foo2ddst: missing "@PJL SET HOLD=OFF" causes Ricoh SP112 to accept job but never feed paper (hangs indefinitely)

[steventran2023]

Environment

• Printer: Ricoh Aficio SP 112 (USB, serial X024M530950)
• OS: Armbian (Debian trixie), aarch64
• Package: printer-driver-foo2zjs 20200505dfsg0-3+b1 (Debian package)
• Driver/PPD: Ricoh SP 112 Foomatic/foo2ddst (recommended), grayscale

Symptom

• CUPS accepts the job, foomatic-rip/foo2ddst-wrapper run without errors,
  all data is sent to the USB device without error.
• The printer receives the job (confirmed by directly writing the raw
  DDST stream to /dev/usb/lp0, bypassing CUPS entirely) but never feeds
  paper. The status LED blinks continuously and the job never completes
  on the printer side.
• The built-in printer self-test (via the physical Start button) prints
  fine, confirming the paper feed mechanism itself is not broken.
• Printing the exact same document from Windows using Ricoh's official
  driver works fine on the same printer/cable.
• Printing via an alternative community CUPS filter
  (https://github.com/vaginessa/ricoh-sp112-ppd, "pstoricohddst-gdi")
  also works fine on the same printer.

Diagnosis

Comparing the PJL header emitted by foo2ddst.c against the working
pstoricohddst-gdi script, I found that foo2ddst never sends
"@pjl SET HOLD=OFF", while pstoricohddst-gdi always does:

@PJL SET COVER=OFF
@PJL SET HOLD=OFF      <-- present in the working driver, missing in foo2ddst

I confirmed via grep -n HOLD foo2ddst.c that this PJL command is never
emitted anywhere in foo2ddst.c.

Fix

Adding this single line after the existing COVER=OFF line in foo2ddst.c
(~line 693, function that emits the PJL job header) resolves the issue
on my unit:

fprintf(fp, "@PJL SET COVER=OFF\r\n");
fprintf(fp, "@PJL SET HOLD=OFF\r\n");   // <-- added

After rebuilding foo2ddst with this change, the printer feeds paper and
prints correctly through CUPS as expected.

Notes

• This may only affect certain SP111/SP112 firmware revisions that
  default job HOLD state differently; other users may not see this
  since foo2ddst apparently works for other SP112 units.
• I also tested disabling JBG_LRLTWO in the JBIG encoding options as an
  earlier hypothesis; that change alone did NOT fix the issue. The
  HOLD=OFF fix above is what actually resolved it.
• Happy to test further patches or provide additional captured DDST
  streams if useful.