| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
| |
At the very least, give control of uniques spawning in Orc over to
unique.des.
|
| |
|
|
|
|
|
| |
random_trap_for_place defaults to NUM_TRAPS; none of the functions that
it calls currently do, but others might.
|
| |
|
| |
|
| |
|
|
|
|
| |
Requires overall weight chance in unique.des and new dummies per-branch.
|
| |
|
|
|
|
|
| |
Fixes MonPropsMarkers not being removed-on-activation when a vault is
placed via &L.
|
|
|
|
|
|
| |
dgn_reset_level(), which resets you.unique_creatures, was being called
before temp_unique_creatures (where you.unique_creatures was being
backed) was actually initialised.
|
|
|
|
|
|
|
|
|
|
|
| |
Rebranded monsters (ie, "kobold name:Durwent name_replace") who are
created with Lua and dgn.create_monster while the player is resting will
generate a "XYZ comes into view" warning message.
However, as this message is triggered by handle_seen_interrupt, which is
triggered before the additional flags are applied, it will disregard
MF_NAME_SUFFIX, MF_NAME_REPLACE, MF_NAME_ADJECTIVE, etc. Applying these
flags earlier fixes this.
|
|
|
|
|
|
|
|
|
|
| |
dur:1-6, marks the abjuration duration of a monster. sum:<str>, any of
"clone", "animate", "chaos", "miscast", "zot", "wrath" or "aid", or the
relevant spell we're mimicking; nas:<str> equivalent to the non-actor
summoner.
Not setting the non-actor summoner will mean that kills by these pseudo-
summons are unmarked as such.
|
| |
|
|
|
|
|
|
|
|
| |
Removed inclusion of travel.h from most .h files to reduce the number
of .cc files dependant on it. This involved moving the level_pos
declaration to externs.h, moving the flood_find template to it's own
header file, and moving two typedefs from travel.h to travel_defs.h
because typedefs can't be forward declared (argh).
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Placing them after connectivity checks causes issues when (larger)
vaults split the level or clobber staircases. This, in combination with
setting it to non-clobber, should solve all of these issues.
It may mean that some uniques are placed less than others (such as
Yiuf's vault), and thus may require that his vault has its weight
increased slightly to compensate.
However, it also means that it's now safe to create other, similar
vaults.
|
| |
|
|
|
|
| |
0 to <= 3 is the same as 0 to < 4, so it's actually 4 chances, not 3.
|
|
|
|
|
|
|
|
| |
Vaults can now include other vaults as a part of their definition. These
subvaults are currently included by tag only and replace glyphs in their
parent vault. See documentation for more details.
Vault:8 has been modified to use subvaults.
|
|
|
|
|
|
| |
If the map doesn't get placed, give it a second and third try at it.
Give up after these, though; this means that uniques should probably
generate at the same rate as they were at before.
|
|
|
|
|
| |
This should help with broken levels created by Yiuf's cottage. It may
mean that we won't be seeing much of Yiuf anymore.
|
|
|
|
|
|
| |
Move the set_vault_mon_list code out of dgn_register_place, as that
function is called after the random monsters are picked. Hopefully, this
also does not cause any future problems.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Both values override the default class values. Placing "rat hd:20" will
generate a rat with a hit dice of 20, while placing "rat hp:20" will set
its hp to 20.
|
|
|
|
|
|
|
| |
_fixup_interlevel_connectivity may fail otherwise.
This fixes the dungeon builder failing when it doesn't choose
an entry vault (say if entry.des is wiped).
|
| |
|
|
|
|
| |
This hopefully fixes recent disconnected level problems.
|
|
|
|
|
| |
Fixes the "add stair at pos" message appearing in normal builds
(BR 2901951).
|
|
|
|
| |
This mostly puts && and || on the proper lines, per the style guide.
|
|
|
|
|
|
| |
_builder_basic. Instead, there is a chance for a trail to be created with a shaft at the end. Shaft generation rate upped from ~2% of all traps to ~5%. Shafts have 50% chance of being created as revealed, this chance decreases as dungeon depth increases.
Signed-off-by: David Ploog <dploog@users.sourceforge.net>
|
|
|
|
|
|
| |
directly writing feature to start/end points.
Signed-off-by: David Ploog <dploog@users.sourceforge.net>
|
|
|
|
|
|
| |
This removes a couple of unnecessary includes from header files.
I may have missed adding some tiles includes.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also move dungeon_feature_by_name into terrain.cc, as it does not make
sense it being in l_dgngrd.cc.
The Lua function "set_border_fill_type(feature)" will replace the
default DNGN_ROCK_WALL fill of a portal vault with the specified
feature. This could be abused, and should probably only be limited to
floor, water and sea, different types of walls, trees and lava; there
are no sanity checks for endless traps, statues, stairs, etc.
Finally, apply this to the island Trove, giving endless water effect.
|
|
|
|
| |
None of the _plan_* functions used the 'do spotty' return values.
|
|
|
|
|
| |
Replacement for replace_area_wrapper, and a lua accessor for the
function. Returns true if the player saw the feature being changed.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
The macro DEBUG_TEMPLES can be defined to get diagnostics messages
regarding the main temple and overflow temples.
|
|
|
|
|
|
|
|
|
| |
Specialized single-altar overflow temples which also have the tag
"uniq_altar_GODNAME" can be placed randomly even if that altar has been
chosen at new-game time to be an overflow temple on a specific level,
and if it's randomly placed before being generated specifically then it
will only be placed once, rather than also being placed a second time at
the location chosen at new-game time.
|
|
|
|
|
| |
For overflow temples with a single altar, look for a map tagged with
"temple_overflow_GODNAME" and use that if available.
|
|
|
|
|
|
| |
Randomly choose the main temple map at new game time, and count the
altars in it to determine how many altars have to go into overflow
temples.
|