FAQ

Is Tuxery an app store or an installer?

No. Tuxery is a catalog: it finds every way an app can be installed on Linux and points you to the right one. Every Install button hands off to the real source — Flathub, the Snap Store, your distribution's software center or package manager, the developer's website. Tuxery never hosts a package or runs installer code.

Why isn't my app listed?

Usually one of these:

It's in a source Tuxery doesn't read yet — see the sources list, and propose one.
The filter took it for a non-app: libraries, fonts, documentation, plugins and test packages are left out on purpose, and the rules sometimes get it wrong.
It's listed under another name, merged into an app it belongs to: editions, other builds and add-ons appear on their main app's page. Search for the main app.

If it should be there, open an issue with the app's name and where it's packaged.

An app is merged wrongly, or shows up twice. How do I report it?

Open an issue on the catalog repository with the address of the app's page on Tuxery and what looks wrong. Merges are fixed with a reviewed rule, so the fix stays. An in-site "flag this match" button is planned. See how sources are merged for why mistakes happen.

The category is wrong. Can I fix it?

Yes, the same way: an issue on the catalog repository. Categories come from the app's own metadata when it has some, otherwise from rules and language models — see the data philosophy.

Why do versions differ between sources?

Because each source ships its own. Stable distributions (Debian, Ubuntu LTS) freeze versions for the life of a release and only add security fixes; Flathub, Snap and rolling distributions follow new releases closely. Tuxery shows each source's version as that source reports it, so you know what you'd actually get. See which format to choose.

Where do ratings come from?

From the people who rate apps on their own stores: GNOME's review service (behind GNOME Software's and Flathub's stars) and GOG's reviews. When an app has ratings from several sources, they are combined, weighted by number of votes, and the per-source ratings are shown on the app's page. Tuxery has no ratings of its own yet.

By figures sources publish — never by payment. Details, and why an app might not be trending, in how rankings work.

How often is the data updated?

Every source is refetched weekly, and the catalog is republished from that data. The date of the data currently shown is on the Status page.

I'm the developer of an app. How do I fix its page?

Most of what Tuxery shows comes from your app's own metadata — particularly its AppStream file (name, description, screenshots, categories, license, release dates). Fixing it upstream fixes it on Flathub, in software centers and on Tuxery at the next refresh. For anything else — a wrong merge, a missing source, an edition shown the wrong way — open an issue on the catalog repository. A way to claim your app's page is planned; see claim.

Can I submit an app or game?

Not through the site yet — a submission form is planned. For now, open an issue; the best long-term route is getting the app into a source Tuxery reads (Flathub, a distribution, the AUR...).

Does Tuxery track me?

No analytics, no advertising, no account. Your settings (theme, operating system, preferred formats) are stored in your own browser and never sent anywhere. Like any website, the server that hosts it (Cloudflare) keeps technical request logs, for at most 7 days. Everything is spelled out in the privacy policy.

Can I reuse Tuxery's data?

Yes — see open data. Everything Tuxery shows is built from public data: open feeds and package indexes, plus public listings we read the way a browser or an app store client would. Nothing comes from a private agreement or a paid feed, so nothing can be withdrawn from you: what Tuxery gathers stays available, and the code to gather it again is open source.

Where does the data come from?

Most sources publish their whole catalog as a file or a feed meant to be read by machines: Flathub's AppStream feed, the distributions' package indexes, the AUR's metadata dump, Nixpkgs' channel file. Reading those is simple and cheap for everyone.

Some don't, and Tuxery has to work around it:

The Snap Store has no full listing — only a search that stops at 100 results. Tuxery sweeps it by category and by ever-longer search prefixes, about 100,000 requests per refresh, and still can't be sure it has everything.
GOG has no documented catalog API; Tuxery reads an undocumented one that launchers such as Heroic and Lutris also rely on.
Lutris ignores the filter for native Linux installers, so every page of installers is read and filtered afterwards, then each game is looked up for its genres and description.
GitHub search returns at most 1,000 results per query, so the search is split by creation date until each part fits, and each AppImage in the community feed is checked against its repository.

If you run one of these stores, a downloadable export of your catalog (a file, an AppStream feed, a paginated API) would make Tuxery's listing of your apps complete and lighter on your servers. You'd be very welcome — get in touch.

Does Tuxery use AI?

Yes, in two places, and always with a person in the loop:

Writing code. Language models help write and review Tuxery's code; every change is reviewed by a maintainer before it lands.
Cleaning the data. For apps no source categorizes, a model suggests whether a package is an app, a game or a library, and its category. Each verdict records the model and a confidence; low-confidence ones are checked by hand, and upstream categories always win. Links between apps are never left to a model. Details in the data philosophy.

Models never write the descriptions, screenshots or ratings you see: those come from the sources.

How can I help?

See how to contribute: reporting bad data is the most useful thing anyone can do, and the per-source to-do lists concrete work on each source.

FAQ