Release history:
Version 3.00 beta pl11 release: April 14, 1996
Version 3.00 beta pl10 release: March 11, 1996
Version 3.00 beta pl9 release: February 6, 1996
Version 3.00 beta pl8 release: May 23, 1995
Version 3.00 beta pl7 release: March 9, 1995
Version 3.00 beta pl6 release: March 6, 1995
Version 3.00 beta pl5 release: February 23, 1995
Version 3.00 beta pl4 release: September 28, 1994
Version 3.00 beta pl1-3, internal releases for beta-testers.
Version 3.00 alpha: Ran on net for testing. Code not released.
Version 2.20 release: November 17, 1993
Version 2.11 release: September 19, 1993
Version 2.10 release: September 1, 1993
Version 2.02 release: Late August 1993
Version 2.01 release: Early August 1993
Version 2.00 release: July 16, 1993 (Initial public release)
Legend
------
JE = Jeremy Elson (jelson@circlemud.org)
--- CircleMUD 3.0 patchlevel 3
8/31/94 JE Added a GET_REAL_LEVEL macro which returns the actual
level taking switched characters into account. Used
in CAN_SEE so that switched immorts can't see invis
gods.
8/31/94 JE Added restriction to interpreter.c so that immort commands
cannot be used by NPCs. Removed all the IS_NPC checks from
the individual god commands since they are now redundant.
8/31/94 JE Made newsign.c into sign.c and deleted the old DIKU sign.c.
newsign.c forks to handle its kids and is generally more
concise and robust.
8/31/94 JE Changed the type argument of affect_from_char and
affected_by_spell to sh_int (from char) to match the fact
that the 'type' member of the affected_type structure is
now a sh_int instead of a char. Spotted by Jean-Jack
Riethoven.
8/31/94 JE Fixed MIN statement using uninitialized 'water' var in
spell_create_water. Spotted by JJR.
8/31/94 JE Ken Cavness noted that MOBs were getting past guildguards.
Not surprising, since I changed the relationship between
do_move/do_simple_move/perform_move so many times. The
discipline has now changed again; MOBs calling perform_move
must call with 3rd arg of 1 instead of 0. Changed the
arg from 'following' to 'need_specials_check' so its purpose
would be clearer.
8/31/94 JE Added 'worn_on' short integer to obj_data structure to go
with worn_by char_data pointer. Indicates which eq pos
eq is being worn on. (Will be used to facilitate a cleaner
version of objectmagic.)
8/31/94 JE extract_obj now automatically unequips the obj first if it
is being used as equipment (uses worn_by and the new worn_on).
do_use no longer unequips. objectmagic didn't have to be
changed since extract_obj automatically does the work now. :)
8/31/94 JE Fixed cityguard routine with a CAN_SEE check so that blinded
guards no longer banzai people, and so that guards don't
banzai invis people etc. Spotted by Ken Cavness.
8/31/94 JE Fixed do_reply to make sure the person you're replying to
is still in the game before calling perform_tell. Note, I
think it still may be possible for your reply to go to
someone else if the recipient logs out and someone else logs
in. In particular, this will break in a big way if I ever
implement some scheme where it keeps a pool of char_data
structures for reuse.
8/31/94 JE CHA added to display when rerolling a character.
8/31/94 JE Changed TOUNGE to TONGUE throughout.
8/31/94 JE Made the '*' (ignore) case in zone reset tables (used for
invalid vnums) set last_cmd = 0 for proper resetting.
8/31/94 JE Enhanced (added?) NeXT compatibility with FOPEN_MAX, my_signal,
and O_NDELAY.
8/31/94 JE Changed boards back to listing lowest to highest since people
seemed to hate it the other way.
8/31/94 JE Added sanity checks to perform_move since king_welmar seems
to cause crashes when he's moving on occassion, and I'd rather
remove that area than try to debug it.
8/31/94 JE Changed 'SEC' to 'RL_SEC' in structs.h to avoid conflicts
with many OS's #defines of SEC
9/1/94 JE Good God, is it September already?
9/1/94 JE Changed list_skills to accumulate in a buffer instead of
using send_to_char, and use page_string. This system really
sucks. There should be a "page_to_char" command, or something.
It shouldn't even require too much conceptual difference now
that output buffers are limited!
9/1/94 JE Added a 'sort_spells()' function and changed list_skills
to use sorted spell list instead of arbitrarily ordered list.
sort_spells() called once on boot from db.c.
9/1/94 JE Finally finished making the background story paged. Now that
process_input and command_interpreter, etc., have been
rewritten, all it took was changing it from SEND_TO_Q to
page_string. Whoo hoo!
9/1/94 JE Jeff Fink suggested having perform_act do some run-time
sanity checking in order to make it more robust. Thus,
I created the CHECK_NULL macro for it in comm.c.
9/1/94 JE Made mobs hit you if you try to cast an offensive spell
at them and fail.
9/1/94 JE Completed line-by-line zone-parsing routines. Instead of
using scanf() to read from disk, Circle now uses fgets().
This allows MUCH more robustness, more accurate and sensitive
error-checking of zone files, both at boot-time and run-time.
In addition, the error reporting now gives you a specific
_line number_ in the zone file of the error. :)
9/2/94 JE Made a 'qecho' command, combined with qsay in do_qcomm
of act.comm.c.
9/2/94 JE Updated the README file.
9/2/94 JE Circle 3.00 beta patchlevel 3 released
-- Patchlevel 4
9/9/94 JE Added IS_NPC check to do_score to prevent crashes.
9/9/94 JE Added #undef MAX and MIN for NeXT compatibility.
9/19/94 JE Started work again after 10 days of real life (school's
starting up again, ya know.)
9/19/94 JE Changed lantern-handling in update_char_objects in handler.c;
now gives warning one tick before the light expires (idea
from Chris <ttl6u@fulton.seas.virginia.edu>) and correctly
decrements world[ch->in_room].light as it should have from
the very beginning.
9/19/94 JE Changed the build_house parser in house.c to allow for
a house which has more than one exit (i.e., only one of
which is access-controlled) because people seem to want
multi-room access-controlled areas.
9/19/94 JE Fixed multi-room-shared-atrium bug in house_control_house,
house.c (I haven't heard of this bug ever manifesting itself
but looking at the code I realized the error.)
9/19/94 JE Changed 'long' to 'time_t' where appropriate.
9/19/94 JE Added graceful handling of missing playerfile in db.c
-- Patchlevel 5
10/4/94 JE ban.c:98: long int format, time_t arg, fixed with cast
10/4/94 JE Made default action of hcontrol to be a help screen;
added 'show' subcommand.
10/4/94 JE Added reni's NeXT function prototypes to utils.h
10/5/94 JE Fixed !IS_NPC in do_score (act.informative.c)
10/9/94 JE Added charisma-to-25 code to do_restore (act.wizard.c)
-- Sanity break
2/20/95 JE Added max length check to set_title in limits.c -- all
title setting should go through this function
2/22/95 JE Added check for null name in 'show player'
2/22/95 JE included unistd.h in house.c, boards.c and objsave.c
for unlink prototype
2/22/95 JE Low-lev imms can no longer see invstart people above their
level entering game
2/22/95 JE Changed 'color' and 'syslog' to show current levels if typed
without arguments, instead of help text.
2/22/95 JE fixed castle by adding terminator to end of move string
reported by Ed Mackey (emackey@BIX.com)
2/22/95 JE If you try to return to your original body, and someone else
has connected into it, disconnect them. act.wizard.c,
do_return().
reported by Aaron Mills
2/22/95 JE interpreter.c - changed the code to sense if a switched
persona of yourself is connected when you log in. Should
make switch more stable.
reported by Aaron Mills
2/22/95 JE new timediff in comm.c to work on systems for which
tv_sec and tv_usec are unsigned
reported by Steven Serocki <ss@sirocco.cup.hp.com>
2/22/95 JE change RLIMIT code for systems that don't have RLIMIT_INFINITY
reported by Steven Serocki <ss@sirocco.cup.hp.com>
2/22/95 JE changed title_type[4][35] to [NUM_CLASSES][LVL_IMPL + 1]
throughout
reported by Darel Cullen
2/22/95 JE changed "his arm" to "$s arm" in act.obj.c
reported by Chris Epler <cepler@ucsee.eecs.berkeley.edu>
2/22/95 JE act.other.c, do_quit() -- kill off all sockets connected to
the same player as the one who is trying to quit. Helps to
maintain sanity as well as prevent duping.
duping reported by Chris Epler <cepler@ucsee.eecs.berkeley.edu>
2/22/95 JE installed new version of Jeff Fink's shop.c code
2/22/95 JE Patched bug in Jeff Fink's shop code reported by him.
Patch sent by Jeff Fink
2/22/95 JE created player_specials structure for new players coming in
after having been deleted. Yet another example of a bug that
could have been avoided completely if we'd been using C++.
Reported by Jeff Fink
2/22/95 JE Added code to support 'R' command in zone resets (remove
obj from room). Code sent by Jeff Fink.
2/22/95 JE Changed plain 'stat' to make better assumptions about what
you're trying to stat.
Code sent by Jeff Fink
2/22/95 JE Added go_gen_door code written by Jeff Fink to eliminate
duplication of checks, etc. for open/close/lock/unlock/pick
2/22/95 JE Fixed shop code -- changed keeper = 0 to keeper = me
2/22/95 JE Added magic_alter_objs code submitted by Jeff Fink
2/22/95 JE Added break statement to mag_damage, energy drain
Spotted by Jeff Fink
2/22/95 JE Added ascii flag system to db.c for world, obj, mob files
2/22/95 JE Brought in the new 3.0 world files from Furry
2/23/95 JE Created 'random.c' as Circle's new, portable random number
generator.
2/23/95 JE Added 'POSIX_NONBLOCK_BROKEN' option to work around OSs
such as AIX which don't implement POSIX nonblocking I/O
correctly, causing the MUD to hang in places such as the
Password: prompt.
2/23/95 JE Added a case for sgi (IRIX) in utils.h OS-specific header
section for cleaner compiles
-- Patchlevel 6
2/24/95 JE Fixed asciiflag_conv in db.c for uppercase flags (typo,
'a' instead of 'A').
2/24/95 JE Added max_bad_pws to config.c.
2/24/95 JE (logged retroactively) Added max_filesize to config.c to
set max limit on size of bug, typo and idea files to prevent
bombing
2/24/95 JE Moved 'pfilepos' from descriptor_data to char_data where is
belongs. More logical, plus fixes the switch-into-a-PC-and
save bug reported on rec.games.mud.diku by
shruew@amiserv.xnet.com (Johnny Mnemonic).
2/24/95 JE Similarly moved password from descriptor to character.
2/24/95 JE Changed nonblock to OR-in O_NONBLOCK to the current flag
bitvector of descriptors instead of setting directly.
2/24/95 JE Changed 'last_tell' (used by reply) to use player ID number
instead of a pointer to the character. Much safer because
I don't have to worry about pointer reuse, plus it'll still
work if the person you're replying to logs out and back in
again.
2/24/95 JE Changed distribution-default of nameserver_is_slow to NO.
2/24/95 JE Fixed src/util/showplay.c
2/24/95 JE Added MAG_ALTER_OBJS to SPELL_INVISIBILITY in spell_parser.c
2/25/95 JE Fixed PC/NPC checks in act.wizard.c, do_set.
Reported by Eric Green and Stefan Wasilewski
2/25/95 JE Fixed carried-by/worn-by reporting of objects
act.wizard.c, do_stat_object.
Reported by Eric Green and Stefan Wasilewski
2/25/95 JE Deleted 'equipment status' from do_stat_obj in act.wizard.c
--didn't seem particularly useful and took up valuable screen
space
2/25/95 JE do_wiznet, act.wizard.c, wiznet chops off first letter of
message when levels are used. Reported by Eric Green and
Stefan Wasilewski
2/25/95 JE do_ungroup, act.other.c, fixed so that disbanding a group
doesn't un-charm charmed mobs following you.
Reported by Eric Green and Stefan Wasilewski
2/25/95 JE magic.c, changed so you can't sanct sancted mobs
Reported by Eric Green and Stefan Wasilewski
2/25/95 JE Brought in fixed world files from furry and corresponding
castle.c and spec_assign.c for new specproc assignments
2/26/95 JE Removed minor output display but on object statting, do_stat,
act.wizard.c Spotted by Eric Green.
2/26/95 JE Added max string length checks in perform_subst(), comm.c
Reported by Eric Green and Stefan Wasilewski. Also added
comments to the function.
2/26/95 JE Similar length checks in perform_complex_alias, interpreter.c
2/26/95 JE Made buffers MAX_INPUT_LENGTH size to handle very large output;
tmp_name in interpreter.c, *_search in do_who and do_users
in act.informative.c
2/26/95 JE Rewrote most of do_group, act.other.c
2/26/95 JE do_ungroup, act.other.c - disbanding a group also ungroups
yourself.
2/27/95 JE Added olc.c and olc.h.
2/28/95 JE Added bank card specproc to spec_assign.c
2/28/95 JE Updated FAQ and README-BETA
3/3/95 JE Patchlevel 6 released
--- Patchlevel 7
3/4/95 JE Added PFILEPOS assignment in interpreter.c so deleted
characters are not replicated in plrfile. (near "Did I get
that right, X?")
3/4/95- JE Extensive rewrite of documentation
3/7/95
3/8/95 JE Changed qecho to LVL_IMMORT (interpreter.c)
3/8/95 JE Created GET_PFILEPOS macro and used it throughout.
3/8/95 JE Changed extern declaration of thaco in fight.c from [4][35]
to [NUM_CLASSES][LVL_IMPL+1]. Spotted by Marcin Zaranski.
3/8/95 JE Problems reported with stop_fighting called when char not
fighting. Most probably the call of stop_fighting() in
hit() when ch->in_room != vict->in_room. Added check for
FIGHTING(ch) before calling stop_fighting. Also made
stop_fighting tolerant of being called at the wrong time.
Problem spotted by Marcin Zaranski.
3/8/95 JE Also changed stop_fighting to use REMOVE_FROM_LIST macro.
3/9/95 JE Implemented ROOM_TUNNEL.
3/9/95 JE Fixed a bug causing the MUD to crash under pathological
input-overflow conditions (such as someone holding down a
key for 60 seconds.)
Reported by Carl Tashian <TASHIACM@ctrvax.Vanderbilt.Edu>
3/9/95 JE "set class" output corrected.
Spotted by Ryan Baker <ed209@free.org>.
3/9/95 JE Added Jeff Fink's do_gen_door patch for picking. Also
initialized obj and vict to NULL, might have been causing
double messages.
3/9/95 JE Fixed find_door to give a better message with a null arg.
3/9/95 JE Brought in new version of world files from Furry
Deleted that stupid '#99999' from the end of all the world
files since the new parser doesn't need them.
3/9/95 JE Changed perform_group so "group all" doesn't group people
you can't see.
--- Patchlevel 8
5/5/95 JE Changed my email address to jelson@jhu.edu throughout.
5/5/95 JE Implemented AFF_WATERWALK flag.
5/5/95 JE Implemented MOB_AWARE flag (mobs that cannot be backstabbed).
5/5/95 JE Implemented MOB_NOBASH flag.
5/5/95 JE db.c, added support for type E (Enhanced) mobs
5/5/95 JE Implemented ROOM_PEACEFUL flag (rooms in which fighting
is not permitted.)
5/5/95 JE Implemented ROOM_NOTRACK flag.
5/5/95 JE Implemented AFF_NOTRACK.
5/5/95 JE Added some sanity checking to obj_to_obj.
5/5/95 JE Neatened up the saving throw table in magic.c.
5/10/95 JE Implemented bless and curse spells for objects.
Code submitted by Jeff Fink.
5/10/95 JE Implemented poisioned objects and remove-poison for objs.
Code submitted by Jeff Fink.
5/10/95 JE Implemented MOB_NOCHARM flag.
5/17/95 JE Completely new building.doc world builders' documentation.
5/20/95 JE Imported latest version of Furry's world files
5/21/95 JE handler.c, extract_char()... if a char is extracted and
has no descriptor associated with it, free the char.
Prevents a memory leak if players get purged.
Suggested by Eric Green.
5/21/95 JE Set mob's default charisma to 11.
Spotted by Jeff Fink.
5/21/95 JE Changed obj num loaded for 'create food' to 10, waybread.
5/21/95 JE Removed strange code from db.c, reset_zone, 'O' command,
that seemed to prevent two objs of the same vnum from being
loaded in the same room.
Suggested by Jeff Fink.
5/21/95 JE handler.c, affect_join() - fixed bug preventing spells
with multiple effects from coexisting.
First reported by Thomas Katzlberger.
5/21/95 JE boards.c - changed board so it will respond to 'examine'
as well as 'look'
Suggested by Chris Epler.
5/21/95 JE Initialized GET_LAST_TELL to NOBODY.
Spotted by Eric Green and Stefan Wasilewski
5/21/95 JE Changed around wiz-inivisibility for better compatibility with
the 'visible' command, and to give better messages.
Suggested by Chris Epler.
5/21/95 JE Fixed mag_groups so caster is only hit once per cast.
Spotted by dodger@WPI.EDU (DoDGeR)
5/21/95 JE Fixed the bug causing crashes on some systems if an
incorrect class is entered when a new character is created.
For the record I think this was happening on architectures
on which the 'char' type is defined as being unsigned.
5/21/95 JE Made is_abbrev asymmetrical as it should be; fixes
"look ingerwal" as being interpreted as "look in"
Reported by Mike Higuchi and others.
5/21/95 JE Changed spells_to_learn from byte to int; used one of the
spares in the playerfile so that people already running
MUDs with 3.0 won't get screwed (grrrrr....).
Spotted by Marcin
5/21/95 JE spell_parser.c, say_spell() - target of spell now only
sees caster casting if they're in the room together.
Spotted by Marcin
5/21/95 JE Also changed say_spell so that object messages will be
correctly reported if target obj is in caster's inventory
as well as in same room.
5/21/95 JE Changed mag_alter_objs to give the room a message as well
as the caster.
5/21/95 JE handler.c, char_from_room(), changed so that if char is
fighting when removed from room, stop_fighting is called
immediately to stop the fight.
Problem found by Marcin
5/21/95 JE Corrected errors with RUSAGE section in comm.c.
Spotted by Thomas Knight
5/21/95 JE Fixed the ancient "pour canteen out" bug -- the problem
was that any drink container whose weight was less than
its quantity could go negative since the MUD assumes that
all drinks weigh one pound per unit of drink. parse_room,
db.c, now checks drink containers and fountains to make
sure that weight is greater than quantity.
5/21/95 JE Fixed the cleaning woman in Welmar's castle so that she
only picks up trash.
Fix submitted by Christopher Dickey
5/21/95 JE Fixed up show_string so that the display works properly;
fix submitted by Michael Buselli
5/21/95 JE Player killers' damage no longer set to 0 if they're inflicting
damage on themselves.
Reported by Julian Fong.
5/21/95 JE Fixed do_gen_door to correctly sense when no arg is given.
Problem spotted by Eric Green
5/21/95 JE *** WAIT_STATE set for all uses of magic items ***
Suggested by Eric Green
5/21/95 JE Found irrefutable clues & figured out who shot Mr. Burns.
5/21/95 JE Bug in mobact.c prevented alignment-aggressive monsters from
working if they were set not in conjunction with the normal
AGGRESSIVE bit.
Problem spotted by Chris Epler
5/21/95 JE Problem with HELPER mobs getting extra hits fixed.
Problem spotted by Chris Epler
5/22/95 JE Created the CircleMUD License. Changed Makefile so that
the license is displayed the first time the MUD is compiled.
Sorry if this is a dickish thing to do, but it's a Derek
Snyder countermeasure.
5/22/95 JE Added logging for whole-world-resets using zreset command.
5/22/95 JE fight.c, hit() -- Changed declaration of all intermediate
vars used to calculate hitroll and damroll to be int to
avoid problems with unsigned bytes on certain architectures.
Same change to 'percent' vars in act.offensive.c.
Suggested by Al Thompson
5/22/95 JE Rewrote mag_affects in magic.c to be more readable and easier
to use, as well as correcting some old bugs (i.e., being able
to un-affect mobs by re-affecting them and then waiting for
the affect to fade.)
5/22/95 JE fight.c, damage() - WIMPY mobs do not flee if they are
damaging themselves.
5/22/95 JE comm.c, close_socket(). When a player loses link, all
players with same idnum are also disconnected. Fixes a
duping bug.
Found by Clay Hardin
5/22/95 JE Added "NOEFFECT" char constant to config.c.
5/22/95 JE Fixed bug in zone-reset 'R' command where second arg was
being resolved as a room vnum instead of an obj vnum.
Fix sent in by BUG.
5/22/95 JE Added two prototypes to comm.h so that comm.c will compile
cleanly under ultrix.
5/22/95 JE spell_parser.c, cast_spell() - prevents ungrouped chars from
casting MAG_GROUPS classes of spells (returns error msg and
does not take mana).
5/22/95 JE Added the GET_EQ macro for character equipment and SECT
macro for sector types; changed code throughout to use
the new macros.
5/22/95 JE act.movement.c, has_boat - added has_boat function so that
unwearable boats in inventory and boats worn will work as
boats.
5/22/95 JE Added sun386 (SunOS 4.0.x) compatibility by defining
sigaction as signal for it.
Patch sent by Tim Aldric.
5/22/95 JE Implemented detect poison as a manual spell.
5/22/95 JE Fixed up mag_areas (generally) and earthquake (specifically).
5/22/95 JE Changed 'invis' command so it will not work on mobs.
Bug spotted by v932459@si.hhs.nl.
5/23/95 JE Fixed get_char_room_vis so that you can no longer access
PC's throughout the world with the '.' syntax. (Oops..)
5/23/95 JE Changed GET_POS==FIGHTING check to if(FIGHTING) in
cityguard specproc, spec_procs.c
Bug spotted & fix submitted by Jason Fischer
5/23/95 JE fight.c, perform_violence() - Implemented wait states for
mobs, so that mobs who are bashed etc. are actually taken out
of the violence for some number of rounds (depending on the
skill), like it is supposed to be.
5/23/95 JE limits.c, gain_exp*(), changed level message so you aren't
spammed if you gain more than one level at once.
5/23/95 JE Added bash as a level 12 warrior skill.
If successful, your opponent loses 1 round.
If unsuccessful, you lose 2 rounds of combat.
You cannot try to bash again for 2 rounds.
5/23/95 JE Fully implemented the bash-class of skills; i.e., for skills
where it tells you you fall on your ass and have to stand up
again, you actually lose the rounds of combat while you're
still sitting.
5/23/95 JE Brought in new lib/misc/messages file from Furry with some
additional attack types.
5/23/95 JE Wrote the E-Spec parser and brought in Furry's world files
that use E-Specs.
5/23/95 JE Fixed all the utilities, including autowiz which now works
again.
5/23/95 JE Took out the scheck utility and replaced it with the -c
option on the server (Check syntax only - do not run game).
Another fantastic idea from Gekke Eekhoorn, the source of many
fantastic ideas.
5/23/95 JE Deleted some extraneous lines from the Makefile (rules for
compiling utiltities that no longer exist).
Found by Jared McDonald
5/23/95 JE Released patchlevel 8.
--- Patchlevel 9
(No dates on this patchlevel... sorry!)
-- Made Circle self-configuring using the GNU 'autoconf' package; created
new header files conf.h and sysdep.h. Most standard library files are
now conditionally included from sysdep.h and have been removed from the
.c files. -lsocket/-lnsl libraries are automatically used if needed.
Circle's built-in 'random' function is used only if there is no random()
on the system.. many other autoconfigured aspects as well.
-- Rewrote much of the code for the login/load sequence to prevent duping.
Duping should hopefully be impossible now. Fixed the dumb bug that allowed
people to cut off players by logging in and out without a password.
-- Added 20 second limit to entering a password before you're cut off.
-- Fixed the braindead way LOADROOM works. Syntax is now simply "set ras
loadroom <vnum>" or "set ras loadroom off". Loadroom now properly set
when someone quits out of their house, so they go back there by default.
-- Truncate message board titles to 80 characters or less - fixes a crash
bug.
-- Changed tunnels so that they are "full" only if filled with a PC, not
just any char.
-- Redid the code for setting the minimum levels classes need to be before
they can use spells and skills, and moved it to class.c. It's now *much*
easier to add new classes--- stupidly, under the old system, every time
you added a new class all of the existing spello() calls had to be
changed! See class.c to see how spell-level setting is done now.
-- Fixed the bug which popped up in systems such as FreeBSD which incorrectly
reported the number of available file descriptors if the number was
unlimited (RLIM_INFINITY).
-- Fixed a bug causing crashes if text files (motd/imotd/news/etc) are empty.
-- FINALLY fixed the dumb bug causing the first char (implementor) to log
in to come in with all stats of 0.
-- Changed magic.c so the number of affects a spell can have is not
arbitrarily limited to 2.
-- Immortals can walk past guildguards.
-- Fixed up some spells (create water works now).
-- Added 'show houses' for Mr. Christopher Epler. :)
-- Fixed a small error in one of the comments in boards.c describing how
to add a new board.
Found by Stefan Rensing
******** Patchlevel 10 ***************************************************
-- Fixed all the (dumb) definitions of str_app, int_app, etc. -- these
used to be arrays of 36 because a long time ago I stupidly thought those
arrays had to be extended to the number of levels on the MUD. They have
been correctly changed to reflect the possible attribute values of
characters (3-18 for PCs, 3-25 for NPCs and Gods).
-- Added Michael Buselli's patches for an enhanced "more" pager that lets
you go back, forward, refresh, etc. Nice work, Michael.
-- Made Circle compile under HP/UX with the help of Ben Denckla (required
an #ifdef for RLIM_INFINITY). Made some other fixes to get_max_descs()
in comm.c as well. I hate this function.
-- Removed the REUSEPORT setsockopt() call from comm.c when opening the
mother connection. This never made any sense in the first place since
REUSEPORT is a UDP and not a TCP thing, plus it broke Circle under IRIX
5.3+.
-- Incorporated David Carver's OS/2 patches. Circle should now compile
under OS/2 more or less out of the box (instructions included).
-- Added error checking to number() in case its called with the arguments
backwards.
-- Added the skill_name() function to safely resolve spell numbers to
spell or skill names. Now used to stat potions, staves, etc.
-- Fixed the but in the usage of is_abbrev in find_skill_num that prevented
spell names from being abbreviated. Reported by numerous people.
-- Fixed a bug which prevented mag_materials() from working at all (although
it isn't used in standard Circle code, anyway.)
-- Changed a str_cmp to isname in the FIND_OBJ_EQUIP case of generic_find()
in handler.c for better location of objects.
Sent by one of our best bugfinding teams, Eric Green and Stefan Wasilewski,
KEEP IT UP, GUYS! :-)
-- Fixed the operation of peaceful rooms so that bash, etc, also don't work.
-- Fixed a dumb bug where aliases (which use player_specials) were freed
*after* player_specials were freed. Reported by numerous people.
-- Added values for movement loss for flying and underwater sector types.
-- Bug fixed in do_set which made the PC/NPC checks not work.
-- Changed backstab multipliers from a table to a formula to make adding
new levels easier. (Adding new levels is now trivial; you just have to
change the #defines in structs.h for the LVL_xx constants, and then
define the titles and exp required for each level in constants.c).
-- do_stat_obj (act.wizard.c) updated to reflect building.doc (very old
bugs lurking in here) -- stat display changed to reflect real meanings
of values[0] - values[3] of objects.
-- do_steal changed to simplify the no-psteal code (psteal simply always
fails on muds where it is not allowed).
-- You can't wake up someone who's dying. (do_wake)
Patch sent by Lincoln Chan <linc@CSUA.Berkeley.EDU>
-- Message seen when a door opens from the other side fixed. (do_open)
Patch sent by Lincoln Chan <linc@CSUA.Berkeley.EDU>
-- All utilities ported and fixed.
-- Some changes to Makefile template Makefile.in
-- Help system completely rewritten to use split help files and an
'index' file just like the world files; help files are loaded into
memory entirely instead of keeping file positions and an open file.
New copy of help files imported from Chris Epler (cepler@cambot.res.jhu.edu)
-- Rewrote one_word() in interpreter.c (used to be modify.c) as part of the
help system rewrite.
-- New world/ hierarchy from Furry (furry@cambot.res.jhu.edu)
-- New 'messages' file from Furry (furry@cambot.res.jhu.edu)
-- New 'configure' script which should detect nsl and socket libraries
more reliably.
-- All scripts rewritten (autorun, automaint, purgeobjs, etc.) to use sh
instead of csh. They've been fixed, too (most of them were broken because
they had not been modified since the location of certain files changed).
Autorun rewrite by Stuart Lamble
-- New FAQ by Furry (furry@cambot.res.jhu.edu)
******** Patchlevel 11 ***************************************************
-- Patches to make Circle compile under MSCV++ 4.0 under Windows 95:
Created a 'conf.h.win' to act as Windows' conf.h. Created a new
constant, CIRCLE_WINDOWS, to use for #ifdef's. Changed OS2's constant
from OS2 to CIRCLE_OS2 for consistency.
Changes to some of the types throughout the code (e.g. changing some
int's to size_t's.
Changed the type of all sockets from int to socket_t; sysdep.h
typedef's int as being socket_t under UNIX and SOCKET to socket_t
under Windows.
Changes to some of the code in comm.c (with ifdef's) to account for
some of the differences between WinSock and Berkeley Sockets.
Wrote a "fake" gettimeofday() function which is an interface to the
Win32 GetTickCount() and a WinSock-compatible nonblock().
Renamed 'act.obj.c' to 'act.item.c' to defeat an apparent bug in
MSVC++ that refuses to properly object files that have the word
"obj" in their name (i.e. act.obj.obj).
Changed the boot-up procedure so that the max player calculation is
done before socket initialization (this is done so that the max can
be reduced from the MAX_PLAYERS value in comm.c if WSAStartup() tells
us we have less than that number of max sockets available.
Created a CLOSE_SOCKET macro: under UNIX, this is #defined as
close(); under Windows, it's defined as closesocket.
#defined log(x) as basic_mud_log(x) to prevent namespace collision
with the mathematical log() function.
-- Various fixes to the 'configure' script:
The '-lsocket' library should hopefully be detected more reliably
(by detecting -lnsl first; some systems need -lnsl before -lsocket
will link successfully).
The -cckr option is given to the compiler if we are not using gcc and
it is determined that your system's cc can take the -cckr argument.
-fno-strict-prototypes given if we're using gcc (this should hopefully
alleviate the continuing headache of getting function prototypes to
work).
Include a prototype for random and srandom if it's determined that
we're using Circle's random/srandom functions and not libc's.
Check for the presence of arpa/telnet.h, and use our own local copy
(telnet.h is now included with the Circle distribution) if it does
not exist.
Check for sys/types.h before including it.
Configure recreated with GNU autoconf v2.9 (previously was 2.7).
-- Made some fixes to the autorun script: most notably, it was not passing
command-line parameters to the MUD, meaning that you had to change the
port in config.c in order for it to be changed if you were using autorun.
This means that port-number changing now works the way it was documented
(autorun setting takes precedence over config.c).
-- Made some changes to conf.h.os2 to reflect the new configure script and
the changing of the constant from OS2 to CIRCLE_OS2.
-- Moved heartbeat functions out of game_loop and into their own function
which is called by game_loop (seems more logical this way to me).
-- game_loop, comm.c: The code which controls the tick-timing was rewritten;
the timer is now much more accurate (does not drift with time), and
gracefully handles making up for lost ticks if the MUD is suspended for
more than a single tick-length (e.g. if the scheduler keeps it waiting
for a long time under heavy load). Added a 'timeadd' function to add
two timevals for use with the calculations.
-- Added a timeout to the "What is your name?" prompt; should prevent denial
of service attacks whereby people use up all available descriptors by
repeatedly telnetting into the game and staying idle. (Though I have not
received reports of this happening, I'm sure someone has thought of it.)
-- Fixed various typos in the code, help files, and libraries.
-- do_advance, act.wizard.c: Does not allow someone to be advanced to the
level they already are.
-- invalid_name, ban.c: Changed the buffer size from MAX_NAME_LENGTH to
MAX_INPUT_LENGTH; was causing crashes when very long names were given.
-- Fixed look_in_obj, act.informative.c, to prevent crashes when looking into
fountains or cups with invalid values.
-- Changed the semantics of top_of_helpt so that it points to the top element,
not the number of elements, in the help table (done to be consistent with
the other top_of_xxx variables). Fixed do_help in act.informative.c to
prevent crashes if "help zzzzzz" is typed.
-- Reshuffled the position of alias freeing in free_char yet again (I think
I've finally gotten this one right).
-- handler.c, affect_from_char(): Changed the list traversal over to the
classic method used when traversing a list and deleting items as we go
(i.e. keeping around a temp variable with hjp->next). (This bug was found
courtesy of Win95's completely unforgiving memory manager.. unlike most
UNIX memory managers, it causes crashes if you try to use memory that
was just freed.... Win95 is almost as good as Electric Fence! :-))
-- handler.c, extract_char(): Changed "if (ch->desc != NULL)" to
"if (!freed && ch->desc != NULL)" because ch was potentially freed in the
previous statement. (This bug was also found by Win 95...)
-- limits.c, point_update(): gain_condition called first rather than last;
causes crashes when gain_condition() is called on a character that has just
bled to death (i.e. killed by point_update()). (This bug was also found
by Win 95...)
-- Some similar cases of structures being used after they were freed in
shop.c fixed (found by Win95...)
-- Fixed the pet shops so pets are no longer priced at 0 coins; pet prices
are now 300 coins times the level of the pet.
-- spell_parser.c, get_skillname(): If the skill is -1, return UNUSED instead
of UNDEFINED (makes stat and scroll of identify output look better).
-- Added more sanity checks to spell_parser.c (spell_level() and cast_spell())
to prevent crashes if scrolls are created with wild spell numbers.
-- spell_parser.c, spello() and unused_spell(): Changed to use a loop
instead of enumerating all classes; should make adding new classes easier.
-- spells.h: Changed types of mana_{min,max,change} to int to correctly handle
spells which take large mana requirements.
-- handler.c, generic_find(): Silly bug introduced in pl10 fixed whereby
you can't look in objects if you are equipped.
-- mail.h: Changed a '1' to a 'sizeof(char)' in the MudMail system (caused
assertion failures on architectures where sizeof(char) != 1)
-- utils.c, sprinttype(): Cleaned up the code. Looked silly before.
-- Fixes to autowiz.c: Added an fflush() at the end so that wizlists are
correctly updated. Also removed some old bzero() calls that were still
lurking from the old days (modern SYSV systems don't like bzero() at all.)
Took out the SO_REUSEPORT socket option.
-- Fixed wld2html.c: The parser code was quite old and could not handle
the 3.0 world files with ASCII bitvectors. wld2html now works correctly.
-- Changed "++col" to "col++" in Michael Buselli's pager code (patch sent by
him).
-- Brought in the new world files and 'messages' file from Furry (scrolls
of identify and the 'harm' spell now work) and new help files from
Chris Epler.
-- Took out the very old bug, typo, and idea files from the standard
distribution.
******** Patchlevel 12 ***************************************************
-- I changed the logic of prompt generation so that prompts are generated
as plain output, the same as all other output (e.g. using send_to_char()),
specifically so that both the regular output and the player's new prompt
can be written to the operating system using a single write() system call.
In many cases, this cuts in half the number of TCP packets generated
and received! This is a big win for slow machines and machines with
slow Internet links (e.g. a modem)!
To illustrate this point, I'll show typical TCP exchanges between MUD
clients and servers under pl11 vs. pl12. These exchanges were all
recorded using Van Jacobson's tcpdump tool, which can be downloaded from
ftp.ee.lbl.gov. The window advertisements have been removed from the
output to make it more readable.
Here's a typical packet exchange between a MUD client and server under
pl11 when a player enters a command (in this case, "look"):
16:17:35.287187 client.32877 > mudserver.4000: P 41:47(6) ack 2579
16:17:35.333721 mudserver.4000 > client.32877: . ack 47
16:17:35.334197 mudserver.4000 > client.32877: P 2579:3132(553) ack 47
16:17:35.376089 client.32877 > mudserver.4000: . ack 3132 win 8760 (DF)
16:17:35.376162 mudserver.4000 > client.32877: P 3132:3134(2) ack 47
16:17:35.426039 client.32877 > mudserver.4000: . ack 3134
The first packet is the 6-byte command, "look\r\n", sent from client to
server. The second packet is the server's TCP acknowledgement (ACK) of
those 6 bytes. The third packet is the 553-byte output of the command
"look" sent from server to client. The fourth packet is the client's TCP
ACK of those 553 bytes. The fifth and sixth packets are the 2-byte prompt
sent from client to server followed by the client's TCP ACK.
Note that the second and third packets are often compressed into a single
packet, if the MUD happens to service that player before the server TCP
stack's delayed ACK timer expires.
This is a much bigger waste of resources than you might expect; keep in
mind that each of these packets usually carry at least 40 bytes worth of
overhead, and IP packet processing is typically takes a lot of CPU time.
Here's the same exchange under pl12:
16:20:15.445465 client.32878 > mudserver.4000: P 29:35(6) ack 2016
16:20:15.492610 mudserver.4000 > client.32878: . ack 35
16:20:15.504223 mudserver.4000 > client.32878: P 2016:2571(555) ack 35
16:20:15.549806 client.32878 > mudserver.4000: . ack 2571
The same exchange now requires only 4 packets instead of 6. (Or, if we
service the client before the delayed ACK timer expires, we could do it in
3 packets.) Note that the third packet is now both the 553-byte output of
"look" *and* the 2-byte prompt. This causes only a single TCP ACK from
the client, instead of two ACKs as in the previous example.
Here's a typical round of combat under pl11:
16:19:02.673794 mudserver.4000 > client.32875: P 4976:5067(91) ack 120
16:19:02.717357 client.32875 > mudserver.4000: . ack 5067
16:19:02.717409 mudserver.4000 > client.32875: P 5067:5069(2) ack 120
16:19:02.767418 client.32875 > mudserver.4000: . ack 5069
The first packet is the combat message (e.g., "The beastly fido hits you
very hard.") The second packet is the client's TCP ACK of those bytes.
The third and fourth packets are the 2-byte prompt and its ACK.
Under pl12:
16:21:33.572363 mudserver.4000 > client.32876: P 8418:8505(87) ack 131
16:21:33.620336 client.32876 > mudserver.4000: . ack 8505
Each round of combat is reported using only a single TCP packet, which in
turn elicits only a single ACK from the client. The exchange requires
only 2 packets instead of 4. If you have a large MUD, where many players
might simultaneously be in combat, the savings can be very significant.
MUDs with slow links or slow CPUs will probably benefit the most from this
change.
-- comm.c, and structs.h, various functions: replaced the "prompt_mode"
variable with the less obfuscated "has_prompt", which has the opposite
meaning.
-- mobact.c:mobile_activity() - What a terrible bug!! Mob spec-procs
were never called if the mob was fighting or asleep. The mayor would
never wake up, but more importantly, mobs with special specprocs for
fighting (e.g. mobs that use magic during battle) were never getting
called. The check for FIGHTING and !AWAKE has been moved to after
the specproc calls but before the mob default actions.
Found by realm@styx.ph.msstate.edu.
-- act.offensive.c:do_flee() - Added check to make sure you're at least in
POS_FIGHTING (not injured) to flee; necessary because wimpy calls
do_flee() directly and doesn't go through the interpreter.
-- spells.c:spell_summon() - Changed so that a MOB_NOSUMMON flag prevents
mobs from being summoned instead of MON_NOCHARM. Found by Daniel
Burke.
-- magic.c:mag_affects() - Changed the protection against re-affecting
affected mobs to check all possible spell affects instead of only the
first two (an oversight from when mag_affects was changed from 2
affects to an arbitrary number). Found by Daniel Burke.
-- shop.c, shop.h: Moved 3 arrays from shop.h to shop.c so that shop.h can be
multiply included if necessary. Suggested by Anil Mahajan.
-- act.wizard.c:do_advance(): Changed to show the diminishment message to
the victim instead of the imp. Found by glennmc@ccsi.com.
-- act.wizard.c:do_stat_character(): Added error checking to aff->type
array indexing to prevent crashes. Suggested by Anil Mahajan.
-- act.offensive.c:check_killer(): Using 'murder' automatically sets your
KILLER flag (if pk is not allowed), even if you miss the first hit.
Suggested by Alex Fletcher <furry@circlemud.org>.
-- magic.c:mag_points(): Added update_pos to the end to correctly upgrade
(or downgrade) the position of your victim after healing. Found by
Johan Dustler.
-- act.wizard.c:do_set(): Broke this monster into two functions, which
makes it easier to read and plugs the 'set file' memory leak.
-- spells.c:create_water(): Only remove the name from the drink container
if it wasn't previously empty. Prevents permanent keywords from being
removed. Found by Eric Green.
-- fight.c:group_gain(): Made the max possible exp gain when killing
a player equal to 2/3 the max_exp_loss. This prevents illegal exp
creation on pk muds. Found by Apua Cesar de Miranda Paquola
<apua@dcc.unicamp.br> and Dirani (spharion@dcc.unicamp.br).
-- fight.c:damage(): Cleaned up code, fixed similar bug to prevent
exp gain by bleeding to death.
CHANGED: You no longer gain xp when you're damaging yourself.
-- shop.h; shop.c:trade_with(): Clarified the error message given to
people attempting to sell a worthless item to a shop. Submitted by
Chris Epler <cepler@circlemud.org>.
-- act.wizard.c, do_dc(): If you try to disconnect a God you can't see,
you are given "No such connection" to hide the presence of the God.
Suggested by Fred Herbert <herb8590@swampfox.fmarion.edu>.
-- act.wizard.c, do_gecho(): Added delete_doubledollar() to prevent
$$'s on gechos. Suggested by Jestyr Saaven <jestyr7@the-hermes.net>.
-- act.wizard.c, do_show(): Changed 'show godrooms' to traverse the world
print rooms marked GODROOM instead of printing rooms in a particular zone
number. Suggested by Jestyr Saaven <jestyr7@the-hermes.net>.
-- act.wizard.c, do_stat_room(): Changed to display the zone's vnum, not
rnum. Found by Jay Vanderwood <infoteq@dragon.azstarnet.com>
-- Tried to use the typedefs mob_rnum, mob_vnum, room_rnum, room_vnum,
obj_rnum and obj_vnum consistently throughout the code. Don't know
if I found them all, but here's hoping...
-- fight.c:damage() - Added die() after an "attempt to damage a corpse",
so if someone does manage to get into the POS_DEAD state, they'll
die for real. Suggested by root <root@cthulhu.dorm.umd.edu>.
-- ban.c:Read_Invalid_List() - Changed so that names are read with one
pass through the file instead of two.
-- act.item.c:perform_drop_gold() - Changed hide_invis from FALSE to TRUE
on act() so room observers cannot see invis people dropping gold; changed
to be consistent with item dropping which already works this way.
Suggested by Chris Epler <cepler@circlemud.org>
-- fight.c:damage() - Changed handling of AFF_SANCTUARY so that damage
is only cut in half if it is 2 or more (i.e. to prevent 1 from turning
into 0). Indirectly suggested by dodger@WPI.EDU (DoDGeR) who reported:
"If you bash a sanct'ed mob, you always see the failure message because
the sanct turns the damage from 1 to 0 and that tells skill_message to
print the miss message."
-- autorun.pl: Made part of the standard distribution (perl version of
autorun), by D. Hall of Vi'Rage Studios
-- Updated FAQ and PORTING documents
-- Made many changes to configure.in and regenerated 'configure' script
with most recent version of autoconf (2.12). configure now checks
which libc functions and system calls are correctly prototyped, and
only prototypes those functions that aren't. Also, it doesn't include
the malloc, nsl, socket, and crypt libraries unless those functions
are missing from libc. This solves problems that autoconf was having
with IRIX, where it would use lsocket and lnsl, unneccesarily generating
linker warnings when those libraries would shadow libc functions. On
the other hand, this will cause a problem on OS's where a dummy
crypt() is shadowed by a real crypt in libcrypt. Future releases may
check crypt() to see if it actually works.
-- constants.c: changed experience and title tables to be functions
returning values instead of arrays, to avoid crashes when newbie imps
try to add levels by changing the LVL_xx constants without extending
the arrays appropriately. SYSERR message is generated to indicate
incorrectly set up xp tables.
-- comm.c, various functions - Made some changes to make porting to other
platforms easier; introduced the CIRCLE_UNIX constant.
-- db.c, various functions: tried to make the parser a bit better at
reporting error messages (and not crashing.. ugh.)
-- Various files: replaced ">> 2" with "/ 4". I can be a real idiot
sometimes.
-- Added preliminary Amiga support with the help of Jack Patton and
Steffen Haeuser.
-- act.wizard.c:do_dc() and structs.h:descriptor_data -- Added a
"close_me" flag to descriptor_data that dc uses to indicate a descriptor
should be killed off; just having dc() call close_socket() crashes if
you're disconnecting the person below you in the descriptor list.
-- The script I use to pack up Circle into a tar and zip file was changed
to add token files to all empty directories to work around buggy unzip
programs that don't bother creating empty directories.
-- act.other.c:do_display(): Now complains about unknown arguments.
-- act.wizard.c:do_shutdown(); db.h: Added the constants FASTBOOT_FILE,
KILLSCRIPT_FILE, and PAUSE_FILE to the header file, and used them
instead of the literals in the code.
-- olc.c:do_olc(): Fixed the warning given by gcc. Just FYI, (for whoever
"Y"ou are), I left that warning in on purpose, sort of as a subtle
reminder that OLC is not done. However I finally decided to change it
after the 10,000th "bug report".
-- Gave the utilities their own Makefile (Makefile.in, actually; Makefile
is generated by configure)