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 2020-10-15 00:00:00Z and 2020-10-16 00:00:00Z

hi guys

01:56

check this i maked wiki

01:56

https://en.wikipedia.org/wiki/Draft:DDraceNetwork#cite_ref-1

Draft:DDraceNetwork

01:59

Anyone want fix my wrongs or additions anything let him/her do that (edited)

thanks developers for take my suggestion

brownbear

02:35

Image attachment

02:35

for the second time troll

troll

[ddnet-maps:master] 1 new commit

00bcdf8 M garden, A FreeDoom2 - ddnet-maps

[ddnet-web:master] 1 new commit

1120107 Add donation - def-

i really hope windows puts out some update soon

07:45

the guy in general also thought that the older clients worked

07:45

he also has a intel from 2019

07:45

so not some old stuff xd

[ddnet-web:master] 1 new commit

c74cd0d Focus input fields on maps/players pages - def-

@Deleted User honestly these issues occur way too often. I don't think we can just keep telling people to update drivers

@deen so whats ur suggestion? block windows update, you must be aware that these reports magically happened like 1-2 weeks ago

08:05

before that nobody said anything

oh, it's a problem with the driver version windows delivers? I didn't follow closely, I assumed it was some gl code change we did using more features

well he has a CPU from 2019, even if i use the most exotic opengl 4.5 function it has to work xD but yeah, appearently its since only 1-2 weeks, bcs then it started, and all had the same bug

then maybe let's write a website news at least?

i've no proof, but i know that alot of ppl think that 14.7 and newer worked and the new GL code is since 14.6.2

08:12

well you can do it, sure, wouldn't hurt

could you report it at microsoft?

well since a driver update yet always solved the issue, its probably known

08:15

but yeah can see if there is some report for it

would be nice if you could write the news post too

https://bugs.documentfoundation.org/show_bug.cgi?id=95639

08:16

this has happened

08:16

in the past

08:16

look this xD

08:16

Image attachment

08:16

ppl get black screen right?

but they are on windows

08:16

is that a forum for windows?

yeah this is windows

ok

u didnt even read the title

but 2015

08:17

doesnt sound new xd

justatest

08:17

but they may have made the same issue

08:17

now

intel is known to have alot of problems always

08:17

not even only opengl

08:17

when i searched the internet a bit yesterday i found so many bug reports

08:19

https://support.microsoft.com/de-de/help/4021566/windows-10-send-feedback-to-microsoft-with-feedback-hub-app

08:19

they have a builtin app, but cant find any other way

ah german

08:19

yes i understands

good, bcs u have to learn german anyway

no

i do c++ with german keywords

08:19

and force it into the world

german words are stupidly large

08:20

xdd

falls(x == 0) { } mache { } solange (x == 0)

english is the best

08:20

u dont have to press multiple keys

08:20

to do stuff like ü

08:20

cuz it doesnt exist

äöü are one key each :p

no

08:24

ñç are

well on a german keyboard at least

@deen i am not good at writing, just write something like this, but in your own words xd:

---
layout: post
title: "Problems with Intel processors"
permalink: /news/intel-driver-problems/
tag: current
---

There were a lot of reports about for bugs and crashes with intel processors under Windows lately.

It might have been created by a windows update, which also updates the intel GPU's driver.

A driver update seemed to help with this situation (downloaded the latest driver for intel and update it with help of [this](//https://forums.lenovo.com/t5/Lenovo-Yoga-Series-Notebooks/Update-to-Latest-Intel-Graphics-Driver/m-p/4457633?page=1#4457786))

We hope that intel can solve their issue quickly, until then you can just try to follow the advice above.

Hi, I was wondering if it would be possible to have a 2nd HD option especially for a foreground in my case (so it remains an option)

08:54

Is it possible or not ?

theoretically yes xd

08:56

if u really just mean that foreground would e.g. show the high details

08:56

lets say it like this: the client knows what is foreground and what is background and it could disable highdetails for one, while the other has it on (edited)

I mean it could be like hd, you check yes on the wanted layers

the client can already do this? it just doesnt split foreground and background

09:03

so the only chance is to split foreground with background option, but not in the edtior

09:03

in the editor you know whats foreground, bcs of the gamelayer ofc

09:04

but if you mean a 2nd option for the same layer

09:04

that is not possible

09:04

only a second option to split between foreground and background layers

ok let me show you

09:09

https://youtu.be/pvR7KPCl2M0

Lavender Forest Foreground

09:10

I'd like the foreground, not playground, to be an option

aslong every layer after the gamelayer would be your definition of "foreground" then yes

09:13

else no

yep everything above it

09:13

or actually no the game layer graphics are above game as well

then it won't work probably, bcs in the client everything after the gamelayer is the foreground

so another HD (graphics/high details) option isn't conceivable?

it would break vanilla compability

09:16

so i guess ddnet wont allow it

hm I thought about it but in that case, wouldn't it just display it for vanilla players ?

09:17

only ddnet users would be able to show/hide it then

Image attachment

09:18

an option for these two is possible

09:18

but not single layers inside these groups

09:19

well maybe it is possible, but i dunno if they want it in ddnet

09:19

i mean everything is possible somehow xD

yeah, maybe it's possible without breaking compatibility

09:22

I thought I might just be the only one to use it but actually it'd be great to have a bar going from 0 to 2 (for HD 0, HD 1, HD 2) where you could decide how many details you want

looking in the editor yes

and not just nothing/everything as it currently is

it'd be possible to add it for ddnet only, i dunno how sensitive the vanilla client is, but probs it will just ignore the flag

09:23

you could add another ~31 high detail steps as far as i can see

Ok, I'll wait for more opinion about it

09:23

anyway I can't code

[ddnet-web:master] 1 new commit

12766a6 Update funding - def-

@Pipou I think HD on/off is enough, wouldn't add more. It also means all have to be tested

09:39

I think the foreground stuff looks cool

09:39

maybe make it a bit transparent to make it clear it's just graphical (if that's your worry)

I think it might be too much that's why I wanted to leave it as an option

I understand, but we have too many options

yeah it's ok then

09:41

here is what it looks like without hd

09:41

Image attachment

i dont even understand whats the problem of having alot of options

09:41

we only save non defaults anyway

it's complexity

09:42

every combination of options can fail

only if its used more than 1 time tho

09:43

if you do int Detail = g_Config.m_HighDetails and that is the level of detail that doesnt add complexity

09:44

0 = no level of detail in this case

09:44

32=all

09:44

smth like that

ok, in this case. but still annoying for testers having to check all 32 levels

[ddnet/ddnet-web] New branch created: pr-news

09:51

[ddnet-web:pr-news] 1 new commit

2efda59 Post Intel GPU news - def-

09:52

[ddnet/ddnet-web] Pull request opened: #104 Post Intel GPU news

can't it be less than 32?

yeah was just an example

so people can adjust the design to their liking or pc performances

i think performance doesnt differ that dramatically. using a new feature to adjust the design to ones liking is not possible on any map released up to this point, would be another thing future mappers have to take care of

Yes of course

[ddnet/ddnet] Issue opened: #3103 Optimize nameplate rendering

Which still calls SnapFindItem for every player every time

11:44

[ddnet/ddnet] Issue opened: #3104 Editor bug

Reported by Ravie, better ask him for more details !image

@Ravie which version is it broken on

12:16

I will pay whoever fixes the damn cmake issue on macOS where it obsessively links to our bundled libs

12:19

@Ravie fwiw the texture UV works perfectly fine on master

Suggestion Preparing a simplified map creation program like this greenthing

greenthing

13:24

Image attachment

[ddnet/ddnet] Pull request opened: #3105 Optimize nameplate rendering. Fix #3103

Removes the unnecessary SnapFindItem (which is a linear search, ew)

@Learath2 @heinrich5991 What do you think about the option I mentioned earlier ? (concerning the HD option)

[ddnet/ddnet] Issue closed: #3103 Optimize nameplate rendering

14:36

[ddnet/ddnet] Pull request closed: #3105 Optimize nameplate rendering. Fix #3103

14:36

[ddnet:master] 2 new commits

85f385b Optimize nameplate rendering. Fix #3103 - Learath2 5a8fe3d Merge #3105 - bors[bot]

cockblock @Learath2

np

:p

14:53

took me whole day to find it xd

14:55

Embedded image

14:56

i should have registered first before making a pr i guess

14:56

u ow me a infinite supply of shirts now learath xd

I'm not the one submitting shitPRs

its good

14:57

spaces are the worst

@Ravie I don't really see why we need another option here. performance isn't a problem

The only reason I could think of would be to let pros get a cleaner view of the map like they like but without stripping the entire design. Though I don't think anyone would use it

I have 2 people unable to update to the latest client version (15.1.1 to 15.1.2)

15:17

non steam version

Any errors?

https://i.imgur.com/sVfHeEN.png

Embedded image

15:20

the other person also said something about ddnet info error

yeah that fails sometimes, maybe we should disable the wwarning again or rate limit it

any way to solve this? or do they have to update to the newest version manually?

thats not about the version actually is it?

Well, it'd be nice to see the console if they don't mind

15:21

I'd like to see why the curl request failed

all I know is, they can't update to the newest version bcs the update button doesn't appear

15:21

i'll ask them @Learath2 abit difficult bcs they don't have discord

its in the main menu now

@murpi they can hop on irc btw no need for discord

the other person also said something about ddnet info error

@murpi

Sometimes there is a popup "ddnet-info error" showing up. Moon, translate it to him

@murpi https://webchat.quakenet.org/ #ddnet

@Learath2 I updated about 2 days ago, just to be clear the bug happens when you select different combinations of quad corners

Embedded image

15:30

is there a way to write the entire console output to a file?

15:30

or is the screenshot suffcient?

f1, logfile client.log

15:30

and then restart cleint

where does it save the file?

dunno

aniblobsweat

and the file keeps growing, mine is > 1 year old already

15:32

the screenshot doesn't show an error

15:34

maybe there's an error that we don't log

Could you review my PR on chat update pls?

me?

Idk, somwone who can finally check it, refactor if needed and merge

I haven't really looked into it, saw that there was a discussion already

there is no need to rush is there

we also haven't finished the discussion about what to do with the options yet

Already done "old chat" config option

15:36

No need to discuss

15:37

Don't like the new style - switch to completely same old

The problem with introducing more and more config options to selectively disable certain features is that it becomes hard to impossible to change the existing code without breaking something, hindering the exploration of new features such as your PR. I'm trying to solve that problem by offering less configuration options, but still allowing players to go back to a version they liked.

15:37

citing myself

15:37

that's what is to discuss

15:37

adding yet another config option is not sustainable

but we could actually move that discussion out of the PR, couldn't we?

It is, because it's only one options to make ppl that don't like an update satisfied

@deen not if players become used to those config options and annoy us due to that

I mean only 1 new config option that solves the problem

@heinrich5991 then open a separate issue, make a milestone 15.2, and tag that issue with the milestone so that we decide on it before release

You could, btw, disable parts of an old style one by one in future updates

15:39

So old players get used to new style slowly

@deen okay, I can do that

@Ravie how do you select multiple corners of a quad?

@Learath2 shift?

what do we do if it's not solved by then? revert the PR @deen

shift + left click

15:40

@heinrich5991 not release until it's solved

15:41

i mean the same issue existed for every configuration option we introduced in the last years. so it's a bit unfair to suddenly hold back one change because of it

8.76 KB

15:41

entire console output

k, I'm fine with that

@murpi is that from the run where the error occured?

Eeeeh, I don't really know what the behaviour would be if it behaved "correctly"

15:43

If you select multiple corners and the U/V value are not meant to be the same, editing them together makes a mess as expected

@deen I have no idea

[ddnet/ddnet] Issue opened: #3106 Solve the issue of exploding amounts of config options around chat

Needs to be solved before a release with #3093.

Is there a way to check if theres an update available via console?

15:44

Embedded image

15:44

lol

what console?

isn't he on the latest version?

bottom says one is available. Is that DDNet client from website or steam?

that's the reason I didn't have enough motivation to make any updates for ddnet. Because you spend a lot of time on useless discussions about "Old players may not like the update, so we will not release it. But we can make a config option to revert an update. But then there becomes to many config options, so we better don't release it" and developments stays for months.

no, he's on 15.1 and 15.1.2 is available

Honestly, the fact that we can't ever to a visual update is pissing me off. We shouldn't have to maintain options to keep the game looking the "same"

it's the client from the ddnet site

s/to/do/

Just stop being scared to improve the game already please

I think it's understandable that it's hard to change UI when people have been used to it for > 10 years

it's just one click to change it

tbf, whether it's an improvement is debatable, I don't really think the tees bring much to the table

15:47

it does look cute though

yeah, should have discussed previously if this is even wanted :/

should I just tell him to manually update then? :S edit: them (edited)

it is wanted

how do you know?

@murpi That log doesn't show an actual error, so I don't really see us being able to help them. They'll need to update by hand

is the auto updater off?

oki

shouldnt it always show the update button if a update is there

Oh wait yeah

yes, i'm checking the windows build right now

I wonder why autoupdate is off

[ddnet/ddnet] Pull request opened: #3107 Log close error to console

damn, somehow the official build is without autoupdater

15:49

@murpi yeah, manual download

The way you can deal with old unpleased players is to slightly remove old features one by one, making them get used to new styles/features/updates

15:50

if there is at least more then 75% players who likes an update. If there is around 50% maybe it's better to: 1. Improve new feature, so more ppl want it 2. Make a config variable for choosing between new/old and support both

i think the chat thing is cool as long as it's not buggy, ofc some ppl will not like it and they always complaain about it in discord but u can just add a "show chat backdrop" or w/e in the options

I already did cl_chat_old option

15:51

so, as I said already, there is nothing to discuss

15:52

move forward, update the game, stop discussing each update for months

not just cl_chat_old u should have a button in options

15:52

Image attachment

15:53

also why is this srv at top when i have both player+ping sorting

15:53

isnt it above 100 ping

it's already a button in options

hmmm @louis

@Дядя Женя Which update have we discussed for months?

15:54

Your change is 2 days old and some of us have other things to do as well

I thought each update which is not bug-fix

it's probably since it has over 64 players @heinrich5991, just a guess

I left ddnet at that moment, but how long did it took you to accept and implement "showing map name instead of "Scoreboard" text in [tab]"?

15:57

just curious

https://github.com/ddnet/ddnet/pull/1819

Show mapname in scoreboard by ChillerDragon · Pull Request #1819 · ...

Not sure if this feature should exclude non ddrace gametypes. Or exclude standard maps. Or if the truncation limit should increase with bigger scoreboards.

15:58

took 20 days to get merged, but I see no discussion

#861 took 1 month

https://github.com/ddnet/ddnet/pull/861

Pull Request using OpenGL 3.3 by Jupeyy · Pull Request #861 · ddnet...

A port to opengl 3.3. I will explain what i did here. To stay in the coreprofil i used the opengl 3.3 core profile you will find here: https://www.khronos.org/registry/OpenGL/specs/gl/glspec33.core...

xD

lol yes

that was a larger thing and we still have troubles ^^

idk what it included, but that shouldn't be so large

u cannot always make small prs

just new message from server on map load + simple text change on client side

16:00

1-2 days feature

@Дядя Женя I was really busy around that time since I moved to a new apartment and had lots of other things private things going on. I don't think you can expect responses in X days

@Дядя Женя I was talking about the opengl3 request that @Deleted User mentioend ("larger thing")

I see now. But the previous thing took 3/4 month to merge

16:02

when it's simple af

16:02

and definitely an improvement

Image attachment

which previous thing?

lol i noticed the boxes dont line up btw

"showing map name instead of "Scoreboard" text in [tab]"

they someone added a useless margin somewhere xd

Which I suggested some years ago

It took 19 days, didn't it?

So 3 weeks is 3/4 of month?

@Дядя Женя anyway, you pr changes the whole chatting, and we need to discuss anyway if we want this default or not

ooh, I misread that as 3-4 months

even if you like it doesnt mean others will like it

If someone don't like it turn this thing off omg (edited)

Image attachment

thats not how it works

but this is a maintenance burden on us

16:05

we now have to support two chat renderings (probably more, I think you added more than one option) going forward

16:05

if you say: remove old options, then great, let's remove as many options as you have introduced

you don't need to support old chat. Have some time to see how many people like and play with new style

great, so tell us which old option we can remove

i dont like the background color

there are a lot of old options already @Дядя Женя

the roundness looks bad in teeworlds

then, if its enough, force others to switch for new if there is a new update coming up that will fuck up old version

we have lots of old options. which can we remove?

the roundness looks bad in teeworlds

@Deleted User agreed, don't worry, I will move on UI soon, after the chat update xD

e.g. show friends in chat, show repetition counters, show IDs

16:07

I guess IDs isn't a maintenance burden (haven't checked) if it just changes the name

we have lots of old options. which can we remove?

@heinrich5991 show me the list of them pls

@Дядя Женя i have no problem with the UI changing, but chat is an ingame element

16:07

its not comparable with the UI

or name the most popular ones

16:08

e.g. show friends in chat, show repetition counters, show IDs

@heinrich5991 force everyone to turn this on, except for IDs. IDs should be kept as debug feature

16:08

ez, solved

k, if that's fine with everyone and you only added two options, then I'm satisfied

if you're talking about chat now, then I've added only 1 new option

great

16:10

@deen @Learath2 what do you say about removing the repetition option and the friend option and adding the new one?

there was no repetition option I saw when writing chat code (edited)

So the only problem now we don't have space for a new toggle in settings?

but was cl_friend_messages, which should be removed if you're not ok with keeping that

the problem is also,

@Deleted User no, space is not the issue I'm talking about

But you discuss it about 2-3 screens already.

16:11

i like the alignment of the new text, but i hate the background, so we need 2 options or force alignment for old text too

1

@heinrich5991 ids are really useful for mods

@deen didn't propose removing that one

what's a repetition option?

I thought we had an opton for the [1] counters

16:12

but we don't

there is no such option

so it's just removing the friend option

there is an option for hearts for friend's messages

16:13

if you don't like that you have a lot of options now - remove that one

@Deleted User the discussion is about having an ever-increasing amount of config options

should be turned on by default 'cause why you even use friend list without it

16:13

btw increasing config options is not bad

@Deleted User which is a problem for code maintenance

if it's easy to maintain them

16:14

cl_friend_messages have no problems with maintance

@heinrich5991 i love an infinite amount of config options that i can change via the ui and not the console.

it had a problem with the repetition counter already IIRC. was hard to find a design that works properly with and without @Дядя Женя

@Mikey. I'll make it for ya, my friend ❤️

@heinrich5991 in web development world we do it in phases 1. Merge a PR with new variant and a switch 2. Test it and collect stats. This needs an adaptation period as most users are always conservative even if new variant is better. 3. Remove switch and remove less favourite variant.

@Mikey. the problem is the development cost, not that it's nice to have them

This way we clean-up options at 3rd stage as they're no longer needed.

it had a problem with the repetition counter already IIRC. was hard to find a design that works properly with and without @Дядя Женя

@heinrich5991 because your counter code sucks, not because it's hard to maintain

Yeah, actually there are features that interact with each other and there are ones that don't. maybe we should differentiate a bit more.

So you fucked up the alignment of the text if cl_friend_messages

@deen but hearts do interact with the rendering of chat messages. you need to think about aligning everything properly with and without

in web development world we do it in phases

@Deleted User already told them to do so

@Дядя Женя

you fucked up

your counter code sucks

This is certainly not a constructive way to propose changes.

(yes. please tone down your language @Дядя Женя)

@Mikey. the problem is the development cost, not that it's nice to have them

@heinrich5991 is there any development cost if he already came with the finished "product"

Yeah, I'm sorry, that's just the way I'm talking, I have nothing against you and I'm not angry or smth

Even if your code is better dealing with this type of discussion and arguments might be the reason of rejecting code at all. As accepting this code means accepting this attitude. Can you change it?

@deen but hearts do interact with the rendering of chat messages. you need to think about aligning everything properly with and without

@heinrich5991 you always need to make a code such way that it will align any message properly, no matter if you add new text to it or remove something

yes

1

16:19

that's why I told before that it's not a problem with options

16:19

it's problem in you

16:19

and your coding <

generic code ftw ( i am drunk and i have no ideea what i am saying) (edited)

Just notice that people who are you talking to deserve some respect at least because they've maintained this product for a very log time and invested vast amount of their time. Also finding cons in your PR is the way discussion generally work. Even if you see your solution as perfect there could be other concerns that matter most in a long run that needs to be discussed too. Don't expect that major changes will be just accepted very fast and without attempts of finding flaws. Finding flaws in your solution and evaluating it is the way of making codebase better and not attitude to your efforts. just the way things usually work with team coding.

yes

1

I think @Дядя Женя got it already, I don't think we need to sidetrack further

2

That's not about code flaws we're talking rn

16:22

But ok, what we should do now?

16:23

If we will not talk about it nothing will change

the message from me was just about the meta discussion

There is nothing that we need time for right now. Just discuss it and move on. We don't have any solution and plan for now. Just waiting for something

You've been working on this part of code for weeks now. You're having it in your brains. Other people are saw your code less than a day ago (or even haven't yet). I think they just need some time.

I've done it in 2 and a half days

@Ryozuki halp I did yay -Syu --devel --timeupdate and my ddnet client is still 15.1.1 and yells at me that 15.1.2 is out

Half day was for understanding the source code I'm working with

@ChillerDragon why don't you just use pacaur -S ddnet-git?

16:27

and you havent even changed what i requested

bcs i use yay instead of pacaur

so your pr isnt ready anyway

You've been working on this part of code for weeks now. You're having it in your brains. Other people are saw your code less than a day ago (or even haven't yet). I think they just need some time.

@Deleted User As I mentioned before, where do you see any code discussion here?

@Learath2 what do you think about removing the hearts config in favor of the new option?

Half day was for understanding the source code I'm working with

There are really no deadlines on merging. People work on this project when they have time.

so your pr isnt ready anyway

@Deleted User of course it's not, it will probably require some refactor to meet ddnet source code standards

Don't expect full-time paid project speed here.

16:28

i gave you a hint on how to deal with it

fulltime paid project speed can actually be slower

16:29

xd

the problem is, i am the one that needs to clean it up later

@Deleted User it's better if we discuss the high-level stuff before we get to the low-level stuff

ddnet is so active o.o

so if we don't want to have this at all (see #announcements), we should not put effort into getting the code style right

fulltime paid project speed can actually be slower

Can be slower. But at least you can demand more speed there. Or try to

the problem is, i am the one that needs to clean it up later

@Deleted User and yes, we will wait until you do that, but after we'll see some accept from authorized users

@heinrich5991 yeah u are right, i keep out of this for now

me food now

So if they say: "Clean up the code, then we can merge it with config option to change for the old style, see, how many ppl use each style and then remove unwanted one later"

16:32

That would be ok, but there is nothing now

we have a vote in #announcements anyway

2

Ok, nice

@Дядя Женя Oh, and I'm really looking forward to your future improvements. You seem quite productive if you managed to get into our code base so quickly

That's true. If you're curious I'm planning to finally implement a color-picker UI component next (edited)

we have one in the editor btw

Where exactly?

well, not a proper one, but at least it shows rgb and a preview

16:39

right click on a quad corner

Oh, yeah, but it's not what I'm looking for (edited)

Why you want to rush so much

But that will bw helpful

instead you want one where you can click inside of it?

Yeah

16:40

As well as type rgb and hex (edited)

16:40

Or hsv mb

sounds good to me

16:41

also sounds less controversial than changing chat

hey

16:43

so i have a problem with my client O.o

@Дядя Женя fill feature in editor would be really nice. Like in mspaint.

16:44

It doesn't matter for ddrace maps as they're mostly lines and rectangle blocks, but filling hollow structures in a map is a pain now.

16:45

I feel like this all the time wasting about 15 minutes of my time on a map where it could be done for 10 seconds with fill:

16:45

Image attachment

@Luan.499 What's wrong?

Ok, why not. But I'm mostly focused on features I want myself and I'm not a mapper unfortunately. And, unfortunately, I want a lot of design changes over there

16:47

So that's only a start of <old-new> war now

When I go my tea is sometimes very slow and my when i open my scoreboard it flickers

get ready for more. I'll split this server on two halves lol

When I played yesterday it was all fine

@Luan.499 any update inbetween?

16:49

ddnet update, windows update?

[ddnet/ddnet] Pull request opened: #3108 Make GL 1.5 default

If ppl see white textures its related to #2125 and you need to revert it, bcs thats not fixable, there is no extension for GL_GENERATE_MIPMAP or smth

@Learath2

Image attachment

Hmmm no not a ddnet update but some windows updates

@Luan.499 might have been an intel gpu driver update? Do you use an Intel GPU?

@deen hmmm wait i take a look

honestly it might be a good idea to make a revamped intuitive way to cut quads

@deen i think i did like a communication update or something i cant remeber

[ddnet-scripts:master] 1 new commit

58f26a8 Fix autoupdate in releases - def-

@Luan.499 f1, gfx_opengl_major 1 and restarting client helps?

[ddnet:15.1-devel] 1 new commit

84be9b8 Version 15.1.3 - def-

@deen what is f1, gfx_opengl_major 1 ?

[ddnet:master] 1 new commit

2825089 Version 15.1.3 - def-

In client press F1, write gfx_opengl_major 1 and press enter, restart client

Ahh okay

17:02

Nope it didn't work

Maybe your keys don't register correctly?

17:03

and it flickers because your key isn't pressed down all the way? is the keyboard working fine?

@deen gfx_opengl_major 1

17:04

@deen is that right?

17:04

@deen yes the keyboard is fine i tried to but some new binds but it didnt work

Is saving a top1 rank play ghost on server is a good idea? Then making any player able to turn top1's ghost on, so he can compete in speedrun or see how to make a part maybe. It may make a speedrun funnier and better? Or it's not, just throwing in an idea. For maps like puzzleboxes disable it in sv_settings

@Ravie eeh, how should it behave though. It's setting the UV exactly as you ask it to

Ok, looks like stupid idea, forget it now xD

It just doesn't make sense for the top corners U coordinates to be same

But maybe it's good for solo maps

@Дядя Женя We do something like that with teehistorian, and have recordings of every gameplay on ddnet servers in the last years. We are missing the teehistorian 2 demo converter

17:07

(or to ghost)

sounds easy enough

@Luan.499 I don't have other ideas then. Maybe try if older DDNet versions work.

It makes implementation a lot easier, but what about the general idea?

@Дядя Женя teehistorian just records inputs

is it good to show top1 play to everyone

I'd love that

17:07

yes

17:08

Would really make the servers more competitive. pros will complain that their tricks get out, but so be it

@deen I have a video from it maybe it helps u a bit more

problems here: 1. everyone can see a skip you found 2. breaks puzzleboxes (solution: disable it in map settings)

https://github.com/ddnet/ddnet/issues/2937

Teehistorian Replayer · Issue #2937 · ddnet/ddnet

I haven't really started yet, if someone wants to help tell me: Replay modus on server that opens a teehistorian file as inputs Should run faster than realtime, no sleeps, instantly go to n...

@deen I still can't get if C++ deterministic or not

Is saving a top1 rank play ghost on server is a good idea? Then making any player able to turn top1's ghost on, so he can compete in speedrun or see how to make a part maybe. It may make a speedrun funnier and better? Or it's not, just throwing in an idea. For maps like puzzleboxes disable it in sv_settings

@Дядя Женя i think it's cooler if you can't steal speedrun techniques

what's going on there, can you just completely reproduce player's vision of his play while having only inputs

17:10

I had an Idea of public top1 ghost for years now, but it always had problems and disagreements in my mind

yes, that's the idea

17:11

and since we have the skin downloader we could also do something similar with ghost downloader to get the top ranks

btw btw btw

17:12

Another good idea is to download player's emoticons

2

17:12

from database

@Learath2 some combinations work as expected (for example 2 vertically adjacent corners) while other combinations spaz out when you start changing the value and you end up with some mess of coordinates

like skins exactly do now

17:12

who no @louis

not everyone needs custom emotes

17:13

emotes are on same level as gmeskin and particles

17:13

+ not a wide vaariety of good emotes

agree with louis, would be too distracting to see to many emoticons, i like the current selection

@deen it would be cool to see skins in chat

@Ravie I think it's more that you don't understand the concept of UV. If you want to stretch it towards the top you need to mess with the V coordinate

@ᶰ°Konͧsti #announcements

ik

if you drag corners themselves it will behave as expected

Im just saying that even me is not against it

17:15

Which is kinda rare Since im against anything new feature normally

@louis @deen you right, but then force everyone to use kitties emotes xD

justatest

@Learath2 I do understand it and yes, linear stretching works but for example selecting all corners bugs out instead of shifting the whole texture around without distorting it (like it used to) (edited)

[ddnet-web:pr-news] 1 new commit

07af5b5 Version 15.1.3 - def-

I don't get how it could have ever worked, before the last patch the U and V didn't even work

[ddnet-web:master] 1 new commit

e19b097 Version 15.1.3 - def-

it did work cause I've been using it for a while

17:22

and every time I need to shift around a texture without distorting it or changing the aspect ratio I would select all corners and use U/V to just move the texture around, in the new version it just bugs out

17:22

I remember the numbers shown in the fields were acting funky, but in the end everything worked as expected

I see what you mean

17:23

I'll fix it in the next patch

17:23

Actually I don't know how to fix it

17:23

We can't display a number if it's not all the same

Can someone tell me what's the difference between these?

m_aClients[id];
m_Snap.m_aCharacters[id];

it would be nice if someone could code a better way to cut quads, most people don't understand how it works because it's such a crap system

17:27

it would make much more sense to drag a selection over the area that you want to crop a quad to

4

17:27

while maintaining the aspect ratio of whatever was in that area

@Ravie why not cut the png into parts?

you need to have a separate .png and layer inside the editor for each part then

17:29

and there are other reasons to crop quads

yes quad chopping sucks i just make a separate png even when chppping is a better option just because idk how to chop

chop chop

it's very useful to keep related objects or frames of an animation in the same image and layer

17:32

so far I've been using a workaround to get this result but it's kinda fucky and a lot of extra work

Can someone tell me what's the difference between these? ```c m_aClients[id]; m_Snap.m_aCharacters[id]; ```

@Deleted User m_aClients is an array of players iirc, players keep existing if you die, m_Snap.m_aCharacters is an array of characters that the snapshot has, note: a player has a character if he is alive, when he dies the character gets deleted, also one thing is the character in the snapshot and the other is the one owned by m_aClients[i]

1

17:34

well this is from my head

17:36

actually m_aClients maye be smth else

17:37

github search rly sucks

m_aClients is for CClientData

17:38

https://github.com/ddnet/ddnet/blob/28250897d3017fae1a1c5c10bd966064e3b1d8c0/src/game/client/gameclient.h#L240

DDraceNetwork, a cooperative racing mod of Teeworlds - ddnet/ddnet

@deen i have a video from it

17:52

@deen maybe it can help u a bit?

How can I fetch and pull master branch from ddnet source and start working on other updates when leaving chat branch untouched, if the chat was pull requested from master branch of forked repo?

18:24

May I close pull request, make a new local branch and pull request from it?

18:25

Should have done that before...

@Дядя Женя the local and remote branches don't have to correspond. in the git command line interface you can say git push origin localbranchname:remotebranchname to push in arbitrary mappings

I can make a new branch in my repo and it will not touch the master with chat until I merge it

18:28

But then I can't pull anything from real ddnet repo

18:28

Since it will be pulled to master

you can do git fetch upstream git merge upstream/master

@heinrich5991 and yes, I need a remote branch, not local, so I can push in it ans it will save my changes in cloud

pull is an "automated" combo of fetch and merge

is an 'upstream' name for the new branch as an example here?

no, upstream is an example name for the remote

18:30

can you give me the output of your git remote -v?

18:30

or tell me what git command you usually use for pulling

18:30

is it just git pull?

18:31

or git pull https://github.com/ddnet/ddnet?

Is that it, I'm noob, you know

Image attachment

18:32

I'm using fork program for windows

do git remote add upstream https://github.com/ddnet/ddnet

18:32

sry, I don't know how to deal with that program :/

so I can click bUttons bUttons, not typing cmds in console

18:32

added

then you can do git fetch upstream

done

have you already started working on your new feature or do you want to create a new branch for that now?

2nd

git checkout -b <newbranchname> upstream/master

18:34

this creates a new branch named <newbranchname> based on the current state of upstream/master

I see

18:36

Can I somehow fix the fact I fucked up by making commits to master?

18:36

instead of staging

do you want to fix it on github's side or just in your local repository

18:38

for github, you'd either need to wait until the PR is accepted or close this one and create a new one

Idk, but Im probably doing something wrong

catJAM

Image attachment

18:39

it's total mess now

18:40

and I can't use my chat update when making new features in color-picker branch

ah, do you want to use that?

18:40

we can also base your new PR on your old PR

ofc I want, it's perfect (edited)

you haven't done any changes yet?

I guess not

then you can do git reset --hard origin/master in the command line

18:41

this resets all the work that was done without committing, and puts you at the state of your origin/master

18:41

this is the state of your chat update branch

yes, I see

18:42

So I should create new branch now

18:42

and pull-request from it probably

18:42

help me help me :3

@deen can I make +fire_dummy command to avoid that stupid bug with switching weapons on dummy since the dummy firing is already used a lot as +toggle cl_dummy_hammer 0 1?

19:25

Also It will allow dummy to fire you with weapons when holding hook on tiles without releasing it

19:26

(also already used as mentioned before bind + hook hold in needed direction) (edited)

19:26

Just making a hacks less buggy and easy to use

i today making map on 60hz monitor and watching tv-series on 144hz . teeworlds didn't crashed no single time ! Could it be reason with 144hz problem ? (edited)

Im using the same setup and have no bugs at all

@Deleted User can you use a debugger to catch the crash?

We got overall idea. Some people don't like chat update, but around 2 times more people likes it. What shall we do now? xD

21:17

fix bugs and release it with fallback to old style config

yeah, sounds good

a type of "envelope" in the editor where u could animate camera movement while rendering would be amaznig

21:42

amazing*

@deen i found out that it is the same with other games

Exported 580 message(s)