🕛Neo Angband: Cutting Room Floor v0.3.0

Changed

• Restore Teleport Other: the restored spell is now priced for the game as it is, not as it was. The three rows previously carried Angband 4.1.2’s values verbatim. Angband 4.2 repriced spells to cost the same in every class that has them, so a 4.1.2 price landed in the current game as a penalty rather than a restoration: it charged a Priest 20 mana at 80 percent failure for a spell the Mage buys at 10 mana and 30 percent.

  Class	Before	After
  Priest	20:20:80:16	18:10:30:20
  Paladin	25:25:80:12	24:10:30:50
  Ranger	31:25:70:3	30:10:30:50

  Mana 10 and fail 30 are measured: both surviving copies of the spell sit there. Level and exp follow the model for the caster class, the Mage for the Priest and the Rogue for the Paladin and the Ranger. How far to soften the Priest’s level cut, and the exact exp for the Paladin, are judgment calls and are labelled as such in the README. The Ranger’s result is identical to the Rogue’s current row, which is the expected outcome given the two classes had identical 4.1.2 rows.

  The effect and description text are unchanged; they were already copied from the Mage’s current copy.

Added

• Two tests that pin the restored rows to the published content pack rather than to a constant in the test file: one asserting mana and fail still equal what both surviving copies of the spell cost in core, and one asserting the Ranger’s row still matches the Rogue’s. A core release that reprices either surviving copy now fails the suite instead of leaving this mod quietly stale.

Documentation

• README gained a “Where these numbers come from” section carrying the measurement: the cross-class pricing convergence counts, the two model deltas, the per-axis basis, and an explicit split between what is measured and what is judgment. It replaces the earlier claim that nothing in the table was derived.
• manifest.json’s section description no longer states that every value comes from Angband 4.1.2.