Teeworlds - development

Teeworlds

discord.gg/teeworlds / development

For discussions around the development of the official Teeworlds

Between 2020-09-06 00:00:00Z and 2020-09-07 00:00:00Z

[teeworlds/teeworlds] Issue opened: #2714 Prediction for 0.8

I'm glad that prediction for other players is now official, and it is no doubt a highly demanded feature. Since the PR is now closed, I would suggest keeping a discussion regarding a future version of prediction. Here's the future works stated in the PR:

[ ] Predict actions on other tees

Currently when you hammer another tee with cl_predict_players on, the force on the other tee isn't predicted
[ ] Predict grenade jumps
[ ] Projectile prediction (for grenade, shotgun, pi...

[teeworlds/teeworlds] Pull request opened: #2715 Show skin names inside skin buttons.

!screenshot_2020-09-06_06-40-07 Search and save features kinda lose half of their purpose when you can't see all skins names.

[teeworlds/teeworlds] Pull request closed: #2715 Show skin names inside skin buttons.

07:18

[teeworlds:master] 2 new commits

112c070 Show skin names inside skin buttons. - TsFreddie 45233c5 Merge pull request #2715 from TsFreddie/0.7-ski... - oy

You on fire @TsFreddie

EeveeShy

https://streamable.com/kad4z1

09:12

The in game cursor speed is inconsistent compare to menu cursor. (edited)

09:12

I would think 100 for both sens should have roughly the same result.

09:14

oof, sorry for the music blasting at the end of that video.

i think its affected/depends by inp_grab (ingame)

both menu sens and in game sens are related to inp_grab i think

now i dunno if it changes behaviour in menu too

~~and of course with inp_grab 1 there is no way to test it.~~ (edited)

09:20

inp_grab 0 allows me to have windows cursor as a reference.

09:26

https://streamable.com/fxlumz

09:26

it is quite accurate in both mode, but i do wish the sens can be matched.

@TsFreddie with inp_grab 0, the mouse is handled by the operating system I think

09:29

windows has mouse acceleration by default

09:29

no way to match that in teeworlds, currently

i turned that off ofc

well

10:05

https://streamable.com/ooi98l

10:05

i did it

But I did switched cursor drawing from world space to screen space

10:09

~~although I don't know why would you draw cursor in world space in the first place.~~ (edited)

¯\_(ツ)_/¯

10:09

@mind here?

I guess some mods could use world space cursor position to do stuff. and matching the targetpos with server is more important (edited)

10:22

EeveeShy

There must be some way to scale the sens to match menus.

@Sonix yes

[teeworlds/teeworlds] Issue opened: #2716 Remove Filter may be clicked out of bounds

In the Server Browser menu, one may click outside the server list and still hit the "remove filter" button outside the range. The filter that is off bounds: !20200906122335_1 Before Click: !20200906122353_1 After Click: ![20200906122355_1](https://user-images.githubusercontent.com...

10:35

[teeworlds/teeworlds] Issue closed: #2709 Consistent scrolling speed with mouse wheel

10:35

[teeworlds/teeworlds] Pull request closed: #2712 CScrollRegion: m_ScrollY now ties to content height instead of slider height

10:35

[teeworlds:master] 6 new commits

de85395 Scrollregion: m_ScrollY now ties to content height - TsFreddie c876164 Remove excessive empty lines and a unused enum - TsFreddie e7886fc Merge branch 'master' into 0.7-scrollregion - TsFreddie 809d384 Optimize slider grab experience - TsFreddie 6b611b5 Merge branch '0.7-scrollregion' of github.com:T... - TsFreddie

I fixed the shadows in the grass tileset, there should be no cuts or at least not visible ones without zooming super in. I hopefully fixed the cuts you mentioned too @mind

11:14

Image attachment

11:15

https://i.imgur.com/yq8YV3L.png if u wanna test it out

Embedded image

11:15

before i PR to github

[teeworlds/teeworlds] Pull request opened: #2717 Smooth scrolling

closes #2713 I hope this make scrolling feels better.

@Sonix I marked the spots that have a missing line of shadow

Image attachment

11:43

And the grasstiles on jungle_main have cut-off shadows, too.

jungle and desert are way trickier and they will get updated in 0.8 so

‍♂️

EeveeShy

or sooner than 0.8 dunno

tried my best @mind, there are some that are still there but they're impossible to fix without reworking everything. The important thing is that there are no plain visible cuts ingame

12:19

https://i.imgur.com/dPq4Ray.png

Embedded image

2

[teeworlds/teeworlds] Pull request opened: #2718 Fix button and hovering logic, cleanup and refactoring

Change CUI::DoButtonLogic to only use left mouse button.
- I don't know why right mouse button was enabled for buttons, but it doesn't feel right and this would free up the right mouse button to be used for other things in the future, like contextual menus/infos.
Represent all the Checked-Values using bool instead of int.
Consolidate mouse button handling from Gameclient and Editor in CUI.
Make CButtonContainer inherit CUIElementBase and move and simplify the ButtonFade handling...

EeveeShy

oof conflicted a bit.

@mind @Sonix yep jungle has been completely redrawn and desert is fixed with parts of it redrawn and some extensions, currently I'm working on improving some small details on the grass main and layout standardisation for all mains in general + new themes including grass tiles for all biomes

Please preserve layouts, we're still dealing with the chaos of the 0.7 layout changes, that corrupted a lot of maps

is it possible to release the source files in order for other designers to use them for new tiles?

sorry but that will not happen Dune layout standardisation is a big thing that Landil wanted to see in 0.7 already but there was no reason to go for it as there were no new tilesets introduiced but for 0.8 it will happen and all tileset will have the same layout

The benefit is tiny compared to the absolute mess it creates

14:22

It was hard to foresee, but now we have observed it

the benefit will outweigh it in the long run think long term

14:25

it will be the last compatibility break from then on all tileset have the same layout allowing for easy tileset swap

the tileset change has create an ambiguity for all old maps

14:25

I have about 50k maps in my map archive

14:25

I don't know, maybe 1k of them use the jungle tileset

14:26

for each and every one of those, it's unclear which map tileset is right

14:26

if we change layout again, I think we should name the tileset differenlty

14:26

maybe jungle2_main

14:27

or somehow change the editor code in 0.8 that clearly marks the new jungle tileset in the editor, so that no ambiguity arises

I'd prefer the 2nd approach

start by versioning maps

1

wdym @jxsl13?

14:47

do you mean by teeworlds version?

14:48

what happens when you open a 0.7 map in the 0.8 editor and save

14:48

does it get marked as 0.7 or 0.8 map?

@Dune this is that original file Landil showed me years ago he wanted to adopt it's layout and that is gonna be the core layout for all tilesets in 0.8

Image attachment

I mean maps having a version like Teeworlds itself.

14:59

you do not open old maps in new editors, but it would allow to actually properly distinguish map versions.

14:59

and properly convert between formats/versions. (edited)

15:00

ofc new formats should not be released in between major versions of teeworlds (current opinion).

@jxsl13 so you'd artificially forbid opening old maps in new editors?

15:01

currently it's possible, I think, even opening 0.4 maps in the 0.7 editor (edited)

yes, I would just return an error that the map version is incompatible. at most maybe having a conversion tool (could be in the client itself, preferrably third party tools) the conversion tool, if part of the client, should at most support the previous map version.

15:04

in order not to bloat the codebase unnecessarily.

15:05

If the map format does not change, why forbid it, a teeworlds version jump would not enforce a map version jump. (edited)

15:06

a map version jump would,on the other hand, enforce a tw version jump.

15:07

as far as I understand the current situation, the maps can be opened, because there is no actual map versioning. they can even be opened, even tho the format or the default tilesets changed. (edited)

15:08

a change of the default tilesets might also enforce a map version jump.

I don't see how that map version thing particularly helps

15:09

grass maps from 0.6 and before are still fine in 0.7

15:09

the map version thing would break that

15:09

however, tile versions would be almost perfectly suited for this

converting between maps from 0.6 to 0.7 seemed to be problematic, or am I wrong here? a lot of maps look broken on 0.7 that were ported from 0.6? (edited)

yes, due to missing tileset versioning

15:12

the jungle tileset changed

hm, I wonder how you'd version tilesets?

15:13

might be the same as map version

https://github.com/teeworlds/teeworlds/issues/1147#issuecomment-687810682

Include a Image CRC for external images in the map files · Issue #1...

This would make it a lot easier to check whether map has 0.5 vs 0.6 vs 0.7 grass tiles and that stuff.

15:13

I just wrote down my thoughts there

hm, well the versioning of either default tilesets/maps might be some 1 to 6 bytes value in the map itself.

15:19

the assumption with pre-versioning maps seems fine.

15:20

crc value for external tilesets also seems good, but would one want to enforce that for default tilesets?

the grass doodads changed btw

@jxsl13 should fit into one integer, the versioning

people might wanna swap those, maybe there is not that much of a demand there, but maybe.

and i say it will be common to bring additions to doodads @heinrich5991 and reposition stuff

default tilesets that is

@Sonix that's what the proposal is for

15:21

fixing the effects of that in the code

15:21

every time you reorder stuff, you increase the tileset version in the code

I would call that map versioing instead of tileset versioning

1

(which is only possible between major versions anyway)

15:22

idc what you call it, I care about the effects

every time you change tilesets itself, you update the version.

that would make grass_main incompatible, even if it stays compatible

if the game gets new actual map entities, one might also increase the counter

[teeworlds/teeworlds] Issue closed: #2716 Various Out of Bounds GUI issues

15:48

[teeworlds/teeworlds] Pull request closed: #2718 Fix button and hovering logic, cleanup and refactoring

15:48

[teeworlds:master] 2 new commits

435fb05 fix button and hovering logic, cleanup and refa... - Robyt3 a88f769 Merge pull request #2718 from Robyt3/buttonLogi... - oy

this happened after Robyt's commits? (edited)

uh, it's from the windows artifact from the robyt's PR

[teeworlds/teeworlds] Issue opened: #2719 Raindrop mapres

Add a raindrop mapres similar to snow or stars that can be used to animate a map.

alright, it doesn't happen in my master with the PR merged in

thats weird

strange.. strange indeed

16:14

but considering it works in the proper build, it is no issue. might be the github actions doing some wonky stuff o.o

16:19

deleted my screenshots to avoid further confusion

[teeworlds/teeworlds] Issue closed: #2713 Add smooth scrolling to CScrollRegion

17:58

[teeworlds/teeworlds] Pull request closed: #2717 Smooth scrolling

17:58

[teeworlds:master] 6 new commits

c50795d Smooth scrolling - TsFreddie 8401446 No highlight in browser & listbox while scrolling - TsFreddie 161ff2f Fix highlight hiding when the list doesn't actu... - TsFreddie 0187e54 Style fixes - TsFreddie 2f730aa Merge branch 'master' of https://github.com/tee... - TsFreddie

oh. that was fast.

17:59

i keep making simple format mistake when i'm tired.

[teeworlds/teeworlds] Pull request opened: #2720 Use absolute instead of abs

bigflush

can i +1 for clang-format

18:01

or add .clang-format to .gitignore

18:03

lol, i checked my phone. apparently i typed this and falls asleep...

Image attachment

wow

wow

1

18:12

hard worker

nah. was about to take a nap. just tried to reply a message and got denied by the comfy bed.

1

y_okie

1

[teeworlds/teeworlds] Pull request closed: #2720 Use absolute instead of abs

19:11

[teeworlds:master] 2 new commits

fe5daf8 use absolute instead of abs - Robyt3 51ef8ca Merge pull request #2720 from Robyt3/absoluteIn... - oy

last few days have been killer in terms of the PRs

1

[teeworlds/teeworlds] Pull request opened: #2721 Minor improvements on GitHub workflow

Exported 139 message(s)