|
|
| (34 versions intermédiaires par le même utilisateur non affichées) |
| Ligne 1 : |
Ligne 1 : |
| = Changements du code de la semaine= | | = Changements du code de la semaine= |
| * Commit "cosmétique" [https://bitbucket.org/opensimulator/opensim/commits/] : il semble que seul les commits soient à jour sur Bitbucket ce qui n'est pas le cas sur Github ou sur Git principal. | | * Commit "cosmétique" [https://bitbucket.org/opensimulator/opensim/commits/] |
| | * Il semble que seuls les commits soient à jour sur Bitbucket ce qui n'est pas le cas sur Github ou sur Git principal. |
|
| |
|
| = Avertissement = | | = Avertissement = |
| Ligne 11 : |
Ligne 12 : |
| * Autre exemple : [https://wiki.secondlife.com/wiki/LlRezObjectWithParams llRezObjectWithParams] | | * Autre exemple : [https://wiki.secondlife.com/wiki/LlRezObjectWithParams llRezObjectWithParams] |
|
| |
|
| = Base de données =
| |
| = Modules =
| |
| = Bugs = | | = Bugs = |
| * Quelques discussions sur Mantis [http://opensimulator.org/mantis/my_view_page.php]
| | === Quelques discussions sur Mantis [http://opensimulator.org/mantis/my_view_page.php]=== |
| | * [http://opensimulator.org/mantis/view.php?id=9100 0009100 : Les opérateurs LSL ne sont pas conformes aux spécifications] |
| | * [http://opensimulator.org/mantis/view.php?id=9098 0009098 : Specified Cast is not valid on region restart (La caste spécifiée n'est pas valide sur le redémarrage de la région)] Voir le chapitre [[Réunion du 14-11-2023#Problème de migration de la base de données | Problème de migration de la base de données]] de la réunion du 14-11-2023. Résolution et mise en place d'une routine de vérification des tables afin de s'assurer qu'elles ont le bon schéma, sinon arrêt du démarrage. |
| | |
| | ===Bugs de textures === |
| | * Problèmes de textures à la connexion ou textures corrompues invisibles. |
| | * '''Ce sont des textures corrompues parfois importées et recopiées dans tout le Métaverse. Les assets avec un problème peuvent être des JP2 ou des meshes '''qui peuvent même provoquer des crashs de région ou de viewer. |
| | * Essayer de nettoyer cela est un peu une bataille perdue d'avance, une importation aléatoire par quelqu'un et tout revient. Il serait probablement plus facile de les remplacer par des fichiers vierges si les données sont mauvaises. |
| | * '''En général, si ce n'est pas rouge, on peut l'ignorer, les avertissements jaunes sont moches, mais pas souvent nuisibles.''' |
|
| |
|
| = Tests =
| |
| = Projets en cours / Infos= | | = Projets en cours / Infos= |
| === Nouveautés et informations Second Life === | | === Nouveautés et informations Second Life === |
| Ligne 26 : |
Ligne 32 : |
| ==== SLVoice.exe==== | | ==== SLVoice.exe==== |
| * SL a commencé à placer les derniers développements de SLVoice.exe derrière une URL api.github réservé aux développeurs de Viewers tiers (TPV) enregistrés. | | * SL a commencé à placer les derniers développements de SLVoice.exe derrière une URL api.github réservé aux développeurs de Viewers tiers (TPV) enregistrés. |
| | |
| | ===[https://fr.wikipedia.org/wiki/Rendu_Physique_R%C3%A9aliste Rendu physique réaliste (PBR)]=== |
| | * Le développement semble être au point mort. |
| | |
| | ===Script PHP de recherche de textures === |
| | * '''Kayaker Magic a écrit un script PHP''' pour trouver et localiser (coordonnées et face) toutes les textures cachées dans les prims d'une région. Il aimerait ajouter les textures '''bump et shiny'''. Ces textures seraient dans '''un asset "material"'''. '''Seul l'UUID de l'asset est présent dans le blob de la texture'''. Donc il y a un UUID dans Texture qui pointe vers un autre asset dans la base de données. |
| | |
| | ===Mise à jour de [http://opensimulator.org/wiki/BulletSim BulletSim] === |
| | * '''BulletSim est le module pour OpenSimulator qui implémente la physique''' du monde virtuel en utilisant le moteur physique Bullet. Ce module fournit une physique de haute performance ainsi qu'une performance physique des véhicules compatible avec Second Life. |
| | * '''Intervenants''' : Cuga.Rajal et MisterBlue |
| | * '''Une mise à jour devrait être faite avant l'OSCC''' ( début du mois de décembre ). Beaucoup de travail de développement a été réalisé, principalement pour corriger des bogues. |
| | * Cuga.Rajal et MisterBlue ont construit et testé la version Bullet 3 - V 3.25 depuis un certain temps. |
| | * Il est prévu de construire un nouvel ensemble complet de librairies Bullet non gérées pour toutes les plateformes. Elles auront toutes la même fonctionnalité. La dernière version du tronc Bullet3, qui est la 3.25 et la dernière mise à jour du wrapper Bullet de MisterBlue 1.1) seront utilisées. |
| | * La bibiolthèque MacOs est prête et testée. MisterBlue travaille sur les autres. |
| | * MisterBlue a créé un script qui automatisera la construction pour toutes les plateformes. Il a même créé une action GitHub pour les construire toutes. |
| | * Le wrapper montre aussi la version actuelle de Bullet maintenant, ce qui n'était pas le cas auparavant. La version utilisée sur OpenSim de Bullet est la v2.86, qui a au moins 12 ans. |
| | * La version 3.25 pourrait être la dernière mise à jour, à moins qu'il n'y ait plus de mouvement de ce côté-là. |
| | * Les bogues qui existent encore sont dans le connecteur, pas dans la librairie Bullet. Ils pourraient être corrigés sans mettre à jour la librairie non gérée. Ubit Umarov et MisterBlue pourraient s'entendre sur ce sujet. |
| | * Mantis : [http://opensimulator.org/mantis/view.php?id=9086 0009086 : llDetectedLinkNumber()] La fonction rapporte un numéro de lien incorrect lors d'une collision avec un jeu de liens activés par la physique, Bullet uniquement, anomalies liées. |
| | |
| | ===[https://github.com/jrouwe/JoltPhysics Jolt Physics]=== |
| | * Moteur physique assez moderne qui a des liens C#. Il alimente certains jeux de catégorie A. OpenSim pourrait peut-être s'en inspirer ? |
| | |
| | ===OSCC === |
| | * Consultez la page web pour vous inscrire à l'événement et pour vérifier le programme de la conférence : https://conference.opensimulator.org/ |
| | * '''N'oubliez pas que certains événements sont généralement organisés avant le week-end de la conférence.''' |
|
| |
|
| = Viewers= | | = Viewers= |
| | === Dayturn === |
| | * Gavin.Hird se demande si SL ne veut pas contrôler les développeurs de Viewer et s'il n'aura pas à développer lui-même toutes les bibliothèques comme le fait Firestorm. |
| | * Depuis le 18 novembre 2023, openjpeg est passé à la version 2.5.0 sur le viewer test et sur le viewer Mac. |
| | |
| = Source= | | = Source= |
| http://opensimulator.org/wiki/Chat_log_from_the_meeting_on_2023-11-21 | | http://opensimulator.org/wiki/Chat_log_from_the_meeting_on_2023-11-21 |
|
| |
| <!--
| |
| [12:11 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Things still being changed on it so before that calms down no point in adding, but will be considered at some point once SL has fallen on its face so we don't have to
| |
| [12:11 PM PT] Vincent.Sylvester @hg.zetaworlds.com: In other news we had a cosmetics commit this week and some rumbling on mantis
| |
| [12:12 PM PT] Ubit Umarov: wel it was supposed to be the last week one..
| |
| [12:12 PM PT] Gavin.Hird @grid.xmir.org:8002: They usually have a feature freeze from this week till over new years
| |
| [12:12 PM PT] Gavin.Hird @grid.xmir.org:8002: at least for what is released to the users
| |
| [12:12 PM PT] Ubit Umarov: for some reason only got into bitbucket repo, not github or main git
| |
| [12:12 PM PT] Gavin.Hird @grid.xmir.org:8002: for SL you mean?
| |
| [12:13 PM PT] Gavin.Hird @grid.xmir.org:8002: SL repos*
| |
| [12:13 PM PT] Ubit Umarov: i meant last week code changes on opensim..
| |
| [12:13 PM PT] Gavin.Hird @grid.xmir.org:8002: oh, ok
| |
| [12:13 PM PT] Gavin.Hird @grid.xmir.org:8002: I still fetch it from bitbucket
| |
| [12:14 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Ubit's git musta contracted OpenSim voodoo :D
| |
| [12:14 PM PT] Andrew Hellershanks: For prim edit rights there is already a mechanism for handling that unrelated to friends. Of course, you normally will only give a friend rights.
| |
| [12:14 PM PT] Ubit Umarov: the "cosmetics commit" was not there except on bitbucket
| |
| [12:14 PM PT] Ubit Umarov: guess the dev forgot to push :p
| |
| [12:14 PM PT] Ubit Umarov: coff coff coff ;)
| |
| [12:15 PM PT] Ubit Umarov: yeah SL usualy stops new releases during xmas period
| |
| [12:15 PM PT] Andrew Hellershanks: :)
| |
| [12:16 PM PT] Gavin.Hird @grid.xmir.org:8002: looks like LL has started hiding the latest SLVoice.exe builds behind an api.github url, so they may start to reserve ot for registered TPV devs as they do with some of the other libs
| |
| [12:16 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Mantis summary is mostly me finding something off with the logical operators not being evaluated to spec and after the ticket about migrations breaking I been working on a checking routine for the tables to make sure they are of proper schema else halt startup
| |
| [12:17 PM PT] Ubit Umarov: did read somewhere that ll is compiling the dep libs.. only for 64bit
| |
| [12:17 PM PT] Ubit Umarov: not only slvoice
| |
| [12:17 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Perhaps they are moving Vivox versions after all, which could mean actual shutdown of v4, in the worst case scenario
| |
| [12:17 PM PT] Gavin.Hird @grid.xmir.org:8002: I was not talking about that
| |
| [12:18 PM PT] Gavin.Hird @grid.xmir.org:8002: there is no point in building 32-bit viewers anyway these days
| |
| [12:18 PM PT] Ubit Umarov: yeah all conspired to kill 32bit
| |
| [12:18 PM PT] Ubit Umarov: .. on intel...
| |
| [12:19 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: if they are doing that, does that focus us a little more on voice development?
| |
| [12:19 PM PT] Gavin.Hird @grid.xmir.org:8002: it could be, but it can also be they want better control over TPV viewer builds
| |
| [12:19 PM PT] Gavin.Hird @grid.xmir.org:8002: moving all the lib and viewer builds to github is a major pita
| |
| [12:19 PM PT] Ubit Umarov: not sure what control they will have on the rust one :)
| |
| [12:19 PM PT] Gavin.Hird @grid.xmir.org:8002: that they do I mean
| |
| [12:19 PM PT] Ubit Umarov: on the unity one.., etc
| |
| [12:20 PM PT] Ubit Umarov: or..
| |
| [12:20 PM PT] Gavin.Hird @grid.xmir.org:8002: might have to start building all the libs myself like FS do
| |
| [12:20 PM PT] Ubit Umarov: on the other hand others even asked ll to sign their code... a very strange concept..
| |
| [12:21 PM PT] Gavin.Hird @grid.xmir.org:8002: ... actually they don't do all, but many
| |
| [12:21 PM PT] Ubit Umarov: think fs does does ll ones also
| |
| [12:21 PM PT] Ubit Umarov: does use...
| |
| [12:21 PM PT] Gavin.Hird @grid.xmir.org:8002: they use ll libs, but build some themselves too
| |
| [12:21 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Some of the media related ones I know they roll their own on
| |
| [12:22 PM PT] Ubit Umarov: well at least hacd , you need to glue yourself
| |
| [12:22 PM PT] Ubit Umarov: and openjpeg
| |
| [12:22 PM PT] Ubit Umarov: since sl uses payed libs
| |
| [12:23 PM PT] Gavin.Hird @grid.xmir.org:8002: just updated openjpeg from 1.5.2 I built myself to teh ll 2.5.0 version
| |
| [12:23 PM PT] Andrew Hellershanks: brb
| |
| [12:23 PM PT] Gavin.Hird @grid.xmir.org:8002: so that is out on the test viewer and the Mac viewer since Saturday
| |
| [12:24 PM PT] Ubit Umarov: :)
| |
| [12:25 PM PT] Ubit Umarov: did you looked to the pbr things?
| |
| [12:26 PM PT] Bill's Chair: This is Bill's Chair
| |
| [12:26 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: hi jamie
| |
| [12:26 PM PT] Bill's Chair: This is Bill's Chair
| |
| [12:26 PM PT] Ubit Umarov: yeah Bill's chair i still reserved for him :)
| |
| [12:27 PM PT] Ubit Umarov: is...
| |
| [12:27 PM PT] Kayaker Magic: Speaking of PBR things, I'm trying to find the old bumpyness and shinyness textures in the databases. I can read the Texture blob but cannot find bump and shiny. Anybody know where I should look?
| |
| [12:28 PM PT] Jamie.Jordan @grid.kitely.com:8002: hey sorry i'm late, wasn't paying attention to thw chair my bad
| |
| [12:28 PM PT] Ubit Umarov: they are on a material asset
| |
| [12:28 PM PT] Gavin.Hird @grid.xmir.org:8002: who looked at teh pbr tthings?
| |
| [12:28 PM PT] Kayaker Magic: I've written a PHP script that finds all unreadable textures in all prims in a region and tells you where they are and which face. I'd like to add bump and shiny.
| |
| [12:29 PM PT] Ubit Umarov: i was asking if you did look to it for dayturn :)
| |
| [12:29 PM PT] Gavin.Hird @grid.xmir.org:8002: I ahve looked at it for months, yes
| |
| [12:29 PM PT] Kayaker Magic: I'm tired of getting 100 bad texture errors from region startup and map generation.
| |
| [12:29 PM PT] Ubit Umarov: the texture blod has uuids for such "materials"
| |
| [12:29 PM PT] Gavin.Hird @grid.xmir.org:8002: but it also seem their development has stalled
| |
| [12:29 PM PT] Kayaker Magic: I've created objects with known UUIDs for bump and shiny, and I don't see them in the Texture blob.
| |
| [12:30 PM PT] Vincent.Sylvester @hg.zetaworlds.com: They are not necessarily missing textures, they could be dynamic textures too or some residual bakes stuff
| |
| [12:30 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: hi cuga
| |
| [12:30 PM PT] Ubit Umarov: kayaker i just told that those are stored on a material asset
| |
| [12:30 PM PT] Kayaker Magic: Yeah, I test for missing and for errors parsing the JP2 data.
| |
| [12:30 PM PT] Cuga.Rajal @rajal.org:9000: Hi all! Sorry late
| |
| [12:30 PM PT] Ubit Umarov: only that asset uuid is present on the blob
| |
| [12:31 PM PT] Kayaker Magic: So there is a UUID in Texture that points to another asset in the database?
| |
| [12:31 PM PT] Ubit Umarov: yes
| |
| [12:31 PM PT] Kayaker Magic: OK, I'll look for that!
| |
| [12:31 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If you have one of those uuids I'll see if I can get you an image
| |
| [12:32 PM PT] Kayaker Magic: ?if I get errors decoding JP2 files, you think you can recover them Vincent?
| |
| [12:33 PM PT] Vincent.Sylvester @hg.zetaworlds.com: If they exist as assets they should be available to convert, things getting corrupted on asset end would mean some bad data entry somewhere
| |
| [12:33 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Given all the textures are still there on most regions I visit that can't be happening that often
| |
| [12:34 PM PT] Vincent.Sylvester @hg.zetaworlds.com: I see texture errors whenever I login too, no idea what they are from they show even if I reset my appearance
| |
| [12:34 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Probably some bakes or something non critical else I think it would have come up on mantis a lot more
| |
| [12:35 PM PT] Kayaker Magic: I'm finding old Arcadia Asylum builds brought over from SL that have corrupted textures on faces that aren't visible. Then people copied those all over the Metaverse and I see dozens of errors on map generation.
| |
| [12:35 PM PT] Ubit Umarov: well that happens with stolen things..
| |
| [12:36 PM PT] Kayaker Magic: That also.
| |
| [12:36 PM PT] Lyr Lobo: Hi Jamie
| |
| [12:37 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Trying to clean that up is a bit of losing battle, one random import by someone and it all comes back. Probably be easier just to replace them with blank ones if the data is bad
| |
| [12:37 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Which it should already do tho
| |
| [12:37 PM PT] Cuga.Rajal @rajal.org:9000: Are the assets with issues only JP2?
| |
| [12:38 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Nah mesh can be bad also, even cause hard crashes with regions and viewers
| |
| [12:38 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Not much input validation going on
| |
| [12:38 PM PT] Ubit Umarov: all complain the little one added to meshes :p
| |
| [12:39 PM PT] Ubit Umarov: soem even did publish articles on how to remove that code..
| |
| [12:39 PM PT] Ubit Umarov: whatever..
| |
| [12:39 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Ignorance is a hell of drug
| |
| [12:39 PM PT] Cuga.Rajal @rajal.org:9000: I had recent issues with very old hair mesh, which is why I changed my hair this year. Older viewers didn't do much import valiidation.
| |
| [12:39 PM PT] Cuga.Rajal @rajal.org:9000: So issues not limited to stolen atuff
| |
| [12:40 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Mesh I uploaded a couple days ago gives zero length errors too, cause I committed the cardinal sin to mess with physics, I take full responsibility for playing with fire heh
| |
| [12:41 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: lol
| |
| [12:41 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Usually if it ain't red you can ignore it, yellow warnings are ugly, but not often harmful
| |
| [12:41 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: ok - here goes - any progress on the voice issue?
| |
| [12:42 PM PT] Andrew Hellershanks: I'm back and caught up with chat.
| |
| [12:42 PM PT] Vincent.Sylvester @hg.zetaworlds.com: You'll probably be asking that for a couple months, something that complex takes time to even research properly
| |
| [12:43 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: so be it
| |
| [12:43 PM PT] Lyr Lobo: welcome back
| |
| [12:43 PM PT] Cuga.Rajal @rajal.org:9000: I have some new news about Bullet
| |
| [12:43 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Do tell
| |
| [12:44 PM PT] Cuga.Rajal @rajal.org:9000: So I nudged MisterBlue last week, because he aand I were promising some Bullet uypdates at the last OSCC
| |
| [12:44 PM PT] Cuga.Rajal @rajal.org:9000: And we have been having some discussions
| |
| [12:44 PM PT] Cuga.Rajal @rajal.org:9000: He fees (and I agree) that we should make an update before OSCC
| |
| [12:45 PM PT] Gavin.Hird @grid.xmir.org:8002: what will de update do?
| |
| [12:45 PM PT] Cuga.Rajal @rajal.org:9000: There has been a lot of dev work, the main hangip was bugs found
| |
| [12:45 PM PT] Gavin.Hird @grid.xmir.org:8002: the*
| |
| [12:45 PM PT] Cuga.Rajal @rajal.org:9000: But the bugs are not new ones, I'm just testing it more thoroughly than it has been in the pasdt
| |
| [12:45 PM PT] Cuga.Rajal @rajal.org:9000: he and I have been building and testing the latest version of Bullet 3 - v 3.25 for some time now
| |
| [12:46 PM PT] Cuga.Rajal @rajal.org:9000: and he also updated the wrapper with some improvements
| |
| [12:46 PM PT] Vincent.Sylvester @hg.zetaworlds.com: You mentioned something about detected faces and the falling asleep of physics last time, was that resolved?
| |
| [12:46 PM PT] Cuga.Rajal @rajal.org:9000: He was reluctant to release then with more testing, but really I've been testing them all year
| |
| [12:46 PM PT] Cuga.Rajal @rajal.org:9000: after some discussion, he is comfortable doing a release build sometime before this OSCC
| |
| [12:47 PM PT] Cuga.Rajal @rajal.org:9000: So, we plan to build a new complete set of Bullet unmanaged libs for all the platforms
| |
| [12:47 PM PT] Cuga.Rajal @rajal.org:9000: they will all have feature parity
| |
| [12:47 PM PT] Cuga.Rajal @rajal.org:9000: Will use the latest from Bullet3 trunk, which is 3.25
| |
| [12:47 PM PT] Cuga.Rajal @rajal.org:9000: and the latest updates in his Bullet wrapper
| |
| [12:47 PM PT] Cuga.Rajal @rajal.org:9000: which I think he numbered 1.1
| |
| [12:48 PM PT] Cuga.Rajal @rajal.org:9000: I already have the macOS lib ready and tested
| |
| [12:48 PM PT] Cuga.Rajal @rajal.org:9000: he is working on all the others
| |
| [12:48 PM PT] Cuga.Rajal @rajal.org:9000: He made a script that will automate the build for all the platforms, and even created a GitHub actions to build them all
| |
| [12:49 PM PT] Cuga.Rajal @rajal.org:9000: I think he is only struggline with the Windows one
| |
| [12:49 PM PT] Cuga.Rajal @rajal.org:9000: I did ask him what date I could tell this group about when they would be ready
| |
| [12:49 PM PT] Cuga.Rajal @rajal.org:9000: \he said, sometime before OSCC, ike early December-ish
| |
| [12:50 PM PT] Cuga.Rajal @rajal.org:9000: Bullet 3 did get some updates in recent months although they did not bump the version number
| |
| [12:50 PM PT] Cuga.Rajal @rajal.org:9000: presumably bug fixes
| |
| [12:50 PM PT] Gavin.Hird @grid.xmir.org:8002: brb
| |
| [12:50 PM PT] Cuga.Rajal @rajal.org:9000: I have a macOS lib ready if anyone wants to test it
| |
| [12:50 PM PT] Cuga.Rajal @rajal.org:9000: the otherrs are coming soon
| |
| [12:50 PM PT] Cuga.Rajal @rajal.org:9000: I think thats all on Bullet, unless anyone has questions
| |
| [12:51 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Did that fix the bugs you had with it?
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: No
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: But those bugs are not new ones
| |
| [12:51 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Ah so there'll be potentially another version to address those later?
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: Just only discovered and documented recently
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: Cirrest
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: Correst*
| |
| [12:51 PM PT] Cuga.Rajal @rajal.org:9000: Correct
| |
| [12:52 PM PT] Cuga.Rajal @rajal.org:9000: He looked at the bugs and they are difficult ones that will require a deep dive
| |
| [12:52 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Godspeed to him then, sounds fun heh
| |
| [12:53 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: will this update beefit end-users?
| |
| [12:53 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: benefit
| |
| [12:53 PM PT] Cuga.Rajal @rajal.org:9000: Yes, it should have a number of bug fixes in the core Bullet, and the wrapper is a bit more effecient
| |
| [12:53 PM PT] Cuga.Rajal @rajal.org:9000: The wrapper also shows the actual Bullet version now, previously did not
| |
| [12:54 PM PT] Cuga.Rajal @rajal.org:9000: The Bullet currently released is v2.86, which is like 12 years old at least
| |
| [12:54 PM PT] Cuga.Rajal @rajal.org:9000: I mean, the version being used in Opensim
| |
| [12:54 PM PT] Cuga.Rajal @rajal.org:9000: So an update has been a bit overdue I think
| |
| [12:55 PM PT] Ubit Umarov: 3.25 is like 8yrs old?
| |
| [12:55 PM PT] Cuga.Rajal @rajal.org:9000: Not that old
| |
| [12:55 PM PT] Cuga.Rajal @rajal.org:9000: For some reason they stopped bumping the rev number some time ago, but there has been active work and a number of bug fixes even in the last 2 yearsd
| |
| [12:55 PM PT] Ubit Umarov: ok pybullet 3.25 is last year
| |
| [12:56 PM PT] Cuga.Rajal @rajal.org:9000: It's even received updates less than 6 mos ago
| |
| [12:56 PM PT] Cuga.Rajal @rajal.org:9000: So we were thinking of documenting the got branch ID or pull date, to document the version
| |
| [12:57 PM PT] Cuga.Rajal @rajal.org:9000: git*
| |
| [12:57 PM PT] Cuga.Rajal @rajal.org:9000: To be honest I can't see any obvious visual difference between old and new on use cases
| |
| [12:58 PM PT] Cuga.Rajal @rajal.org:9000: But I have not collected any detailed metrics, CPU time, etc
| |
| [12:58 PM PT] Ubit Umarov: seems pretty much dead, with exception of a few prs
| |
| [12:58 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Speaking of, I did recently hear that there is a pretty modern physics engine that has some C# bindings, JoltPhysics, which powers some triple A games. Possible something we could learn some tricks from. :)
| |
| [12:59 PM PT] Cuga.Rajal @rajal.org:9000: Yeah, 3.25 might be the last version update unless there is more movement over there
| |
| [12:59 PM PT] Ubit Umarov: there are several c# enginers now
| |
| [12:59 PM PT] Tess JL: Time for me to go. Have a lovely week everybody!
| |
| [12:59 PM PT] Ubit Umarov: ppl love to use wrong tools :p
| |
| [12:59 PM PT] Cuga.Rajal @rajal.org:9000: The bugs that still exist are in the connector, not the Bullet lib
| |
| [1:00 PM PT] Cuga.Rajal @rajal.org:9000: So potentially could be fixed without updating the unmanaged lib
| |
| [1:00 PM PT] Ubit Umarov: one is beuphysics
| |
| [1:00 PM PT] Ubit Umarov: one is bepuphysics
| |
| [1:01 PM PT] Cuga.Rajal @rajal.org:9000: Ubit I hope you and MisterBlue can connect on that
| |
| [1:01 PM PT] Cuga.Rajal @rajal.org:9000: The C## stuff is a little beyond my capabilities
| |
| [1:01 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Well bindings and native are different things, C# for physics is not a good fit yeah, c strapped with C# works better
| |
| [1:01 PM PT] Ubit Umarov: well i have no news on those changes to bullet
| |
| [1:01 PM PT] Cuga.Rajal @rajal.org:9000: Hopefully I dicumented the bugs well enough in Mantis
| |
| [1:02 PM PT] Cuga.Rajal @rajal.org:9000: Seems I'm the on;y one who is testing physics engines that thoroughly
| |
| [1:02 PM PT] Ubit Umarov: Alicia did use that bepu asn seems worked ok
| |
| [1:03 PM PT] Cuga.Rajal @rajal.org:9000: Some issues might be due to XEngine - YEngine changes
| |
| [1:03 PM PT] Cuga.Rajal @rajal.org:9000: MrBlue never went back to look at that
| |
| [1:03 PM PT] Ubit Umarov: ofc call c# to what that really uses is abuse
| |
| [1:03 PM PT] Cuga.Rajal @rajal.org:9000: *shrugs*
| |
| [1:03 PM PT] Ubit Umarov: uses all the *unsafe* nonsense dotnet5 added
| |
| [1:04 PM PT] Vincent.Sylvester @hg.zetaworlds.com: C flexibility producing a cursed mess heh
| |
| [1:04 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Don't remind me, cringe*
| |
| [1:05 PM PT] Cuga.Rajal @rajal.org:9000: So thats the scoop on Bullet.. expect an update around early Dec, but not really anything major
| |
| [1:05 PM PT] Cuga.Rajal @rajal.org:9000: other than the big jumo in version #
| |
| [1:06 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: yEngine got tighter in nessie too
| |
| [1:06 PM PT] Cuga.Rajal @rajal.org:9000: If there's somebody who is really passionate about Bullet, and has the programming chops
| |
| [1:06 PM PT] Andrew Hellershanks: I recently watched a video about someone who provides comments on someones code and optimizes it. One thing I learned from it is that just because a language has some (new) feature doesn't mean you should use it. Sometimes sticking to older (and often simpler?) ways of doing things works better.
| |
| [1:06 PM PT] Cuga.Rajal @rajal.org:9000: We could use the heslp
| |
| [1:06 PM PT] Cuga.Rajal @rajal.org:9000: I agree Andrew
| |
| [1:07 PM PT] Cuga.Rajal @rajal.org:9000: But I wouldn;t say that is a hard rule, depends on the case and all the considerations
| |
| [1:07 PM PT] Ubit Umarov: a lot of new c# things are pure sugar
| |
| [1:07 PM PT] Andrew Hellershanks: Cuga, right. The best option is try it both ways and see which works best.
| |
| [1:07 PM PT] Ubit Umarov: compiler tricks only
| |
| [1:08 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Even if you cut the amount of machine instructions down doesn't mean you save time
| |
| [1:08 PM PT] Ubit Umarov: but a few are pure hacks around the main rules of c# like string invariance or array bounds checks etc
| |
| [1:08 PM PT] Ubit Umarov: those are nonsense incoerence
| |
| [1:08 PM PT] Andrew Hellershanks: The video I watched was related to ray tracing code. It may have used lambdas (or some other C feature) and the code took 7 minutes to run. Changing one bit to a simple for loop and it ran in 23 seconds without other changes.
| |
| [1:08 PM PT] Vincent.Sylvester @hg.zetaworlds.com: The thing that broke the mysql connectors a while ago, pure silliness
| |
| [1:09 PM PT] Ubit Umarov: but in use now on their codem and part of reason why dotnet seems faster now
| |
| [1:10 PM PT] Vincent.Sylvester @hg.zetaworlds.com: Trimming things out does make it run faster heh that's just physics
| |
| [1:10 PM PT] Ubit Umarov: oops rl calls
| |
| [1:10 PM PT] Ubit Umarov: cya ppl
| |
| [1:11 PM PT] Vincent.Sylvester @hg.zetaworlds.com: mono died so dotnet could be faster lol
| |
| [1:11 PM PT] Andrew Hellershanks: ok, Ubit. See you next week.
| |
| [1:11 PM PT] Andrew Hellershanks: Does anyone else have a question they wanted to ask but didn't get to ask yet?
| |
| [1:12 PM PT] Andrew Hellershanks: Lyr, is registration open for this years OSCC?
| |
| [1:14 PM PT] Kayaker Magic: I'm pretty sure it is open.
| |
| [1:14 PM PT] Jamie.Jordan @grid.kitely.com:8002: have a great week yall
| |
| [1:14 PM PT] Andrew Hellershanks: Lyr must be afk at the moment. We are just a couple, or so, weeks away from this years OSCC. Check the web page to register for the event and to check for the conference schedule.
| |
| [1:15 PM PT] Andrew Hellershanks: Keep in mind that there are usually some events held in advance of the actual weekend of the conference.
| |
| [1:15 PM PT] Clifford.Hanger @alternatemetaverse.com:8002: ty everyone
| |
| [1:15 PM PT] Andrew Hellershanks: With that, I'll wrap up todays gathering.
| |
| [1:15 PM PT] Andrew Hellershanks: Thank you all for coming. See you again next week.
| |
|
| |
| -->
| |