YACC1-D's video card V3: the netlist and the timing check P32W's video card is held to

Two files of Ken's YACC1-D repository (github.com/ksr/YACC1-D), copied byte for byte, each under the path it has there. Do not edit them.

Here	In YACC1-D, as of its commit
hardware/cards/video/kicad/v3.0/v30_netlist.py	db48aa5 (2026-11-11)
tests/video/v30_netsim.py	12565a7 (2026-11-11)

(The paths are kept because the second file finds the first by its own.)

P32W's video card, revision 2 (docs/video-card-acrtc.md, D103), keeps 35 ICs of that card as they are: they are simulated gate for gate and timed with their data-sheet delays there (tests/video/v30_netsim.py). generators/card_video.py transcribes those parts and kept_check() proves the transcription equal to this file: every kept part, every pin, every net, but for the 11 pins of the seam that the generator lists with their reasons. make test runs that check, and the netlist's own (python3 hardware/cards/video/kicad/v3.0/v30_netlist.py).

generators/sim_video_timing.py then runs v30_netsim.py's two scripts, with that file's models and checks unchanged, on P32W's netlist: the kept part's timing, at eight corners (D107). It also borrows that file's engine for P32W's own bus side. v30_netsim.py run by itself here stops at its first step, which reads YACC1's microcode: that step is YACC1's bus side, not kept.

The copies are here so that the checks run wherever P32W is checked out. If YACC1-D's netlist changes, the check's line in make test says so (on a machine that has both repositories); taking the change is then a decision, since the kept part's margins belong to the parts wired this way.

YACC1-D's V3 is not being built for the YACC1 (D103).