00:33 < bridge> <AssassinTee> did something change with the new version regarding ghosts?
00:46 < bridge> <Izanagi> it doesn't work like that, prediction margin adds delay to your already existing ping delay
00:47 < bridge> <Izanagi> and anti ping will try to compensate that
00:47 < bridge> <Izanagi> instead of the other way around, that it adds delay to anti ping itself like i wanted to
01:58 < bridge> <Cipy29> Bruv it only takes like 2-3 weeks to get used to antiping then you're better forever, stop silly idea of adding extra delay lol
02:01 < bridge> <Izanagi> it's not silly
03:28 < bridge> <triki> i use this and i feel diff when hammerwall with 48 and 8 ping. its not same.
03:28 < bridge> <triki> https://cdn.discordapp.com/attachments/293493549758939136/1071618233221779597/image.png
03:29 < bridge> <triki> and it is not work if tee on other side unfrezzed.
04:16 < bridge> <louis> its silly
04:17 < bridge> <louis> i remember antiping having some sort of 'minimum ping' cutoff, but they removed it
04:18 < bridge> <louis> also game just looks smoother during drag parts etc when simulating 0 ping , there's no time delay when hooking, it's instantaneous and feels fast
04:27 < bridge> <Izanagi> I’ll have to completely rewire my monkey brain for that
04:27 < bridge> <Izanagi> I guess I’ll try
04:28 < bridge> <Izanagi> What can i lose
04:28 < bridge> <Izanagi> Except for lots and lots of fails in the first few weeks
04:29 < bridge> <louis> i remember when jawsh first switched he couldnt finish tsunami, now he is a huge gores pro
04:30 < bridge> <Izanagi> So i will also transition to a QR code player? :justatest:
06:30 < bridge> <KDaniel> Since I don't have all of you as friends on discord and I can't find some of you on GitHub... I just start the discussion here, where also not mentioned people can form an opinion.
06:30 < bridge> <KDaniel> @deen @Avolicious @ReiTW @heinrich5991 @ChillerDragon @fokkonaut
06:30 < bridge> <KDaniel> Sorry for the mass ping 😦
06:30 < bridge> <KDaniel> https://cdn.discordapp.com/attachments/293493549758939136/1071664187467649074/TeeworldsWorldCupConcept.pdf
06:35 < bridge> <KDaniel> good night 🙂
08:42 < bridge> <Ryozuki> mass ping spam  :troll:
09:47 < bridge> <<~{Barsik}~>> mass ping spam  :troll:
10:00 < bridge> <Skeith> Ya, swapping to antiping was very worth it, imo. Took me awhile to get used to it, swapped about a year ago now I think
10:02 < bridge> <Skeith> The fact that it makes catching tees as simple as hooking them directly, is one of the better reasons to swap over
10:03 < bridge> <Skeith> Never have to play that little guessing game again, unless it's a part that requires movement from the tee as you hook them 😅
10:05 < bridge> <Skeith> 13 years of not using antiping, if I can do it, anyone can. Just accept that you're going to play worse for a bit
10:06 < bridge> <Skeith> 13 years of not using antiping, if I can do it, anyone can. Just accept that you're going to play worse for a bit, you should end up better than you were
10:07 < bridge> <fokkonaut> Stop trying to convince everyone to use Antiping, thanks
10:07 < bridge> <chillerdragon> Antiping propaganda
10:08 < bridge> <Skeith> I used to hate antiping too :kek:
10:08 < bridge> <Skeith> I was fucking dumb
10:09 < bridge> <fokkonaut> missspelled my name xd
10:09 < bridge> <chillerdragon> „All Four disciplines“ \:triggered\:
10:09 < bridge> <chillerdragon> (@KDaniel)
10:10 < bridge> <fokkonaut> Point is I dont hate it, but you hate it now to play without.
10:10 < bridge> <Skeith> I can play without still
10:11 < bridge> <Skeith> It's just worse, far less consistent. Why would I choose to do that to myself. I can freely play on any server I want
10:11 < bridge> <chillerdragon> Let’s fix bad servers and ddos for the sake of tournaments @Learath2 how come you never thought of tournaments to fix ddos?
10:21 < bridge> <chillerdragon> So you need mappers for multiple new ddrace,gores,fng and block maps.Devs to build a website with accounts and stats tracking and it has a api to publish game scoresCustom server modifications to publish these scores (and thus also to track the games)ddos protection (lol)And randomly expect a bunch of people to cooperate. I can see some of the random selection of people you pinged being not interested.That’s quite big
10:21 < bridge> <chillerdragon> (@KDaniel)
13:32 < bridge> <KDaniel> Mappers doing maps all they long.. like every year this would be 7 maps for ddrace and gores..
13:32 < bridge> <KDaniel> 
13:32 < bridge> <KDaniel> The website thing is that what I want to build. It just tracks during the tournement like all other tournement right now.. no one talked about accounts.
13:32 < bridge> <KDaniel> Yea ddos(lol) I know it.
13:32 < bridge> <KDaniel> 
13:32 < bridge> <KDaniel> - 7 new maps per year for big event 
13:33 < bridge> <KDaniel> - it's just a simple website where you can register and other can see the current state of the tournement and provides API to fill the lists with information.
13:33 < bridge> <KDaniel> - yea the idea to bring the gamemmodes together and show  some blocker that gore's exists. 
13:33 < bridge> <KDaniel> - the only thing is some help with hosting or just some mental support. 
13:33 < bridge> <KDaniel> 
13:33 < bridge> <KDaniel> The whole thing can done by just 1 person because everything is already done except of the website but this project would not work without people.
13:33 < bridge> <KDaniel> But thanks for your opinion:)
13:34 < bridge> <KDaniel> Mappers doing maps all they long.. like every year this would be 7 maps for ddrace and gores..
13:34 < bridge> <KDaniel> 
13:34 < bridge> <KDaniel> The website thing is that what I want to build. It just tracks during the tournement like all other tournement right now.. no one talked about accounts.
13:34 < bridge> <KDaniel> Yea ddos(lol) I know the struggle of none. 😄
13:34 < bridge> <KDaniel> 
13:34 < bridge> <KDaniel> - 7 new maps per year for big event 
13:34 < bridge> <KDaniel> - it's just a simple website where you can register and other can see the current state of the tournement and provides API to fill the lists with information.
13:34 < bridge> <KDaniel> - yea the idea to bring the gamemmodes together and show  some blocker that gore's exists. 
13:34 < bridge> <KDaniel> - the only thing is some help with hosting or just some mental support. 
13:34 < bridge> <KDaniel> 
13:34 < bridge> <KDaniel> The whole thing can done by just 1 person because everything is already done except of the website but this project would not work without people.
13:34 < bridge> <MilkeeyCat> do everything yourself :gigachad:
14:47 < bridge> <chillerdragon> Okay maybe no accounts with password but some kind of user tracking for teams and leaderboard
14:48 < bridge> <chillerdragon> But yea I could host stuff if you want. Got no fancy ddos protection tho
14:55 < bridge> <Ryozuki> @Jupeyy_Keks i got my new keyboard
14:55 < bridge> <Ryozuki> i like the feeling
14:55 < bridge> <Ryozuki> but i gotta get used to it
14:55 < bridge> <Jupstar ✪> with speed mx?
14:55 < bridge> <Ryozuki> yw
14:55 < bridge> <Jupstar ✪> yeah
14:55 < bridge> <Jupstar ✪> its epic but i know what u mean
14:55 < bridge> <Jupstar ✪> its weird first time u use it
14:55 < bridge> <Ryozuki> i also mean the layout
14:55 < bridge> <Jupstar ✪> oh ok 😄
14:56 < bridge> <Ryozuki> xd i have the other imprintend into memory muscle
14:56 < bridge> <Ryozuki> this one is a bit different
14:58 < bridge> <Ryozuki> the keyboard itself
14:58 < bridge> <Ryozuki> is beautiful
15:22 < bridge> <DefaultO> are there ways to make ddnets network traffic footprint less?
15:22 < bridge> <DefaultO> The internet speed was halved for me. and now this game is unplayable.
15:23 < bridge> <DefaultO> Something like not subscribing to garbage information about people outside your team or something (for example)
15:24 < bridge> <DefaultO> (World of Tanks runs smoothly for reference)
15:25 < bridge> <DefaultO> are there ways to make ddnets network traffic footprint less?
15:25 < bridge> <DefaultO> The internet speed was halved for me. and now this game is unplayable. anti ping makes it just worse.
15:25 < bridge> <DefaultO> are there ways to make ddnets network traffic footprint less?
15:25 < bridge> <DefaultO> The internet speed was halved for me. and now this game is unplayable. anti ping makes it just worse (my ping is "normal").
15:26 < bridge> <Jupstar ✪> what is your new internet speed?
15:26 < bridge> <Jupstar ✪> do you have cable internet now?
15:28 < bridge> <DefaultO> My PC supports both. But I use cable.
15:28 < bridge> <DefaultO> https://cdn.discordapp.com/attachments/293493549758939136/1071799472419516456/image.png
15:28 < bridge> <DefaultO> should be around 6MB/s
15:28 < bridge> <DefaultO> should be around 6MB/s download, upload is garbage
15:29 < bridge> <Jupstar ✪> i dont mean a LAN cable
15:29 < bridge> <Jupstar ✪> i mean cable, instead of DSL or fiber
15:29 < bridge> <DefaultO> no clue
15:29 < bridge> <DefaultO> but sounds like it doesnt it
15:30 < bridge> <Jupstar ✪> anyway with ctrl + shift + d u see the network footprint, but your internet should be good enough
15:30 < bridge> <DefaultO> yea but I lag nonstop
15:30 < bridge> <DefaultO> it really is unplayable
15:30 < bridge> <Jupstar ✪> but your upload is really trash
15:35 < bridge> <DefaultO> so I can't do anything and its the upload that causes lagspikes nonstop?
15:37 < bridge> <Jupstar ✪> no i dont think so
15:37 < bridge> <Jupstar ✪> but if you have cable internet. the upload is defs a problem
15:38 < bridge> <Jupstar ✪> cable internet by design is a problem
15:44 < bridge> <DefaultO> I assume DSL is cable
15:44 < bridge> <DefaultO> https://cdn.discordapp.com/attachments/293493549758939136/1071803520803672144/image.png
15:44 < bridge> <Jupstar ✪> yeah
15:45 < bridge> <Jupstar ✪> but the upload is completely different to what u posted before
15:46 < bridge> <DefaultO> yea but what I posted before is what I have on my PC
15:46 < bridge> <Jupstar ✪> i dunno maybe this is a temporary problem for you/vodaphone
15:46 < bridge> <Jupstar ✪> 
15:46 < bridge> <Jupstar ✪> have u tried to restart your router and computer? xD
15:46 < bridge> <Jupstar ✪> fixes everything
15:46 < bridge> <DefaultO> i could try orinoco speedtest, but that should give similar results
15:46 < bridge> <DefaultO> i restarted my PC a couple of times already
15:46 < bridge> <DefaultO> tried to install hackintosh on my highend PC
15:46 < bridge> <DefaultO> to dualboot into
15:47 < bridge> <DefaultO> but after I did my research over an installation issue, I found out that MacOS doesn't support 3090 and won't support it
15:47 < bridge> <DefaultO> i could try ooklaspeedtest, but that should give similar results
15:47 < bridge> <DefaultO> i could try ookla speedtest, but that should give similar results
15:48 < bridge> <Jupstar ✪> what why macos?
15:48 < bridge> <DefaultO> looks fancy
15:48 < bridge> <Jupstar ✪> anyway, you seem to have a fritzbox
15:48 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1071804592892936192/image.png
15:49 < bridge> <Jupstar ✪> lets face it:
15:49 < bridge> <Jupstar ✪> macos is complete bullshit and just another useless OS that doesnt solve any issue but creates many
15:49 < bridge> <Jupstar ✪> no OS in this world is so bad
15:49 < bridge> <DefaultO> thats the linux user talking out of you
15:49 < bridge> <DefaultO> i love macos
15:49 < bridge> <DefaultO> just not interested to waste 2k on a macpro
15:49 < bridge> <Jupstar ✪> bro even windows is not so bad
15:50 < bridge> <DefaultO> you want me to do what it tells me and slide it to the left?
15:50 < bridge> <DefaultO> https://cdn.discordapp.com/attachments/293493549758939136/1071805032284041277/image.png
15:50 < bridge> <DefaultO> feels like it slows my computer down
15:50 < bridge> <Jupstar ✪> you could at least try it
15:51 < bridge> <Jupstar ✪> pf in no way macos has as good GPU drivers as windows has
15:51 < bridge> <DefaultO> yea that's why I wasn't able to install it duh
15:51 < bridge> <DefaultO> it doesn't support my GPU
15:51 < bridge> <DefaultO> too stronk
15:51 < bridge> <Jupstar ✪> macos is designed to work for what apple wants it to work for
15:51 < bridge> <Jupstar ✪> it has so many limitations and bad design flaws
15:52 < bridge> <DefaultO> so intel and in the future their M chips
15:52 < bridge> <DefaultO> i have got intel
15:52 < bridge> <DefaultO> so intel and in the future only their M chips
15:56 < bridge> <DefaultO> you mean that bad design flaw that malware basically has got no chance, privacy and other shit?
15:57 < bridge> <Jupstar ✪> yeah closed source and privacy
15:57 < bridge> <Jupstar ✪> who doesnt trust american companies
15:57 < bridge> <Learath2> There is no malware for macOS simply because it’s not profitable
15:57 < bridge> <DefaultO> fake news
15:58 < bridge> <DefaultO> Windows 75%, MacOS 15%, Linux 1%
15:58 < bridge> <Learath2> Why spend your time developing a malware for about 5% of computers?
15:58 < bridge> <DefaultO> Windows 75%, MacOS 15%, Linux 3%
15:59 < bridge> <Learath2> 15% is including iOS iirc btw
15:59 < bridge> <DefaultO> no
15:59 < bridge> <DefaultO> specifically MaxOS X
15:59 < bridge> <DefaultO> iOS is not included
16:00 < bridge> <Jupstar ✪> anyway, since u want to do gaming, steam hardware survey is moreinteresting
16:00 < bridge> <Learath2> Ok, even that is 1/5th of all computers
16:00 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1071807626658201651/image.png
16:00 < bridge> <DefaultO> that's what I would use dualboot windows for
16:00 < bridge> <Learath2> Just spend your time developing a malware for windows and you’ll make more money
16:01 < bridge> <Avolicious> gRPC in the ddnet codebase works good so far. I'd thought about using the CJobPool idea for the grpc calls
16:04 < bridge> <Avolicious> Any input on the CJobPool idea?
16:04 < bridge> <Jupstar ✪> https://store.kde.org/p/1400424/
16:04 < bridge> <Jupstar ✪> 
16:04 < bridge> <Jupstar ✪> here xd
16:05 < bridge> <Jupstar ✪> i mean cjobpool can do single async threaded tasks, if that is what you aim for then its good enough
16:05 < bridge> <Jupstar ✪> but the tasks are blocking
16:05 < bridge> <Jupstar ✪> so if the theads are in use and smth is IO bound they are in use
16:06 < bridge> <Jupstar ✪> its not comparible to smth like futures
16:06 < bridge> <Learath2> This is what I did, but with the new multi pool approach I introduced in #5842
16:06 < bridge> <DDNet> https://github.com/ddnet/ddnet/pull/5842
16:06 < bridge> <DefaultO> that's not the same feeling
16:06 < bridge> <Jupstar ✪> xd
16:07 < bridge> <Avolicious> What problem would it solve?
16:07 < bridge> <Avolicious> Running multiple jobs in parallel?
16:07 < bridge> <Avolicious> or reduce context switch?
16:08 < bridge> <Learath2> I haven’t really dug into the grpc api, but I think the issue was that I needed special handling for completion
16:08 < bridge> <Jupstar ✪> anyway about your original problem, is your internet connection new? or did it simply suddenly happen
16:08 < bridge> <DefaultO> after our internet speed was reduced
16:08 < bridge> <Avolicious> grpc offers different operation types, I'd not use the completionqueue, because it would add complexity
16:08 < bridge> <Jupstar ✪> ah yeah, why was it reduced, did u change plan?
16:08 < bridge> <DefaultO> was some special offer
16:09 < bridge> <Avolicious> So I decided to go with normal blocking calls
16:09 < bridge> <Learath2> I used the completion queue yeah that was the name
16:09 < bridge> <Learath2> With normal blocking calls I think the standard jobpool would work just fine
16:10 < bridge> <Jupstar ✪> i can only tell you my experience with old DSL cables that come from the street to the house:
16:10 < bridge> <Jupstar ✪> 
16:11 < bridge> <Jupstar ✪> make sure they are as short as possible (from street to your router)
16:11 < bridge> <Jupstar ✪> make sure they are well isolated (they are very bad with signal problems)
16:11 < bridge> <Jupstar ✪> for me personally: disable ipv6 support in the router (telekom)
16:11 < bridge> <Jupstar ✪> 
16:11 < bridge> <Jupstar ✪> 
16:11 < bridge> <Jupstar ✪> but i'd say you have a different problem, maybe just call your ISP and ask why your upload sucks so hard
16:12 < bridge> <Avolicious> Okay, we are using local on machine gRPC proxy/interceptor to cache some stuff like bans. So the jobs are being processed very fast & if needed they are delegated to the gRPC node behind the proxy
16:13 < bridge> <Learath2> Oh, you are much further than me it seems. I had a similar plan to make sure the calls ended in reasonable time
16:13 < bridge> <DefaultO> yea sucks then, 246 metres away
16:14 < bridge> <DefaultO> yea sucks then, 246 metres away
16:14 < bridge> <DefaultO> ipv6 was disabled already
16:14 < bridge> <Avolicious> Its just a PoC. Dont wanna upgrade the servers & setting up the gRPC proxy for all servers, if not tested 😄
16:15 < bridge> <Jupstar ✪> mine is apparently even longer
16:15 < bridge> <Jupstar ✪> 
16:15 < bridge> <Jupstar ✪> but as said, it must be a different problem
16:16 < bridge> <Jupstar ✪> 
16:16 < bridge> <Jupstar ✪> maybe vodaphone cheating on you bcs they think u stupid/old whatever
16:16 < bridge> <Jupstar ✪> 
16:16 < bridge> <Jupstar ✪> just call the ISP and it will magically fix
16:16 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1071811422289399888/image.png
16:37 < bridge> <DefaultO> @Jupeyy_Keks fixed it
16:46 < bridge> <Jupstar ✪> what was it
16:52 < bridge> <DefaultO> not sure
16:52 < bridge> <DefaultO> it still isn't what the dns box says I should have
16:52 < bridge> <DefaultO> but 2-3MB/s upload is an improvement
16:57 < bridge> <DefaultO> not sure, a couple restarts, disabling ip6 completely
16:58 < bridge> <DefaultO> not sure, a couple connection restarts (increasing stability to the max), disabling ip6 completely, getting rid of unused devices
16:59 < bridge> <DefaultO> but 2-3MB/s upload is an improvement. also no lagspikes anymore in ddnet.
17:04 < bridge> <DefaultO> not sure, a couple connection restarts (increasing stability to the max), disabling ip6 completely ( we weren't connected using it, but it still supported it), getting rid of unused devices
17:05 < bridge> <heinrich5991> why disable ipv6? 😦
17:05 < bridge> <heinrich5991> ^^
17:05 < bridge> <heinrich5991> probably because it fixed some issue
17:05 < bridge> <Jupstar ✪> it tunneled ipv4 for me
17:06 < bridge> <Jupstar ✪> and ipv6 support isnt really existing for tw, except maybe now with 1 server xD
17:09 < bridge> <cauldron> Why
17:11 < bridge> <cauldron> https://images.app.goo.gl/ivvF7sKXUBQ6qFnL8
17:11 < bridge> <cauldron> https://cs.brown.edu/~adf/cerf/Cerf_ipv6_poster.jpg
17:12 < bridge> <DefaultO> IPv6 is good and so
17:12 < bridge> <DefaultO> but if everyone keeps using IPv4
17:12 < bridge> <DefaultO> it will never take over
17:12 < bridge> <DefaultO> IPv6 is good and so (no subnetting to create "enough" ip addresses)
17:12 < bridge> <DefaultO> IPv6 is good and such (no subnetting to create "enough" ip addresses)
17:13 < bridge> <cauldron> I‘m not an admin.
17:17 < bridge> <Jupstar ✪> ofc u are
17:17 < bridge> <Jupstar ✪> u are admin of your PC
17:18 < bridge> <cauldron> I‘m not an admin, I can‘t say it in this specific case, but I think it‘s the trend of many things that they get deprecated. Until… well I actually are not in a possession of a personal computer right now haha believe it or not. Writing from my iphone
17:19 < bridge> <cauldron> Until someone big makes it mandatory and drops support for the old technology
17:19 < bridge> <cauldron> And others have to follow
17:19 < bridge> <cauldron> I‘m not an admin, I can‘t say it in this specific case, but I think it‘s the trend of many things that they get deprecated. Until… well I actually am not in a possession of a personal computer right now haha believe it or not. Writing from my iphone
17:22 < bridge> <cauldron> By admins I mean people that earn money with it.
17:56 < bridge> <Izanagi> Too late they convinced me
18:17 < bridge> <Nouaa> Fokkount
18:17 < bridge> <Nouaa> :kek:
18:27 < bridge> <fokkonaut> fakof
18:28 < bridge> <Nouaa> u
18:29 < bridge> <Fokkount> u
18:29 < bridge> <Nouaa> ahaha
18:30 < bridge> <Nouaa> Did you rename yourself or did someone else do it?
18:32 < bridge> <Nouaa> for many reasons, IPv6 headers are simpler and more efficient than IPv4 headers, leading to faster and more efficient routing of packets
18:33 < bridge> <Nouaa> they also have built-in support for mobile devices and mobility, allowing for seamless connectivity as devices move from one network to another
18:34 < bridge> <Nouaa> & it's more secure
18:38 < bridge> <Ryozuki> i wish everyone had ipv6
18:38 < bridge> <Ryozuki> ipv4 are expensive rn
18:39 < bridge> <Ryozuki> cuz not enough
18:39 < bridge> <Ryozuki> ipv6 are free
18:39 < bridge> <Ryozuki> and u get a /64
18:39 < bridge> <Ryozuki> xd
18:40 < bridge> <Jupstar ✪> yeah
18:41 < bridge> <Izanagi> tell that to the isps that don't even know that fiber exists
18:41 < bridge> <Ryozuki> my isp, movistar has fiber
18:41 < bridge> <Ryozuki> but no ipv6
18:41 < bridge> <Ryozuki> yet
18:41 < bridge> <Ryozuki> xd
18:41 < bridge> <Ryozuki> they stupid
18:41 < bridge> <Nouaa> it's free but organizations still need to pay for the hardware, software, and technical support necessary to deploy and use IPv6 xd
18:42 < bridge> <Ryozuki> hardware and software supports it since ages lol
18:42 < bridge> <Ryozuki> the only thing remaining is the actual users, clients of isps, to get a ipv6 address and be able to use ipv6
18:43 < bridge> <Nouaa> true
19:28 < bridge> <судный день.> top isp in georgia doesnt have fiber optiom
19:50 < bridge> <JOHN William> Do you need financial support? Are you tired living from paycheck to paycheck… wanna say goodbye to 9to5? Wanna make $2000 daily $5000 weekly before your next pay check arrives? 
19:50 < bridge> <JOHN William> 
19:50 < bridge> <JOHN William> THEN ASK HOW....📥
19:50 < bridge> <JOHN William> https://t.me/+svoBgWzahsxjZWQ8
20:49 < bridge> <Neldyk Hazretleri> ı need help
20:49 < bridge> <Neldyk Hazretleri> https://cdn.discordapp.com/attachments/293493549758939136/1071880139308269598/image.png
20:49 < bridge> <Neldyk Hazretleri> I downloaded the game a long time ago, I am facing such an error
20:49 < bridge> <Neldyk Hazretleri> Tried on a different steam account, same error.
20:58 < bridge> <Izanagi> DM me
21:17 < bridge> <Neldyk Hazretleri> @Vy0x2
21:49 < bridge> <Patiga> in antiping, there isn't adifference in that occasion. if there were, the entire reason for it to exist would just break apart. I would assume that you are just used to non-antiping makes it feel weird for you
21:49 < bridge> <Jupstar ✪> i wouldnt be so sure about it tbh
21:49 < bridge> <Jupstar ✪> 
21:49 < bridge> <Jupstar ✪> if the ping calculation for antiping the one from scoreboard?
21:50 < bridge> <Jupstar ✪> else there is probs a good chance that the interpolation is only tick accurate not ms accurate
21:51 < bridge> <Jupstar ✪> then there is probs a good chance that the interpolation is only tick accurate not ms accurate
21:55 < bridge> <Jupstar ✪> @nuborn give us insights
22:26 < bridge> <trml> that's strange, yes, at least with dummy it should be the same (if both 8 and 48 ms are with antiping on)
22:29 < bridge> <trml> the prediction doesnt calculate any ping but instead has two clocks, one for gametime or and one for predicted time, that are synchronized with the server independently (the first one using timing offset of snapshots and the second using timing offsets of input messages)
22:30 < bridge> <trml> or, at least it doesnt use the concept of ping directly
22:31 < bridge> <<~{Barsik}~>> isn't there also a delay between main tee and dummy
22:32 < bridge> <<~{Barsik}~>> its approximately = ping
22:33 < bridge> <trml> there isn't supposed to be any, if both cl_predict_dummy and antiping are enabled
22:34 < bridge> <<~{Barsik}~>> wow
22:35 < bridge> <Jupstar ✪> but arent these timing package per tick?
22:35 < bridge> <Jupstar ✪> if they are part of the snapshots
22:35 < bridge> <<~{Barsik}~>> since when ddnet client have cl_predict_dummy?
22:35 < bridge> <Jupstar ✪> a 10ms difference during interpolation would already look different
22:36 < bridge> <<~{Barsik}~>> so it means i can finish justcopyfly on chinese servers?
22:36 < bridge> <trml> oh, that one was actually supposed to be enabled by default!
22:37 < bridge> <<~{Barsik}~>> i remember i tested it every year
22:37 < bridge> <<~{Barsik}~>> 2023 is first year i can finish justcopyfly on 80+ ping server
22:37 < bridge> <<~{Barsik}~>> thanks you so much if its your work xd
22:39 < bridge> <<~{Barsik}~>> thank*
22:39 < bridge> <trml> kind of, but for the input timings the server additionally tells you the milliseconds offset, and for both timings there is a "SmoothTime" function that is supposed to smooth out jitter gradually when adjusting the prediction/game time
22:40 < bridge> <Jupstar ✪> does the server know the exact time when an input happened?
22:40 < bridge> <Jupstar ✪> does it not delay it?
22:41 < bridge> <trml> glad you liked it, thats code from some time ago:)
22:41 < bridge> <Jupstar ✪> why dont we use this information to show the real ping in the scoreboard?
22:41 < bridge> <<~{Barsik}~>> the next stage is predicting main and dummy in team like its a local server for both
22:41 < bridge> <<~{Barsik}~>> cuz you don't need to check collision and events for other players
22:46 < bridge> <trml> the server responds with the difference between when the input when received and the start of the tick the input was meant for (so basically how early / late the input was vs what the client thought), so it shouldnt affect the client if it is delayed by the server (as long as the server calculates the offset immediately when it arrives)
22:46 < bridge> <trml> the server responds with the difference between when the input was received and the start of the tick the input was meant for (so basically how early / late the input was vs what the client thought), so it shouldnt affect the client if it is delayed by the server (as long as the server calculates the offset immediately when it arrives)
22:46 < bridge> <Jupstar ✪> alright
22:47 < bridge> <trml> oh, when I think about it its actually use for show_pred (the one that shows ping besides the fps)
22:47 < bridge> <trml> oh, when I think about it its actually use for show_pred (the one that shows ping next to the fps)
22:48 < bridge> <Jupstar ✪> but that also doesnt show the real ping
22:49 < bridge> <Jupstar ✪> around 4ms off for me
22:50 < bridge> <trml> doesnt it? oh, it adds the prediction margin as well probably
22:50 < bridge> <Jupstar ✪> i set it to zero
22:50 < bridge> <trml> lol
22:52 < bridge> <trml> ok, I guess 4 ms means its not completely accurate, because of how some things are calculated perhaps
22:52 < bridge> <Jupstar ✪> question is, what if it differs between two servers ^^
22:53 < bridge> <trml> is it a constant error on each server?
22:54 < bridge> <trml> guess its something with those timing calculations that are off or not completely taken into account then:P
22:54 < bridge> <Jupstar ✪> tried on 45.141.57.22:8306 
22:55 < bridge> <Jupstar ✪> ger10
22:55 < bridge> <Jupstar ✪> 
22:55 < bridge> <Jupstar ✪> ping command shows 18ms
22:55 < bridge> <Jupstar ✪> 
22:55 < bridge> <Jupstar ✪> ingame prediction shows 30
22:55 < bridge> <Jupstar ✪> so in this case its quite far off
22:55 < bridge> <Jupstar ✪> even on LAN it shows 2
22:55 < bridge> <Jupstar ✪> my ping timing is 0.02ms
22:56 < bridge> <Jupstar ✪> but i dunno, i really dunno how exactly the calculations work, but 10ms would be feelable
22:58 < bridge> <trml> yes. I remember there was something added to make ping calculations more accurate one time, not sure if ddnet has that now? also things like vsync or network jitter can show up as more ping (its kind of possible to see it if you zoom really in on the network graphs)
23:00 < bridge> <trml> but 10 ms is a lot yes
23:02 < bridge> <Jupstar ✪> is there actual time synchronization between server and client?
23:02 < bridge> <Jupstar ✪> 
23:02 < bridge> <Jupstar ✪> e.g. if server sends a snap, will the client reset the internal prediction time to that exact tick (if we assume no lags)
23:04 < bridge> <Jupstar ✪> what i basically mean is
23:04 < bridge> <Jupstar ✪> 
23:04 < bridge> <Jupstar ✪> lets say clietn and server has 0ms ping
23:04 < bridge> <Jupstar ✪> 
23:04 < bridge> <Jupstar ✪> server does ticks at 0ms 20ms etc
23:04 < bridge> <Jupstar ✪> 
23:04 < bridge> <Jupstar ✪> client does at 10ms 30ms etc.
23:04 < bridge> <Jupstar ✪> 
23:04 < bridge> <Jupstar ✪> would they somehow get synced again?
23:05 < bridge> <trml> yes, but it's "smooth". so instead of resetting its clock offset immediately the client instead sets the new time offset as a target and interpolates between the old and new offset over time
23:05 < bridge> <Jupstar ✪> mh i see
23:05 < bridge> <trml> so yes, they will
23:06 < bridge> <Jupstar ✪> yeah i assumed so, but that indeed makes it strange
23:06 < bridge> <Jupstar ✪> since 10ms is indeed quite a lot
23:06 < bridge> <Jupstar ✪> but even 2ms on LAN confuses me actually
23:06 < bridge> <Jupstar ✪> i'd have thought its 0 or 1 from rounding
23:08 < bridge> <trml> that code is not perfect though (especially for jitter) and could probably be improved for many cases. it's supposed to work well when you have stable ping though, but even with stable ping things like vsync or something happening a few ms off might cause some jitter
23:08 < bridge> <Jupstar ✪> i have almost no jitter 😄
23:09 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1071915489078542406/image.png
23:09 < bridge> <trml> oh, yes:D, I mean like the ddnet code having jitter (but I suppose thats not too likely if your computer is also fast)
23:09 < bridge> <<~{Barsik}~>> is there any problem with "edgy" sync?
23:09 < bridge> <Jupstar ✪> ping -i 0.1
23:09 < bridge> <Jupstar ✪> ah yeah 😄
23:10 < bridge> <Jupstar ✪> but yeah the thing smoothes out after a while
23:10 < bridge> <Jupstar ✪> can i make it less smooth? 😄
23:11 < bridge> <Jupstar ✪> when i tab out and in, its dead for atleast 20 seconds
23:11 < bridge> <Jupstar ✪> wait does the prediction margin reset every time i close the client?
23:12 < bridge> <trml> it would make things suddenly jump a little bit forward/backward in time. so the smooth part instead makes time a bit faster or slow motion for a little bit to hide the fact that the corrections happened
23:12 < bridge> <Jupstar ✪> also the minimum is 1
23:12 < bridge> <Jupstar ✪> and it resets to 10 every time
23:12 < bridge> <Jupstar ✪> well that might explain all mistery
23:12 < bridge> <trml> oh, yes, I forgot about that
23:12 < bridge> <<~{Barsik}~>> so you mean server and client can't sync one time and then work properly
23:14 < bridge> <trml> not really, since latency can change over time or be slightly different from packet to packet
23:14 < bridge> <<~{Barsik}~>> xxi century made internet a global (stable) local net xd
23:14 < bridge> <<~{Barsik}~>> okey, got it
23:14 < bridge> <Jupstar ✪> does that mean i get 9ms better reflexes in fng if i make it 1?
23:15 < bridge> <Jupstar ✪> anyway the config misses `CFGFLAG_SAVE`
23:15 < bridge> <Jupstar ✪> 
23:15 < bridge> <Jupstar ✪> is that intended?
23:15 < bridge> <Jupstar ✪> or a bug
23:16 < bridge> <trml> yes. my internet connection is horrible, so I use it to add 100-200 ping to smooth out jitter
23:16 < bridge> <Robyt3> @ChillerDragon I commented on https://github.com/teeworlds/teeworlds/pull/3194. Can you also port this PR to ddnet after making the change?
23:18 < bridge> <trml> it should probably be CFGFLAG_SAVE, originally I didnt add it just in case someone just used if for debugging (for example with a higher value) and then accidentally forgot to reset it
23:20 < bridge> <trml> also only some servers support it, so it will only work on fng servers that are based on newer ddnet-versions (not sure if there is any good way to check if a server has it besides setting a very high value and seeing if that adds latency)
23:32 < bridge> <Jupstar ✪> I see thanks for the details
23:35 < bridge> <Finder> i have kinda a problem with like the ping it usualy sticks right around the server ping, my wifi is okay i just dont know whats going on
23:39 < bridge> <trml> np, I also thought about perhaps looking more into some of these things at some point