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 2020-12-14 00:00:00Z and 2020-12-15 00:00:00Z

[ddnet/ddnet] Pull request opened: #3410 Color picker is ready

Kinda lost track of all the updates this PR contains, so... It's just huge. For in-game changes it changes HUD settings tab and adds Color Picker UI feature with preview, editboxes on RMB and so on... https://cdn.discordapp.com/attachments/295908390956433410/787843848810135582/unknown.png (Github does not want to upload my image here...)

Checklist

[x] Tested the change ingame
[x] Provided screenshots if it is a visual change
[x] Tested in combination with possibly rela...

[ddnet-maps:master] 1 new commit

cf9a82a M Justice 2, A Glade - ddnet-maps

Hi, I'm not sure where should I write this but, I would like to suggest to put map Nyan Cat in Special Maps section of solo server, and increase Aufnahmetest 3 to solo 4 stars, cause is one of the hardest 3 star solos and Aufnahmetest is like the easier version and it's 3 star

map's rating system is fk ⏫

06:40

so don't worry

@Дядя Женя so better not improve it? what's that attitude?

06:57

try to be more constructive with your criticism, please. non-thought out criticism just makes life harder, and doesn't improve anything

reporting one map is not the way you fix something so screwed

06:59

you should have some solid rules about map ratings and rerate everything

06:59

as well as filter some maps to another or even new categories

@Дядя Женя it's resources, I don't think we have resources to re-rate every map

accounts solves everything (edited)

accounts sux

07:09

stop trashtalk

as an discord moderator could you warn or punish him for such behaviour in my direction please @heinrich5991

@Дядя Женя I don't see obviously bad behavior. I see "accounts good", and "accounts bad"

Maybe I'm misunderstanding some english

https://en.wikipedia.org/wiki/Trash-talk

Trash-talk is a form of insult usually found in sports events, although it is not exclusive to sports or similarly characterized events. It is often used to intimidate the opposition, but can also be used in a humorous spirit. Trash-talk is often characterized by use of hyperbole or figurative language, such as "Your team can't run! You run like...

bro you are not the first person suggestion accounts its a topic with strong opinions and long history so expect some rude language if you bring it up and claim "solving everything"

07:53

Embedded image

I wouldn't consider yes/no sayers "strong opinions"

08:21

I haven't seen a single person elaborate his opinion why accounts would hurt the community

08:23

in fact, there isn't even a discussion/issue on github about accounts

yea ik

08:32

its more a heated topic with ppl flaming ingame and so on :D

08:32

it became almost an tabu

08:33

is tabu even an english word idk

in english? no, german? yes

08:34

taboo in english

Accounts could be bad if names/points can be stolen.

Eh?

[ddnet/ddnet] Issue opened: #3411 Accounts

If anyone wants to summarize the discussion, feel free to. I just wanted to create the issue now.

I guess Im 'corneum says that a downside of accounts could be points being stolen by renaming someone's account without their permission

murpi

I haven't seen a single person elaborate his opinion why accounts would hurt the community

Everyone elaborated, but it's all scattered around so you need to have been here for every discussion to know all the opinions. Even I don't remember who owns what argument anymore

10:17

But one major concern was namestealing, which we said we could handle with "steam-like" accounts where usernames and nicknames are separate. The counter to that is that then the accounts are pointless since it's too hard to identify who is real who is fake

heinrich5991

I guess Im 'corneum says that a downside of accounts could be points being stolen by renaming someone's account without their permission

I guess this depends on how accounts are implemented

Another issue that comes up is legacy ranks. How do we attribute legacy ranks to accounts if we attribute them at all. Every single solution I came up with has a disadvantage. Though the most promising solution seems to be not attributing legacy ranks at all, but instead encouraging people to make new ones with accounts

Learath2

Everyone elaborated, but it's all scattered around so you need to have been here for every discussion to know all the opinions. Even I don't remember who owns what argument anymore

ah well, thanks for opening the issue.

10:22

oh heinrich opened the issue, oops

After that the arguments get a little less nuanced and thought out but privacy and ease of use sometimes get mentioned

10:28

Most of my attempts at is has been thwarted by me wanting way too much, Zooko's Triangle starts to get in the way

blockchain is the solution to zooko's triangle!!1

blockchain so good

aPES_WTFEyes

i went all in bitcoin yesterday

blockchain

Why if its in a all time high

isn't bitcoin at an all time high rn?

ikr

10:29

yes!

Chillee pepeg

guess someone needs to pay the profits of the others

10:30

Well you can indeed make a construct with blockchain that satisfies all 3 desired properties and thus I guess "refute" Zooko's Triangle

yes!

monkalaugh

So it is a solution, but definitely not a trivial one

@Learath2 what was ur technicao aproach to accounts?

https://www.statista.com/statistics/647523/worldwide-bitcoin-blockchain-size/ it's 300 GB already

Bitcoin blockchain size 2009-2020 | Statista

How big is the Bitcoin blockchain?

Shit mobile

(blockchain constructs are kinda cheating by skirting around "Secure" though, teeworlds is small enough that a Sybil attack is not that inconceivable e.g.)

10:32

@Ryozuki I basically had asymmetric cryptography. It's Secure and Decentralized, but the Identifiers are not human meaningful

10:32

My idea was to rely on a local map that maps identifiers to names

Is it really worth to make it decentralized?

(and we also distribute a sort of address book so we are a central authority but as we distribute the address book, people don't need us to verify eachother)

the ideas sound really good!

10:34

more thought out than mine were, so far

@Ryozuki I was kind of hoping to make it more palatable to the teeworlds people too, but I guess it's not a necessity. I had a centralized idea too. Basically just stealing kerberos

10:37

A simplified version of it is enough for us, so we'd only need to issue a sort of ticket proving that the client is who they say they are

10:38

(the ticket granting ticket construct in kerberos is mostly to do authorization, which is not really something we care about, in our construct every client would be authorized to play/log in on every server)

10:39

((one problem with a central construct like this is that we can't guarantee that someone you meet on a non ddnet server is who they say they are, atleast not trivially))

10:40

(((because I didn't figure out how to efficiently do it doesn't mean it's impossible, but the best I could come up with was an n-way exchange of tickets)))

10:41

Oh actually came up with one right now, the server could broadcast an encrypted blob per client, which can be submitted to our AS to check whether someone is who they say they are

10:42

(though the server can allow anyone on the server to impersonate another, now that I think about it)

10:42

Designing Authentication Protocols are just too difficult

is*?

[ddnet/ddnet] Pull request opened: #3412 Allow reordering in the settings editor

Requested by Ravie on discord !t

Checklist

[x] Tested the change ingame
[x] 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
[x] Considered possible null pointers and out of bounds array indexing
[x] Changed no physics that affect...

@heinrich5991 is* indeed

11:18

[a-z]+s are is just muscle memory

11:19

I wonder if the brain utilizes state machines somewhere deep down

@Learath2 well i would parse that as designing (x) is...

11:24

think_bot

@Ryozuki while reading I would too but while writing apparently my reflex to do [a-z]+s are is stronger than my actual english parsing subsystem

english parsing subsystem

11:26

POGGIES

the context free (possibly even FSM?) muscle memory being faster to act than the probably turing complete english parser makes sense

cyberpunk 2077

11:26

real

I doubt you can parse english with a pushdown automata

11:28

I wonder if a computer can ever parse natural language. If you think about it the amount of things we can understand is baffling. A sentence can make no sense at all and we can still parse it

well the brain is rly complex

11:31

"100 Trillion Connections" (edited)

The future is in an NPU natural processing unit. Just a brain in a jar with a BCI

monkalaugh

Tensor cores and stuff are just soo inefficient. Brains are sooo optimal

well

11:34

they are optimal for x things

11:34

but for example

11:34

you cant remember 100% of ur life

11:34

every detail

11:34

i guess its a optimization itself that

It does all this with 20W

did u know we sense things in a logarithmic scale

11:35

thats why when ur older time goes faster

It'd make sense in some way

https://spectrum.ieee.org/podcast/biomedical/bionics/does-the-brain-work-logarithmically

11:35

Does the Brain Work Logarithmically? New research suggests it does, when it’s the efficient way to process information

11:36

@Learath2 imagine ur in danger and there are 90 tigers, if there were 91 u wouldnt even notice

11:36

but imagine there is 1 tiger

11:36

if another appears

11:36

u would be alarmed 2x

11:36

kinda

11:36

my explanation sucks

11:36

monkaSteerS

11:37

https://www.youtube.com/watch?v=Pxb5lSPLy9c

1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,2...

Ah I knew I read this somewhere before https://en.m.wikipedia.org/wiki/Weber%E2%80%93Fechner_law

Weber%E2%80%93Fechner_law

this video explains it

11:37

rly well

11:37

yea

This is also probably the reason we measure things that relate to our senses in log scale

11:38

Like sound in decibels

11:39

Or light intensity

@Learath2 i found out the top 1 competitive coder in youtube

11:42

https://www.youtube.com/watch?v=97tieEKfvBs

Gennady Korotkevich

touristream 012: Codeforces Round 672 (Div. 2)

11:42

https://codeforces.com/profile/tourist

tourist - Codeforces

11:42

this guy is insane

11:42

probs a genius

11:42

the video goes from ez things to things u wont understand

11:42

magic

Depends on your definition of genius

he looks rly young to have studied a lot

11:42

idk

Competitive math, competitive chess, competitive programming are all very impressive but they are soooo memory heavy

11:44

Time constraints mean you need to have so many lookup tables drilled into your head

idk if all the contests are time constrained

11:44

are they?

And they prefer it that way btw, look how much more popular blitz is in chess, or how much more popular shorter form golfing became in the recent years

cmon

11:45

blitz is popular because ppl dont want to waste 8 hours in classical chess

It looks extremely impressive when you instantly know that you need to use “x algorithm”

i dont think its a fair comparision

11:46

i doubt they know everything

11:46

or that every problem is not unique

Honestly I prefer both playing and watching classical chess. It’s just so much of a hassle to know everything, you can’t follow anything in blitz without a commentator

11:47

And to even play blitz and not lose 100 games in a row you need to dish out the first 6-7 moves instantly and correctly

grandmasters play superbullet e.g 15 seconds

11:47

amazingf

doubt that you need to play the first 6-7 moves instantly and correctly in each of these 100 games @ blitz

I can usually play/watch rapid when I’m at my peak awakeness and have coffee running through my veins

Image attachment

11:48

is youtube down

11:48

???

11:48

lol

doubt that it can be down globally

11:49

I don't want to click any of the videos on the yt front page

@heinrich5991 well most of your opponents above a certain rating certainly will play openings optimally

but that loads, at least

https://www.youtube.com/ itself

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

11:49

doesnt load

@Learath2 doesn't really matter in blitz. you can be down 1 piece and still win

wtf

11:49

on private browsing it works

are you logged in in main?

Well sure, but you are definitely at a massive disadvantage against someone who knows the lines

ok deleting all cookies fixed it

I'm not logged into yt

Image attachment

11:50

i cant sign in back

11:50

xd

11:50

Image attachment

11:50

something is definitly happening at google

11:52

test

E.g. it’s extremely tough to play (IMO) queens gambit declined for me on blitz

11:53

Took me like 10 games of queens gambit to decide that I just don’t have the processing power to play qgd in blitz

google meet down, ppl doing online school monkalaugh

monkalaugh

But I almost always will decline in classical or rapid since I have time to think the moves

declining is the way

11:54

taking puts u in a disadvantage later

11:54

queens gambit is not rly a gambit due to that

11:54

xd

Accepting is much more viable for me in blitz as it’s more straightforward to play

11:55

But ignore chess for a while, it’s the same in competitive programming and competitive math

11:57

You need to recognize the questions instantly, you need to instantly refer to your book of tricks to simplify the problem and you need to do that while actually starting to lay down the scaffolding for your solution

11:58

(I mean if you want to be competitive at all, you can just have fun and/or learn something and take it slow)

11:59

It’s a different set of skills is my point, one requires more problem solving and the other great memory

there may still be problems that u cant solve with ur book of tricks

12:00

and i think there is where smart ppl come to play

Not in competitive anything, you can’t expect people to creatively solve problems in limited time. Maybe sometimes they tack on an extremely difficult question that is more suited for this but that’s only for people that are insanely smart, they both need the book of tricks to get thru the other questions AND have time to think about the extreme question

12:02

Math is better at this (probably because competitive math has been around longer)

12:06

Anyway, as I said at the beginning, depends on how you define genius. I associate it more with the creative process (though that might have to do with the turkish word for it having hidden connotations)

Mh can't upload images, discord is having trouble too, a massive botnet flexing probably

12:14

The api response time is all the way up to 130ms

Is there a clever bitmasking trick for todays part 2?

@timakro I ended up setting the bits manually if you mean that

I think I'm going for a recursive thing

12:44

I believe that will give the shortest code

I did a loop to 2**num_bits_to_vary

and then a lots of shifting

12:45

Did you write rust?

just one shift, I manipulated the number as a string

12:45

no, python

** == ^?

** is "to the power of" in python

12:49

^ is XOR in python

oh

@heinrich5991 Can I see your code? Heres mine (spoilers) https://termbin.com/2lyg

13:11

And feel free to give me rust tips :D (edited)

https://paste.mozilla.org/U63AofjQ

13:13

(also spoilers)

Were you the one learning python?

no

13:13

I'm using python to code as fast as possible, trying to get on the global leaderboard

Was learath then

Tis me polishing up my python yes

your rust code looks good to me @timakro

cool :)

just saw that, not relevant for AoC: from a variable name perspective, mask_0, mask_1 and mask_x are inconsistent ^^

how?

mask_0 doesn't have 1s where the 0s are in the string

right ^^

[ddnet:master] 2 new commits

bd1c847 Allow reordering in the settings editor - Learath2 32bbf0f Merge #3412 - bors[bot]

13:54

[ddnet/ddnet] Pull request closed: #3412 Allow reordering in the settings editor

@Comrade TIL Bitcoin is smaller than DDNet Teehistorian recordings

yes, but teehistorian recordings are not something that people normally work with (edited)

Learath2 scans through all of them for some statistics

still, i think blockchain is a fail actually

14:05

it's not robust solution

why is it not robust?

because it's a pack of data with endlessly increasing size

14:19

i mean there are some usecases where it makes sense, but in Bitcoin for example: why do you need to care about some 10 years old transactions that no longer matter?

The 10 year old chain is what creates security

14:20

There are already attempts at making it more usable, like lighter wallets, but if the chain is gone there is no security at all

yes, but as an end-user, you want transfer some money and not waste 300 GB of your storage

Which is where lighter wallets like electrum come in

isn't that just a workaround?

well yes but it's the only way this whole thing works

14:22

it's either the huge chain securing us or trusting a central authority to secure us

14:23

I mean as far as we know for now, maybe someone will come up with something even more clever

so we should use some centralized systems instead for now

14:24

but Bitcoin is a nice experiment

well there is more value to bitcoin than just as an experiment

14:26

it's very hard to trace cryptocurrency transactions, it gives people atleast a bit of privacy

14:28

bitcoin is not the best at privacy but you still can't know the recepient or sender of any funds, barring out of band attacks like forcing exchanges to comply by using powers of government

i think it's like Tor - used mostly for illegal stuff so when you try to use it, you're instantly suspicious

Good thing there is no way to prove you use either Tor or Bitcoin as long as you do it correctly

14:37

Besides this "suspicion" and "stigma" is pushed by people that have great interest in these technologies failing. e.g. the banking industry, governments, intelligence agencies, huge datafarms like google/amazon. So you are in a sense right that these things will most probably never be used widely

14:37

If they were used widely it would be disaster for all of the above

well, both Tor and Bitcoin has significant disadvantages

14:39

just imagine playing Teeworlds over Tor

14:39

*have

Tor I see, bitcoin I don't with light wallets the storage requirement is minimal, with lightning transaction times are not an issue

Bitcoin already requires terawatts of energy to do seemingly pointless things

14:45

all those standard bank systems written in Java are still much more effective probably

Hm, the energy usage is an interesting point but a comparison to traditional banking is not really fair, since well there are "significant disadvantages" to using traditional banking

14:47

e.g. trusting that the "banks don't fail", trusting that the bank will give you your money back, trusting that the bank will let you use your money as you want, letting the bank know of every transaction you make (and by virtue of the bank the government)

yes, that's true

before you say that's all conspiracy, it's all happening in turkey, there is now soft capital control in place, I can't take out more than a certain amount in usd or eur without paying a decent chunk of tax

14:49

or lebanon, where people were limited to 100$ a week during the harshest part of the crisis there, or in greece where the banks literally ran out of money

well regarding privacy when it comes to bitcoin is also meh. Because in the end most people buy btc over companys or have their wallets in the cloud. And the goverment can force companys to collect user data.

14:49

blockchain.com for example forces you to upload a document and shit before you can trade

14:50

and france also went wild on regulations for crypto

ChillerDragon well you can always mix up your bitcoin or use another coin which does have great privacy properties like zcash or monero

yea

14:50

i guess

14:51

i actually do not know how zcash and monero enforces priuvacy so i cant comment on that

well it's very complicated math that goes waay above my head aswell, but I trust that people have checked that the claims are not insane

14:52

You know how people get about other people saying something wrong on the internet

being one of those math/crypto magicians must be fun

that is one of the few skills I'm not really jealous of, idk I don't think I enjoy the theory part of math

14:54

it's soo detail heavy, especially discrete mathematics, so many edge cases to consider

14:56

Or analysis where you have to consider things like "everywhere continuous but nowhere differentiable functions", I mean sure they exist and all but what is the point, it's so academic

14:58

Fractal curves exist irl even, but are they reallly necessary to do anything in real life? You can approximate things quite well, infinite precision is rarely needed

idk if fun but i would be proud of knowing such stuff

15:13

i heard its true that bitcoin is rly bad for the enviroment due to energy cost

15:14

the environment, with areas that use mostly fossil fuels for electricity, such as Inner Mongolia, China, contributing more to the carbon footprint than regions using renewable resources, such as Sichuan, China.

15:14

its mainly from miners and heavily depends where they are located tho

15:15

@Learath2 do u need to keep mining for bitcoin to keep working?

15:15

https://www.sciencedaily.com/releases/2019/11/191120080246.htm

Estimating the environmental impact of Bitcoin mining

As an alternative to government-issued money, the cryptocurrency Bitcoin offers relative anonymity, no sales tax and freedom from bank and government interference. But some people argue that these benefits have an enormous environmental impact, particularly with regard to Bitcoin mining -- the process used to secure the cryptocurrency. Now, rese...

Bitcoin uses like 50% of all datacenter electricity iirc, that’s just insane

15:16

Yeah if noone mines bitcoin can’t work

meh

15:17

is the market cap for euro and dollar always increasing too?

Mines are the ones that secure the blocks, and for their reward they get transaction fees for the block + the miner reward

15:17

As the miner reward decreases, there is actually a limit to how many bitcoin there will be in circulation ever

15:18

Euro and dollar can be arbitrarily printed by the eu and the us, but both have good incentive to keep inflation low

yeah

well teeworlds city servers also keep printing money and infaltion isnt too bad :D

15:34

but i rly like the idea of bitcoins having a limit and also keep in mind over time more and more bitcoins will be lost forever due to ppl lossing their keys

15:35

so it is unlikley that there is classical inflation

15:35

as long as ppl believe in it the value can only increase

15:35

it doesnt matter how many ppl believe in dollar or euros if the goverments keep printing it it loses value

15:36

someone mentioned lighting network i havent rly looked into bitcoin since the last hype :D so idk what it is as far as i remember bitcoin was pretty useless as payment due to the low capacity of transactions possible

ChillerDragon the value of bitcoin depends on the value of euro or dollar

Well both traditional money and bitcoin have value because we all collectively say it has value

15:45

There is a demand for us dollars by other governments and businesses because they both allow trade with us companies and they are a relatively stable currency (a store of value)

15:47

Ditto for euros. And only the first one for the chinese yen. However, bitcoin has no real demand outside people messing around with it

15:48

If everyone accepted bitcoin, then it’d have no dependence on the euro and dollar. Though that’d require governments to tax people in bitcoin, which they won’t ever do because they stand to lose sooooo much from losing control over the flow of money

15:50

Btw the lightning network is a pretty neat idea, in which the bitcoin network is only used to settle the transactions every once in a while between large nodes. So it’s sort of a decentralized version of the banking and payment systems used in traditional banking

hm so less live syncing i see

16:07

doesnt it make sibling attacks easier?

16:07

and what even is a large node?

16:07

isnt every node of the same size

You mean a sybil attack? For that you need to control a large amount of full nodes. The lightning network is built on top of the bitcoin network, it’s not the same nodes

16:27

‘Larger’ would have been a better description. The idea is that A and B open a channel, they both put x bitcoin into it. A smart contract chain guarantees that either side would be paid off in case of one of them not complying. Now since the “trust” is established, they can instantly make transactions of up to 2x bitcoin between them, instantly. Say I want to send you money with lightning, but we don’t have a channel yet, the network tries to route the money through other channels that already exist

16:31

How the network is designed is really beyond the point. It could be a mesh, it could become a hub and spoke network. What matters is that the transactions are instant and only depend on the bitcoin network for safety

ethereum did some nice stuff, to combat the energy intensiveness of its network

i need friends

17:41

Good night

hello friend

17:42

I don't feel like ordering food and I don't feel like cooking. Do I just starve?

Instant noodles

That sounds like a great idea actually, hope I have some in the cabinet

I only seem to have a packet of instant ramen lying around and I don't have anything to top it with

17:57

untopped instant ramen could be like the saddest meal known to mankind

17:57

okay maybe after plain toast

bowl of cereals can also do a good meal

[ddnet-web:master] 1 new commit

afbb4a1 Remove Brazil1, KSA - def-

cereals ❤️

Exported 318 message(s)