [12:02] Ubit Umarov: i don't look to SL viewers code, for features [12:02] NevenNate Howell: Their are viewers that are optimized for opensim because they did not like the changes in SL [12:02] Ubit Umarov: those do not work on opensim [12:02] Ubit Umarov: i only look to code of viewers "for opensim" [12:03] Kayaker Magic: I try to go to the TPV meetings, but missed the last few. Have they moved to a different time or location? [12:03] Ubit Umarov: wel and only a few :) [12:03] Ubit Umarov: no idea [12:03] Ubit Umarov: looked on the movies site? [12:04] Ubit Umarov: https://www.youtube.com/c/PanteraPolnocy [12:04] Ubit Umarov: keeps a record of meetings like that [12:04] Ubit Umarov: and other things :) [12:04] Ubit Umarov: last there seems to be may 13 [12:05] NevenNate Howell: Firestorm does not work on standalone out of box by the way I had to download cool viewer [12:05] Ubit Umarov: works fine for mw [12:05] Ubit Umarov: well or did several long weeks ago [12:06] Andrew Hellershanks: Really? I haven't used my Standalone that recently but I have never noticed an issue accessing it with FS. [12:06] NevenNate Howell: I tried mine yesterday new OS and Firestorm and it just hangs [12:06] Ubit Umarov: last time i did run a standalone :) [12:06] NevenNate Howell: I just accepted everything that did not require an entry [12:06] Andrew Hellershanks: NevenNate, What happens when you try using FS with a standalone? [12:07] NevenNate Howell: It just hangs [12:07] Andrew Hellershanks: On startup or when trying to login? [12:07] NevenNate Howell: Cool logs right in but firestorm does nothing but says cannot connect even if I run both at the same time [12:07] NevenNate Howell: login [12:07] Andrew Hellershanks: ok [12:08] Andrew Hellershanks: I'm starting up my Standalone copy of OS. [12:09] Vincent.Sylvester @hg.zetaworlds.com:8002: If you believe there to be an issue with OpenSim in that regard putting a mantis ticket with log files in would help, difficult to debug these things without seeing what goes on [12:09] Vincent.Sylvester @hg.zetaworlds.com:8002: A connection failure and stuck login generally hints at bad network configuration [12:09] NevenNate Howell: Ill try to figure it out I dont think its OS I think its FS [12:09] NevenNate Howell: like I said Cool works fine [12:11] NevenNate Howell: I just don't use SL anymore and see no Firestorm stuff posted for OS so thought I would mention it [12:11] Selby.Evans @grid.kitely.com:8002: Scenegate works fine [12:12] Andrew Hellershanks: Check the OS console for any messages that might indicate the source of the problem. [12:12] Ubit Umarov: working fine for me [12:13] Ubit Umarov: just did update a standalone, fired it and did log fine with fs 66095 [12:13] Ubit Umarov: 6.5.5 [12:13] Selby.Evans @grid.kitely.com:8002: must go -- bye all [12:13] Ubit Umarov: cya [12:13] NevenNate Howell: Ill run stuff on my end and see If I can find the problem then [12:14] NevenNate Howell: ill let you all know [12:14] Andrew Hellershanks: ok, Selby. See you next week. [12:14] NevenNate Howell: good day everyone [12:14] Ubit Umarov: ports at opensim.ini and regions.ini are the usual [12:14] Ubit Umarov: noobs issue :p [12:14] Ubit Umarov: ( and less noobs also ) [12:15] Andrew Hellershanks: I logged in to my standalone using OS that I last built on May 4 [12:17] Ubit Umarov: Oh [12:17] Andrew Hellershanks: NevenNate, let us know what you find. I'm testing it now with OS that I built just now. [12:17] Ubit Umarov: seems metro gird will close [12:17] Ubit Umarov: grid.. [12:17] Ubit Umarov: they have it on their news site [12:18] Vincent.Sylvester @hg.zetaworlds.com:8002: Surprised me quite a bit [12:19] Andrew Hellershanks: I don't think I've ever been there. [12:20] Ubit Umarov: one person that used to be part of the admin, used to be here on al meetings [12:20] Andrew Hellershanks: Hm... my Standalone is throwing out some errors from ubODE -> invalid degenerated mesh [12:20] Ubit Umarov: sheera? [12:21] Andrew Hellershanks: Oh, Sheera. Haven't seen Sheera in a long time. [12:21] Ubit Umarov: yeah guess years even :( [12:21] Ubit Umarov: used to sit here to my left [12:21] Ubit Umarov: hm and did the logs no? [12:22] Andrew Hellershanks: I just logged in to my Standalone using FS 6.5.3 without any problem, [12:23] Vincent.Sylvester @hg.zetaworlds.com:8002: Kinda sad to see them go, they been a constant for a long time [12:23] Andrew Hellershanks: I remember the tiger(?) avatar. [12:23] Ubit Umarov: yes the avatar was like a tiger [12:23] Ubit Umarov: humanoid tiger :) [12:23] Andrew Hellershanks nods [12:24] Andrew Hellershanks: My friends list showed that she came in to this meeting via metro.land. [12:24] Andrew Hellershanks: I should pop over to metro to leave her a message before that grid closes down. [12:25] Andrew Hellershanks: On that note, I think it is time to close down this meeting. [12:25] Ubit Umarov: think she was about to leave metro also [12:25] Ubit Umarov: rl things [12:25] Andrew Hellershanks: We were a small group this week. Thank you all for coming. See you again next week.