A repository called fmhyplay/android has appeared on GitHub — an app the authors describe as "any streaming site, but with a real player." The repo is fresh: 13 commits, a CI workflow, a CHANGELOG, documentation, and built APKs in Releases. Judging by its structure, this isn't a one-evening prototype but a project with a reproducible build and signed releases from a tag.
Here's how it works: the app opens a site in a WebView and catches the moment the page launches an HLS or DASH stream. The video then goes into ExoPlayer with hardware decoding, with no overlays on top of the picture. So that players that verify the source don't break, FMHYPlay replicates the page's Referer, Origin and cookies. The Servers and Quality buttons in the app press the site's own buttons — including inside cross-domain iframes. Separately: a poster and ratings catalog from Cinemeta, voice search, a channel list from iptv-org with background checks that a channel is actually live, built-in OpenSubtitles, and gesture-based brightness/volume controls on the phone. The code is pure Java and SDK command-line utilities, no Gradle, roughly 11 MB per APK.
The reason anyone would build this is obvious without hints: browsers on TVs are terrible. A remote instead of a mouse, pop-up windows, a built-in player on weak hardware — that's the standard experience of anyone who has tried to watch anything more complex than YouTube directly on a TV. FMHYPlay doesn't try to replace the site; it takes from it exactly the part that works poorly. Our assessment: this is a more honest engineering approach than yet another "content library."
The second interesting piece is WARP inside the app. It's a userspace WireGuard tunnel only for the app's own traffic: no system VPN permission, no root, and the local network stays directly accessible. Technically, it's a neat loophole in Android's rules: the phone doesn't know the app is tunneling anywhere. Plus domain-based blocking of ads, trackers, pop-ups and redirects. From experience, it's the redirects on such sites that are the most infuriating.
Now for the unpleasant part — and this is our interpretation, not a quote from the README. The "FMHY" in the name is no random acronym: it's a well-known catalog of links to "free" streaming sites, and its legal standing is nil. The authors carefully write "sites you choose" and "no accounts, no telemetry" — the app doesn't store content and keeps no logs. But that removes the question from the developer, not from the user. The question is whether you, as a reader, are ready for it — not whether the code works.
The writeup contains no version number, no release date, and no list of sites where the interception was actually verified. We don't know how many streaming pages break when their layout changes — and they will break, because the app latches onto the buttons and streams of someone else's player. There's no data on speed or battery drain on weak set-top boxes, and no independent reviews: the only source is the project's own README. There is a LICENSE file in the repository, but its text isn't quoted in the writeup, so the distribution terms need to be checked separately.
Prediction: within a quarter, the issues tracker will be dominated by complaints about broken sites and Servers/Quality buttons — stream interception depends on the layout of specific pages, and without regular fixes the project will start falling apart.
Disclaimer: the data comes only from the README and the repository structure; there are no independent checks, and no one has officially claimed support for specific sites.
Source: GitHub — fmhyplay/android
Source: github.com · post in Telegram