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 2020-02-03 00:00:00Z and 2020-02-04 00:00:00Z

[ddnet-maps:master] 1 new commit

436d4a2 A Canyang - ddnet-maps

https://github.com/ddnet/ddnet/blob/master/src/game/server/gameworld.cpp#L228 does this mean that the real client id k's fakeid is now Mapc for the current pMap?

11:18

https://github.com/ddnet/ddnet/blob/master/src/game/server/gameworld.cpp#L236 and this, that for pMap the fake id rMap[k] gets reset to -1 now because hes too far away?

11:18

Can anyone explain the playermap updating? :D

[ddnet/ddnet] Issue opened: #2056 Physics Engine Issue: Quick dropping tee triggers tele tiles abnormally

Dropping tee accumulates vertical speed and after a specific point, it could sometimes trigger outer tele tiles before inner ones. It is quite an old problem but I do not see any fix or discussion. Reproducing steps with hammerfly map DDmax-Hell Fly: This map does not have any checkpoint to limit teleport, purely depends on tele number. !QQ图片20200203200324 ![QQ图片20200203200547](https...

@fokkonaut rMap is a map from real ids to ids in the players idmap

12:47

Mapc iterates through the "fake" idmap to find an empty slot to put the client in

12:47

if there isn't any, Demand is increased

12:52

Demand signifies how many slots we need in the fake map to fit all the closest 16 tees after fitting what we can

12:53

Then we remove Demand-many clients from the idmap so that in the next iteration of updateplayermaps the required clients fit

12:53

The idea with the weird algorithm is that client ids don't change while the client is in view, as that causes weird behaviour

I am trying to port 128 player to 0.7, its working fine but there is one problem, because in 0.7 you cant just change the client ids in the player's snap, or their name. you would need to send disconnect/connect packets to client x individually. I couldnt figure out yet where to send those correctly...

Huh, what do you mean you can't change client ids

12:59

In the pMap, if a slot is set to -1, that means it "disconnected", if it's set to something else that means it "connected" could work

the client receives connect/drop messages

12:59

and there, the client id , name and other information is passed

13:00

so you cant change the client ids anymore, or names and stuff without sending reconnect packets

13:00

(disconnect, then connect)

I see, what an ingenious idea to move all that out of the snap

yes

why bother with 128p anyway, it's hardly viable

well, because my server is filled with 64p sometimes

13:02

currently 45 again

anyway, you can try what I said

I tried a lot yesterday, couldnt get the info correctly. So players 0-63 were correct, but the ones with higher client ids would just apply the information from the fake id they have for you

13:03

Because all that is not in the snap anymore :/

Send a disconnect about rMap[pMap[i]] when setting pMap[i] to -1

and a connect when it gets set?

13:06

i tried exactly that. But then the problem was, disconnect packets have to get sent before the other connect packets can come of course. so i stored them in a vector and after both loops i just sent the disconnect ones, then connect

13:06

but also that didnt work

13:06

No idea why tbh

13:06

i could push my current progress so you can take a look, if youd like

well first of all verify that you get the packets in the correct order, teeworlds network stack doesn't really guarantee order iirc

When it's a vital packet order should be the same

I'm not sure, what if packet a is dropped, packet b arrives, packet a will surely be resent as it's vital

13:09

but it'd arrive after packet b

The received will ignore b and the sender will resend both. When I remember correctly

13:10

Highly inefficient but it works

13:11

receiver*

Seems that's right @redix

I want to map at Nintendo switch or mobile phone , will keyboard work on mobile phone (displaying all stuff example space bar)

depends on the phone

Example if press space bar *

16:07

Huawei

and either way you'll need a build for the device, we have an old android build

16:08

I actually wanted to do a nintendo switch build but libnx is just not good enough yet in the homebrewsphere

Oh also its hard to publish there :( i think its costs money :(

16:09

Will i have problems because of old android verision ?

I don't think you can map with the older version

Bruh

[ddnet/ddnet] Issue opened: #2057 Logic warning from clang [-Wtautological-bitwise-compare]

`[build] ../src/game/collision.cpp:885:247: warning: bitwise or with non-zero value always evaluates to true [-Wtautological-bitwise-compare] [build] if(m_pDoor[TileOnTheRight].m_Index == TILE_STOPA m_pDoor[TileOnTheLeft].m_Index == TILE_STOPA ((m_pDoor[TileOnTheRight].m_Index == TILE_STOPS || m_pDoor[TileOnTheLeft].m_Index == TILE_STOPS) && m_pDoor[TileOnTheRight].m_Flags|ROTATION_270|ROTATION_90)) [build] ...

Lol that Spoiler shouldn’t the GitHub bot message be a markdown free area?

Probably forgot to escape the text there

nobos

find me a markdown escaper for php and i'll add it

But I mean the whole thing looks like it is made for this

23:39

U?

fixed the comment

Wait? Who made the bot?

@Learath2 filters the events, but the actual bot is by one of discord or github I think

Yea I thought so that it is something official

my filter pretends to be the github webhook

That’s why I wonder it does not have proper markdown escape

23:41

Hm?

I get the events from github, then i pretend I'm the actual github bot to discord

Ah so you send the actual messagesb

23:41

Ah i see

23:41

And you wrote it in php I guess

yes, that's the one language I'm quite efficient with when writing web stuff

Did you try bash

?

23:42

Oh yea and sorry I just wanted to rant on Microsoft I didn’t expect the bot to be built by u

oh yea, please give me a user account on your machine by writing a internet-facing service in bash

23:43

*an

eh, discord only supports a subset of markdown, so i'll need to code an escaper on my own

bash is bullet proof :p

every ctf has atleast one bash challenge, very easy to get bash wrong

I wish I wasn’t too nobo

23:50

One day ima join a ctf

Exported 78 message(s)