What the Linux driver exposes

July 30, 2026 · View on GitHub

If you drive a Brother P-touch from Linux, you are almost certainly using ptouch-print. It is a good tool, but it is a print-only driver. It has no notion of device settings, and no amount of digging through its options will turn one up.

This page records the actual surface, so you can stop looking.

Verified 2026-07-30 against ptouch-print v1.8.r15.gf7cce68+ talking to a Brother PT-P710BT over USB on Ubuntu.

The full option list

--copies=<number>          --debug
--font=<file>              --font-margin=<size>
--font-size=<size>         --force-tape-width=<px>
--timeout=<seconds>        -w, --write-png=<file>

print commands:
-a, --align=<l|c|r>        --chain
-c, --cutmark              -i, --image=<file>
-n, --newline=<text>       --precut
-p, --pad=<n>              -t, --text=<text>

other commands:
--info                     --list-supported

That is everything. Every one of those is about this print job. None of them touch the printer's stored configuration.

What `--info` can tell you

--info is the only read path into the device, and it reports the tape, not the settings:

PT-P710BT found on USB bus 3, device 4
printer has 180 dpi, maximum printing width is 128 px
maximum printing width for this tape is 128px
media type = 0x01 (Laminated tape)
media width = 24 mm
tape color = 0x01 (White)
text color = 0x08 (Black)
error = 0x0000

Useful: tape width, media type, tape and text colour, and an error word. Not available: remaining tape length, auto-power-off timeout, Bluetooth state, or anything else in NVRAM.

Auto power-off: not settable from Linux

This is the one that costs people an afternoon. The printer powers itself off after a period of inactivity (60 minutes from the last print job on the P710BT — status reads do not defer it), and it cannot be woken over USB, because when it powers off it leaves the USB bus entirely. Someone has to physically press the button.

The setting exists. You just cannot reach it from Linux. Confirmed dead ends:

Route	Why it fails
`ptouch-print`	No device-settings support at all — see the option list above
Bluetooth	Cannot carry device settings, by design
Brother's mobile apps (Design&Print, iPrint&Label)	Do not expose the setting
Brother's Printer Setting Tool under Wine	The tool reaches the printer through a Windows kernel-mode driver; Wine implements Win32 user space only, so it launches and sees no printer

The fix

Brother's Printer Setting Tool → Device Settings → Basic → Auto Power Off, set to None. Values are None / 10 / 20 / 30 / 40 / 50 / 60 minutes.

The tool is Windows/macOS and USB-only. The choice is written to the printer's own NVRAM, so it is a one-time change — it survives reboots, re-cabling, and swapping the host machine entirely. You need Windows or macOS once, not permanently.

On a headless Linux box, the practical route is a throwaway Windows VM with the printer passed through by USB vendor:product id (04f9:20af). That is written up in ptouch-cube-print-bridge, along with the hardware notes this page summarises.

A factory reset of the printer (hold power while inserting the cassette, or Reset in the Setting Tool) reverts it to 60 minutes and you get to do it again.

Don't write a keep-alive

It cannot work. On the P710BT the idle timer is reset by printing, not by USB activity — a status-poll loop looks like it is doing something and is not. Measured across three power-on windows with a 300 s --info poll running throughout: the printer still powered off at ~60 minutes each time, including in a window with zero print jobs.

Consequences for this MCP

printer_capabilities returns this list, so the agent does not offer to change a setting it cannot reach.
printer_status failing usually means the printer is off, not that anything is broken. The honest response is "press the power button", not a retry loop.
Copies are passed to ptouch-print --copies N as a single multi-page job, with --precut so the auto-cutter separates them into individual labels. (Printing N separate one-page jobs also works and was the original approach, but it re-runs the ~25 mm leader trim every time and wastes tape.)
Tape length remaining is unknowable, so batch runs cannot warn you before you run out. Check the cassette window before a long run.