DDraceNetwork

Kaffeine

Weird work with m_JumpedTotal in CCharacterCore::Tick(): https://github.com/ddnet/ddnet/blob/1dea02d4562ef6028db47da1a3fb3fd0c4efdfc8/src/game/gamecore.cpp#L156-L211

if(m_Input.m_Jump)
{
    // ...
    if(Grounded)
    {
        m_JumpedTotal = 1;
    }
    // ...
}

// No any read from m_JumpedTotal

if(Grounded)
{
    m_JumpedTotal = 0;
}

I do not know about the intention so I can't propose a fix. If there is no bug then at least m_JumpedTotal = 1; can be removed without behavior changes. (edited)

m_JumpedTotal and m_Jumped needs to be refactored/reworked anyway because it is incorrect if you have 1 jump. Also we have a special case with -1 availible jumps that also creates strange results xD ... But yes your point is correct, the check if you are grounded should be done before the check if you make a ground jump. so that if you do a ground jump m_JumpedTotal is 1 and not 0. But this was always like this xD but we could finaly fix it. Do you want to try fixing it?

08:11

m_JumpedTotal is always one off because of this... It laggs one jump behind because ground jump is not counted. but because of this, you have to care about the other physic logic and not just count ground jump

08:12

It was always like this, so take care to break no physics

Kaffeine

Weird work with m_JumpedTotal in CCharacterCore::Tick(): https://github.com/ddnet/ddnet/blob/1dea02d4562ef6028db47da1a3fb3fd0c4efdfc8/src/game/gamecore.cpp#L156-L211

if(m_Input.m_Jump)
{
    // ...
    if(Grounded)
    {
        m_JumpedTotal = 1;
    }
    // ...
}

// No any read from m_JumpedTotal

if(Grounded)
{
    m_JumpedTotal = 0;
}

I do not know about the intention so I can't propose a fix. If there is no bug then at least m_JumpedTotal = 1; can be removed without behavior changes. (edited)

If there is no bug then at least m_JumpedTotal = 1; can be removed without behavior changes.

I would prefer the correct value of m_JumpedTotal, so ground jump should also be counted. So please do not just delete m_JumpedTotal = 1; Instead fix the complete logic xD

APFFF

Only happened once so far: Rendered Video is sped up gameplay, Sound is normal speed. Was rendered after recent Render changes

I can not reproduce it, what version do you use? can you share the demo? I tested it with a demo on 3x speed, sound is correct

see https://streamable.com/linefz and https://streamable.com/ci9ue2

c0d3d3v

If there is no bug then at least m_JumpedTotal = 1; can be removed without behavior changes.

I would prefer the correct value of m_JumpedTotal, so ground jump should also be counted. So please do not just delete m_JumpedTotal = 1; Instead fix the complete logic xD

it is being counted

09:32

also, JumpedTotal is being used in CCharacter

09:33

all correct

fokkonaut

all correct

no it is not correct, at least not in the predicted code

what is incorrect?

i never understood why the prediction code is almost a copy paste of the real code

09:36

why didn't they just share it to avoid such problems?

Chairn

i never understood why the prediction code is almost a copy paste of the real code

to predict it? xD

Chairn

why didn't they just share it to avoid such problems?

was too complicated on the fly

yeah, but you don't need to copy paste it, just makes the interface usable by both

not entirely easy

well, i guess some data is only server side

thats why the charactercore is shared, and the character itself isnt

Chairn

well, i guess some data is only server side

thats why we predict the missing data xD lol

fokkonaut

what is incorrect?

I can make a video one moment xD

just explain

I already explained it

where

c0d3d3v

m_JumpedTotal and m_Jumped needs to be refactored/reworked anyway because it is incorrect if you have 1 jump. Also we have a special case with -1 availible jumps that also creates strange results xD ... But yes your point is correct, the check if you are grounded should be done before the check if you make a ground jump. so that if you do a ground jump m_JumpedTotal is 1 and not 0. But this was always like this xD but we could finaly fix it. Do you want to try fixing it?

just read it

ah

c0d3d3v

thats why we predict the missing data xD lol

i think my brain turned off this morning

Chairn

yeah, but you don't need to copy paste it, just makes the interface usable by both

I would also prefer one interface, so that we share the code for server and client. But I also think we have in the client partially also vanilla physic predictions ... so it is not very easy to do that

c0d3d3v

I can not reproduce it, what version do you use? can you share the demo? I tested it with a demo on 3x speed, sound is correct

Ver 16.0.3 (I think). But as I said, only happened once so far

Mult0map_2022-04-24_20-36-35_replay.demo

817.25 KB

APFFF

Only happened once so far: Rendered Video is sped up gameplay, Sound is normal speed. Was rendered after recent Render changes

if u find a way to reproduce tell us ^^

sure

nice