OPEN FEED

OPEN FEED is a first-person horror game built around public camera feeds, ordinary late-night locations, and the feeling that watching is no longer a passive act. The player moves from a supermarket to a night drive, then into a domestic space and finally to a monitor, where the in-game openfeed website becomes part of the story.

This document is a process diary. The entries describe what I made during development, what problems came up, how I solved them, and what still needed to be done in the next stage.

OPEN FEED started from a fairly simple feeling that is difficult to describe precisely. It is the feeling you can get when browsing the internet late at night and finding an old website, a public camera feed, or a strangely empty place. Those places are not directly frightening, but they feel too real and somehow unsafe at the same time.

I have always been interested in the quieter side of the internet. Not so much social media, but forgotten pages, public systems, and anonymous places where people probably no longer think about the fact that someone might be watching. That is where the first idea for OPEN FEED came from: what if the player finds something they were never supposed to see?

The game began more from atmosphere than from mechanics. I wanted to make an experience where fear does not immediately come from a monster or a jumpscare, but from uncertainty. For example, the moment when you are alone at night looking at an old camera image and nothing is really happening, but you still feel like you should close the monitor. OPEN FEED is not meant to scare the player constantly. I want to make a game that lingers a little after it is over, where the player still thinks back to an empty camera frame or a broken signal.

In the earlier pre-production notes, the core themes were already clear: surveillance, paranoia, loneliness, and the strange moral discomfort of seeing something that is technically public but still feels private. One idea that stayed with me from the beginning was that if something is visible online, that does not automatically mean it should be watched. That tension between curiosity and guilt became one of the most important parts of the concept.

Visually, I am inspired by PS1-era low-poly games, old CRT monitors, analog noise, VHS artifacts, 1990s internet design, amateur television signals, broken UI elements, and surveillance aesthetics. I like the idea that the whole game feels like a found system that should not have been opened.

In the core concept, the player sits at a computer and browses public or poorly protected camera feeds. At first everything is fairly ordinary: stores, parking lots, warehouses, apartment buildings, and empty streets at night. The more the player watches, the more the feeds begin to connect. The same people, symbols, numbers, places, signal errors, and strange interruptions can start repeating across the cameras.

The intended player experience is not fast action but a shorter atmospheric horror experience built around observation, interpretation, and escalating unease. The player mostly acts as a distant observer, noticing small changes, connecting clues, listening to radio frequencies and warnings, and deciding whether what they are seeing should be ignored, investigated, or reported. That distance is important, because the fear comes partly from not being able to step directly into the scene and fix what is happening.

Even at the concept stage, the structure already leaned toward two repeating moods: investigation through feeds and quieter reflection through movement and sound. The player watches camera feeds, notices anomalies, and follows strange details, then moves through more personal spaces where radio, ambient sound, and silence reshape the same tension in a different form. That rhythm helps keep the game from becoming one-note and makes the later, more personal scenes feel stronger.

I continued working on the later story moment of the openfeed website. After the NO SIGNAL part of the hallway archive feed, I added an anonymous messenger window that replaces the earlier phone call idea. From there, the player receives a link to a Westfield Herald article, which now opens in the shell browser as a separate tab next to openfeed.icu. After the article is closed, the conversation continues with an interruption: the user warns the player, the connection cuts off, and after a while the openfeed page starts glitching.

This required solving several technical problems, because Unity WebView and the file:// environment behaved more strictly than a normal browser because of sandboxing. Reading parent-frame functions directly could cause a SecurityError, so I made the calls safer and made the glitch start inside the page itself first. I also locked the shell browser after the offline state: refresh, back, forward, home, and the address bar can no longer bring the page back. This helps the ending feel like the system has really collapsed, not just temporarily frozen.

By the end of the evening, I focused on connecting the final part of the house scene more clearly. The TV's EAS message is no longer accessible from the beginning. It only unlocks after MonitorSite has fully fallen offline after the glitch sequence. After that, the television gives the player a new signal: a blinking red light and a beeping sound that draw attention to the next step.

I also improved smaller interactions around the desk. When holding the Nokia phone, the screen is now easier to read and the phone sits better in the camera. If the player cuts off the relay call too early, the phone still shows the forum.html hint so the game flow does not accidentally get stuck. I also configured the Still.mp3 audio so it starts at the beginning of the house, but fades and stops when the monitor is being used, so the website moments do not compete with the background music.

Finally, I fixed the energy drink bug where after several sips the can model could visually become flat. The problem came from the empty-can state changing the scale too aggressively. Now only the material and color change, while the model keeps its shape. It is a small fix, but details like that help keep the world believable and prevent funny technical bugs from breaking the atmosphere.

The final part of the house became much more playable in this stage. The original idea was that after the TV's EAS message, the player would simply go to the car and start the escape. Instead, I added a new moment between those beats: the TV tells the player to leave, but before reaching the car, the player sees the creature moving past the driveway-side window. From that moment a countdown starts and the player has to quickly find a place to hide.

At first the text was very direct and told the player to go under the bed. Later I made it more general: FIND A PLACE TO HIDE. NOW. This fits better because the bed is currently the first working hiding place, but other options could be added later, such as the shower or under the desk. Going under the bed currently happens through a short fade-to-black cut. I did not make a separate crawling animation because the goal of the moment is a fast panic reaction rather than a long staged animation.

When the player is under the bed, the camera is low and shakes slightly to create the feeling that the character is trying to stay still. The creature enters the house through the door, follows the recorded housepath route, checks the rooms, and finally reaches the area near the bedroom. I used paths recorded in play mode so I could walk the route myself and then have the creature follow it later.

The technical side of the creature needed many fixes. It now uses a textured biped model and a limping walk animation, but getting that to work required solving scaling, materials, animation looping, and movement speed problems. I also added spatial footsteps, a door creak, and whispers that come from the creature's position.

If the player does not hide in time, a separate failure path triggers. The player sees the warning, sees the creature, has to hide quickly, and only after that can run to the car again. This gives the ending a stronger rhythm and connects the EAS message, the creature, and the house space into one clearer sequence.

I worked on supermarket polish and the main menu's black-screen behavior. I cleaned up the shelf interactions: after three products have been placed in the cart, the system no longer allows the player to add more items. I removed Space to Skip, so the black screen now has to wait until the time limit or flow itself ends. This creates a more logical flow for the player and avoids the chance of accidentally skipping important text.

May 19 was the final day of development, so I spent it tightening OPEN FEED's ending flow and fixing the last issues that could break the experience. In the supermarket, I slowed down the intro text so it is easier to read, restored the missing "walk in when you're ready" prompt, made the cart objective clearer, and expanded the checkout interactions so milk cartons, hockey puck variants, and the dollar bill payment all behave as part of the same cashier sequence. I changed the cashier line from "five twenty" to "five dollars please," and added the dollar bill as a world-space 2D object on the counter, scaled and trimmed so it no longer has the huge white border. I also adjusted counter placement so items face the player more naturally while keeping the previous hot dog orientation.

I also polished the transition from the supermarket into ForestDrive. Instead of showing the last supermarket frame and then jump-cutting, I changed the flow so it fades to black first, shows the "you don't have anywhere to be" style text over black, and then slowly fades into the driving scene. This made the handoff feel much smoother and more intentional. In the house, I made the monitor prompt easier to read despite the PSX shader, and I changed the Unity monitor launch so the Open Feed site always starts in the reset/reloaded state with nothing unlocked. That keeps the web sequence reliable for a fresh playthrough every time.

A lot of the final day went into MonitorSite stability. I fixed several white-page dead ends in the shell browser, including returning to the feed index after the echo cipher tools popup, closing the tools popup, selecting the Lima, Peru node on the map, and opening the Westfield Herald news article. The site now keeps its shell and content flow intact instead of dropping into a blank browser view. I also made sure all MonitorSite audio cuts out when the final shutdown state is reached, so the collapse feels clean and final instead of leaving stray sounds behind.

The last major part of the day was the house and EAS audio flow. I removed the old intensemoment.mp3 and replaced it with boneaudio.mp3. I made it start early enough that its 34-second mark lines up with the TV's "LEAVE NOW" message, added a fade-in so it does not pop in, lowered the volume so it does not overpower the other sounds, and made it fade out when the post-escape message begins. I also added a test hotkey so I can trigger the EAS message directly during development. When the "you were seen" failure path plays and I click try again, the player is restored back to the TV checkpoint, the cutoff warning is reset, and boneaudio.mp3 is reset to the same 34-second alignment. Finally, when endingmessage.mp4 finishes playing, the game returns to the main menu, which gives the ending a cleaner full-circle close.

By the time I zipped the final build, the project had been in development for several months, from the first rough Unity prototypes in February to the finished version on May 19, 2026. I cannot know the exact hour count anymore, but considering the amount of scene work, scripting, testing, website development, audio editing, bug fixing, and late polish, it is easily one of the most time-intensive projects I have worked on, likely well over a hundred hours. Finishing it feels strange: partly relief, partly exhaustion, and partly disbelief that all of these separate systems finally became one playable thing. After so many small fixes and moments where one solved problem created another, zipping the build felt like closing a very heavy tab in my brain.

Teacher feedback helped me understand which parts were already working. For example, the main menu background, the driving scene atmosphere, the store visuals, and the house idea received positive feedback. That was useful because in an atmosphere-based project it can sometimes be hard to judge whether the feeling is reaching other people.

I also received the suggestion to add more small interactive objects to the desk scene. That idea influenced how I began to see rooms: not only as places where the story happens, but as environments where small details help make the world believable.

I am also grateful for the guidance I received during the process. It helped me stay focused, notice what was already working, and make better decisions about where to push the project further.

From my experience, I can say that game development is tough work. There is always some anxiety around the possibility that the game will disappoint people, break at the wrong moment, or fail to communicate the feeling I had in my head. That anxiety can be crippling at times. Maybe my anxiety disorder makes it heavier, but after working on OPEN FEED I can much more clearly imagine how stressful professional game development can be, especially in a bigger studio with a release date getting closer, bugs still appearing, and one fix sometimes creating a completely new problem.

One of the biggest things I learned is how easily scope can grow. A small idea turns into a scene, a scene needs a transition, a transition needs audio, the audio needs timing, and suddenly one "simple" feature has five other systems attached to it. OPEN FEED started as a concept about cameras and atmosphere, but over time it became a supermarket, a driving sequence, a house, a web browser, a fictional website, a phone, a TV broadcast, a hiding sequence, and an ending. I am proud of that, but I also understand much better now why scope control matters so much. Every new idea costs time, energy, testing, and emotional patience.

This project also gave me much more respect for game developers. Before this, I understood in theory that games are complicated, but now I have felt it directly. A game is not just art, code, sound, writing, or level design. It is all of those things happening at the same time, and they all have to work together. Even a short game can contain an overwhelming amount of invisible work. The player might only see a small interaction, but behind it there can be hours of debugging, adjusting, rebuilding, testing, and trying to make the moment feel natural.

Another difficult part was juggling this project between my other school assignments and responsibilities. OPEN FEED needed a lot of focus, but I could not give it unlimited time. There were moments where I wanted to keep polishing, but I also had other deadlines and tasks waiting. That made the project feel even heavier sometimes, because I had to switch between creative work, technical problem-solving, school pressure, and personal stress. Still, returning to the project again and again also showed me that I genuinely care about this kind of work.

After finishing OPEN FEED, my feelings about pursuing game design are more complicated, but also more serious. This project did not make game development seem easier. If anything, it showed me how demanding it really is. But it also showed me how meaningful it can feel when separate pieces finally come together into an experience that someone else can play. I still feel drawn to game design, especially atmosphere, interaction, horror, and the way games can create feelings that other media cannot. OPEN FEED has made me more aware of the stress and responsibility that come with this path, but it has also made the desire feel more real. It feels less like a distant dream now and more like something I have actually started doing.

As a closing note, I want to thank anyone who takes the time to read through this diary or play the final build of OPEN FEED. This project was made through a lot of uncertainty, trial and error, late fixes, and moments where I was not sure if everything would come together. If the game manages to leave even a small feeling behind, whether it is unease, curiosity, or just the memory of staring at a strange screen too late at night, then I feel like the journey was worth it.

Martin