• src/doors/termgfx/sgrmouse.h

    From Rob Swindell (on Debian Linux)@VERT to Git commit to main/sbbs/master on Tue Aug 11 14:13:52 2026
    https://gitlab.synchro.net/main/sbbs/-/commit/d16b8d40c42f3d63d02f4d53
    Modified Files:
    src/doors/termgfx/sgrmouse.h
    Log Message:
    termgfx: note that SyncTERM's mouse motion encoding was fixed

    The 96-and-97 hover encoding this decoder works around is no longer what
    a current SyncTERM sends. bbf7c4c79b (dice-12-carpet, 2026-07-11) stopped routing the no-button sentinel through the wheel remap, so unpressed
    motion now arrives as xterm's 35 and a wheel notch as a clean 64 or 65.

    Captured live from CTerm 1.332 to confirm it rather than trusting the
    diff: twelve motion reports at b=35, wheel notches at b=64 and b=65, no
    96 or 97 anywhere.

    No code change. A door serves whatever client dials in, including
    releases built before that fix, so both values must keep decoding exactly
    as they did; what changes is that the comment described the broken
    behavior in the present tense.

    Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

    ---
    ■ Synchronet ■ Vertrauen ■ Home of Synchronet ■ [vert/cvs/bbs].synchro.net
  • From Rob Swindell (on Debian Linux)@VERT to Git commit to main/sbbs/master on Tue Aug 11 14:13:52 2026
    https://gitlab.synchro.net/main/sbbs/-/commit/5ed64d120647284ab1b1dc40
    Modified Files:
    src/doors/termgfx/sgrmouse.h
    Log Message:
    termgfx: name the second SyncTERM fix that finished off the 96 encoding

    The note added a moment ago credited the whole thing to bbf7c4c79b (dice-12-carpet, 2026-07-11), which is only half of it. That commit fixed motion with NO button held. A wheel notch left its button latched in the
    button state -- a wheel "button" is momentary and no release ever follows
    -- so motion after a scroll was not the no-button case and still came
    through the remap as 96, reading as a drag rather than a hover.

    a34c0f4e62 (exchange-33-sign, 2026-08-09) clears it, which is what
    finished the job. Recorded because the two fixes are four weeks apart and
    a reader who finds only the first will conclude, wrongly, that anything
    still reporting 96 must be a build older than July.

    Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

    ---
    ■ Synchronet ■ Vertrauen ■ Home of Synchronet ■ [vert/cvs/bbs].synchro.net