Kontrol S4 Mk3 ES6 mapping #4056

Be-ing · 2021-07-04T03:15:22Z

This mapping is not compatible with Mixxx 2.3. It requires the main branch to work at all and #2618 for full functionality. I recommend making use of Git worktree to use both branches simultaneously.

This is complete though the jog wheel handling could be improved.

Turn the left encoder to beatjump. Press and turn the left encoder to adjust the beatjump size. Turn the left encoder with shift can be used to nudge the track.
Turn the right encoder to adjust the loop size. Press the right encoder to (de)activate a loop. Press the right encoder with shift to reactive an existing loop if looping is disabled, or if looping is enabled, jump to the beginning of the loop and stop playback. Turn the right encoder with shift to beatjump by 1 regardless of the configured beatjump size.

Shift + HOTCUES and record pad mode buttons can be used for manually setting the loop in and out points.

The button with the play icon adjacent to the library encoders (top right of the decks) loads the selected track to the preview deck and plays it. While the preview deck is playing, the library encoder beatjumps forward/backward 16 beats in the preview deck. The star button acts like pressing shift + tab and the playlist icon acts like pressing tab. The view button maximizes the library.

The buttons beside the gain knobs store the current gain of the deck as the track's default gain (#4031).

The default pad mode is mapped with the top row controlling intro & outro cues and the bottom row controlling hotcues 1 - 4. Press the HOTCUES pad mode button for hotcues 5 - 12 and the button next to it with the recording symbol for hotcues 13 - 20. Press the button for the active pad mode to switch back to the default pad mode. Note that the record pad mode button's LED is only red and does not support other colors, so it will not always match the deck color.
The pads can be switched to control samplers by pressing the SAMPLES button and a second page of samplers is available by pressing the mute icon. Press an unlit pad to load the selected track in the library to that sampler. While a sampler is loaded, it will be dimly lit if it is paused and brightly lit if it is playing. Pressing a loaded sampler plays from the cue point. Pressing a sampler pad with shift pauses it if it is playing and ejects it if it is paused. The script assigns the samplers to the crossfader by default, which can be disabled by setting samplerCrossfaderAssign at the top of the script to false.
The STEMS pad mode button toggles keylock.

Effect chain preset selection (FX SELECT buttons) requires #2618 to work. To load presets, press FX SELECT 1-4 buttons. Each selector button is lit in the color representing the preset it will load. Pressing and releasing a selector button will load that preset to all decks unless one of the QuickEffect enable buttons is pressed before the selector button is released. In that case, the selected preset will only be loaded to the selected deck. 16 presets can be selected with combinations of selector buttons, which is very powerful. While holding a selector button, the other buttons will change colors indicating more presets that can be selected by pressing a second selector button before releasing the first. For example, to load preset 5, hold button 1 then press button 2.

The effects mapping is currently basic with the knobs controlling the metaknobs and the buttons controlling the enable buttons. The leftmost knob of each effect unit is the mix knob. There is not yet any effect focusing or control of individual parameters.

The deck colors are set to match my Chroma Caps. Feel free to change them to your liking by editing the deckColors array at the top of the script.

uklotzde · 2021-07-04T22:09:55Z

Btw, I also purchased an S4 MK3 some time ago but didn't find the time to test it yet 🙈 So count me in ;)

res/controllers/Traktor-Kontrol-S4-MK3.js

Be-ing · 2021-08-30T06:03:15Z

I have expanded the foundation of the Components library for HID and used it to map the play, cue, and deck toggle buttons of the left deck, now including LED output.

@Swiftb0y I know we had discussed wanting to decouple the Components library from I/O but this turned out not to be practical without ugly hacks. For HID we need to compare each field in an input packet to its state in the prior packet and only execute Components' input callbacks if their state has changed, similar to the common-hid-packet-parser.js library. Otherwise, each Component would need to track this state.

Without this, logic errors will occur excessively executing callbacks. For example, I ran into a logic error when holding the cue button for previewing, pressing the play button, then releasing the cue button. What should happen in this case is that the play Control gets set to 1, then cue_default gets set to 0 when the cue button is released. Without comparing input data to the old state, the input callback for the play button was inappropriately executed when releasing the cue button because they are in the same input packet, so immediately after cue_default was set to 0, play was toggled to 0 and the track stopped.

Be-ing · 2021-08-30T06:13:56Z

 warning [0x5580786fc000] ControllerScriptHandlerBase: "Uncaught exception: file:///home/runner/work/mixxx/mixxx/res/controllers/Traktor-Kontrol-S4-MK3.js:173: TypeError: Cannot call method 'trigger' of undefined\nBacktrace: outTrigger@file:///home/runner/work/mixxx/mixxx/res/controllers/Traktor-Kontrol-S4-MK3.js:173\nconstructor@file:///home/runner/work/mixxx/mixxx/res/controllers/Traktor-Kontrol-S4-MK3.js:377\n%entry@file:///home/runner/work/mixxx/mixxx/res/controllers/Traktor-Kontrol-S4-MK3.js:472"
/home/runner/work/mixxx/mixxx/src/test/controller_mapping_validation_test.cpp:122: Failure
Value of: testLoadMapping(mapping)
  Actual: false
Expected: true
Error while validating /home/runner/work/mixxx/mixxx/res/controllers/Traktor Kontrol S4 MK3.hid.xml

This test failure is an artifact of the test setup. There is no problem when running Mixxx.

Swiftb0y · 2021-08-30T12:19:56Z

For HID we need to compare each field in an input packet to its state in the prior packet and only execute Components' input callbacks if their state has changed, similar to the common-hid-packet-parser.js library. Otherwise, each Component would need to track this state.

Yes, that was my understanding as well. We already have a MidiMapper (as a module currently unused because its only exposed to the ModuleScriptingEngine) and we need a complementary HIDMapper as well.

I have never taken a look at our common-hid-packet-parser.js library myself so I can't really say much about it except that I heard others say its terrible...

Be-ing · 2021-08-30T17:03:08Z

I mapped the QuickEffect enable buttons and FX SELECT buttons which can be tested together with #2618. The QuickEffect enable buttons' colors represent the currently loaded preset. Press them toggle the enable switch of the QuickEffect chain. Long press and release to toggle the state again when releasing the button ("power window" behavior).

To load presets, press FX SELECT 1-4 buttons. Each selector button is lit in the color representing the preset it will load. Pressing and releasing a selector button will load that preset to all decks unless one of the QuickEffect enable buttons is pressed before the selector button is released. In that case, the selected preset will only be loaded to the selected deck. 16 presets can be selected with combinations of selector buttons, which is very powerful. While holding a selector button, the other buttons will change colors indicating more presets that can be selected by pressing a second selector button before releasing the first. For example, to load preset 5, hold button 1 then press button 2.

I was initially planning on including the FILTER button as a 5th preset selector button which would allow selecting up 25 different presets. However, the LEDs only support 16 different colors. Also, by design, there is no API to load a hardcoded effect. 16 presets is a lot to choose from, so I'm not really upset about this... but now I don't know what to do with this button. Maybe map it to the headphone split switch because it's right by the headphone mix and volume knobs? 🤷

Be-ing · 2021-08-31T02:37:53Z

Both decks and the mixer are now mostly mapped. The top row of the pad grid is mapped to intro & outro cues. The bottom row is hotcues 1 - 4 (with colors supported :) ).

Be-ing · 2021-08-31T05:49:08Z

This is a basically usable state now. It is already more useful than my Xone K2. I briefly attempted to get the jog wheels working but I have not figured out how to interpret the signals from the controller for them. As a quick, temporary hack I mapped the turntable and jog buttons to seek backward/forward (faster with shift) until the jog wheels are working. I took a look at the S3 mapping and the signals for the jog wheels seem similar but I don't think they're identical. The S3 has a timer indicated 3 bytes. The S4 Mk3 seems to have 4 bytes that continually vary for each wheel. @uklotzde could you test this and try to make sense of the jog wheel signals in input report 3?

Swiftb0y

I only bothered to look at the components code for now. I wouldn't even bother to fix most of the issues I pointed out for now since there are more broad issues I'd like to discuss first. For example, I still don't believe that coupling components to the actual protocol is a good idea, though this version of components is already a much better starting point than its midi counterpart for further decoupling efforts.

Swiftb0y · 2021-09-11T18:29:50Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+    constructor(reportId) {
+        this.reportId = reportId;


I dislike the assumption that the packet structure changes based on the reportID. For example, the pioneer protocol abuses the reportID just indicating which deck (or generally unit) on the controller receives the data.

I think Pioneer is breaking the HID standard with that, so I'm not really concerned about that.

Yeah... as I said, they don't care about that because they're actually not using anything actually related to the HID standard. They might as well not be using HID since their entire driver is just implemented as part of the DJ software.

But at the same time, they still have standard packets and parsing them using this library could be quite useful, so thats why I dislike the dependency on that. In the end, we write the same user-space device drivers as Pioneer

I think it would be up to the scripts for Pioneer devices that abuse the HID protocol to figure out how to identify packets some other way.

I don't think that's really possible.

I still believe it is, but I have yet to explore my ideas with some concrete experiments.

My rough idea is to create another abstraction that sits between IO and Components and irons out the differences in the protocol. This would also do the parsing. Components would then only used for modeling the actual controller and tracking the state of the individual components.

We already do this ironing a bit in the part that compares values between packets and then only call input handlers when the value actually changed. We just need to abstract this for ComponentsJS so the components don't have to care about how midi or HID works.

I'd gladly discuss this on Zulip some time. Though I'm aware that an actual PoC would make more sense, I don't think I could find the time for that currently.

I would need to see a proof of concept to understand how this would work.

I'll try to come up with something.

Swiftb0y · 2021-09-11T18:32:10Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+        if (bitOffset === undefined) {
+            bitOffset = 0;
+        }
+        if (typeof bitOffset !== "number" || bitOffset < 0 || !Number.isInteger(bitOffset)) {


Suggested change

if (typeof bitOffset !== "number" || bitOffset < 0 || !Number.isInteger(bitOffset)) {

if (!Number.isInteger(bitOffset) || bitOffset < 0) {

works just as good I think.

Swiftb0y · 2021-09-11T18:33:48Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+        if (bitLength === undefined) {
+            bitLength = 1;
+        }


JS supports default values, use that instead. (pretty sure that applies to ES7 as well)
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Functions/Default_parameters

Swiftb0y · 2021-09-11T18:37:03Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+                this.fields = this.fields.filter((element) => {
+                    return element !== field;
+                });


Suggested change

this.fields = this.fields.filter((element) => {

return element !== field;

});

this.fields = this.fields.filter((element) => element !== field);

also, capturing a reference to the field here could cause some memory leaks in the long term.

Capturing a reference to the field is required for this to work at all.

I know, thats why I'm saying we should consider finding an alternative solution.

I can't think of a better way. Do you have any suggestions?

Please remind me tomorrow.

On second thought, the field should get destroyed as soon as its removed from the fields array and the handler to the callback goes out-of-scope. So its probably not a memory-leak potential.

Swiftb0y · 2021-09-11T18:39:45Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+            return;
+        }
+
+        for (let field of this.fields) {


Suggested change

for (let field of this.fields) {

for (const field of this.fields) {

This is applicable everywhere generally. This would be enforced by the prefer-const eslint rule. We should probably update our ESlint rules for main anyways.

It's not always applicable, but I'll switch to it where possible.

Well at last when iterating over a container, I have not spotted any location in this code where a for (let [...] of/in [...] would be justified.

Swiftb0y · 2021-09-11T19:55:00Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+    output(value) {
+        if (value) {
+            this.send(this.color + this.brightnessOn);
+        } else {
+            this.send(0);
+        }
+    }
+    outConnect() {
+        if (undefined !== this.group) {
+            this.outConnections[0] = engine.makeConnection(this.group, this.outKey, this.output.bind(this));
+            this.outConnections[1] = engine.makeConnection(this.group, this.colorKey, (colorCode) => {
+                this.color = this.colorMap.getValueForNearestColor(colorCode);
+                this.output(engine.getValue(this.group, this.outKey));
+            });
+        }
+    }


this makes a lot of assumptions that need to be worked over.

Can you elaborate?

Well thats largely related to my previous comment that this only supports half the usecase that the ColoredHotcueButton was actually made for.

but a quick example that comes to mind is the this.color + this.brightnessOn part because thats device specific.
Also this.send(0); should rather be this.send(this.off);

Swiftb0y · 2021-09-11T19:56:45Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+    input(pressed) {
+        if (!this.shifted) {


this is very different to how we usually handle shift with other components.

Yes, that is intentional. The old way with MIDI Components of reassigning the input method IMO ended up being rather awkward and clumsy in many cases. I do not recall who it was, but I do remember a mapping developer giving feedback on the MIDI Components library and mentioning that approach being overcomplicated. With this system it would be even worse because the input callback would need to be deregistered from the HIDInputPacket and the new one would need to be registered every time shift is pressed.

Mhmm. I'll think about that. I find the overwriting input method a bit more elegant, but it also forces the JIT to deoptimize code so it might not be that good idea after all...

My apologies for jumping in here, I just read your discussion and would like to add my 2 cents.

Both of you already had a look into the custom library that I contributed with Behringer devices in mind. I chose a different approach for the shift logic there and want to share my thoughts; maybe some details may be applied to other mappings, too.

I decided against the idea of handling the different behaviors on shifted/unshifted state inside a JS Component. Instead, the two behaviors are split into two components which are registered/unregistered according to the state of the shift button, so they are unaware of shifting.

In my implementation, you'll find a shiftable Component called "layer manager" which holds 2 component containers (= layers) named Default and Shift. JS Components may be added to either layer. This layer manager is the only component that knows about shifting. When it is shifted or unshifted, all affected components on the corresponding layer are registered / unregistered. Components that are not shiftable stay untouched.

The benefit of this approach is a good separation of concerns for the cost of more frequent component registrations. In my mappings, the number of shiftable components is small (measured against the total component count), and thus the number of registration changes is low. This approach works fine for me: the components look uncluttered and I didn't notice any effect on the performance.

While that is another valid approach, that's not what I want to do. That would require considerably more boilerplate and complexity. Most Components have some alternative functionality with shift so that complexity would add up.

I consider the layered approach as the more versatile and cleaner architecture. But it may depend on the use case. Probably both in parallel are needed.

The layered approach also better resembles how Shift is actually implemented, i.e. it alters the feedback of all controls and not only a single one.

I think it does depend on the use case and it is good to support both approaches.

The layered approach described seems valid. The actual physical control surface is still a single physical component, so using multiple components seems a bit wrong. We should think about adjusting terminology or maybe introduce a new concept for this.
Using another container class seems like an interesting concept actually. If that container class is responsible about connecting components based on whether they're shifted or not, it could actually have customizable behavior depending on the controller. For example for taking advantage of the controllers firmware shift feature.

Swiftb0y · 2021-09-11T19:58:09Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+                this.send(this.color + this.brightnessOn);
+            } else {
+                this.send(this.color + this.brightnessOff);


I didn't even know the samplerbutton came with any color feature until reading this method...

Yeah, this is hardware specific.

Swiftb0y · 2021-09-11T20:03:09Z

res/controllers/Traktor-Kontrol-S4-MK3.js

+        }
+    }
+    outConnect() {
+        if (undefined !== this.group) {


Suggested change

if (undefined !== this.group) {

if (this.group !== undefined) {

res/controllers/Traktor-Kontrol-S4-MK3.js

wastez · 2022-02-09T00:48:04Z

Wow this is really Great.
I buyed a s4 mk3 and i'm so unhappy with it because the lock of 8% +- pitch in tt mode. (Which idiot Design some stupid thing? Yes i know they want a 1210 feeling but that is no reason to take the one Bad thing from them)
I buyed it because the tt Mode. (I'm comming from turntables)
Is there a working Version out there for mac?
Otherwise i will sell this fucking Controller after one day.
I would like it but this pitch is a no go.
Could you please help me?????

Be-ing · 2022-02-09T00:51:47Z

I haven't tested this myself on macOS but AFAIK there's no reason it shouldn't work. The arbitrary +/- 8% limitation in Traktor also really irritated me, among other strange behavior of the controller mapping with Traktor. You can set the tempo fader range in Mixxx's Deck preferences.

Be-ing · 2022-02-09T00:52:51Z

I buyed it because the tt Mode. (I'm comming from turntables)

I kinda got the motorized wheels working but there are lot of rough edges with that. I'd be grateful if someone picked this up to improve that!

wastez · 2022-02-09T01:24:03Z

So just compile your Version for mac should work?
I'm so angry about this controller , i cant describe my frustration.
Got it today and now i'm think about to sell it immedatly.
Would be really cool of your solution could help me.

Be-ing · 2022-02-09T01:32:07Z

You don't need to compile Mixxx. You can use a development build and download the mapping files from this branch:
https://raw.githubusercontent.com/Be-ing/mixxx/kontrol_s4_mk3/res/controllers/Traktor-Kontrol-S4-MK3.js
https://raw.githubusercontent.com/Be-ing/mixxx/kontrol_s4_mk3/res/controllers/Traktor%20Kontrol%20S4%20MK3.hid.xml

and put them in ~/Library/Containers/org.mixxx.mixxx/Data/Library/Application Support/Mixxx/controllers

ex0nuss · 2022-11-16T20:13:47Z

Hi there,
sorry for reopening this pull again. I am currently looking for an Upgrade to my DDJ-400 and the S4 MK3 looks pretty neat. With this Upgrade I also wanna switch to Mixx :)
What is the state of this mapping and does it work with the main branch?

Cherrs!

daschuer · 2022-11-16T22:27:34Z

I think this is in a "working" state and it looks like the prerequisites are merged to main. However this PR needs some work and testing to be merge-able to the main branch. Do you have interest to continue this work? You need the controller and some basic Java script skills. Your help is welcome.

Serphentas · 2023-05-25T18:13:32Z

Coming in late but I also acquired the controller and have skills and time to spare. I'll check out what's been done to date and expand on it. Any pointers, feel free to drop them here.

Many thanks for what's been done until now !

JoergAtGithub · 2023-05-25T18:30:10Z

Have a look at #11284 which requires #11326 to work.

Serphentas · 2023-06-02T21:43:43Z

I've compiled Mixxx from the 2.4-beta tag successfully, but upon enabling the mapping the console shows:

warning [Controller] Failed. Trying to open with make, model & serial no: 6092 5920 "3A47C8C5"
warning [Controller] Unable to open specific HID device "Traktor Kontrol S4 MK3 C8C5_3" Trying now with just make and model. (This may only open the first of multiple identical devices.)
warning [Controller] Unable to open HID device "Traktor Kontrol S4 MK3 C8C5_3"

Here is lsusb with the Controller setting view:

For reference, loading any mapping works in Mixxx 2.3.5 (Arch).

JoergAtGithub · 2023-06-02T21:48:53Z

This PR is no longer maintained. Please try the latest version of the mapping in #11284
Regarding the problem - you might change your udev file on Linux, because 2.4 use the hidraw kernel module instead of the libusb userspace driver. Full installation of 2.4 should fix this.

Serphentas · 2023-06-02T22:43:26Z

You're right, sorry about that. I've used an appropriate udev rule and no read error anymore. Thanks !

Be-ing marked this pull request as draft July 4, 2021 03:15

github-actions bot added the controller mappings label Jul 4, 2021

Be-ing commented Jul 4, 2021

View reviewed changes

res/controllers/Traktor-Kontrol-S4-MK3.js Outdated Show resolved Hide resolved

Be-ing commented Jul 4, 2021

View reviewed changes

res/controllers/Traktor-Kontrol-S4-MK3.js Show resolved Hide resolved

Be-ing requested a review from Swiftb0y July 5, 2021 02:21

Be-ing force-pushed the kontrol_s4_mk3 branch 3 times, most recently from bec512b to 9235022 Compare July 5, 2021 02:24

Be-ing mentioned this pull request Jul 5, 2021

Mapping for Traktor Kontrol S4 MK3 #3900

Closed

Swiftb0y reviewed Jul 5, 2021

View reviewed changes

res/controllers/Traktor-Kontrol-S4-MK3.js Outdated Show resolved Hide resolved

res/controllers/Traktor-Kontrol-S4-MK3.js Outdated Show resolved Hide resolved

res/controllers/Traktor-Kontrol-S4-MK3.js Show resolved Hide resolved

Be-ing requested a review from Swiftb0y August 30, 2021 05:43

Be-ing force-pushed the kontrol_s4_mk3 branch from f79a1e5 to bafa042 Compare August 30, 2021 05:45

Be-ing force-pushed the kontrol_s4_mk3 branch from 61813a0 to f712759 Compare August 30, 2021 07:44

Be-ing force-pushed the kontrol_s4_mk3 branch 6 times, most recently from 7606bde to 0698ec8 Compare August 31, 2021 02:36

Be-ing force-pushed the kontrol_s4_mk3 branch from 0698ec8 to b6a5e84 Compare August 31, 2021 04:51

Be-ing force-pushed the kontrol_s4_mk3 branch from b6a5e84 to 526e255 Compare August 31, 2021 06:10

Be-ing requested a review from uklotzde August 31, 2021 06:13

Be-ing force-pushed the kontrol_s4_mk3 branch from 526e255 to c8c0060 Compare August 31, 2021 06:18

Be-ing force-pushed the kontrol_s4_mk3 branch 2 times, most recently from e916bfe to 3b6c856 Compare September 11, 2021 00:11

Swiftb0y requested changes Sep 11, 2021

View reviewed changes

Be-ing force-pushed the kontrol_s4_mk3 branch 3 times, most recently from 555e09a to a0f03c1 Compare September 26, 2021 17:39

Be-ing force-pushed the kontrol_s4_mk3 branch from a0f03c1 to e4e5d1c Compare October 3, 2021 20:56

Be-ing force-pushed the kontrol_s4_mk3 branch from e4e5d1c to 70ea195 Compare October 12, 2021 19:27

Be-ing mentioned this pull request Oct 14, 2021

create [Master], maximize_library in C++ #4424

Closed

Be-ing force-pushed the kontrol_s4_mk3 branch from 70ea195 to f9758de Compare October 26, 2021 01:51

Be-ing force-pushed the kontrol_s4_mk3 branch from f9758de to e802327 Compare November 5, 2021 20:11

Be-ing force-pushed the kontrol_s4_mk3 branch from e802327 to ab636db Compare November 14, 2021 01:17

Be-ing mentioned this pull request Nov 14, 2021

HidController: change getInputReport to automatically call script #4520

Closed

Be-ing force-pushed the kontrol_s4_mk3 branch from ab636db to 2271ab1 Compare November 16, 2021 01:33

Kontrol S4 Mk3: initial commit

a3fd96d

Be-ing force-pushed the kontrol_s4_mk3 branch from 2271ab1 to a3fd96d Compare November 22, 2021 14:54

Be-ing closed this Nov 30, 2021

rah2501 mentioned this pull request Oct 3, 2022

Kontrol S4 Mk3 mapping #10937

Closed

	if (typeof bitOffset !== "number" \|\| bitOffset < 0 \|\| !Number.isInteger(bitOffset)) {
	if (!Number.isInteger(bitOffset) \|\| bitOffset < 0) {

	for (let field of this.fields) {
	for (const field of this.fields) {

	if (undefined !== this.group) {
	if (this.group !== undefined) {

Kontrol S4 Mk3 ES6 mapping #4056

Kontrol S4 Mk3 ES6 mapping #4056

Conversation

Be-ing commented Jul 4, 2021 • edited Loading

uklotzde commented Jul 4, 2021

Be-ing commented Aug 30, 2021

Be-ing commented Aug 30, 2021

Swiftb0y commented Aug 30, 2021

Be-ing commented Aug 30, 2021 • edited Loading

Be-ing commented Aug 31, 2021 • edited Loading

Be-ing commented Aug 31, 2021

Swiftb0y left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wastez commented Feb 9, 2022 • edited Loading

Be-ing commented Feb 9, 2022

Be-ing commented Feb 9, 2022

wastez commented Feb 9, 2022 • edited Loading

Be-ing commented Feb 9, 2022

ex0nuss commented Nov 16, 2022

daschuer commented Nov 16, 2022

Serphentas commented May 25, 2023

JoergAtGithub commented May 25, 2023

Serphentas commented Jun 2, 2023

JoergAtGithub commented Jun 2, 2023

Serphentas commented Jun 2, 2023

Be-ing commented Jul 4, 2021 •

edited

Loading

Be-ing commented Aug 30, 2021 •

edited

Loading

Be-ing commented Aug 31, 2021 •

edited

Loading

wastez commented Feb 9, 2022 •

edited

Loading

wastez commented Feb 9, 2022 •

edited

Loading