00:21 <+bridge_> <learath2> Should have made Turkish the global language, we have no gendered grammar at all
00:42 <+bridge_> <yuphio> Hey is this the spot to request a rename for my account?
00:42 <+bridge_> <gorp_tw> #✉-create-a-ticket Rename
00:42 <+bridge_> <gorp_tw> must have 3k points among other requirements, list is in the channel
01:01 <+bridge_> <haunt_1337> https://media.discordapp.net/attachments/1069667540151844884/1104070676719083622/E3A60D56-D2D2-455B-9C2C-C5A036C5DCAF.gif
01:01 <+bridge_> <vitalya32goda> https://cdn.discordapp.com/attachments/293493549758939136/1465496921232642149/dlc.rpf?ex=697951dd&is=6978005d&hm=ae2fdf8be749a0c66a0fa9cce1bae8a90032bb35b0bbd0022d2ed1c212c9ee1f&
01:01 <+bridge_> <vitalya32goda> https://cdn.discordapp.com/attachments/293493549758939136/1465497016825155725/dlc.rpf?ex=697951f4&is=69780074&hm=98b649a25d01e2a0dc0248c6c88a3fa77bca5c2f2948345b3adc2fd7b838cbb1&
01:02 <+bridge_> <haunt_1337> https://tenor.com/view/alice-in-wonderland-curtsy-gif-10714016786536042726
01:03 <+bridge_> <vitalya32goda> https://cdn.discordapp.com/attachments/1355143911265538098/1396222272649302058/image.gif
01:03 <+bridge_> <vitalya32goda> https://cdn.discordapp.com/attachments/1351295179998625873/1398689711811334174/a72826b45ffaea84ba31a2743417b5df.gif
01:03 <+bridge_> <haunt_1337> https://media.discordapp.net/attachments/879122843160440903/923982054314635364/image0.gif
01:03 <+bridge_> <haunt_1337> https://cdn.discordapp.com/attachments/660629930845143050/1371825779901599895/IMG_9207.gif
01:04 <+bridge_> <haunt_1337> https://media.discordapp.net/attachments/1183101985491255475/1197230234068598885/8175dc9eb380419d.gif
03:38 <+bridge_> <gorp_tw> ts guy is not ready for code tutorial, he is ready to sail a ship
03:38 <+bridge_> <gorp_tw> https://cdn.discordapp.com/attachments/293493549758939136/1465536438773878867/image.png?ex=697976ab&is=6978252b&hm=c93b10d2b3ce1bdb77873a3da22c0b7618e6204b961e9e3f012a0529ddd219b6&
04:02 <+bridge_> <gorp_tw> that was not fun
04:02 <+bridge_> <gorp_tw> https://cdn.discordapp.com/attachments/293493549758939136/1465542506140799119/image.png?ex=69797c51&is=69782ad1&hm=1b6e9386b7be648b8a01182069b12da4a4f51817b8b4cfc710614120647fc121&
04:03 <+bridge_> <gorp_tw> uses whole range of svg box, math is hard
06:55 <+bridge_> <alw5> XD D
07:23 <+bridge_> <remakepower> I think we can remove most of those pronouns as they could be removed.
07:25 <+bridge_> <remakepower> For example,
07:25 <+bridge_> <remakepower> ```
07:25 <+bridge_> <remakepower> // The player has only one ground jump, so their feet are always dark
07:25 <+bridge_> <remakepower> ```
07:25 <+bridge_> <remakepower> could be
07:25 <+bridge_> <remakepower> ```
07:25 <+bridge_> <remakepower> // The player has only one ground jump, so the feet are always dark
07:25 <+bridge_> <remakepower> ```
07:31 <+bridge_> <remakepower> Anyway, 'they' is often considered as plural around the world, so I think it would better use plurals instead.
07:32 <+bridge_> <remakepower> -#  I agree that pronouns are trash.
07:54 <+bridge_> <chillerdragon> Yes I noticed that too!
07:55 <+bridge_> <chillerdragon> I don’t think it’s fun .\_.
07:56 <+bridge_> <chillerdragon> irc uses mattermost and matrix changes every 2 weeks ask zwelf
07:56 <+bridge_> <chillerdragon> matterbridge\*
08:23 <+bridge_> <Assa> and @kebs I can explain, there are multiple types of draggers, they can have different speeds and they can be hook-through-walls
08:24 <+bridge_> <Assa> multiple speeds are done by making the head tick faster so the head rotates faster
08:24 <+bridge_> <Assa> and the hook-through-walls draggers rotate counter-clock-wise by making the headtick go negative
08:25 <+bridge_> <Assa> @kebscs
09:17 <+bridge_> <Solly (bun bun)> this is an interseting thing
09:18 <+bridge_> <Solly (bun bun)> i like the old green but i guess an icon is better (colorblind)
09:41 <+bridge_> <gorp_tw> ok time for bed
09:41 <+bridge_> <gorp_tw> https://cdn.discordapp.com/attachments/293493549758939136/1465627730836979712/image.png?ex=6979cbb1&is=69787a31&hm=f75ebd62e768c7576cff2aed12ff53cb070ddb20aacede6120a6db6eb21c50dc&
09:42 <+bridge_> <gorp_tw> gotta do line guides tomorrow
09:47 <+bridge_> <PRO_Gaming1921> ey
10:01 <+bridge_> <gorp_tw> i guess for now this is more readable without any guides
10:01 <+bridge_> <gorp_tw> https://cdn.discordapp.com/attachments/293493549758939136/1465632712445530235/image.png?ex=6979d054&is=69787ed4&hm=29f4086294a21f2190c3aceef8309084a748c8dba730538666bbed830d6e9dc3&
10:08 <+bridge_> <manikoo> need scale
10:08 <+bridge_> <manikoo> of points
10:18 <+bridge_> <Assa> I think this is unreadable without a comment in the code 🙈
11:19 <+bridge_> <heinrich5991> none at all? 😮
11:19 <+bridge_> <heinrich5991> not even pronouns?
11:20 <+bridge_> <learath2> As in personal pronouns? Nope, we have 1st person, 2nd person and 3rd person, singular and plural
11:20 <+bridge_> <heinrich5991> interesting
11:20 <+bridge_> <learath2> He and She both map to `O` in Turkish
11:23 <+bridge_> <learath2> We also don’t have articles really. Sometimes the number 1 gets used as a sort of pseudo indefinite article
11:23 <+bridge_> <Assa> I got envelope sounds working correctly 🥳
11:23 <+bridge_> <heinrich5991> the protocol or the tiles? 🙂
11:24 <+bridge_> <Assa> the protocol is already there - I just need to test it, which is actually the last step before I am finished 😄
11:25 <+bridge_> <Assa> should I add a console command to trigger envelopes? Otherwise this is currently unused in ddnet
11:25 <+bridge_> <Assa> should I add a server console command to trigger envelopes? Otherwise this is currently unused in ddnet
11:27 <+bridge_> <heinrich5991> not sure how we best go about testing it
11:27 <+bridge_> <heinrich5991> maybe add a command for testing, but we should probably remove it eventually
11:28 <+bridge_> <Assa> I can also only make it available in a debug build
11:28 <+bridge_> <Assa> (I know we hate this branching code)
11:28 <+bridge_> <heinrich5991> it's okay, as long as we plan to remove it again, you can add it for testing
11:34 <+bridge_> <Assa> using a netobject makes it totally not trivial
11:53 <+bridge_> <Assa> I know this is probably not how I should send a netobject, but how do I set the message id?
11:53 <+bridge_> <Assa> 
11:53 <+bridge_> <Assa> ```
11:53 <+bridge_> <Assa> void CGameContext::ConTriggerEnvelope(IConsole::IResult *pResult, void *pUserData)
11:53 <+bridge_> <Assa> {
11:53 <+bridge_> <Assa>     CGameContext *pSelf = (CGameContext *)pUserData;
11:53 <+bridge_> <Assa>     if(pResult->NumArguments() != 2)
11:53 <+bridge_> <Assa>         return;
11:53 <+bridge_> <Assa>     dbg_msg("dbg", "envelope object send");
11:54 <+bridge_> <Assa>     CNetObj_EnvelopeTrigger *pTriggerMessage = pSelf->Server()->SnapNewItem<CNetObj_EnvelopeTrigger>(pResult->GetInteger(1));
11:54 <+bridge_> <Assa>     pTriggerMessage->m_StartTick = pSelf->Server()->Tick();
11:54 <+bridge_> <Assa>     pTriggerMessage->m_Type = pResult->GetInteger(0);
11:54 <+bridge_> <Assa> }
11:54 <+bridge_> <Assa> ```
11:54 <+bridge_> <Assa> 
11:54 <+bridge_> <Assa> shouldn't the snap item number do the trick?
11:54 <+bridge_> <Assa> I know that it should undo itself after 1 tick, but I don't receive this at all at the client side
11:56 <+bridge_> <Assa> no this is correct hmm
12:07 <+bridge_> <Assa> putting this into the gamecontroller snap makes it working :3
12:12 <+bridge_> <01000111g> it might be cool for tournament "loading screens" though
12:13 <+bridge_> <Assa> totally possible rn, already working - you could even add a loading sound
12:13 <+bridge_> <Assa> totally possible rn, already working - you could even add a loading sound which stops when finished
12:21 <+bridge_> <Assa> hmm `2026-01-27 12:21:34 I console: Command 'env_trigger' cannot be executed from a map.
12:22 <+bridge_> <Assa> 2026-01-27 12:21:34 I console: Command 'env_trigger' cannot be executed from a map.`
12:22 <+bridge_> <Assa> 
12:22 <+bridge_> <Assa> I think there is a console flag missing maybe?
12:22 <+bridge_> <Assa> I think I made the first command that only needs `CFGFLAG_GAME`
12:23 <+bridge_> <Assa> if you do this the server crashes
12:24 <+bridge_> <Assa> no, the server crashes because I set the callback to nullptr and this is not checked 🙁
12:26 <+bridge_> <manikoo> nullptr and nullptr.sg are great maps
12:34 <+bridge_> <blazu da apol> https://github.com/ddnet/ddnet-rs-rfc/issues/2 blehh
12:34 <+bridge_> <blazu da apol> thoughts on this?
12:35 <+bridge_> <blazu da apol> (this is an idea for ddnet-rs though, not ddnet)
12:36 <+bridge_> <patiga> I also believe ddnet should have music at some point. I could imagine it like the game "Human fall flat" does it. I guess it is similar to minecraft, but to me the vibe feels more appropriate
12:37 <+bridge_> <blazu da apol> I will take a peep at how HFF does it
12:37 <+bridge_> <blazu da apol> I used to play it but it was so long ago I didn't even remember it has music x)))
12:54 <+bridge_> <Assa> ```
12:54 <+bridge_> <Assa> 2026-01-27T11:52:00.4974735Z   File "/home/runner/work/ddnet/ddnet/scripts/export_settings_commands_table.py", line 38, in parse_arguments
12:54 <+bridge_> <Assa> 2026-01-27T11:52:00.4975838Z     raise RuntimeError(f"Failed to parse {name} arguments: {arguments_line}") from e
12:54 <+bridge_> <Assa> 2026-01-27T11:52:00.4977589Z RuntimeError: Failed to parse command arguments: "env_trigger", "i[zone] s[trigger_type] i[env]", CFGFLAG_GAME, [](IConsole::IResult *, void *) {}, this, "Client side only: Set a trigger type for an env in a trigger zone"
12:54 <+bridge_> <Assa> ```
12:54 <+bridge_> <Assa> 
12:54 <+bridge_> <Assa> ja gg, the argument parser crashes on valid C++ code, I'll need to add an empty command
13:58 <+bridge_> <learath2> You probably figured out by now but NetObjects aren't things you just randomly send whenever, it needs to happen during snap
13:58 <+bridge_> <learath2> So your command would set some state and when snapping that state would be checked to add the netobject
14:19 <+bridge_> <Assa> yes I figured it out and updated the description in the PR
14:19 <+bridge_> <Assa> I honestly thought I can do this whenever and the snap would just do it's thing
14:24 <+bridge_> <learath2> Did it error out btw? Maybe we should assert it if it doesn't, you should never do snapnewitem if a snap isn't happening. (Probably should even be reworked so it's an explicit object that you need to be passed, like a `CSnapHandle` so you can't even call `SnapNewItem` without it)
14:25 <+bridge_> <Assa> it did silently do nothing
14:26 <+bridge_> <learath2> Definitely should assert it, I'll keep it in mind
14:38 <+bridge_> <Assa> hoi, do you know why chiller marked you for review on the envelope trigger tiles or do you want to be informed for every possible new tile?
14:39 <+bridge_> <heinrich5991> marked for review due to code owners file. @patiga wanted to be informed about map file changes
14:39 <+bridge_> <Assa> yes - but not in this case. I didn't change mapio, chiller requested a review explicitly
15:48 <+bridge_> <kebscs> https://youtu.be/3kddz-ZHRDY?si=LzOEOyzgVKFHcMWQ
20:58 <+bridge_> <kebscs> https://github.com/ddnet/ddnet/pull/11682
20:58 <+bridge_> <kebscs> any review? 🙂
21:28 <+bridge_> <alw5> is the weak rocket bounce used in any maps?
21:28 <+bridge_> <alw5> is the weak rocket jump used in any maps?
21:28 <+bridge_> <alw5> dont think so right?
21:39 <+bridge_> <kebscs> its same height as just jumping
21:50 <+bridge_> <gorp_tw> I'm thinking about Rank stars
21:51 <+bridge_> <gorp_tw> so for this graph https://discord.com/channels/252358080522747904/293493549758939136/1465632712567160833 i may want to display a line where Brutal 1* started, maybe where Moderate 5* started if the graph is low enoough
21:51 <+bridge_> <gorp_tw> so for this graph https://discord.com/channels/252358080522747904/293493549758939136/1465632712567160833 i may want to display a line where Brutal 1* started, maybe where Moderate 5* started if the graph is low enough
21:51 <+bridge_> <gorp_tw> I think this would be better than some 1600 1650 1700 etc scale
21:52 <+bridge_> <gorp_tw> since it's less human readable
21:58 <+bridge_> <totar> @kebscs did you profile the new event prediction system?
21:58 <+bridge_> <totar> it looks sorta slow
22:01 <+bridge_> <kebscs> not really
22:01 <+bridge_> <kebscs> is fps significantly lower or whats slow
22:02 <+bridge_> <totar> it just looks scary
22:03 <+bridge_> <totar> CreatePredictedEvent could scale very poorly
22:04 <+bridge_> <kebscs> yea idk what else to do here
22:04 <+bridge_> <kebscs> it shouldnt have many sounds anyway, so its not big problem
22:04 <+bridge_> <totar> idk
22:05 <+bridge_> <kebscs> most if not all predicted sounds are usually main+dummy
22:05 <+bridge_> <totar> you must be careful tho
22:05 <+bridge_> <totar> the client already lags from prediction quite noticably at like 1000ping
22:05 <+bridge_> <totar> the client already lags from prediction quite noticeably at like 1000ping
22:06 <+bridge_> <totar> what was wrong with the old trick?
22:07 <+bridge_> <totar> it seemed fine
22:09 <+bridge_> <kebscs> what trick?
22:09 <+bridge_> <totar> the one that the old predicted events used
22:09 <+bridge_> <totar> ``m_aLastNewPredictedTick[Dummy]``
22:10 <+bridge_> <kebscs> so the jump sound?
22:10 <+bridge_> <totar> these
22:10 <+bridge_> <totar> ```cpp
22:10 <+bridge_> <totar>             if(g_Config.m_ClPredict && !m_SuppressEvents)
22:10 <+bridge_> <totar>                 if(Events & COREEVENT_AIR_JUMP)
22:10 <+bridge_> <totar>                     m_Effects.AirJump(Pos, 1.0f, 1.0f);
22:10 <+bridge_> <totar>             if(g_Config.m_SndGame && !m_SuppressEvents)
22:10 <+bridge_> <totar>             {
22:10 <+bridge_> <totar>                 if(Events & COREEVENT_GROUND_JUMP)
22:10 <+bridge_> <totar>                     m_Sounds.PlayAndRecord(CSounds::CHN_WORLD, SOUND_PLAYER_JUMP, 1.0f, Pos);
22:10 <+bridge_> <totar>                 if(Events & COREEVENT_HOOK_ATTACH_GROUND)
22:10 <+bridge_> <totar>                     m_Sounds.PlayAndRecord(CSounds::CHN_WORLD, SOUND_HOOK_ATTACH_GROUND, 1.0f, Pos);
22:10 <+bridge_> <totar>                 if(Events & COREEVENT_HOOK_HIT_NOHOOK)
22:10 <+bridge_> <totar>                     m_Sounds.PlayAndRecord(CSounds::CHN_WORLD, SOUND_HOOK_NOATTACH, 1.0f, Pos);
22:10 <+bridge_> <totar>                 if(Events & COREEVENT_HOOK_ATTACH_PLAYER)
22:10 <+bridge_> <totar>                 {
22:10 <+bridge_> <totar>                     m_PredictedWorld.CreatePredictedSound(Pos, SOUND_HOOK_ATTACH_PLAYER, pLocalChar->GetCid());
22:10 <+bridge_> <totar>                 }
22:10 <+bridge_> <totar>             }
22:10 <+bridge_> <totar> ```
22:11 <+bridge_> <kebscs> they arent predicted they just spawn on client-side from gamecore
22:11 <+bridge_> <totar> wdym
22:11 <+bridge_> <totar> that doesn't change how the deduplication logic works tho
22:12 <+bridge_> <kebscs> i dont get what you mean, the COREEVENT, were spawned on client-side, not predicted or sent from server
22:13 <+bridge_> <kebscs> hook attach was the only one that was sent from server, because it relies on other players so it wasnt spawned client-sidde
22:13 <+bridge_> <kebscs> hook attach was the only one that was sent from server, because it relies on other players so it wasnt spawned client-side
22:14 <+bridge_> <kebscs> so if it worked like these coreevents, i would need to mute all sounds from the server, and that is not possible as not everything is predicted
22:18 <+bridge_> <kebscs> i dont remember exactly, the pr was open for 2 months, but ive had problems with these coreevents
22:19 <+bridge_> <totar> the reason that double jump and hook attach don't trigger multiple times is because it checks 
22:19 <+bridge_> <totar> 
22:19 <+bridge_> <totar> ``Tick > m_aLastNewPredictedTick[Dummy]``
22:19 <+bridge_> <totar> and then sets
22:19 <+bridge_> <totar> ``m_aLastNewPredictedTick[Dummy] = Tick;``
22:19 <+bridge_> <totar> 
22:19 <+bridge_> <totar> Why can't you do this for all the predicted events?
22:19 <+bridge_> <totar> instead of the complex duplication checker
22:19 <+bridge_> <totar> would it not work?
22:20 <+bridge_> <kebscs> havet tried that
22:20 <+bridge_> <totar> I don't really see any reason it wouldnt
22:21 <+bridge_> <totar> I don't really see any reason it wouldnt work
22:21 <+bridge_> <totar> also infinitely faster
22:21 <+bridge_> <totar> and simpler
22:22 <+bridge_> <kebscs> so same logic, but HandlePredictedEvents in `Tick > m_aLastNewPredictedTick`?
22:22 <+bridge_> <totar> I haven't read your code enough to know where to put it
22:23 <+bridge_> <totar> but if you apply the same idea I think it should work
22:23 <+bridge_> <totar> btw how do you prevent events that the server sends from causing duplicate sounds?
22:24 <+bridge_> <kebscs> comparing position and tick
22:25 <+bridge_> <kebscs> if theres a missmatch or smth itll be a duplicate
22:25 <+bridge_> <kebscs> `CheckPredictedEventHandled`
22:25 <+bridge_> <totar> hmm
22:26 <+bridge_> <totar> did the server not even send sounds for the previous predicted events?
22:26 <+bridge_> <kebscs> previously predicted events are coreevents?
22:27 <+bridge_> <totar> the logical distinction between core and gameworld isn't really important
22:27 <+bridge_> <kebscs> if you mean `COREEVENT_AIR_JUMP` then the server doesnt send it
22:27 <+bridge_> <kebscs> idk why air jump needed the extra m_aLastNewPredictedTick
22:28 <+bridge_> <totar> because otherwise you would cause an air jump every time OnPredict() runs
22:28 <+bridge_> <kebscs> thats why its broken in demo
22:28 <+bridge_> <totar> right
22:28 <+bridge_> <kebscs> yes but why attach ground didnt need it
22:28 <+bridge_> <totar> what about hook sounds?
22:28 <+bridge_> <totar> did the server send those
22:28 <+bridge_> <totar> those work in demos?
22:29 <+bridge_> <kebscs> no, all coreevents are clientsided in gamecore.cpp
22:30 <+bridge_> <kebscs> yes
22:30 <+bridge_> <totar> but
22:30 <+bridge_> <totar> they need prediction to even function?
22:31 <+bridge_> <totar> does OnPredict run with cl_predict 0?
22:31 <+bridge_> <kebscs> idk
22:31 <+bridge_> <kebscs> this is some jank
22:31 <+bridge_> <kebscs> https://cdn.discordapp.com/attachments/293493549758939136/1465821616398405874/image.png?ex=697a8042&is=69792ec2&hm=bb098a1acc58be84aa5e6af4b1e03fbefd182662da6418b6ae479ad9f40d80df&
22:31 <+bridge_> <kebscs> for other players air jumpo
22:31 <+bridge_> <kebscs> for other players air jump
22:32 <+bridge_> <totar> I mean for hook sounds
22:32 <+bridge_> <totar> like if you recorded a demo with cl_predict 0 you need OnPredict() to run for them to exist in the demo
22:32 <+bridge_> <totar> which implies it always runs
22:33 <+bridge_> <kebscs> doesnt seem like the sounds are tied to cl_predict
22:33 <+bridge_> <totar> I know
22:33 <+bridge_> <totar> this confuses me
22:34 <+bridge_> <totar> we run OnPredict() even in spectate?
22:35 <+bridge_> <totar> https://github.com/ddnet/ddnet/issues/1282
22:35 <+bridge_> <totar> this is so easy to fix
22:36 <+bridge_> <kebscs> idk
22:44 <+bridge_> <totar> why do we check for ``m_SuppressEvents`` in OnPredict
22:44 <+bridge_> <totar> I guess it's fine but it's never used
22:46 <+bridge_> <totar> you could easily fix air jumps in demos if you just make this work for the local player during demo playback
22:46 <+bridge_> <totar> it won't conflict with prediction
22:50 <+bridge_> <kebscs> idk i dont want to touch prediction atm
22:51 <+bridge_> <kebscs> at most i could fix your tele prediction bcs i understand whats going on there
22:53 <+bridge_> <gorp_tw> I might have found the most deranged response to an issue ever
22:53 <+bridge_> <gorp_tw> https://github.com/ddnet/ddnet/issues/10262
22:53 <+bridge_> <gorp_tw> This is how terminals work. DDNet is a terminal.
22:54 <+bridge_> <gorp_tw> All users are smart! They will never paste anything into chat.
22:54 <+bridge_> <gorp_tw> CMClient is patching this forcibly
22:56 <+bridge_> <gorp_tw> someone even points out this is how 0.6.5 does it and literally every other game ive seen XD
22:56 <+bridge_> <totar> I find it annoying that the console auto sends
22:56 <+bridge_> <gorp_tw> It should not read control codes as "sends"
22:56 <+bridge_> <gorp_tw> that is dangerous probably
22:57 <+bridge_> <totar> very often I will copy an IP, and then type "connect ctrl+v" and it will join the server as soon as I paste because there was a new line at the end of the ip
22:57 <+bridge_> <totar> which is not at all what I wanted because I wanted to edit the port
22:57 <+bridge_> <gorp_tw> xDDDDDDDD
22:57 <+bridge_> <gorp_tw> trailing whitespace should be stripped anyway
22:58 <+bridge_> <gorp_tw> when do you ever need it? in chat it's removed
22:58 <+bridge_> <gorp_tw> there are no multiline fields in ddnet that im aware of?
22:59 <+bridge_> <totar> there's another issue that needs resolving
22:59 <+bridge_> <totar> @learath2 when you have time maybe we can chat again 🙂
23:00 <+bridge_> <gorp_tw> Never mind the fact that this can quickly get you close to muted by the server
23:01 <+bridge_> <kebscs> could you put everything into the github pr, so we could help you resolve them?
23:01 <+bridge_> <kebscs> rn seems like pr is abandoned when draft and no updates
23:08 <+bridge_> <gorp_tw> I could see someone hiding a disconnect/shutdown bomb in a long list of F2 commands shutting a server down
23:08 <+bridge_> <gorp_tw> anyone stupid enough to paste into F2 directly though probably shouldn't have pressed F2
23:08 <+bridge_> <totar> @kebscs in a perfect world we would be able to play sound events halfway through their playback, but idk if thats possible. Either way I would probably do this.
23:08 <+bridge_> <totar> 
23:08 <+bridge_> <totar> Record all snap sounds from the server as normal but *dont* play them when  prevented events are active, then all sounds will come from new tick gated prediction. So we will lose some sound effects when mispredicts happen but thats already how it works for hook and core events, plus the ones we lose would be out of sync anyway
23:09 <+bridge_> <totar> @kebscs in a perfect world we would be able to play sound events halfway through their playback, but idk if thats possible. Either way I would probably do this.
23:09 <+bridge_> <totar> 
23:09 <+bridge_> <totar> Record all snap sounds from the server as normal but *dont* play them when  predicted events are active, then all sounds will come from new tick gated prediction. So we will lose some sound effects when mispredicts happen but thats already how it works for hook and core events, plus the ones we lose would be out of sync anyway
23:10 <+bridge_> <kebscs> my solution works too
23:10 <+bridge_> <totar> hide full range ban+shutdown :troll:
23:11 <+bridge_> <kebscs> detecting if sound is already playing is probably more expensive to what i do now
23:11 <+bridge_> <totar> Im not sure its actually better
23:11 <+bridge_> <totar> When did I say we had to do this?
23:11 <+bridge_> <kebscs> "we would be able to play sound events halfway through their playback"
23:12 <+bridge_> <gorp_tw> xD
23:12 <+bridge_> <totar> Thats what fighting games with rollback netcode do but Im assuming we dont have that capability so we just play them on new tick instead
23:12 <+bridge_> <totar> Ideally the entire sound playback state at any given time is reset whenever we do a full prediction run
23:13 <+bridge_> <kebscs> i would say most games play sounds based on whats happening on client side visuals
23:13 <+bridge_> <kebscs> rather than with events from server
23:13 <+bridge_> <totar> Not true
23:13 <+bridge_> <totar> Overwatch does it as ddnet did prior to your pr
23:13 <+bridge_> <totar> They intentionally play bullet hit effects out of sync when ping is over a certain threshold to prevent mispredicted effects
23:14 <+bridge_> <kebscs> how did they solve it
23:14 <+bridge_> <kebscs> so its not laggy
23:14 <+bridge_> <totar> It is laggy
23:14 <+bridge_> <kebscs> me > overwatch
23:14 <+bridge_> <kebscs> 😂
23:14 <+bridge_> <totar> When your ping is <80 they use yout system, when its over 80 they use olf system
23:14 <+bridge_> <totar> They do both
23:14 <+bridge_> <totar> When your ping is <80 they use yout system, when its over 80 they use old system
23:15 <+bridge_> <totar> When your ping is <80 they use your system, when its over 80 they use old system
23:15 <+bridge_> <totar> Because if you hit someone and then it mispredicts the play posts a clip online and says "billion dollar game studio cant figure out working netcode"
23:16 <+bridge_> <totar> Because if you hit someone and then it mispredicts the player posts a clip online and says "billion dollar game studio cant figure out working netcode"
23:16 <+bridge_> <kebscs> true tho
23:16 <+bridge_> <kebscs> if i thought of the solution in my free time without looking up anything
23:16 <+bridge_> <kebscs> and they use same way
23:17 <+bridge_> <totar> Theres a finite number of solutions, lots of games converge to the same thing
23:20 <+bridge_> <kebscs> @totar
23:20 <+bridge_> <kebscs> i would love the tele predict pr be merged :)
23:20 <+bridge_> <totar> I dont want to rush it
23:20 <+bridge_> <totar> Unless you can get them to drop backcompat
23:22 <+bridge_> <kebscs> but were not rushing, youre stalling if you have problems and not share them
23:23 <+bridge_> <kebscs> and old clients dont predict anyway so theres no backcompat
23:23 <+bridge_> <totar> Im not obligated to share my thoughts with you instantly upon having them
23:23 <+bridge_> <kebscs> i have not said that
23:24 <+bridge_> <totar> There will be back compat when we realize that the implementation that got merged is broken and needs fixing
23:24 <+bridge_> <kebscs> yes thats why there are reviews, so not everything is on your shoulders
23:24 <+bridge_> <totar> The reviewers have demonstrated that they are unable to spot the issues I found
23:25 <+bridge_> <totar> So its on my shoulders
23:26 <+bridge_> <kebscs> idk if the projectile tele out is a problem if its mixed with clientid and current tick too
23:27 <+bridge_> <kebscs> just adding direction and its fine
23:29 <+bridge_> <totar> Its really not hard to make a setup that causes issues
23:29 <+bridge_> <totar> Just look up into a looping teleporter and toy can get collisions between two projectiles easily
23:29 <+bridge_> <totar> If done wrong
23:30 <+bridge_> <totar> Just look up into a looping teleporter and you can get collisions between two projectiles easily
23:31 <+bridge_> <kebscs> to me its not a problem that should delay it for months, if its broken in only very specific, unlikely scenario
23:32 <+bridge_> <kebscs> and not really backcompat issue bcs this isnt a netmsg, smth like that could be changed later
23:37 <+bridge_> <kebscs> and its not "broken" if im thinking correctly, just not fully random
23:43 <+bridge_> <totar> you absolutely cannot easily "change later" because if you change the seeding then you need the client to support connecting to servers that use either version of the physics
23:44 <+bridge_> <kebscs> ddnet updates the servers, idk if any server owner will sit specifically on this broken version
23:44 <+bridge_> <totar> not fully random *is* broken. if you go on the map "Avoid" with the first version of the tele PR it's completely broken
23:45 <+bridge_> <totar> over time the bullets will randomly teleport on top of each other and then they never get seperated again which completely changes the map
23:45 <+bridge_> <kebscs> i see
23:45 <+bridge_> <kebscs> what about current version of pr?
23:46 <+bridge_> <totar> it's fixed for that specific case but I want it to be indistinguishable from the previous behavior from a player and mapper perspective
23:49 <+bridge_> <kebscs> was anything wrong with the "real" start tick?
23:49 <+bridge_> <totar> no
23:50 <+bridge_> <totar> the new issue is that we sometimes have player projectiles without the player
23:50 <+bridge_> <totar> which breaks things
23:52 <+bridge_> <kebscs> i have never noticed that
23:52 <+bridge_> <kebscs> just sometimes id is -1 on spawn