The previous entry left one test: compare the count at byte 8 against the
payload extent. Pointer spacing gives the extent, since the packets lie in
ascending order in EE RAM.
Range measured: savestate slot 1, the 17 packets in RAM 0x22a000..0x24037f,
each with its record address, byte-8 count and pointer, sorted by pointer.
Result: eight packets sit exactly count*16 bytes after their predecessor -- a
ratio of 16.00 in every one, no rounding.
0x22a6e8 count 22 ptr 009376f0 gap 352 16.00
0x22a728 count 36 ptr 00939c50 gap 576 16.00
0x22a768 count 23 ptr 0093cc70 gap 368 16.00
0x22a7a8 count 36 ptr 00998880 gap 576 16.00
0x22a7e8 count 25 ptr 009c19c0 gap 400 16.00
0x22a828 count 24 ptr 009cf370 gap 384 16.00
0x22a868 count 41 ptr 009d2da0 gap 656 16.00
0x22a8a8 count 21 ptr 009f5aa0 gap 336 16.00
Eight exact hits at a ratio that must be an integer is not luck. The count at
byte 8 is the record count, and these payloads are 16 bytes per record.
These are exactly the sub-list payloads -- 009376f0, 00939c50, 0093cc70,
00998880 and 009d2da0 were read directly in earlier passes and all open
00 00 00 30 followed by a repeated word.
The other nine have ratios of 168, 419, 559, 31288, 7985, 3458, 764 and 6452
-- no shared factor, far too large to be a record count times anything. Either
their payload is not followed immediately by the next packet in pointer order,
or they are not 16 bytes per record. Ruled out either way: they do not follow
the same rule.
This resolves byte 8, measured two entries ago as "a count, 0x0b to 0x29" and
now shown to size a payload exactly in eight independent cases within one
level. Whether it also sizes the fixed-point payloads is the open half.
Still open: why the fixed-point packets differ, and what the 16 bytes per
sub-list record are -- the repeated word (0xcd, 0xc0, 0xc8, 0x78) differs per
packet, so the records are not identical to each other.
ReaperOfVeriod
f1cf6cd164