DDraceNetwork - developer

though I do have that other tool that already takes screenshots of maps without a window, using that same method would give you exactly that

00:00

its definitely a thing I keep in mind :)

Learath2

the person who kills will have the lowest entity id iirc

But there's a segment in KillTeam(int Team, int NewStrongID, int ExceptID) that purposefully uses the NewStrongID/ExceptID to ignore the person who killed first

The person who hits kill first is already killed, so when killing team we need to ignore the already killed person

00:14

And that person ends up with the lowest entity id, since they killed before the team

Hence, Die(int Killer, int Weapon, bool SendKillMsg) (edited)

00:21

I'm confused at what you guys are confused at <:(

sorry

00:21

I'm not investing the right amount of time to criticize the PR properly, I think that's why I'm being so vague :/

It's my fault.. I should be more patient

Voxel

I'm confused at what you guys are confused at <:(

I'm just confused why you have this bool when you can get team state inside CCharacter::Die

anyway, what's currently a bit weird (I don't have a good solution to this) is that the Character::Die function basically implements the team logic for whether to send a team kill message or just a kill message

Ravie

I'm just confused why you have this bool when you can get team state inside CCharacter::Die

c++
void CGameTeams::KillTeam(int Team, int NewStrongID, int ExceptID)
{
    for(int i = 0; i < MAX_CLIENTS; i++)
    {
        if(m_Core.Team(i) == Team && GameServer()->m_apPlayers[i])
        {
            GameServer()->m_apPlayers[i]->m_VotedForPractice = false;
            if(i != ExceptID)
            {
                GameServer()->m_apPlayers[i]->KillCharacter(WEAPON_SELF, false); // <<< this line of code is connected with the tees dying, and if i don't have the bool then what's the point in squashing it all

                if(NewStrongID != -1 && i != NewStrongID)
                {
                    GameServer()->m_apPlayers[i]->Respawn(true); // spawn the rest of team with weak hook on the killer
                }
            }
        }
    }
}

(edited)

00:26

The false in KillCharacter tells the code if to send the message or not

00:26

And if all the tees send the message and it's squashed it's very repetitive

heinrich5991

anyway, what's currently a bit weird (I don't have a good solution to this) is that the Character::Die function basically implements the team logic for whether to send a team kill message or just a kill message

I mean, that's the thing. When a tee dies, it doesn't even touch up on any team related code. It sends the kill message before anything else

as said I don't have a good solution to it

00:29

but it could be moved out of this function entirely

00:29

e.g. making CTeams send the kill message

00:29

but don't implement that now, I don't see why this is a good solution right now

Voxel

The false in KillCharacter tells the code if to send the message or not

I know but you can decide this INSIDE CCharacter::Die, you have access to team stuff there

okay i think i know what you guys are talking about

00:32

yeah i'll wait until fokko verifies this until I do it

but I also agree with heinrich that team logic inside Die function is kinda unclean

cwh7435

[ddnet/ddnet] Pull request opened: #6475 Update Korean translations by CHaBek

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 (especially base/) or added coverage to integration test
[ ] 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-ad...

bors[bot]

[ddnet:master] 2 new commits

d0d3325 Update Korean translations by CHaBek - cwh7435 98a5c4e Merge #6475 - bors[bot]

06:14

bors[bot]

[ddnet/ddnet] Pull request closed: #6475 Update Korean translations by CHaBek

Ravie

I know but you can decide this INSIDE CCharacter::Die, you have access to team stuff there

I wouldnt do it there

Nice stuff !!! (@Patiga)

Patiga

output.mp4

06:38

Im hyped

Patiga

@Jupstar ✪ I finally actually started on the tee rendering ^^

looks funny with eyes only

def-

[ddnet-web:master] 1 new commit

378d395 Update translation status - def-

fokkonaut

[ddnet/ddnet] Issue opened: #6476 Death effect got bigger?

Reported by @cyberFighter probably related to https://github.com/ddnet/ddnet/pull/6439 ? @Robyt3

hm I just implemented interpolation for the rendering. on the stronghold (also cutted) demo, there are still the stutters as in the video above, but on the new demo I'm trying everything is perfectly smooth thonk

GitHub

Click to see attachment 🖼️

ah that caused it xD i already saw it on kog and thought its some weird new feature

Patiga

hm I just implemented interpolation for the rendering. on the stronghold (also cutted) demo, there are still the stutters as in the video above, but on the new demo I'm trying everything is perfectly smooth thonk

and in ddnet client?

maybe now its more random

Jupstar ✪

and in ddnet client?

smooth

Ryozuki

maybe now its more random

it was bigger radius xd

12:49

randomly bigger radius ez

Patiga

smooth

mhh weird

I simply just take the characters out of the snapshot and read the x, y and angle of the contained charactercore

12:51

ah

i'd start with printing all variables involved and see if its not an interpolation bug

got it, there are some snapshot parsing issues

ah

for some reason I had logging turned off

12:52

interesting that this causes a difference tho

12:53

since obv part of the parsing still works fine, super weird

12:53

were the character snap objects variable sized at some point o.o

12:54

(wild guess)

don't think so

@heinrich5991 u might be interested

13:35

https://github.com/Vanille-N/tree-borrows/blob/master/half/main.pdf

13:35

https://github.com/rust-lang/miri/pull/2785

Tree borrows by Vanille-N · Pull Request #2785 · rust-lang/miri

This PR implements the experimental Tree Borrows (TB) rules for tracking reference aliasing, as an optional alternative to Stacked Borrows (SB). SB and TB are mutually exclusive. Using -Zmiri-tree-...

13:35

https://perso.crans.org/vanille/treebor/

I don't understand it ^^

13:37

ah

damn scholar slang

it's some runtime checker for borrowing rules

for miri

GitHub

Click to see attachment 🖼️

I can fix on the weekend or someone else can if they find jt. Seems like I did a math error when calculating m_Vel even though I checked at least 3 times. Previously it was a random value between 0.1f and 1.1f, now it's a random value between 1.0f and 1.1f.

is there a list of all attributes that can be changed via console?

Simon

is there a list of all attributes that can be changed via console?

https://ddnet.org/settingscommands/

wtf i like the new death particles

Same

What's int To?

It smells like your GameServer() is bad, no idea how you added your new class but have you initialized your gamecontext pointer properly?

Are you sure GameServer() is valid?

Did you initialize m_pGameServer properly?

15:59

e.g. in the constructor of your class or an init function

Btw whats better design, constructor or init function? Or is it just used when its preferred in each situation?

fokkonaut

Btw whats better design, constructor or init function? Or is it just used when its preferred in each situation?

We don't really have a convention for it, but if it can fail it needs to be in an init function since we don't use exceptions

Original message was deleted or could not be loaded.

Ofc not, 0x0 is a null pointer constant. You need a pointer to the working instance of CGameContext

Learath2

Ofc not, 0x0 is a null pointer constant. You need a pointer to the working instance of CGameContext

kek

16:04

try 0xbeef (joke)

0xDEADBEEF

16:05

0xBAADFOOD

16:06

or my favourite 0xBAAAAAAD

16:08

https://www.networkworld.com/article/2222804/microsoft-code-contains-the-phrase--big-boobs------yes--really.html this reminded me of this

Microsoft code contains the phrase 'big boobs' ... Yes, really

UPDATED: If you thought we might be beyond this stuff by now, think again

16:12

Yes, and when calling your constructor make sure you are calling it from somewhere where you have access to a valid game context pointer

Original message was deleted or could not be loaded.

oh actually the signature should be CMyClass::CMyClass(CGameContext *pGameServer)

16:13

You can't just name constructors whatever you want, and you can

16:14

and you don't want to pass CGameContext by value, you want a pointer to it

are you inhereting cgamecontext?

Original message was deleted or could not be loaded.

huh, you don't need to create new instance of gameserver, just pass it on CGameContext::OnInit(...)

17:12

i don't remember which class has this thing but search up for init

17:12

ah wait im wrong again

Original message was deleted or could not be loaded.

normally no

today i started implementing a tagged union in llvm (like rust enum) and remembered that llvm is moving from pointers with types to opaque pointers

17:49

well it already moved on

17:49

https://llvm.org/devmtg/2015-10/slides/Blaikie-OpaquePointerTypes.pdf

17:49

https://llvm.org/docs/OpaquePointers.html

17:49

LLVM IR pointers can be cast back and forth between pointers with different pointee types. The pointee type does not necessarily represent the actual underlying type in memory. In other words, the pointee type carries no real semantics.

>

Historically LLVM was some sort of type-safe subset of C. Having pointee types provided an extra layer of checks to make sure that the Clang frontend matched its frontend values/operations with the corresponding LLVM IR. However, as other languages like C++ adopted LLVM, the community realized that pointee types were more of a hindrance for LLVM development and that the extra type checking with some frontends wasn’t worth it.

Original message was deleted or could not be loaded.

As I said, don't pass CGameContext by value

18:08

This is why I recommend learning C++ properly before trying to do things, it's much more pleasant when you don't have to blindly hammer on things until they work

Marmare314

[ddnet/ddnet] Pull request opened: #6477 Allow multiple selection in filedialog

!multiple_selection When loading a lot of already loaded pictures all warnings are shown and you have to click them away one by one. This is definitely not ideal. One option to solve this would be to only show one warning. The other would be to show a list of warnings, but I'm not sure how hard that is to implement.

Checklist

[x] Tested the change ingame
[x] Provided scr...

Learath2

This is why I recommend learning C++ properly before trying to do things, it's much more pleasant when you don't have to blindly hammer on things until they work

well you learn by doing things, that's kind of a paradox to expect of someone

Ewan

well you learn by doing things, that's kind of a paradox to expect of someone

You learn by doing specific things targeted at learning. Not everything you might want to do works well as learning material

Original message was deleted or could not be loaded.

That part looks correct, how did you declare it in the header?

most folks on the internet who code would not be where they are without screwing around with codebases that are way out of their league

Ewan

most folks on the internet who code would not be where they are without screwing around with codebases that are way out of their league

So did I and I butted my head against random things until I picked up a book and learned the language. You learn 10 times quicker from proper material

as is the case with me but you will never keep interest in the material without having some background with it

18:36

you have to struggle to succeed

Original message was deleted or could not be loaded.

This looks like it has the correct signature too, so that leaves the place you call the constructor. I'm guessing you are trying to construct a new CGameContext tgere

18:36

there*

Ewan

you have to struggle to succeed

Eh, I find that certain kinds of struggle is just not useful, but to each their own I guess. I would never learn a language that way again if I can help it, that's all I know for sure

if it was your first language, you don't have to

18:39

though yeah there are very unproductive forms of struggle that beginners go through

18:39

that they can definitely be made aware of to avoid

Original message was deleted or could not be loaded.

please post the full code so we can better help you

Ewan

that they can definitely be made aware of to avoid

but this is not it. he's asking for help and trying to do things the right way

making things that don't work helps you actually understand how things work rather than reading all the theory and just "doing things how it should be"

agreed

Ravie

making things that don't work helps you actually understand how things work rather than reading all the theory and just "doing things how it should be"

in code and informatics yes, don't try that with chemistry

Ravie

making things that don't work helps you actually understand how things work rather than reading all the theory and just "doing things how it should be"

If it doesn't work due to a failure in your problem solving, yes. If it doesn't work due to you just not understanding how a class and it's instance differ I honestly don't think looking at it over and over will help much

hmmm, pseudocode is just the high level view, doesn't say a lot about c++ implementation

Ravie

making things that don't work helps you actually understand how things work rather than reading all the theory and just "doing things how it should be"

Also why good teaching never consists of "read all theory before doing anything", you are supposed to get bite sized pieces of theory, then doing exercises directed at digesting that information

especially if you're a newcomer to coding

Original message was deleted or could not be loaded.

Uhm, maybe paste it to somewhere like codepad or pastebin?

18:45

broke my keyboard lmao

if ur asking every 10 mins about a problem here, its better u learn some actual tutorial

18:45

cuz at this point u using the ppl time here as a tutorial

Well I don't mind it tbf. I only wanted to say that there is a better way that'll cut the pain short

Learath2

Also why good teaching never consists of "read all theory before doing anything", you are supposed to get bite sized pieces of theory, then doing exercises directed at digesting that information

doing exercises is just as much of a waste of time as figuring things out on your own but way less fun

Hard disagree on the waste of time part. Agree on the less fun part (though well made exercises can actually be fun)

18:51

The whole idea of exercises is that they are targeted. You learn some theory, you soludify the knowledge in an exercise within 30 minutes. You can spend 8 hours trying to code something fun and not learn anything new

i think u see exercises with too much "school" point of view, there are things i consider exercises that are way more fun

18:51

like nand to tetris

18:51

https://www.nand2tetris.org/

Home | nand2tetris

at least for me personally it's much better to understand things in my own ways rather than being fed some information and methods, and I learn fast so I guess there is something to it

Original message was deleted or could not be loaded.

Missing the interesting part where you construct the class

is ur source code closed?

18:53

well in that case i wont help, not that i was doing before anyway

18:54

whats stopping u from uploading to git

Ravie

at least for me personally it's much better to understand things in my own ways rather than being fed some information and methods, and I learn fast so I guess there is something to it

An experienced programmer learning their 6th language is very different to someone learning their first

6th? weak numbers

Ryozuki

6th? weak numbers

ryo with 53 known languages and 7 upcoming

and my own