Fix the Qt code so it compiles wth Qt 2.1 again (Install.Qt still claims
the code works on Qt 2.0 and above). One or more of the changes made this
week broke compilation on my RedHat Linux 6.2 system. Qt 2.1 is still the
most recent Qt shipped by RedHat for this version.
time. I think it's because the modern data.base contains so many quotes that
have nothing to do with Nethack that nobody tries to use data.base to figure
out how to handle monsters any more.
to make that field unconditional, otherwise
NetHack won't compile without TEXTCOLOR defined.
Also provides at least an interim solution for the has_color()
problem that Warwick pointed out.
Lastly, Archeologists know touchstones.
- "Saved game" dialog (could be useful in other ports - plain C code used)
- "splash screen"
- compact mode cleanup for 240x320 displays
- messages-on-map overlay display
- allows online reading of Guidebook in HTML format (coming soon)
Summary of spell changes:
-- wimpiness of 'default' spell fixed by doing half damage for magic resistance
instead of 1 damage, and using half monster level instead of 1/3. It may
still need tweaking, but is much better than before.
-- 'default' spell for cleric monsters is now the wounds spell, by analogy with
wizard monsters.
-- added clerical lightning strike, flame strike, gush of water
-- all spells should now say the monster is casting a spell, and all spells
should have messages. (Side effect: monsters speeding up by other means
also give a message saying so).
-- casting undirected spells is not affected by whether the monster knows
where you are. Monsters that are attacking your displaced image, that are
several squares away, or that are peaceful can use undirected spells.
-- messages should correctly say whether the spell is undirected (a monster
was always casting at thin air or pointing at you and cursing, without checking
to see if the spell wouldn't require pointing)
-- Monsters which are attacking your displaced image, etc. use up mspec_used.
If they are casting an undirected spell, the spell still works.
-- Monsters which are not attacking can cast spells that don't attack.
-- If a monster didn't have ranged spellcasting ability (which most don't),
it would print a curse message from buzzmu() every round it was at range,
creating a useless stream of constant curse messages
I still haven't made spellcasters "smarter" in the sense of noticing whether
you have reflection, fire resistance, etc. That opens a big can of worms
because it would mean giving monsters a memory.
Known bug: the higher level a monster is, the more spells it has; since it
chooses a noncombat spell by randomly picking a spell and casting if it
happens to be noncombat, the higher level the monster is the greater the
chance of getting nothing.
Several long-awaited updates for the Macintosh port, by Dean
Luick and myself. This set affects Mac-only files.
* Update the support for MPW compilers.
* Use new system call names provided for in the latest Apple
Universal Headers.
* Tune up some of the includes for CodeWarrior.
* Define YY_NEVER_INTERACTIVE for the dungeon and level compilers.
* Remove pointless debugging code.
* Clean up some unterminated comments.
* Updated preprocessor conditionals for the MPW compilers.
* Use new system call names provided for in the latest Apple
Universal Headers.
* Tune up some of the includes for CodeWarrior.
* Define YY_NEVER_INTERACTIVE for the dungeon and level compilers.
* Remove pointless debugging code.
magic while wearing dragon scales/scale mail were being turned
into random monsters instead of into dragons.
Also
Two items from <Someone>'s list.
Files patched:
include/obj.h
src/mon.c, muse.c, worn.c, zap.c
Ken:
Fix an error in my fix for compression error messages.
No Makefile.src change should be necessary because files.c depends on hack.h,
which "depends" on wintty.h (actually ifdef USE_TRAMPOLI, but the Makefile
doesn't know that).
auto_credit flag; that's what sell_response is for).
follow-up comment from Pat:
This introduces behavior that I consider to be buggy.
You need something separate from sell_response, otherwise if
you answer (a)ll when the shopkeeper still has money you end
up selling everything for credit without being asked when he
runs out of cash in the middle of the transation. Avoiding
that is the reason for the old behavior in the first place.
cohrs