transfer.submit {action: withdrawBid, listingId} removes the manager's bid
before the seller decides. The orders list of the transfer market does it
with its delete button.
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
transfer.submit {action: withdraw, playerId} ends the manager's own list
entry as it ends without a buyer. The transfer market's first button does
it for a listed player.
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
A new arrival no longer inherits the previous occupant's board flags, and a
player who is already listed is not listed a second time (from OpenBMH
1.0.3).
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
At the national coach's desk the friendly read quotes a match against
another nation on a copy of the Watcom generator (price, payer, ratings,
free dates); friendly.book draws the same ratings and books it. The
opponent bound of the national quote follows the world layout.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Measured with twice the offer, four boards brought a top club about 5 million a year and the dearest camp paid off clearly on average; at 150 % it is about 3.8-4 million, roughly break-even for Dubai with luck on top. Default sponsor_bonus is now 150 (data format F4).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Middle East camps share their grades but differ in price; a stay
can now bring an exclusive sponsor: 15 / 30 / 45 % per stay (Doha /
Abu Dhabi / Dubai), a shirt (about every third offer) or four boards at
twice the club's best regular offer for 12 months. Accepted on a free
place it starts at once, otherwise it is queued and starts at a month
change once the place is free. The offer goes through the regular
advertising acceptance.
Camp fields sponsor_chance / sponsor_bonus / sponsor_months (data
format F4, editable; built-in values for the Middle East camps), state
saved in state/modern.json, read query {kind: "sponsor"}, session action
dispatch_exclusive_sponsor, tests in modern_camp_tests, SIM_SPONSOR=1.
Sponsor artwork does not exist yet (asset: null, documented).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Each column costs the stake (10,000-400,000) and is settled on its own
with the original quotas; all columns stake 400,000 at most, every
column beyond the first adds a 10 % fee, no two columns alike. Column 1
stays on the original coupon; columns 2-6 and the last settlement live
in ModernState::pools and are saved. Session action
dispatch_pools_columns, read query {kind: "toto"}, unit test
modern_toto_tests, SIM_TOTO=<columns> in the season sim.
Measured over two seasons: 1 column returns 30-71 %, 6 columns 12-49 %
of the stakes; more columns win more often, never on average.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Players of the data set keep their age and seven skills at the career
start; the original draws new ones (and OR-ed ages produced players
of 46 and a crash once the draws were skipped).
- Training camps: the session allowed only resorts 0-7; the manager's
club was read as a signed byte (clubs above 127 crashed); at the
national coach's desk no camp can be booked.
- SIM_CAMP=<resort> in the season sim; the human club's final place is
taken from the almanac. Twelve seeds: five stars +1.7 places, six
stars +2.5 (about +0.8 over five stars at three times the price), so
the six-star rules stay as they are.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
M3 - Modern world loader:
- modern/content/world: typed view of the world tables (leagues, clubs,
players, nations, league dates, training camp catalogue).
- modern/content/world_package: opens a Modern world (runtime:modern-v1, or
editor:world-v1 until data revision 1.4), reads only the engine entries and
decodes the Classic payload (import: load_classic_payload).
- modern/content/world_expand: turns the Classic start into the world: club
ids in the order of the world's leagues, per-club tables remapped (added
clubs copy an original club of their league), added players from their
attributes, 24 squad slots per league club, pools kept, layout, league
sizes and league dates installed; UTF-8 names and OBW ids in ModernState.
- Fixes found by playing the Modern data set: referee pool exhausted on days
with German and larger international leagues (Modern deviation), card
table per league club grown, slot-to-team lookup via WorldLayout, form
history stride from the layout, remaining slot literals.
- Classic UI projections used by the session read the layout instead of
executable constants (Classic values unchanged).
- modern_season_sim opens .obw worlds; MODERN_DEBUG / MODERN_WATCH tracing.
M4 - training camps:
- Catalogue in ModernState (built-in 18 fictional camps on three maps, or the
world's extended catalogue); price, booking and effect use it.
- Six-star grades: step index capped, limit +1 from pitch 5 and +1 with pitch
and fitness 6, catering 6 one more point of condition, lodging 6 one week
less injury. Classic data behaves as before. modern_camp_tests.
Verified: 190 tests pass; Classic replay regression identical to the M0
baseline; season_sim and modern_season_sim identical with Classic data; the
Modern data set plays full seasons (leagues 34/38/34 rounds).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- src/modern/game: the season modules as bmh::modern, working on ModernState
(a GameState). Data types stay shared with Classic; Modern code calls its
season functions qualified (modern::...), so a missing Modern function is a
compile error instead of a silent call into Classic.
- src/modern/app/modern_session: the session on the Modern core. It still uses
the Classic UI projections until the Modern projections exist.
- SeasonHooks: the shared match core reaches player statistics and the table
prediction of the season core that owns the game (Classic: unchanged).
- WorldLayout: team ranges, squad slots and pools of the running world. Shared
code and the Modern fork read them; Classic keeps the original numbers.
- Transfer list views moved to game/transfer_offer_view.hpp (shared).
- tools/modern/check_fork_calls.sh, modern_season_sim and world_layout_tests.
Verified: Classic tests and replay regression unchanged; season_sim and
modern_season_sim produce identical output with Classic data.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>