Read system TV-remote input independently, expose launcher actions through arrows and OK, and preserve DS4 shortcuts. Add reconnect handling, explicit Stop confirmation, diagnostics, and reader/navigation regressions.
Prepare the next patch version as a test candidate; PS4 HDMI-CEC reception still needs hardware confirmation before release.
GoldHEN versions before v2.4b18.10 don't run ELF payloads reliably (v2.4b18.8
improved ELF loading, v2.4b18.10 fixed a sceUserServiceInitialize failure in
them) and have no auto-start, and every firmware GoldHEN supports has a
v2.4b18.10 or later. The installation guide now lists each firmware with the
GoldHEN version it needs, marking the ones confirmed on a console (10.01,
13.02, 13.04, 13.52) apart from those that only have GoldHEN and start-up
support so far. The README points to it.
- Requirements list the firmwares Control4Free is tested on (10.01, 13.02,
13.04 and 13.52, with GoldHEN v2.4b18.10 and v2.4b18.12); 13.52 needs 1.1.1.
- Updating says you don't need to stop Control4Free before deleting or
replacing the app. A new "After a restart" section says auto-start brings it
back by itself and the app doesn't get in the way.
- Uninstalling warns that deleting the app alone leaves Control4Free starting
with GoldHEN, and how to clean up when the app is already gone.
- Reporting a problem stops the normal version before starting the diagnostic
one, so the report comes from the diagnostic version, and says how to switch
back.
- The kernel-log message also names old test builds of the diagnostic app.
CI now builds Control4Free-<version>-diag.pkg next to the normal package and
attaches it to each release. The bug report form asks for its screenshots or
report file instead of an optional log, and the troubleshooting guide's
"Reporting a problem" section walks through it: install the diagnostic
package, make the problem happen again, screenshot the 5 pages, then put the
normal package back.
When Control4Free was sent but did not answer, the app locked itself until
the PS4 restarted, so two copies could never run. The payload already takes
care of that: a second copy finds the instance lock and quits at once. So
the app now says to press Cross to try again, and a restart never helped a
payload that cannot start anyway. The lock stays only for the app itself
failing to start ("Reopen the app").
When the PS4 reports a user signed in on a controller, the service asks for
the user's name and adds it to the status as `user`. The PS4 may not have
finished signing the user in by then, so a failed look-up is retried every
half second for five seconds. The tile shows the name in place of
"Controller N" (the ring keeps the number), and the controller screen reads
"Controller 1 · Alex". Names are kept out of the log, since people post logs
in bug reports, and they go out cleaned: quotes and backslashes escaped, and
control characters or anything that isn't valid UTF-8 replaced, because a
browser drops the whole WebSocket on a text frame that isn't valid UTF-8.
An Invite panel, opened from a button next to the settings or from the menu,
shows the page's address as a QR code for friends to scan. The console makes
the code with the QR encoder the app already ships, from the address the page
reached it on, and sends the module grid; the page draws it as SVG, so it still
needs no libraries.
The gamepad picker's tooltip had the same "Controller 1 · Controller 1" doubling
the controller screen had; it now names the user too.
A gamepad's row now says "no rumble in this browser" when the browser has
no way to rumble it: common for a DualSense, while Chrome and Edge rumble
Xbox pads. Before, it just never rumbled.
The heartbeat now reports how many rumble/light-bar reads ran in the last
minute and what they cost on average and at worst, in microseconds: the
longest an arriving input could wait behind one.
A game talks to a virtual controller as it would to a real one, and the system
keeps what it asked for. scePadVirtualDeviceGetRemoteSetting reads it back in
the layout of a DualShock 4's output report: the two motors at [3] and [4], the
light bar at [5..7]. The layout was read on the console while games rumbled and
users signed in, and feedback.c is tested against those exact buffers.
The service reads it every 16 ms for each controller; games pulse rumble for
only tens of milliseconds, which a 100 ms poll missed.
- Rumble goes to the controller's owner as {"method":"v","params":[pad,large,
small]}, once per change, and to whoever takes a controller over, so a page
never keeps buzzing on old news. The page already drove an Android phone's
vibration and a gamepad's actuators from it.
- The light bar goes into the status everyone sees. The PS4 sets player colours
at a quarter strength, so the colour is brightened to full, keeping its hue;
an unlit bar keeps the controller's own colour.
- Without the call, controllers work as before and it is logged once.
Also: the page says "Choose user on PS4", not "on TV", since not everyone plays
on a TV. The test stub takes C4F-TEST-FEEDBACK lines to set what the "game"
wants, and the test client can wait for messages the service pushes.
- README: what Control4Free is and why it reaches menus and sign-in, features,
a quick start, screenshots, requirements, limitations, credits.
- docs/: installation, playing, troubleshooting (every message the page and the
app can show), how it works, and the WebSocket protocol.
- CONTRIBUTING.md: building, testing, code style, releasing.
- Screenshots of the page on phones and a PC and of the app, made against the
real service built for the host, at the example address 192.168.1.20.
- Issue forms, a contact list pointing at the guides and private security
reporting, and FUNDING.yml.
- CHANGELOG.md dated. CI now publishes this version's section of it as the
release notes rather than the whole file.
- THIRD_PARTY.md: the DualShock 4 diagram the buttons are drawn after (CC BY
3.0), and Apollo Save Tool for the package settings.
- Page: the Vibration setting no longer promises to buzz when a game rumbles,
which 1.0.0 cannot do yet.
MoHadiShibli