-
Hey guys,
I thought I'd start keeping detailed changelogs, since there are more of us coding. Hopefully this documentation habit will rub off on the newer coders 8).
A brief categorization, line number(s), and description of the fix/change/addition should do.
Here's my first round.
[code]
10/29/06
* Fix: mob_cmds.c, lines 275-278
Added an ifcheck to make sure that PCs don't execute "mob" commands.
This causes segfaults with things like mob delay, and any mpcommands that
reference the pIndexData.
10/28/06
* Fix: olc.c, lines 2660, 2598, 2661
Added "nocolor" filter so that the objects, mobs and rooms with color in the name
would line up right in the olist/mlist/rlist.
10/27/06
* Fix: hunt.c, line 513
Commented out a redundant ifcheck in the trackless step routine. The ifcheck
check whether or not the huntee was a ranger before going on, thus causing
a segfault for NPCs who have the skill.
* Commented out: comm.c, line 2344-2353
Commented out some ifchecks that were incorrect and prevented new players from seeing prompt
and having nosummon on.
* Addition/Fix: act_obj.c, line 4642
Added an ifcheck to prevent mortals from stealing from immortals.
* Fix: shoot.c, line 629-630
Cleaned up some code (removed debugging), and added an ifcheck so the pcdata of NPCs wouldn't b
checked.
* Fix: fight.c, line 4196
Fixed it so NPCs could intimidate.
* Fix: fight.c, line 281
Removed IS_NPC from the ifcheck on account of giving all mobs a movement value.
* Fix: fight.c, line 4941
Same fix for dirt kick as bash.
* Fix: fight.c, line 4282
Fixed the get skill if-check so that mobs could bash, too.
* Fix: act_info.c, line 1796
Inserted an if-check to make sure that the mob is a PC before
trying to check its pcdata. Has to do with sith sense danger skill.
* Fix: handler.c, line 479
Fixed the get_skill function to return reasonable values for NPC skills, scaling as
40+19 log(level). This will do to at least enable many skills for NPCs for now until we
establish a more specific way of assigning skills to NPCs. The range is from 40% at level 1
to 91% or so at level 500.
10/24/2006
* Fix: mob_cmds.c, line 3112
The "obj remove" function (opremove) was referring to obj->next_content
instead of obj1->next_content when looping through the list in the person's
inv, causing an infinite loop (obj1 is the temporary variable used to cycle
through the list).
10/22/2006
* New function: mob_cmds.c, lines 3205-3234
Prototype added in: mob_cmds.h, line 144
Implemented "obj at" command
[/code]
-
Not that I'm a coder or anything, but since I have access to this forum, I'll see what I can help with…
First off
````
* Fix: hunt.c, line 513
Commented out a check for ranger within a check for trackless step to include the possibility
of others such as NPC rangers having the skill (and causing crashes).
````
Could you clarify that? I'm sure that coders would know what that means just by looking, but it could help with posting change logs. Do you mean that you removed the check for trackless step in the hunt code? Or did you remove a check to see IF the target is/was a ranger, but leaving the skill check for use with the possibility of, say, NPCs having the trackless step skill? Just curious. Seemed a bit ambiguous.
-
~~@Nibelung:~~
> Not that I'm a coder or anything, but since I have access to this forum, I'll see what I can help with…
>
> First off
>
> ````
> * Fix: hunt.c, line 513
>
> Commented out a check for ranger within a check for trackless step to include the possibility
> of others such as NPC rangers having the skill (and causing crashes).
>
> ````
>
> Could you clarify that? I'm sure that coders would know what that means just by looking, but it could help with posting change logs. Do you mean that you removed the check for trackless step in the hunt code? Or did you remove a check to see IF the target is/was a ranger, but leaving the skill check for use with the possibility of, say, NPCs having the trackless step skill? Just curious. Seemed a bit ambiguous.
Oops, no I gave quite a shite summary of that. Here is the offending bit of code:
````
// For trackless step skill
if (get_skill( victim, gsn_trackless_step ) > 0
//&& victim->pcdata->second_sub_class_cleric == CLASS_CLERIC_RANGER
&& ( victim->in_room->sector_type == SECT_FIELD
|| victim->in_room->sector_type == SECT_FOREST
|| victim->in_room->sector_type == SECT_HILLS
|| victim->in_room->sector_type == SECT_MOUNTAIN
|| victim->in_room->sector_type == SECT_TUNDRA ) )
{
if ( number_percent() < get_skill( victim, gsn_trackless_step ) )
{
sprintf(buf, "%s has covered $S tracks too well for you to follow.",
IS_NPC(victim) ? victim->short_descr : victim->name);
act(buf, ch, NULL, victim, TO_CHAR);
return;
}
}
````
The fix I referred to is the commented-out AND statement which (1) crashes the game if the NPC is being checked, since the pcdata pointer is null on NPCs, and (2) is redundant anyway, since only rangers (and immortals, and currently mobs) have the trackless step skill. Hope this clarifies it for you - I have rephrased it in my changelog (see above).
-
[mr. burns]Excellent…[/mr. burns]
-
Backlogged on posting my changes here so I thought I'd fix that.
````
12/21/2006
NOTE FROM SYN
I am going to change the style of my code documentation a little bit to
make it easier on me. First, I will post the same
text in (a) my changelog included with the source code,
(b) the coding board and (c) the in-game changes.
(a) and (b) will also contain further information about the
changes accessible to coders - this is generally a module name
(e.g. fight.c), a function name, and a brief description of
the change. I'm mostly trying to phase out recording line numbers
except where necesarry since they change often with code editing
and our functions aren't gargantuan, so a function name will usually
suffice to locate a change.
Obviously, information about the code will not be posted in-game,
since this is easily accessible to coders via the MB. I will
try my hardest to make my changes consistent, clear and concise to players,
immortals, and coders alike, with additional information posted to
coders to allow them to better understand my code. Of course, I will
comment my changes within the code as well.
* Fixed - NPCs couldn't initiate combat, leading to various issues
involving NPC combat (aggro mobs not attacking, mounts of crusaders not
helping them fight, and probably some others.)
(fight.c, mob_hit(), commented out unnecesarry if-check)
* Removed "Seralia/Athemia" points display, the awarding of military points,
and the "xxx is now a pirate", announcement since they added absolutely
nothing of value to gameplay.
(act_info.c, do_score(), commented out bits about seralia/athemia points
fight.c, player_kill(), commented out bit about pirates
quests.c, do_quest(), commented out bit about military points)
* Fixed - NPCs would wander off even if they were following or grouped
with someone.
(update.c, mobile_update(), a few additions and restructuring of
some code)
12/12/2006
* Fixed - Reduced the hitting power and the skill levels of NPCs
in the newbie area to make it a little easier on players
who aren't used to the game. Also added a hint on
practicing/training to the hints channel.
* Fixed - The regeneration spell had a buggy duration when placed
on objects.
11/06/2006
* Additions
interp.c, 458, cmd_table
interp.h, 441, prototype
merc.h, 4000, #define
olc.h, 36, #define
olc_mpcode.c, 410-464, new function, do_tpedit
olc_mpcode.c 62-62, tpedit command table
11/01/2006
* Fix: magic2.c
Functions: spell_toxin_paralysis, spell_toxin_neurotoxin,
spell_toxin_weakness, spell_toxin_venom
Added echoarounds for when someone resists a Sith toxin bite
so that the sith has an idea of what happened (why their bite
didn't work.) Also fixed the current echos to the victim.
* Fix: fight.c, line 4811
Fixed Sith bug where no matter what type of toxin you were trying to
bite with, it would use paralysis - problem was checking "argument"
instead of "arg2".
* Addition: bit.c, line 344
tables.c, line 1160
merc.h, line 1562
auction.c, line 355
Implemented "noauction" object flag.
* Addition: act_obj.c, lines 890, 1062
Additions to fix the "no_container" object flag.
* Addition: act_comm.c, line 1240
Addition so room no_comm flag would block yells as well.
10/30/2006
* Addition: merc.h, line 1561
bit.c, line 343
locker.c, line 139
tables.c, line 1159
Implemented "no_locker" object2 flag.
* Fix: olc.c, line 131
Inserted oedit_timer into the list of oedit commands. Not sure at all
why it was off, but it's back on now.
* Fix: fight.c, line 3873 and 3881
Changed it so non-imms can't see numerical values of hits unless on
the test port.
* Change: update.c, line
Moved update_has_done (minor function that was sucking 10% of CPU time)
from aggr_update to char_update.
* Fix: const.c, line 3694
db.c, line 449
olc_act.c, line 5673
merc.h, line 215 and 4074
Added a mob skill level table generated at startup so that our processor
won't be crunching a logarithm every time a skill lookup for an NPC
is needed (often).
* Fix: const.c, lines 1894-3688.
Replaced the unused SLOT() field with a "race" field that specifies
if the skill belongs to a certain race. This is to fix the problem of
all NPCs having all racial skills regardless of their race. Right now
unfortunately it doesn't account for racial skills that belong to
more than one race such as shift - but I will replace this with a bitfield
instead of an integer when I have the time. It works fine for now,
since no automatic skills belong to more than one race, and those are
the ones I wanted to correct first.
* Fix: act_obj2.c, line 812
Added an "!IS_NPC()" ifcheck in do_consume to prevent segfaults
when an NPC tries to consume a corpse.
* Addition: merc.h, lines 248-273.
Added a list of #defines for PC race types. These are now used
in skill_table (const.c) to check for racial skills with a simple reference,
rather than looping through the list of pc races and then looping through
each race's list of skills.
* Change: merc.h, line 3386.
Added a "race" field to each skill. This replaced the old, unused
"slot" place. I added this so that NPCs wouldn't use racial skills
unless they were of that race.
* Removal: act_wiz.c, lines 1739-45, 3248-86
Removed "slookup" function. Supposed to be for 'stat skill' which
does nothing.
* Addition: mob_prog.c, lines 2893-2988
Added a p_direction_trigger, almost identical to p_exit_trigger,
except that it allows other trigger types (in this case, open
and close) to fire on a direction.
* Addition: act_move.c, lines 1129, 1224, 1166, 1253
Implemented open and close triggers for objects (first two lines)
and doors (last two lines).
* Addition: merc.h, lines 3454-3455
Added #define for open and close triggers for rooms and objects.
````
