[medoix]

What does this PR change?

Selecting the "Current position" origin (Ruida WCS CURRENT) never reached the Ruida driver.

• MachineController.switch_active_wcs() sent the WCS name to the device with run_raw(). That works where the name is a command (G54 to G59 on G-code controllers), but the Ruida names (MACHINE, ANCHOR, CURRENT, SET_POINT) are not rpascript. The adapter kept MACHINE, so the switch was logged as Device did not confirm WCS switch to CURRENT, got MACHINE.
• After any reconnect, sync_active_wcs_from_device() read MACHINE back from the adapter and reset the machine's selection, so jobs could silently run from the machine origin instead.

Changes:

• switch_active_wcs() calls driver.select_wcs() instead of run_raw().
• The base Driver.select_wcs() now sends the name with run_raw(), so GRBL and other G-code drivers send exactly what they sent before.
• The Ruida adapter starts from the machine's active WCS when it is one of its reference points, so a rebuilt adapter no longer reverts the selection to MACHINE. A Ruida controller does not store this selection, so the machine model is its only record.

Why did you choose this specific architectural approach?

Driver.select_wcs() already exists as the driver hook for choosing a WCS, and the Ruida adapter already implements it, but the UI path went around it. Routing switch_active_wcs() through select_wcs() lets each driver decide what selecting a WCS means, and giving the base method the old run_raw() behaviour keeps the G-code drivers unchanged without touching them.

Alternatives considered:

• Special-casing Ruida in the controller. That puts driver knowledge in the core.
• Making the Ruida run_raw() recognise WCS names. That mixes WCS selection into the raw rpascript path, and read_parser_state() would still be reset on reconnect.

MachineController.select_wcs() also calls driver.select_wcs(); it has no callers today, so the new base behaviour does not change any existing path.

Known limitation, not addressed here: the Ruida frame() sends absolute moves, so framing still ignores the selected origin. That needs its own change.

How did you manually test this locally?

RDLC-V26.60.20 controller over UDP and an RDC8445S over TCP.

1. In the bottom panel, choose the "Current position" origin. Before: log shows "Device did not confirm WCS switch to CURRENT, got MACHINE". After: no warning, and later syncs log "Synced active WCS from device: 'CURRENT'".
2. Place a small design at the canvas origin, jog the head with the controller panel to a clear spot, and run a low-power cut. The captured job starts with REF_POINT_CURRENT (D8 12) and the controller cuts at the head position instead of returning to the machine origin.
3. Switch to another machine and back (reconnect). The origin stays "Current position" instead of reverting to the machine origin.

Tests: new tests check that switch_active_wcs() goes through select_wcs() without run_raw(), that the base select_wcs() sends the name with run_raw(), and that a new Ruida adapter reports the machine's WCS (falling back to MACHINE for G54). The machine and Ruida driver test folders pass apart from test_end_failed_carries_message and test_failed_job_reports_error_message, which fail the same way on unmodified main in my environment.

Checklist

• I am not an AI. I have manually read the code and understand it.
• I take full responsibility for the quality and architectural fit of
  this code.
• I agree to publish this code under the MIT license.

[knipknap]

Thanks for the fix and the clear description - looks good!