Commit 79407e68 (v4.3.11) changed recvclient() to keep failing after
receiving EOF from player. This was bad, because some places getting
input check player->aborted instead of recvclient() failure, and
player->aborted wasn't set on EOF. Bugs caused by this:
* comm_bomb(), ship_bomb(), plane_bomb(), land_bomb() went into an
infinite loop that eventually ate all memory.
* deli(), desi(), dist(), fly(), morale(), zdon(), att_prompt(),
ask_move_in() interpreted EOF as empty input instead of no more
input.
* cmd_sail_ship() dereferenced a null pointer.
Fix by setting player->aborted on EOF, too.
Reading input fails after EOF and while the current command is
aborted. Commands should detect that and fail. If a command neglects
to do that in a loop, the loop can become infinite. This is
especially bad after EOF, because then the client might not read
output anymore. Output gets buffered until memory runs out.
Mitigate such bugs by counting how many calls have failed in a row,
oopsing on the 256th, and sleeping one minute from the 256th on.
fly() reads the carrier, then passes it to pln_dropoff(), which writes
it back. fly() also calls pln_oneway_to_carrier_ok(), which updates
the carrier when its plane summary information is incorrect.
The old code called it between reading the carrier and passing it to
pln_dropoff(). This made pln_dropoff() wipe out the plane summary
update, and triggered a seqno mismatch oops. Broken by introduction
of pln_oneway_to_carrier_ok() in commit 1127762c, v4.2.17.
Fix by reading the carrier right before passing it to pln_dropoff().
Commit db1ac2ed (v4.3.6) introduced a separate territory for deities,
and made the territory command use it by default for deities. Deities
can still access the old default territory as territory number 0.
Selectors terr, terr0 (alias for terr), terr1, terr2, terr3 remained
unchanged, and a new selector dterr was created to let deities access
the deity territory.
Make selector terr virtual, so that it can do "the right thing", just
like the territory command: select territory 0 for mortals, and the
deity territory for deities.
Unfortunately, this requires us to switch xdump to use terr0 instead
of terr, because otherwise the deity xdump would contain the deity
territory twice, and territory 0 not at all.
Missile interdiction leaves behind used up missiles with the
PLN_LAUNCHED flag set. This can lead to a bogus warning from
pln_zap_transient_flags() on server restart.
Change pln_zap_transient_flags() to ignore dead planes.
Change retreat condition prompt to point to help. Before, it listed
conditions rather cryptically, without mentioning how to get help.
Don't provide help when encountering a bad retreat condition
character.
Fail the command when encountering a bad retreat condition character.
Before, they were dropped.
Don't fail the command when the player asks for help at the condition
code prompt.
Retreat condition help failed to explain 'c'.
The old code recognized group retreat only when the first argument was
on the command line. Didn't make a difference, because it was only
used when there were at least two arguments on the command line.
The old code relied on rflags being represented as two's complement.
When given an empty retreat path, the old code deleted the retreat
path and set the retreat flags normally. The new code deletes both.
Neither is nice; it should perhaps keep the retreat path and only set
the flags.
Commit 7ca4f412 (v4.3.12) marked planes flying a sortie with
PLN_LAUNCHED, and made pln_arm() reject planes with that flag set.
This was designed to reject escorts that were already flying as
bombers. It didn't work, because the test for PLN_LAUNCHED used a
stale copy of the plane created by pln_sel(). Fix by getting a fresh
copy.
The bug always existed, but the botched fix in commit 7ca4f412 made it
worse. Before, ac_encounter() dropped escorts that were also bombers,
so the bug merely wasted plane fuel. After, such planes were
effectively duplicated, and damage to one of them, usually the bomber,
was wiped out. Abusable.
The code wrote the swept sector after calling shp_check_one_mines().
This failed to use up the mine that hit the minesweeper, and triggered
a seqno mismatch oops.
The code wrote the minesweeper after calling shp_check_one_mines().
This used freed memory when the minesweeper got sunk there.
Broken in 4.0.17. Fix by moving both calls before
shp_check_one_mines().
ask_olist() let non-light land units board ships that can carry only
light units. If the board succeeds, the non-light unit move onto the
ship and then are stuck there.
Spies shot were only deduced from sector military; land units were
immune to losses; in fact they needn't have any military to spy.
Fix by requiring and using only sector military. Closes#758483.
Make spy() not skip sectors without civilians, military and land
units. There could be other interesting things to report there:
efficiency, gold bars, planes...
The values in these columns were computed by count_sect_units() and
count_sect_planes(), which included land units and planes in the count
that aren't shown by prunits() and prplanes(), namely own and embarked
units. Confusing. Moreover, count_sect_planes() and prunits() rolled
dice separately for spy units. This could leak the presence of spies
even when prunits() didn't show them.
All fixable, but not worth the trouble; just remove the counts.
perform_mission() needs to know whether it is targeting ships or
something else, because the rules differ: submarines interdict only
ships, land units get their damage reduced when interdicting ships,
and different news are generated.
The old code assumed it was targeting ships when the target sector was
sea. Wrong when interdicting ships in harbors, bridges and such.
This has always been broken. Except when checking a submarine's
target: there it tested argument s, which is gross, but at least it
works. That code was added in v4.0.8.
Replace the broken test by the gross hack everywhere. This fixes news
and damage from land units when ships get interdicted in non-sea
sectors.
Remove confusing columns Op-sect and Radius from table of available
missions. Add their code letters.
Fix efficiency required for artillery units to fire.
Firing damage reduction for range is a feature that was always there
and never really documented. Different ways to fire reduced damage
differently for range, or not at all. Fix that by dropping the
reduction everywhere.
The reduction happened randomly, with probability p = (d/m)^2, where d
is the distance to the target, and m is the maximum firing range. The
fire command printed "Wind deflects shells" when it happened.
The old fire command (before MULTIFIRE) either halved damage (50%
chance), or reduced it by a factor of 1-p.
MULTIFIRE's fire command halved damage. v4.0.2 reduced that loss to
10-20%.
Interdiction halved damage, but only when firing from ships.
Other ways to fire (support, return fire, interdiction from forts and
land units) did not reduce damage for range.
People can be delivered since v4.0.0 (the check that disabled that was
commented out), and cutoff shows these deliveries since v4.0.2. The
change was documented in info Empire4, but info pages weren't updated.
Do that, and remove the commented out code.
Commit 092a52f2 (v4.3.4) removed the code to estimate defense, because
the use of the estimate had been disabled since v4.0.0. This
accidentally removed the reporting of defending units, because
get_dlist() reported them when called for an estimate, and not when
called for real.
Fix by removing the unused estimate capability from get_dlist(). It
now reports defending units always.
Before s_commod() attempts to recursively supply a supply unit it
wants to use as supply source, it zaps the unit's load. When
actually_doit is false, it later restores the old load by overwriting
the change with a saved copy of the unit. That triggers a seqno
mismatch oops.
Avoid that by copying the new sequence number to the saved copy.
Change supply_commod() and try_supply_commod() not to call s_commod()
when zero units are wanted.
This isn't just for efficiency, it's also for limiting exposure to
supply bugs a bit.
