No. Your files stay on your machine and are never copied to the server or anywhere else.
There is one thing worth being precise about. When you watch away from home, the video does travel through the server on its way to you: your machine sends it out, the server passes it along to your phone, and that is that. It is relayed, not stored — nothing is written to disk on the way through and there is no copy afterwards. When you watch at home, the video does not go near the server at all; the app fetches it straight from your machine over your own network.
The description of your library, rather than the library itself:
- Titles, years, descriptions, genres and posters
- Filenames and full paths as they appear on your machine, plus sizes, durations, resolutions and codecs
- Which profile watched what, and how far into each thing
- Any subtitles that have been downloaded, since those are small text files
- Your account details
That list is deliberately complete, including the part people do not always think about: the server knows the names of your files and where they sit on your disk. It has to, because that is what makes browsing work when your machine is asleep.
Video is the thing that never lands there.
Nobody. Each account sees only its own library, and there is no browsing between accounts, no shared catalogue, and no way to search across other people's things. Two friends with the same film have two entirely separate entries.
The person running the server has administrative access to it, in the way that whoever runs any server does. They can see that accounts exist, how many titles each has, and whether the machines are online. Being honest about that is more useful than pretending otherwise. If you would not be comfortable with that, this is not the right thing for you.
At home, things keep working. The apps notice the server is unreachable, switch to talking to your own machine directly, and you carry on browsing and watching from the copy of your library that your machine keeps locally. There is a "Local mode" indication so you know where you are.
Progress is recorded while you are in local mode and handed back to the server when the connection returns, so you do not lose your place.
Two limits worth knowing:
- It only works on your home network. Away from home with the server unreachable, there is no path to your files at all.
- It is bounded in time. Local mode keeps working for about three days without the server. After that your device needs to reach the server again before it can play anything. This is a security limit — a device that has been cut off indefinitely has no way to learn that its access was revoked.
The web player has no local mode. Browsers always go through the server.
The same as above: at home, the apps carry on in local mode; away from home, nothing works until it is back.
You cannot sign in on a new device while the server is down, and a television that has never signed in cannot be set up.
One machine per account, currently. Pairing a second one is refused with "This account already has a linked machine. Revoke it first."
If you move your files to a different machine, install the agent there, unlink the old machine under Settings → Linked Machine, and pair the new one. Your library, watch history and profiles are on the server, so they are not affected — the new machine reports the same files and everything reconnects.
One machine can hold as many library folders as you like, across as many disks as it can see, so this is only a real limit if your media is genuinely split across two computers.
Devices you watch on are a different matter — sign in on as many phones, televisions and browsers as you want.
It depends almost entirely on your home upload speed, which is the one number that matters and the one most people have never looked at.
Video has to leave your house at the rate it plays. A 1080p file at 8 Mbit/s needs a sustained 8 Mbit/s of upload. Many home connections upload at a tenth of what they download, so a connection advertised as "100 Mbit" may only manage 10 Mbit going out.
Nothing about this software changes that. What it can do is make the files smaller: the "optimize for streaming" option re-encodes a title on your own machine into something around 8 Mbit/s that plays comfortably over a normal connection. That takes hours of your machine's time per film, but it is done once.
Video sent from outside your home is also capped at roughly 8 Mbit/s when it has to be converted on the fly, so that one person watching cannot use up the whole household's upload. At home there is no cap.
Test this early. Watch something on mobile data on day one, so you find out where you stand before you have built a routine around it.
Three parts, three answers.
The agent updates itself. It checks once a day, verifies a signature on the
new build before installing it, and restarts into it. If a new version fails to
start properly three times, it puts the previous one back on its own. You do not
have to do anything. If you would rather control this, add
"disable_auto_update": true to its config.json.
The Android app does not. There is no store to update it and no in-app
updater yet, so when a new version comes out you download the new .apk and
install it over the old one. It keeps your sign-in and settings. There is no
notification when a new version appears — check the releases page occasionally.
The website is always current. It is a website; there is nothing to update.
To watch from outside your home, yes. There is no cached copy anywhere else — your files exist in one place, and if that machine is off, there is nothing to play.
Browsing still works with the machine off, because the library description lives on the server. You can add things to your list and read descriptions. Pressing play gets you "Your library's machine is offline."
Sleep counts as off. A machine that suspends after 30 minutes will be unavailable exactly when you want it. Turn sleep off on any machine you intend to use this way.
No, and this is deliberate. The agent makes its own connection out to the server and keeps it open; everything reaches you back down that connection. There are no ports to forward, no dynamic DNS to set up, and it works behind the shared addressing that many providers now use, where port forwarding would not be possible at all.
The only networking you might touch is your machine's own firewall, and only for home-network playback on port 8484.
Nothing to you, beyond your own electricity and internet. The server is paid for by whoever invited you.
Worth knowing: every minute you watch away from home uses your home connection's upload allowance and the server's bandwidth. If your internet has a data cap, watching remotely counts against it.
It plays the file as-is wherever it can, which is the fast path — nothing is re-encoded and your machine barely notices.
When it cannot, it converts on the fly. In a web browser that happens for most
.mkv and .avi files, because browsers cannot read those containers, and for
anything in a codec the browser does not support such as HEVC. The app is far
less fussy and plays most things directly.
Conversion uses real processing power. Two at once is the limit; a third is refused rather than making all three unwatchable. If your machine is modest and your library is full of HEVC remuxes, "optimize for streaming" is worth setting up so the conversion happens once, overnight, instead of every time someone presses play.
It makes a version of a file that streams well: 1080p at most, H.264, aimed at about 8 Mbit/s, with the index at the front so playback starts immediately. If a file is already in the right shape it is repackaged rather than re-encoded, which takes minutes instead of hours.
You choose what happens to the original:
- Keep — the new version is written to a hidden
.medialib/optimized/folder next to the original. Nothing you have is touched. This uses more disk. - Replace — the original is moved to a
.medialib/trash/folder next to it and the new version takes its place. The original is not deleted, so you can get it back, but nothing removes it for you either. That folder grows until you empty it yourself.
It refuses to start if there is not enough free space, and it checks that the new file is the same length as the original before it replaces anything.
No, apart from "optimize for streaming" in replace mode, described above, which you have to choose per file.
Scanning only reads. Deleting a title in the web interface removes it from the library listing; the file on your disk stays where it is. If you want a file gone, delete it yourself and it will disappear from the library at the next scan.
If it runs Android TV or Google TV, yes — the app installs on both and has a proper remote-control interface. See install-apps.md.
If it is a Samsung, LG or other non-Android television, there is no app for it. The usual answer is a cheap Android TV box or streaming stick plugged into it.
Yes. They never left your machine, in the folders they were always in, with the names they always had. Delete the agent and you are back where you started. The only thing you lose is the watch history, which lives on the server.
Sign in and go to api.spameri.cz/feedback. Pick whether it is a fault or a request, describe it, and send.
That files a report in the public issue tracker under your display name and the versions you are running. Your email address is never included, and neither is anything about what is in your library. You do not need a GitHub account.
Include which film or episode it was, what you were watching on, and whether you were at home or away. Those three facts distinguish most causes from each other.