The sequencer's run #
The run theorem of DescriptiveComplexity.Draw.seqRule: from the checkpoint
before the first stage at the marker, one abstract stage run per position,
to the exit phase one cell to the marker's right after the last – the
matrix runs its classified atoms with it, the gates their blocks, and
neither needs to know a stage's internals beyond its entry phase and its
control and tape transforms.
The run comes with its cost (seq_reachesIn): a stage's own width plus
the dispatch and the walk back, once per stage, and one step to leave.
seq_run is it with the budget forgotten – the widths recovered from the
stages' runs by DescriptiveComplexity.TMData.exists_reachesIn_of_reflTransGen
and the largest taken, which is why a space-bounded caller need not have
counted anything.
The sequencer's run, on a clock: from the checkpoint before the first stage at the marker to the exit phase one cell to its right after the last stage, at each stage's own width plus the dispatch and the walk back, and one step to leave.
Dependency graph
The sequencer's run, the budget forgotten: what a space-bounded caller reads, its stages' runs carrying no count.
Dependency graph
The sequencer's prefix: from the checkpoint before the first stage to any later checkpoint, given the stages strictly below it – what a run that leaves the sequence early (a failing gate block) composes with.
Dependency graph
The sequencer's prefix, on a clock: as
DescriptiveComplexity.Draw.seq_run_prefix with the stages counted – a stage's
own width, its dispatch and the step back, once per stage below the
checkpoint.