1) consolidate all core usage of `errno' in files.c;
2) give more feedback for any failure by create_levelfile or open_levelfile,
similar to what was being done for problems during level change;
3) include trickery info in paniclog (many instances of "trickery" seem to
be due to disk or quota problems rather than user misbehavior...).
The create_levelfile call in pcmain probably ought to be changed to use
error feedback, but in the meantime this should continue working.
Perhaps error() should be modified to update paniclog too, but I didn't
want to go through all its port-specific incarnations making changes.
Move get_saved_games() functionality to files.c
Use moved get_saved_games() functionality in Qt windowport.
[also some non-enabled perminv code in Qt windowport]
save process, after the save file had already been written by
forcing execution of an interrupt handler at one of the
termination prompts. This resulted in "You were not healthy
enough to survive restoration" and worse.
This clears program_state.something_worth_saving immediately
after the save has been completed.
As Warwick suggested, instead of having fixed tile
sizes as options, allow specification of the size
explicitly.
Also, at Yitzhak's suggestion, provide a hook for
overriding the port's tile file name. That name,
and the contents of the file it points to, will
be window-port specific of course.
to allow common parsing in the core, and direct access to the
results by the window port.
Notes:
o Adds a new field, wincap, to the window_procs
structure for setting bits related to the preference
features that the window port supports. This allows
run-time determination of whether a particular option
setting is applicable to the running window port. A
window-port is free to support as many, or as few,
of the available options as it wants. Ensure that
only the ones supported have their corresponding bit
set in window_proc.wincap. [see chart in
doc/window.doc for help with that.]
o The settings I stuck into wincap for each window
port are almost certainly not accurate, so each port
team should review them. You should only include
the ones that you will actually react to and make
adjustments for if the user changes that option.
Without the setting in wincap, the option won't even
show up in the 'O'ptions menu.
o preference_update() added to the window-port
interface, so that the window-port can be notified
if an option of interest (an option with its
corresponding bit set in wincap field) is
changed.
o provided a genl_preference_update() routine in
windows.c and used it for all the existing
window ports since they don't have a functional
one of their own yet.
o this messes around heavily with iflags and the options
arrays in options.c
o I hope I didn't break any port's existing code. I
tried not to. The Mac however, in particular, should
be looked at because it suffered a namespace collision
with what I was working on around fontname. It had
Mac specific font stuff in options.c. Please test
the Mac.
If you get interrupted while reading a spellbook and then
the book gets destroyed or you change levels, the object pointer
remembered for the book will be invalid and could accidentally
match one subsequently allocated to some other book. That would
result in "you continue your efforts to memorize the spell" when
starting to read that other book; it would also end up bypassing
the reading difficulty check and reuse the old book's delay counter.
I don't remember who reported this. It was quite some time
ago and I have an abandoned patch dated last March from when I
first started to fix it.
Files patched:
include/extern.h
src/save.c, shk.c, spell.c
