A few months have passed since the last report, which means it’s time to sit down and write a new one ;)

My arm has already healed, the cast has been removed, now I wear an orthosis and do exercises. Yes, it still hurts, it’s hard to lean on it or lift weights, but I can already hold the mouse and stylus as before.

To be honest, after that fracture it turned out that there were also problems with my left arm. I pulled some tendons and muscles, so I couldn’t lift my arm above shoulder level. So basically, I injured both my hands🥲 But I went to physical therapy, did exercises, and now I can move it more or less normally. Soon I’ll need to return to the gym and start working out my muscles with a trainer.

And now about the main thing // what about the progress?
Due to problems with my hand, it was difficult for me to hold the mouse and make scenes during the first weeks, so I decided not to waste time and work on the code, since my left hand could work with the keyboard... And to some extent I’m glad that I did, because the progress is simply incredible.

In-game phone // and all related mechanics
How do you remember the old phone from OBD? And from other games?
Something like this probably comes to your mind:

Like a carbon copy, right? What if I say that the updated phone in OBD is something that you’ve never seen before on Ren’Py? Will you believe me?

Okay, I won’t delay, because I recorded a whole presentation for you:
https://youtu.be/UW1I0FTH_AE

I'll pretend that you watched the video 😅. And I’ll continue with standard questions that you might have.

Why does the game need a phone at all?
Initially, when I first started making the game, I followed other developers. I copied their mechanics, took ideas and tweaked the code. At some point I decided to add a phone, although there was no need for it. Now this mechanic no longer feels unnecessary, and on the contrary, it is necessary and, in my opinion, adds more variety to the game.

Why did you decide to rework your old phone?
In short, not long ago I decided to play BaDIK for the first time (I played very few AVNs), and I was confused that the phone interface in all games was somehow the same, bland and inconvenient.

What did the old versions of the phones look like?
Differently. I’ll start with the very first version, when I didn’t have any skills yet... Yeah, I make backups and store all the old versions of the game on hard drives. As a keepsake.

I was just getting used to the engine, trying to look for ready-made code on the Internet, racking my brains over why everything wasn’t working...

From time to time, I took a break from rendering and occupied myself with the phone, doing something in it, experimenting...

It even seemed to me that I was succeeding: the music player was playing, the buttons were working... At some point there was an idea to add an investment mechanic to the game.

Here I already began to think about stickers and the beauty of the interface. Now I can’t actually call it beauty, but then I wasn’t capable of more. By the way, this version appeared in OBD.

Aaand, I think this is how you remember this phone now. Even like this it looks quite nice... But the time has come to bury it.
For those who for some reason didn't watch my video 😤, I'm posting screenshots of the final version of the phone! 

Well, that's it. This version will remain in the game until the end. I’m very pleased with the result, and I almost don’t regret the time spent.

The experience gained from older versions was very useful to me. All the skills I learned and a broken arm allowed me to make a very complex, but at the same time structured and stable system. Each application is a separate module, each account is part of the social network, and each device in the game is independent. Technically, the game has the ability to open absolutely any phone and show what’s in it! But I'm unlikely to reveal the secret of how to do this >:)


How was all this done?
Now, it’s time to tell you how this phone was designed, and what problems and limitations I encountered.

Well... I started by designing it in Photoshop, and it only took two days. At first, I just wanted to remove everything unnecessary from the game interface so that there were no stupid, inconvenient buttons, and when I started working on the phone itself, my friend Gallant Trombe came to my aid. He suggested one idea: remove all the frames and leave only the working screen... Damn, why not?

From that moment on, absolutely everything changed! If earlier the procedure was clear: you make a phone frame and fit pictures in it, as it’s done in all the other games, but now there is enormous freedom for improvisation.

I wrote the shaders that create rounded edges a long time ago, and if I didn’t know how to write them, I would never have achieved such a result.

During my work, I encountered a huge number of problems. How to open and close apps? What to do with screens that overlap each other? How to remove these stupid pages and make news feeds and messages endless? AND HOW THE DAMN SHOULD I OPTIMIZE ALL OF THIS SO AS NOT TO BURN YOUR DEVICES?

Ren’Py can’t render a huge number of objects at once, and it doesn’t have optimization tools, so I had to write them myself.

And now, here are a lot of boring text...

When an application opens, the screen behind it automatically disappears and appears back when needed. Separate optimizers were written for the news feed and messages. They calculate the size of the panels, the position of the current window, and based on this they give the Ren’Py command to hide or render the object. This made it possible to get rid of the "pages" and make the FPS as stable as possible. Yes, there will be drawdowns on weak phones, but for this there is a “Graphics” item in the game settings.

The next problem was how Ren’Py renders. Many things affect how an object is rendered, which can cause text to appear blurry. I had to experiment and find out experimentally which tricks would lead to the result I wanted. Well, I also had to switch from WebP images to SVG icons so that they would be displayed clearly on all monitors.

After solving these problems, I encountered another one - animated pictures. This is the biggest problem of the engine, since it is generally designed for static images, so animation for it is like a side quest. I had to create my own classes of audio channels that would simplify working with audio (this is directly related to animation playback), and also create a class that would store the animation itself, prepare the necessary audio channels for playing, or simply return a static picture, if there was one necessity. This way it turned out that any objects from the gallery can become wallpaper on your phone.

There was also a need to create a database of all social network content. Information about songwriters, tracks, pictures, stickers, emojis and accounts is now obtained by ID, rather than written down directly. Thanks to this, it will be much more difficult for me to break something on my phone in future updates.

End of boring text.


Small but very important bugfix

While working on the phone, I accidentally discovered the reason why the game was not resetting variables when starting a new game and when loading saves, necessitating a full restart. The nature of this was not known to me from the very release... I had to abandon the way I worked with variables before and switch to more “advanced” code. Well, now everything works perfectly :3

iOS support // let’s play on all platforms!

No, this is not an AppStore release. But it will be possible to play the game on an iPhone with a fairly stable FPS! Instructions for launching, setting up and everything else will be available at release and in the game itself :)

I tried very hard to optimize the mechanics for weaker devices and tried to take into account the features of playing on touch screens to make the design more convenient and universal. And if something slows down, you can always lower the game’s "graphics" level in the settings. This will reduce the detail of some animations, but make the game smoother.


So, what now?
Now that I'm done with the phone, I can go back to my regular life as a content creator. Yes, I planned to release it at the end of summer, but, as you can see, force majeure happened and I had to visit many doctors... I would even say that it was not a very pleasant incident, but it prompted very pleasant changes.

As usual, I’m not on radio silence and I’m always here or on Discord. The release is just a stone's throw away.


And many thanks to those who believe in this project and support it! Without your help the game would not be what it is now ❤️

Good day, dear readers of the journal “Notes of the Unpunctual keyclap Developer”. Today we're gonna talk about why the hell has there not been an update for so long. Also, just what the hell the dev has been doing.

Four months have passed since the fateful decision was made to leave Daz and switch to Blender. And now it’s time to take stock of what has changed since. I will compare how it was and how it is now. I’ll say right away that this decision is the second scariest and best decision in my life after leaving university ;)

Phew. So, let's start.

Scene storing // let’s make Blender an AVN industry standard!

Daz: each frame is saved into a separate file weighing 50-200MB.
Blender: all frames (usually about 150) are saved into one file as an animation. 

It would seem that this is not so important and that the only advantage of the new method is less occupied disk space. But how many games have you seen where the new frame is not just the same scene from a different angle, but a full-fledged continuation of the previous frame? The old way of working didn't give me much control over the narrative, and if I needed to change something, I had to open files that Daz opened for 30-60 seconds and edit, being careful not to accidentally move the camera. Because of this, AVN games are just a bunch of incoherent pictures. The old Daz method is simply tedious and slow, it needs to be banned at the legislative level! ;) 

In short, the new method should liken AVN to watching movies.

Character Posing // down with FK, glory with IK!

Daz: no controllers at all, each bone is rotated individually.
Blender: controllers that move several bones and some Shape Keys at once. 

Eight thousand lines of written code in 0.5 seconds create a full-fledged rig that allows me to do absolutely everything I need. This not only speeds up the posing of characters, but also greatly reduces the time needed to create new ones. Much has already been said about this in previous reports, so I won’t focus on it.



Fixing Meshes hammers and screwdrivers in the 3D world

Daz: a mesh was sent to Blender through my old script, fixed and sent back to Daz.
Blender: a Shape Key “layer” is created, on which the mesh is being adjusted, and its value is animated to the desired frames.

It may seem that this is all clear and everyone has it. But I assure you that 90% of the developers on Daz are afraid of this problem like hell, which is why you will hardly see characters who wear anything other than regular clothes:

The Rendering Process // happiness is where everything works as it should

Daz: each frame is manually rendered immediately after the scene is saved.
Blender: preview renders are created, scenes are adjusted based on them, and a few days later the final renders are made at night.

In order to not waste time, I quickly wrote a program that I called “kQueue” in a day, which puts projects into a rendering queue. Yes, there are already similar programs on the Internet, but hey, they either don’t work as I need or have critical bugs.

What is the advantage of the new method? It's simple. The quality of each render is now higher due to the fact that more time is spent on rendering. Also, in the future this will all develop into a render farm >:D

Post-Processing // away with the stash of thousands of thousands of files!

Daz: using Photoshop and manual saving.
Blender: using Blender's built-in compositor and automatic converter.

With the transition to Blender, the dependence on Photoshop disappeared almost completely. All work with color and effects have moved to the built-in compositor in Blender, which works with 16bit images and simply buries Daz with its 8bit. Why is it important? Because before, in order to perform color correction in Photoshop, I had to make sure that color banding didn’t appear in the gradients. Additionally, in order for the animations to be the same color as regular renders, all animation frames had to run through Photoshop, color corrected and saved, which is actually very time-consuming and stupid.

Moreover, in order to resave images in 1080p, 2K and 4K, I had to open Photoshop and do everything manually. The new method assumes that the raw renders are stored in .png and have a 4K resolution. My converter itself takes them, changes the resolution, renames them to a unique name and very quickly saves them with the desired size:

One of the critical advantages is complete control over the size of the game, which puts the optimization head and shoulders above other games that weigh 15GB or more ;)


Our Bright Days // finally time to talk about the most important thing

Well, it's time to talk about the game itself, and since a lot has already changed, why not change something in the plot? And the most important change is that the events are moved to a new place - to the foot of the mountains of the Krasnodar Krai ;)

But no need to worry, Suzy and Lucy will sill work together in a cafe, Mariya will go to university, and Ekaterina will work in the office, but all this will be in the new small fictional city.

But... Why was the decision made to leave Clairton, California? The first reason the survey, in which the vast majority of people were not from the USA:

The second reason is the changing climate, which is more understandable to me. Wouldn't it be cool to see winter scenes in the game?

And one more thing, after listening to an interview with Texic (Russian developer of the game Milf's Plaza) and hearing that he also tried to make his game for an American audience, but something didn’t go according to plan, I decided that it would be right to make my game about what I’m familiar with, rather than trying to mimic a foreign culture.

Some new renders:

Plans for the future // <UwU>

I wanted to do this from the very beginning, but due to the slowness of Daz, the idea was put on hold... until recently. I'm talking about Motion Capture and transferring captured animation to kRig. In fact, half the work is already done, and I was able to find a way to transfer animation to controllers! After this transfer, all Blender functionality for working with animations becomes available.

I recorded this video to show how it all works:
https://www.youtube.com/watch?v=CviesmZfNnU

Regarding the animation capture itself, I have a couple of ideas. I can buy several cheap webcams, “calibrate” them and run them through AI (but not the one on the Internet by subscription, but the one on GitHub). Or I can save up for a Rokoko Mocap suit, but it will take a lot of money ;0

What is all this for? For 18+ scenes? Nope, I have an idea to use animations of dances, movements and generally everything else for the in-game social network, game trailer and cinematics. I don't think many people have done this with their AVN games ;)


Finish!

As you can see, the game is in active development, progress is progressing by leaps and bounds, and in technical terms, keyclap has already overtaken the vast majority of its competitors >:) 

Report #13 🔥 Super Duper Important

Hold on to your seats, this report is gonna be simply bombastic. There has been a ton of changes over these two months. I'll start with the little things and end with the most important news that EVERYONE SHOULD READ! Thank you :)


Discord bots and kitchen hosting

Yeah-yeah, I know, they’ve been around the server for a long time performing simple tasks: monitor spammers, welcome newcomers, delivering hearts, etc. But the problem was that these bots lived on the working computer, which is why they could not always do their job and protect the server. So, that’s why I decided to order a Raspberry Pi 4 and move the bots to it...

And they were moved. To the kitchen. On top of the fridge. On a napkin. Because it makes noise like hell!

But it works and can even send messages :)

Someday they will talk...


Scene Optimizer my lifeline for developers

The moment I started working on a huge beach scene that had to have a lot of details and characters, I got stuck because the people selling assets at Daz Store had no idea what optimization was. Spending a day or two cleaning the stage is just dumb monkey work. So, I spent whole two weeks on a script that does that itself. Genius. Simply genius. But it does it much better and faster than me doing it by hand :}

The official Daz store already sells a well-known Scene Optimizer among developers for $25, but this garbage takes 30 damn minutes to do its job!!! And does not have any usable logic to its features. How does that compare to my optimizer? Well, it is written in two languages ​​(Daz Script 2 and Python). It is based on the LOD system, which is used in game engines like UE or Unity. It works 50 times faster thanks to multi-threading, changes texture resolution depending on distance, removes garbage and does much more. This allows Devs to render huge locations with limited video card memory.

Guess which picture has the original scene and which one has the optimized one? The original one takes up 21GB of memory, and the optimized one takes up 1GB.

And here I am looking at this thing and thinking... I can sell this to other developers so that they can make renders in large locations, not in toilet stalls :} So, I sent an application to Daz to become a Daz Published Artist and publish the script! Hehe! I've been waiting a month for them to check everything and still waiting… Explains a lot about why Daz Studio is so slow.

Just for fun, I made this video demonstrating the absolute advantage of this tool has over the garbage that is sold on the Daz Store >:D Watch with sound!

https://youtu.be/EXDaagYtTiE?si=K2ZR1BksgPqoTUpF

By the way, you may have heard from other developers that their video cards do not have enough memory, and that they have to buy Quadro A6000, which cost almost 5K dollars. So, now you can tell them that they are wasting their money :)


kRigger strategic resource of keyclap

On a previous report, I discussed developing my own rig, which will be used by all characters when posing and creating animations. Well, it's finally finished. Not completely, of course, but I can already work with it. In total, two months were spent on thi− damn, it took a long time... But wait! Watch the video first!

https://youtu.be/L_g2H8_ho-Q?si=ND11GMIvtyk5MB9u

I had to go through rigging course and at the same time transfer all I've leant into code,  and ask for advice from real animators in order to make the most professional rig possible. After all, the quality of the controllers will determine how believable the characters’ emotions will be. All this, of course, is done only for working in Daz. But a month ago a thought came to my mind, which made me sit down and think seriously...


Blender waste of time, a disaster or progress?

Now we've reached the craziest part.

As soon as I somehow thought that I was just wasting time in Daz, I got sad... After all, it’s true that all I’ve been doing for the last year is fighting with the program, and not creating content. Any task in it took an unforgivably long time, and simple animation took the whole day. A specific question arose. Should I stay where I am, create content, earn money and pay off debts? Or, should I switch from smearing charcoal on the wall to painting with brushes and canvases?

It was scary, very scary. I felt like I was about to betray the program I've worked with for so many years. What if nothing works out and I just lost a month? I wanted to switch to Blender only after I finished this game, as all normal developers do. But my desire not to sit still made me try it. There was only one task - transfer one character and adjust the skin shader so that it looks more or less the same. I tried to do this before, but nothing worked. Do you think it worked this time? Of course not. I didn't even take a screenshot because it looked terrible. And I went to bed angry, thinking that I don’t need all these challenges, and that I have to continue to work at Daz, where everything works out for me the first time.

But the process has already begun.

The next day, the first thing I did was not launch Daz, but Blender. I put my anxiety aside and I began to try to do at least something in it. As soon as I managed to create something more or less beautiful, I came to the conclusion that it was time to pack up my stuff and move onto Blender.

For the first two hellish weeks, I kept repeating to myself that the render engine is just a tool. That it really all depends on the person. I had to re-learn the Blender tools and research the information I need... Good luck doing that with Daz xD

I set myself a deadline of 1 month. During this time, I had to transfer the main characters, their clothes, set up shaders (and without ready-made locations this is difficult to do), make sure that the Shape Keys are transferred from the characters to the clothes and the character’s skin does not pass through the clothes. And there are also locations, kRigger’s unfinished facial rig, the vital Scene Optimizer and other scripts needed for work.

Guess what? I managed to do almost everything in time ;)

Before this I only made test renders to check individual details, but today I made three full renders in Blender for the first time.

With the new program came the opportunity to come up with a radically new way of working. In Daz, I did this: put the characters in poses (15-20 minutes), saved the project (30 seconds, yeah-yeah, slooow), rendered the frame (2-5 minutes), postprocess and edit in Photoshop (1-3 minutes).

But now everything will be completely different, and this is how I see the new process: put the characters into poses using the power of Blender, a graphics tablet and a generated rig by kRigger (5 minutes), press the “freeze scene” button and go to the next frame, as if the whole project is animation. This will make it possible to move through frames and easily control absolutely everything at any time.

Blender is fast, allowing you to open and close projects in seconds. The rendering process itself will now take place separately at night, which will increase the clarity of the renders by a factor of two.

Well, was it worth moving? :>

It's been a week since I released the initial version of "Our Bright Days". I received tons of good and bad reviews from people and I'm really happy to see all the support you've shown me! I spent few days on reading all the comments and I took into account all the constructive feedback. I also found some really good people who are ready to help me with the game and my bad Engrish. So, thank you all!

Let's talk about the future release that is going to come up within a few days. Please, don't expect much, as the coming release will be all about proofreading and making the game more friendly.

Release TO-DO (0.1.1):
1. Disable rollback blocking for the timed choices. Block rollback only after the warning message.
2. Add some more instructions about the game GUI.
3. Add a choice to either use the default Ren'Py navigation for dialogues or the custom one.
4. Add skip button and rework auto-forward and hide interface buttons.
5. Make the rollback function save last 100 steps.
6. Make it more ADHD friendly (reduce the number of transitions used in the game).