EverythingYou Hear,One Place.

Think about the last day you spent listening. You probably had music in one app and a podcast in a second. The radio station from home lives somewhere else, on a website with too many adverts. Your audiobook is in a fourth app, tied to a store. A lecture you wanted is on a video site that will not play with the screen off. Each of those has its own account, its own queue and its own idea of who you are. None of them knows what the others know.

We have come to accept this as natural. It is not natural. It is a result of how the industries were divided, and of companies wanting you inside their fence. Your ear has no such divisions. To you, it is all one thing: the sound you live with during a day. vyv begins from that plain observation. All of it should be in one place.

vyv is a minimal, agent-powered player for music, radio, podcasts, audiobooks and video, on the web and on the desktop. In the Kurdish context it carries a second name, KRD PLY, as a national multimedia player. It is working software: live sources, a real library, and a desktop application with installers. It is still in review and transformation, so I will speak about what it is and also about what it is trying to become.

Your ear has no divisions. The industries that sell to it do.

I care about this for a reason beyond convenience. I come from a culture that has been carried in sound for most of its history. Kurdish memory travelled by voice: the dengbêj singing a story for an hour, a radio programme heard in secret, a cassette passed from hand to hand across a border. For a long time the language could not be printed freely, and so it was sung.

So when I look at the modern audio world, I ask a particular question. Where does that sound live now? The honest answer is that it lives nowhere in particular, scattered across services that were not built with it in mind.

Convergence is an easy word to say and a hard thing to build. The lazy version is a screen with six tabs, which merely moves the fragmentation inside one window. A real convergence has to respect that these things are different, and then unify what can be unified.

A radio station is alive. It has no beginning, and you cannot skip it. A song is an object you return to. A podcast is an appointment with a person, and you need to resume where you stopped. An audiobook is a long relationship, measured in days. A lecture is something you learn from, and you may want to slow it down and go back. An archive film belongs to the dead. Treating all of these as tracks in a list would be a kind of disrespect.

What vyv does is keep each nature and share the things underneath. Every source, however different, maps onto one common idea of a playable item. So play, queue, like, bookmark, playlist and history work the same way across all of them. You can put a live radio station, a heritage recording and a chapter of a book in one collection. You can resume all of them. That is the actual breakthrough, and it is invisible when it works.

The sources themselves are open and real. Public directories of radio stations. An open platform of independent artists. The public podcast directories. The great public archives, with their recordings and read-aloud books and out-of-copyright films. Video search through an official embedded player. Most of this runs with no keys at all. I find something honest in a product built on the open parts of the internet rather than on exclusive deals.

A live station, a heritage recording and a chapter of a book can share one playlist.

Educational video deserves a word of its own, because it is where this idea has the most to give. A student in Erbil may learn from a recorded lecture, an audiobook on the same subject and a podcast from an expert, with no tool that treats these as one course of study. A player that keeps your place across all three, and that can follow you between them, begins to behave less like a jukebox and more like a tutor.

And Kurdish goes first. On Radio and Home, Kurdish stations, in Sorani and Kurmanji, appear at the top. That is one line of logic in a sorting function, and it is a political act. Every global service ranks a language like mine last, if it ranks it at all. Here, the default is reversed. The whole world's stations are available, and you begin at home.

There is a design conviction in vyv that I hold strongly. A player is a guest. It lives in the corner of your day, and it should behave accordingly.

Most media applications are engineered to capture you. Autoplay, endless recommendations, red badges, the next thing already loading. Their real customer is the time you were going to spend elsewhere. I want the opposite. The interface here is almost entirely icons, in two themes called Obsidian and Porcelain, with soft curves and very few words. If an icon says what it does, no label is added.

The desktop version takes this idea further than anything else I have built.

The real operating system window changes shape. In Full mode you see the library. In Cover mode the window becomes a square of artwork. Micro turns it into a floating pill. Nano reduces it to a small orb that stays on top of everything and asks for nothing. The player takes exactly as much of your screen as you decide to give it. Most software assumes that it deserves the whole room. This one knows how to sit quietly at the edge of the table.

A small detail shows the seriousness of this. The player dock samples whatever is behind it and adjusts its own ink, so that it stays readable over white, black or a busy image. No one will ever notice that, and that is the point. Good design does a lot of work to avoid being noticed.

There is also an agent. You can tell it what you want in ordinary language: a mood, a year, a question, a command. Something quiet for working. Songs from 1995. I think this is the right use of AI in a media product, and a much better idea than the infinite recommendation feed. The feed decides for you, on the platform's behalf. An agent you speak to works on your behalf. The difference between those two is the difference between being served and being farmed.

A feed decides on the platform's behalf. An agent works on yours.

Here is a piece of the engineering I find beautiful, because it comes from thinking about where people actually live.

Connections in my region are not always good, and internet services can vanish. So vyv is built in layers, each ready to catch the failure of the one above. Live sources come first. Beneath them sits a bundled catalogue, a snapshot of more than a thousand real items that opens instantly and takes over if a live service goes down. Beneath that is your own library. Beneath that is your device: the application works with no account, with nothing stored anywhere but on your machine.

Streams reconnect on their own. If a stall lasts, the player moves on. Sources that refuse certain browser permissions are played through a second path, so that they never go silent, even if the equaliser cannot reach them. All of this is invisible, and all of it comes from one design rule: the sound must not stop.

Accounts follow the same philosophy. There are no usernames or passwords to remember. If you want your library to follow you between devices, you can sign in with an identity you already have, and your likes, playlists and resume positions sync, protected so that only you can touch your own rows. If you do not want that, everything stays local. The application does not hold your music hostage in exchange for your email address.

The application should not hold your music hostage in exchange for your email.

The business model remains an open question, and I would rather say so than pretend. A national player could be funded in several ways. What I will not do is fund it by selling the attention of the people who use it. That is the one route I have ruled out.

Why would anybody call a media player national? It sounds grand for a piece of software.

Consider what a country's sound actually is. It is its radio, its music, the voices of its writers read aloud, its recorded history, its lectures, the old recordings slowly decaying in cupboards. Every nation has these, and most of them have handed the distribution of all of it to a few global platforms. Those platforms are often excellent, but they answer to their own business and they rank by their own logic. They have no reason to care whether a Kurdish lullaby stays findable.

For a people without a state, or with a contested one, this matters more. Our voices have historically had to be carried on whatever channel would take them, and they could be removed at any moment. A home for that sound, under our own care, is a modest form of independence. It does not need to compete with the giants. It only needs to exist, and to work well, and to last.

In the age of generative AI there is a further reason. Synthetic voices are now good, and they will get better. It will soon be easy to produce a convincing recording of nearly anything, in nearly any language, including mine. In that world, a trusted place holding real voices has a new kind of value. The recording of a real dengbêj, attributed and preserved, is an original. Originals matter more when imitation is free.

So I think of vyv as a very simple promise. All the sound you care about, in one calm place, in your own language first, with nothing trying to capture you. The music, the radio, the voices, the lessons and the old recordings, all together.

The idea is not complicated, and that makes me trust it. The people who sang for an hour to keep a story alive did not need a platform strategy. They needed someone to listen. Software cannot do much about that, except make sure that when somebody wants to listen, the voice is there, and it plays.