Entry 206

"The carry's votes checked against content: a name whose table reads like Infection's stays"

2026-09-26volumes

Filestools/main_data.pycrates/piney-data/src/main_data

Since 0202 a unanimous vote from paired code outranked a name tools/xfer.py carried: "the code wins". That was right for Mutation's skillTbl and MailTbl. A look at the others printed for the three later volumes found most of them wrong: the tracker had paired a lui with the wrong low half, or the functions only partly aligned.

Several of these override places were the same address on two or three discs, which no real move would give.

A vote now overrides a name only when it scores at least as high. The score is how many of the global's words read as Infection's at each place, over its first 0x400 bytes: a pointer (Infection's word relocations) compares by the bytes it points at, and a pointer to untexted data counts for neither. A table the later volumes changed a little still reads more like Infection's at its own place.

The overrides left:

• skillTbl, MailTbl: checked right before;
• idolItemList44: neither place reads like Infection's;
• Outbreak's mt: zeros at both.

The sequence pass (0205) applies the same test to the labels it moves.

The misses the source still names: Mutation 5, Outbreak 23, Quarantine 28. All three later discs still go through New Game into the desktop with no error printed.