DDraceNetwork - developer

DDraceNetwork

Development / developer

Development discussion. Logged to https://ddnet.tw/irclogs/ Connected with DDNet's IRC channel, Matrix room and GitHub repositories — IRC: #ddnet on Quakenet | Matrix: #ddnet-developer:matrix.org GitHub: https://github.com/ddnet

Between 2022-01-28 00:00:00Z and 2022-01-29 00:00:00Z

.

https://xkcd.com/1031/

xkcd: s/keyboard/leopard/

09:09

https://twitter.com/onyxaminedlife/status/1334578466851852290

Master of Lillusions (@onyxaminedlife)

Replacing the words "the economy" with "rich people's yacht money" -How can we respond to COVID without sacrificing rich people's yacht money? -Saving the environment sounds nice but what about rich people's yacht money? -Medicare for all would destroy rich people's yacht money

Retweets

28218

Likes

87492

09:09

xd

09:11

if ur ever bored, follow this https://twitter.com/gitlost its a twitter bot that posts about developers swearing in their git commits

09:17

https://news.ycombinator.com/item?id=30107540

pacificat0r

Prime Video Uses WebAssembly

09:17

they are using rust poggers

09:17

which makes sense since rust has wasm as a first class citizen

09:20

Another benefit of using Rust is having access to an ecosystem of high-quality libraries. For instance, we built an application that overlays debugger information on an application scene render using egui, a Rust GUI library. Integrating egui with our Wasm renderer took a couple of hours of work and offers us an easy way to gain insights into the engine’s internals.

09:20

egui is like the imgui of c++ but in rust its also awesome https://emilk.github.io/egui/

09:20

(its native)

Image attachment

10:23

thats cute

Ryozuki

https://blog.benjojo.co.uk/post/lto-tape-backups-for-linux-nerds

well, its the most compact storage technology actually. Some tapes can store up to 500TB in the same space than a classic HDD

14:40

+ the offline thing makes it great for archiving, backup and protection against malware, especially ransomware that encrypts your data

microsd cards are 15mm x 11mm x 1mm; 3.5" hdd slots are 101mm x 26mm, if I assume a length of 150mm (can only find height and width), then I can put about 2300 microsd cards into a hdd slot, which translates into 2300 TB at 1 TB per microsd card

14:45

probably not practical though

monkalaugh

+ the cost would be way more expensive

14:52

and SD cards have retention time of only few tens of years, while magnetic tape are closer to 100 of years iirc (edited)

probably depends a lot on the tape used, and the extreme tapes probably haven't even existed long enough to say anything about the lifespan

you don't need to wait 10 years to know, you just test it in degraded conditions for a few days (usually 85°C) and count the numbers of bit flip

14:58

Then, of course, is just an estimation

ah

ChillerDragon: real gamer :D

15:08

Image attachment

heinrich5991

microsd cards are 15mm x 11mm x 1mm; 3.5" hdd slots are 101mm x 26mm, if I assume a length of 150mm (can only find height and width), then I can put about 2300 microsd cards into a hdd slot, which translates into 2300 TB at 1 TB per microsd card

I've measured 3.5''HD size: 100 mm x 145 mm x 24 mm Good estimation

[ddnet-web:master] 1 new commit

a72d651 Add Pipou as second testing leader - def-

[ddnet/ddnet] Issue opened: #4642 Disallow start_video/stop_video outside of demo viewer

This breaks the game, all particles will be stuck in place when start_video is used while being on a server. Can only be fixed by leaving the server. Quite a cool bug. :D Found by tela on discord.

20:12

[ddnet/ddnet] Pull request opened: #4643 Disallow start_video outside of video player (fixes #4642)

Checklist

[ ] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test if it works standalone, system.c especially
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...

[ddnet:master] 2 new commits

1cf377f Disallow start_video outside of video player (fixes #4642) - def- 62619fd Merge #4643 - bors[bot]

20:54

[ddnet/ddnet] Issue closed: #4642 Disallow start_video/stop_video outside of demo viewer

20:54

[ddnet/ddnet] Pull request closed: #4643 Disallow start_video outside of video player (fixes #4642)

could it be possible to turn switches on / off with rcon ? just like switch_open is a thing for map settings already

Hi, I was wondering if it has ever been considered to add a configuration option for the antiping feature that would allow for disabling prediction for non-frozen tees, basically eliminating the whole jittering issue. This seems kind of obvious from a gores players perspective, so sorry if this has already been discussed.

Okay, I just came across https://github.com/ddnet/ddnet/issues/3549, apperently I spelled antiping wrong in my search

that issue is for when you are frozen

Could be... At any rate I would love to hear some opinions about it

going from frozen to unfrozen or unfrozen to frozen would look really weird i think

without looking at any code, isnt that the same issue as stopping local prediction when the local character is frozen?

this issue seems kind of related, but a different idea (and also trickier to do) https://github.com/ddnet/ddnet/issues/1712

yeah they would both be weird

trml

this issue seems kind of related, but a different idea (and also trickier to do) https://github.com/ddnet/ddnet/issues/1712

Interesting, thank you. One more thing, I vaguely remember hearing about ddnet not wanting to add any new console commands / variables... Would that be related if say I wanted to implement a feature like this?

plsplsplslol

yeah they would both be weird

was that regarding my message or trml's?

it was to yours, but trml's would also be weird

Oh, I thought turning off prediction while frozen was already implemented... Maybe I am mixing something up here.

i think right now, prediction only gets turned off for ppl in solo or when you are spectating

a nice thing about only doing it for frozen tees is that you would at least have a visual cue of when prediction is disabled/enabled, which you wouldn't necessarily have with the other one

22:50

not sure about the console commands

the visual cue would be slightly off though because of prediction

I thought as in ninja skin/no ninja. unless it's also extended to non-frozen non-moving tees

Yea something along the lines of the less configs the better. Reducing complexity and avoid overseeing Config combination complications (@GutZuFusss)

Interesting, thank you. One more thing, I vaguely remember hearing about ddnet not wanting to add any new console commands / variables... Would that be related if say I wanted to implement a feature like this? (@trml)

saddo

1

@trml That is a valid point. Generally I think the idea proposed in 1712 comes with cleaner implementation opportunities, but in terms of practicability I think there are more things related to your argument to be considered. Most notably (again, mind you from the perspective of a gores-only player) the moment the interaction takes place the predicted positions become less relevant anyways, because the situations where we would want a predicted position are typically ones right before we interact with a player (e.G. attempting to catch a tee, hitting a hammer for throwing, ...). Regarding the possibly confusing sudden change in position I am not yet sure if its less confusing to have it happen on player interaction or on player freeze state change. It probably strongly depends on the map type or even situation and should rather be a consideration between having the feature at all instead of choosing the variant. I think I will implement it quick and dirty to see how bad it really is ingame. (edited)

@GutZuFusss prediction isn't turned off if you're frozen, it's just that it correctly predicts taht you can't move while frozen

hey

1

yes, with the other one I think one would have to define some duration that the "interaction" should last for it to be useful, and it still wouldn't handle the case of catching tees. so predicting frozen and/or inactive tees in general makes more sense (to me). the change in position would perhaps look acceptable if it doesn't happen too often, since you usually get some glitchy movements with antiping anyway

One more thing that just came to my mind... How much worse could the transition between unpredicted and predicted really be then the accuracy focused antiping implementation right now? I mean the simplest worst case for the current iteration of antiping that I can think of is a player changes his input direction from 1 to -1. This means, we jump from a prediction that will move the character in one direction to one that will do the exact opposite. Maybe I am missing something, but shouldnt be changing from predicted to unpredicted cause a less noticable jump in positions? That is besides the fact that both proposed variants would occur less frequent than a player input change.

It just sounds far more confusing to have to learn to play with 2 different predictions, and swap between them on the fly. I guess in gores it's probably more appealing, but since the client is specifically for ddrace, I don't see the appeal.

Skeith

It just sounds far more confusing to have to learn to play with 2 different predictions, and swap between them on the fly. I guess in gores it's probably more appealing, but since the client is specifically for ddrace, I don't see the appeal.

Fair enough.

There are too many parts where the tee isn't frozen, so implementing it properly sounds messy

23:54

Both options for antiping kinda suck in their own right, though. As it is now, it's probably less confusing for newer players, at least.

yes. btw, there is cl_antiping_smooth that was an attempt to soften antiping for players that move unpredictably (without making transitions jumpy), but it makes less of a difference than the setting you propose would

https://www.youtube.com/watch?v=PTtbAhQL6kQ&ab_channel=Meskalin cl_antiping_smooth is enabled... I think I missed the smooth part (once I get into race) (edited)

They at least get the visual cue of "Holy shit, this ping sucks" lol

yes, that setting just makes a very marginal difference, since making it more smooth would at some point have sacrificed other things

Exported 64 message(s)