[Announce] Angband 3.0.8 prerelease
Andrew Sidwell
want to put out a prerelease version for testing purposes. I believe
this version to be mostly bug-free and fairly consistent, but you may
find otherwise, and if you do, I want to know! If no problems are
found, I'll release 3.0.8 final without any modifications on 1st July.
Download at:
Source code: http://rephial.org/downloads/angband-3.0.8pr1.tar.gz
Windows: http://rephial.org/downloads/angband-3.0.8pr1-win.zip
RISC OS: http://rephial.org/downloads/angband-3.0.8pr1-ros.zip
(RISC OS version will be available later.)
No Mac OS X build has been made yet, but I'm sure one will be soon.
Please note that the source archive is now ~2.3MB, because I've included
graphics, fonts and sounds.
Quite a number of people have been involved in this version, and thanks
must be given to Hugo Kornelis, Pete Mack, Marco K and Frank Palazzolo,
Christer Nyfalt, Leon Marrick, Andrew Doull, Kenneth Boyd, Iain McFall,
Kiyoshi Aman, Christophe Cavalaria, Nick McConnell, and Brendon Oliver.
Special thanks go to Antony Sidwell for setting up http://rephial.org/,
which is the new home of Angband. It hosts the wiki, downloads, and the
Open Source effort, which centralises all of those bits nicely. Special
thanks also to Yendor for hosting the issue-tracker and source code
repository, which has been invaluable.
All gameplay options are now only alterable at birth. Be careful when
importing savefiles to make sure that the AI and scumming options are as
you want them for the entire game! And please keep backups of old
characters, just in case things do go wrong.
Changes
=======
(All changes from 3.0.6 are mentioned.)
Major visible changes
---------------------
- New splash screen.
- The Ey/Un-style knowledge browser has been imported. You can set
autoinscription settings through that browser now.
- Mouse support a la FAangband added, but nowhere near as advanced or
useful. Most of the menus you encounter can be selected using the
mouse now, as well as using keypresses. Clicking somewhere on the
map will now "walk" there, within certain constraints. It's a tad
bizarre in the town, but seems to work OK otherwise.
- Massive options clearout/reorganisation/redefaulting:
- many options have been removed (20 or so, I think)
- a large number have new default settings
- new option: remaining HP changes the colour of the '@' sign.
- new option: always pick up things matching an item in the inventory
- experience to next level always displayed until clev50
- macros/visuals/colours are now accessible (only) from the options
menu
- Sangband-style object handling, which makes it much easier to sift
through things in the dungeon.
- Item squelching support, EyAngband-style. Instead of being
automatically destroyed, items are instead marked with the inscription
"squelch" (if there is no already existing inscription), and all items
in your inventory and on the floor inscribed with this will be
destroyed if you press '!' at the "destroy" command.
You can set individual squelch settings for only those items which do
not get pseudo-ID'd. For those that do, you can choose to mark things
as "squelch" depending on the feeling you get when you first ID them
in that way. These categories are much broader than those found in
other variants; complain if necessary.
Worthless items and emptied chests are automatically inscribed with
"squelch", saving some micromanagement.
- Two kinds of ego-light added. You'll know when you find one; there's
no ID required to figure out if your lantern is special or not.
- Lights now don't use fuel in the daylight in the town.
- Amnesia (the monster attack) is now a timed effect (a la confusion),
and has makes you forget how to read scrolls/pray/cast spells one
time in two.
Potions of Healing, *Healing* and Life are the only three things which
will get rid of amnesia early, but it goes away within a few turns.
Please tell me if this is still too harsh.
Minor visible changes
---------------------
- Allow weapons and armour to be damaged instead of destroyed by
inventory damage effects, thus making swap weapons more viable.
- Browsing books now displays information about the spells therein, if
requested. As part of this, move spell names/descriptions/locations
out into a new file, spell.txt. (Hugo Kornelis)
- Arrow keys are now distinct from the numeric movement keys, and
because of this you can now edit text at prompts in game properly.
Supported on Windows, Mac (untested), and curses. If you get it
working for anything else, please send the patch!
- Resizing the game window automatically updates its contents, on
platforms which support this.
- Inscribing an item with '!t' will now confirm if you try to take it
off.
- If a stack of rods is recharging, you'll get a message when the first
rod is recharged and another message when they are all recharged.
There is no message when the second, third, etc., rod is recharged.
(Hugo Kornelis)
- The recharge notification will now disturb you if you have set the
option to be disturbed when boring things happen (disturb_minor) to
yes. (Hugo Kornelis)
- Add the recharging fix from V CVS -- things can now explode when you
recharge them once more.
- Added support for improved pluralization of object names. Allow
custom plurals for words which need them (e.g. "Kni|fe|ves" and
"M|ous|ic|e"). Unused at present. (Hugo Kornelis)
- Bugfix: Monsters that resist one or more of acid, lightning, fire,
cold, poison and water *and* resist one or more of nether, plasma,
nexus and disenchantment had two "it resists ..." sentences in the
monster info. (Hugo Kornelis)
- Bugfix: Sustain CON was reported before Sustain DEX in self knowledge.
Since DEX is reported before CON in all other situations, this is now
reversed. (Hugo Kornelis)
- Set wrapping width to 72 for character dumps' item information, to
avoid well-behaved newsreaders messing them up.
- Move back to pre-3.0.6 panel changing behaviour.
- Semi-rewrite of the store code; now uses a scrolling menu with a
slightly more streamlined interface. Bigscreen support in stores.
Ammo is treated in more aesthetically pleasing quantities.
General Store always stocks what it stocks now.
- Remove double-rate spellcasting with the smart_monsters option.
(imported from V CVS; Julian Lighton).
- '[' command displays a monster list in the main terminal window.
- "Enter" key brings up a little window in the middle of the screen,
from which you can select what command you would like to use. This
is inspired by (but not derived from) Hengband.
- Prompt for overwriting savefiles.
- Added echo to macro trigger entry.
- Restore some pre-Ben flavour text.
Major code cleanup
------------------
Most of the cleanup has been related to removing ancient or
non-functional cruft.
- A powerful generic menu interface has been added and used throughout
the code. It reduces code duplication and makes it much much easier
to create menus in various styles; e.g. scrolling or multicolumn.
See ui.c.
- Event handler added. (Menus can listen to each other, etc.)
- Lots and lots of code duplication removed:
- pref file dumping now has much less boilerplate
- the status line, sidebar, and term window update code is much more
streamlined.
- Nuked a lot of old ports: cap, lsl, sla, xpj, ibm, dos, vcs, lfb, ami,
emx, vme.
- Added a new SDL port by Iain McFall, based on the ToME and Sang ports.
- As a result of the above, makefiles considerably simplified. As many
makefiles as possible use a common list of source files in
"Makefile.src", which simplifies maintainance. Makefile.std has also
been rewritten and contains much less cruft.
- Timed player modifiers (blind, confused, etc.) are now kept in an
array, rather than being seperate variables in p_ptr. This simplifies
a *lot* of code, and makes it trivial to add new effects. The
majority of the work was done using sed scripts.
- Use a new autoconf/automake system, called OMK.
- Removed Lua.
- Charges for wands/staves are customisable in the lib/edit/ files.
- Considerably simplify and speed up main-gcu.c. (Thanks to Christer
Nyfalt.)
- Removed trivial-to-hack-around or non-functional compile-time options
(VERIFY_SAVEFILE, _TIMESTAMP, and _HONOR, CHECK_TIME).
- Combine the various h-*.h files into a single header file; remove a
lot of cruft, use C99 types when available,
- Clean up file locking, move file handling functions into z-file.c,
remove usleep() function, add a "portable" directory scanning
interface (works on Windows and Unixes at the moment), and
my_fexists().
- Use the safer string handling functions much more; the only use of
the "unsafe" built-ins are in the various main-* files.
- Split random name code from randart code into its own file with a bit
of a rewrite. Use this code for scroll names.
- Move lots of things out of defines.h to more appropriate places, like
h-basic.h and config.h.
- Replace various magic numbers with constants or N_ELEMENTS().
(Hallvard Furuseth)
- Use C99 types where appropriate (for bools and ints).
- Import Sangband's/Steamband's code to put chance of generating piles
of items in object.txt.
- Use tables of commands instead of a big switch, to allow doing thing
like the new command menu easily without duplication. (See cmd0.c.)
- Switch to new model for producing sound: instead of a TERM_XTRA_SOUND
hook, we now have a simple "sound_hook" which should be set to the
appropriate function to play a sound. Introduce a new modular system
for sound modules for ports that use main.c (for the future, in case
we want to use e.g. gstreamer). Allow sound to be toggled as an
in-game option.
- Remove hardcoded values for p_ptr->noscore.
Platform
--------
- Modernize Mac support; now works on OS X 10.3 onwards, but the old
main-mac has been removed, as has the MPW makefile. (Pete Mack)
- OSX font selection problem fixed.
- Add console mode patch for Windows (Frank Palazzolo). This is
basically playing the game in an old-school terminal window for
those who like uncluttered gameplay.
- main-x11 now reads window placement from x11-settings.prf.
- Fixed Windows bug where you can't move one of the font files after
having played a game this reboot. (Thanks to Leon Marrick.)
- Add platform-specific ifdefs to readdib.c and main-ami.c.
- Allow ^S without freezing the game on the console. (Hallvard
Furuseth)
- Mouse support on RISC OS.
- Sort out licences for the various fonts distributed with the game.
- Import the old DOS/IBM fonts for use in the Windows port.
- Include SDL_mixer sound module, useable on any port that uses the
main.c mechanism. Thanks to Brendon Oliver for the initial code
for this.
--
Andrew Sidwell
http://rephial.org/ -- the home of Angband
My email address changes monthly, and is the first three letters of the
month (in English), followed by the last two digits of the current year,
@entai.co.uk.
Eddie Grove
> - Item squelching support, EyAngband-style. Instead of being
> automatically destroyed, items are instead marked with the inscription
> "squelch" (if there is no already existing inscription), and all items
> in your inventory and on the floor inscribed with this will be
> destroyed if you press '!' at the "destroy" command.
Oh well, no future V for me. I quit playing [Ey] over this. The whole point
of squelch is so that you can ignore junk. I want an implementation where
junk is invisible, not where I have to use ! repeatedly to clear every useless
item produced when I clear a pit with dispelEvil. I don't want messages when
monsters pick up junk. I don't want to be disturbed when I run and that takes
me adjacent to junk or through junk. Junk should not impinge on the game.
I've been wondering whether I should patch NPP or V. Now I know.
Eddie
Timo Pietilä
> - Amnesia (the monster attack) is now a timed effect (a la confusion),
> and has makes you forget how to read scrolls/pray/cast spells one
> time in two.
Uh. Does this mean that "The lock" is no longer "The lock"? This makes
amnesia one of the most dangerous effects in entire game. No more 100%
sure escape. Umbar brothers are now among "avoid at all costs" -group.
AAARRGHHHH! I just counted how many monsters have FORGET-flag: 25!
Including Sauron which cannot be avoided.
In practice this equals blindness-attack without resist except saving throw.
You either need to add resistance or one-turn cure for this effect.
Othervise it can be way too deadly.
Timo Pietilä
Timo Pietilä
> Andrew Sidwell wrote:
>
>> - Amnesia (the monster attack) is now a timed effect (a la confusion),
>> and has makes you forget how to read scrolls/pray/cast spells one
>> time in two.
> You either need to add resistance or one-turn cure for this effect.
> Othervise it can be way too deadly.
Earlier than potion of healing, that is. I don't like the fact that
monster can bypass the lock. Blindness, confusion and sound should be
only ones that can disrupt your spellcasting. This breaks one of the
fundamentals of the game.
Timo Pietilä
roger....@gmail.com
> Angband is soon to have a new official release. Before that, though, I
> want to put out a prerelease version for testing purposes. I believe
> this version to be mostly bug-free and fairly consistent, but you may
> find otherwise, and if you do, I want to know! If no problems are
> found, I'll release 3.0.8 final without any modifications on 1st July.
>
> Download at:
>
> Source code: http://rephial.org/downloads/angband-3.0.8pr1.tar.gz
> Windows: http://rephial.org/downloads/angband-3.0.8pr1-win.zip
> RISC OS: http://rephial.org/downloads/angband-3.0.8pr1-ros.zip
> (RISC OS version will be available later.)
>
> No Mac OS X build has been made yet, but I'm sure one will be soon.
Firstly just a "Well Done", for getting a new release out :) Angband
is alive, progressing, adapting and changing. You've worked hard, and
it shows.
However, I do have to agree with the above posters about amnesia -
it's just a shade *too* deadly at the moment. Perhaps skew the odds a
little more in the player's favour by making it curable by a potion of
Cure Critical Wounds? If that's too much, then perhaps giving a potion
of Cure Critical Wounds a 50% chance to cure amnesia? Or allowing it
to be cured by Potions of Wisdom/Intelligence/Enlightenment.
Still, bloody well done!
Roger Barnett
-----
http://angband.calamarain.net/ - Tales From The Pit
The Angband Webcomic
-----
Twisted
<roger.barn...@gmail.com> wrote:
[snip]
This was bound to be controversial. It will be fun to watch this all
shake out.
I have to agree about the amnesia. If it impacts spellcasting and
scroll reading it should be curable with CCW, and easy to prevent by
the late game. I have several ideas:
* Leave scroll reading alone.
* Have the effect be a chance of forgetting one or two spells, most
likely high level. Your most crucial reliable spells, if any, tend to
be lower level; and then there are warriors.
* Have just the effects on scrolls and/or spells curable with CCW.
* Or just make it a chance to confuse through resist?
And regardless:
* Double the influence of saving throw, so most characters save 100%
of the time against this before they encounter Sauron or any other
deep unique with S_FORGET.
* Or maybe provide a specific temporary resistance. Potions of Resist
Confusion say might provide temp confusion resistance earlier in the
game and amnesia resistance whenever needed. Or co-opt some existing
thing, such as potions of heroism or even bless/chant/prayer.
* Or, resist confusion (from equipment) prevents the "confusion-like"
effects that may occur 1 time in 2.
* One wacky possibility is to co-opt nexus resistance. As it is one
wants it or a perfect save when facing Sauron, lest one gets Sauron
down to one star only to receive the messages "Sauron gestures at your
feet. -more- You rise through the ceiling..."
* One could just give amnesia attacks a chance to randomly do nothing
extra, blind, confuse, cause fear, or cause hallucination, with
blindless, confusion, fear, and chaos resistance respectively
providing protection, as well as a second saving throw against the
additional side effect. (This is after you've already failed a saving
throw vs. amnesia.)
I'm certainly all for making amnesia more interesting while removing
the "have to identify everything again and re-explore the whole level"
tedium (and, for warriors, expense; warriors already have it tough
with a lack of free ranged attacks and free id, as things are!).
camlost
>
> Uh. Does this mean that "The lock" is no longer "The lock"? This makes
> amnesia one of the most dangerous effects in entire game. No more 100%
> sure escape. Umbar brothers are now among "avoid at all costs" -group.
Um, don't play S. Sangband (and other 4GAI?) has brain smashing, which
can cause fear, blindness, confusion, stun, slowed, and/or take energy
from a player. And you basically can't do anything about it.
>
> AAARRGHHHH! I just counted how many monsters have FORGET-flag: 25!
> Including Sauron which cannot be avoided.
>
> In practice this equals blindness-attack without resist except saving
> throw.
>
> You either need to add resistance or one-turn cure for this effect.
> Othervise it can be way too deadly.
I would guess that most casters would be able to mostly avoid such
creatures, and melee types wouldn't suffer as much.
>
> Timo Pietilä
Antoine
> I'm certainly all for making amnesia more interesting while removing
> the "have to identify everything again and re-explore the whole level"
> tedium (and, for warriors, expense; warriors already have it tough
> with a lack of free ranged attacks and free id, as things are!).
What about it instead making you forget one chunk of the level (not
the chunk you're standing in) and one item 'flavour'?
That would maintain the amnesia 'forgetting what stuff is' component
while reducing the overall annoyance?
A.
Phil Cartwright
None of you suggested a chance to drain experience, which also is
in-theme for amnesia and could cause spell forgetting or similar.
Personally I like twister's suggestion of a chance to confuse, blind,
drug, terrify, or whatever, maybe with drain-XP as another option on
that list. Of course we don't want to make it really hugely annoying or
damaging. Might as well just make all those enemies breathe time instead
if that's what you want. Or put it back the way it was. ;-)
--
There's only four things you can be certain of: taxes, change, spam, and
death.
Billy Bissette
@q75g2000hsh.googlegroups.com:
> I'm certainly all for making amnesia more interesting while removing
> the "have to identify everything again and re-explore the whole level"
> tedium (and, for warriors, expense; warriors already have it tough
> with a lack of free ranged attacks and free id, as things are!).
Would it be possible to make Amnesia temporarily reduce the effective
character level? Not by the Drain Experience route of actually dropping
the level, but perhaps some kind of estimated skill reduction? A level
30 Wizard hit by Amnesia might temporarily be treated as if he were CL20.
His spell failure rates might increase, he might lose access to his
highest level spells (now beyond his capability,) etc. A Warrior might
be a less effective fighter. Experience gained would be treated as
normal, as the current level hasn't really been altered.
Heck, making forgetting the whole dungeon level might not be as bad
if Amnesia were a time-out effect. At the moment the character is made
to forget, save the known map. When the effect runs out, combine the
saved map with the current known map, so that the character "remembers"
any forgotten squares that he hadn't re-explored.
Nick
OK, so how about a 4-tier squelching system:
Weak: Like current V squelching. Items always appear, have to be manually
squelched.
Strong: Like current NPP squelching. Items can be made to appear as a purple
dot, and can be silently destroyed on walking on with appropriate (pseudo-)id,
but otherwise act exactly as if unsquelched.
Extreme: As for strong, plus after level generation (and acceptance if using
autoscum) squelched items are silently removed. Possibly the same thing happens
with monster drops - the drop is generated as usual, then silently has
squelching rules applied before the items actually appear.
Eddie: Includes all features of extreme. Everything is generated identified.
There is an automated "player recall" system, something like monster recall,
which decides at what point in the game the player will want to squelch a given
item and immediately silently set it to squelch (this will apply from everything
from Potions of Slowness right up to whether that Mace of Disruption of Extra
Attacks is enchanted enough to be a good swap). After level generation, all
monsters which would be trivial to kill will automatically and silently commit
suicide, leaving any unsquelched drop in a neat pile on the ground in a
player-configured order. Gold is squelchable.
Did I miss anything? :)
Nick.
--
"There is no safety, and there is no end. The word must be heard in silence;
there must be darkness to see the stars. The dance is always danced above the
hollow place, above the terrible abyss."
- The Farthest Shore, Ursula Le Guin
will_asher
Look at it as slightly changing one of the fundamentals of the game. I think it
is an appropriate change. Most monsters that have FORGET are late enough for
people to have healing potions. Making healing potions more common would be
nice to go along with this change, though.
Eddie Grove
> On 2007-06-24 20:32:27, Eddie Grove <eddie...@hotmail.com> wrote:
>
> > Andrew Sidwell writes:
> >
> > > - Item squelching support, EyAngband-style. Instead of being
> > > automatically destroyed, items are instead marked with the inscription
> > > "squelch" (if there is no already existing inscription), and all items
> > > in your inventory and on the floor inscribed with this will be
> > > destroyed if you press '!' at the "destroy" command.
> >
> > Oh well, no future V for me. I quit playing [Ey] over this. The whole point
> > of squelch is so that you can ignore junk. I want an implementation where
> > junk is invisible, not where I have to use ! repeatedly to clear every useless
> > item produced when I clear a pit with dispelEvil. I don't want messages when
> > monsters pick up junk. I don't want to be disturbed when I run and that takes
> > me adjacent to junk or through junk. Junk should not impinge on the game.
> >
> > I've been wondering whether I should patch NPP or V. Now I know.
>
> OK, so how about a 4-tier squelching system:
>
> Weak: Like current V squelching. Items always appear, have to be manually
> squelched.
You mean future V squelch, which will require three keypresses.
Current 3.0.7 squelch is automatic.
> Strong: Like current NPP squelching. Items can be made to appear as a purple
> dot, and can be silently destroyed on walking on with appropriate (pseudo-)id,
> but otherwise act exactly as if unsquelched.
I am opposed to "detect artifact using kill command". Purple dots are bogus.
Otherwise this is barely minimally ok.
> Extreme: As for strong, plus after level generation (and acceptance if using
> autoscum) squelched items are silently removed. Possibly the same thing happens
> with monster drops - the drop is generated as usual, then silently has
> squelching rules applied before the items actually appear.
I am opposed to this too. If you squelch !restoreDex, then get dex drained
by time hound even though you have sustains, when you unsquelch !restoreDex
any potions you have been ignoring should appear on the screen.
*** Squelch should be about ignoring, not about removing or destructing. ***
> Eddie: Includes all features of extreme. Everything is generated
> identified.
This is almost diametrically opposed to what I want. While I am opposed to
the identify spell, I believe in learning by testing, not by generating
identified. Egos should be learned sort of like flavors. To-hit and to-dam
modifiers should be apparent when you pick up a weapon, but I have never
proposed generating things identified. In fact, I don't think you should
see damage dice until you pick up a weapon.
> There is an automated "player recall" system, something like monster recall,
> which decides at what point in the game the player will want to squelch a given
> item and immediately silently set it to squelch (this will apply from everything
> from Potions of Slowness right up to whether that Mace of Disruption of Extra
> Attacks is enchanted enough to be a good swap). After level generation, all
> monsters which would be trivial to kill will automatically and silently commit
> suicide, leaving any unsquelched drop in a neat pile on the ground in a
> player-configured order.
All wrong.
BTW, there is a game where overly wimpy opps commit suicide when you walk near
them at high levels. My friends tell me that that is the best part of the game.
> Gold is squelchable.
Yes, in that you should be able to make it invisible, but even squelched gold
should be picked up silently and added to your purse when you cross it.
I am sorry to clutter this thread with what promises to be a long contentious
discussion of squelch, but I couldn't leave such wrongheaded suggestions of my
opinions on these matters without response.
Eddie
Timo Pietilä
> Timo Pietilä wrote:
>>
>> Uh. Does this mean that "The lock" is no longer "The lock"? This makes
>> amnesia one of the most dangerous effects in entire game. No more 100%
>> sure escape. Umbar brothers are now among "avoid at all costs" -group.
>
> Um, don't play S. Sangband (and other 4GAI?) has brain smashing, which
> can cause fear, blindness, confusion, stun, slowed, and/or take energy
> from a player. And you basically can't do anything about it.
You can have resistances to the effect. Also you can get saving throw
but not for 100% sure case like in vanilla. You also can cure most of
the effects by simple CCW potion (slowing and fear are exceptions).
Also I don't think there are any brain-smashing monsters before
stat-gain in Sang. At least nothing has used brain smash against my
high-elf karateka yet and it is soon at 2000'.
>> AAARRGHHHH! I just counted how many monsters have FORGET-flag: 25!
>> Including Sauron which cannot be avoided.
>>
>> In practice this equals blindness-attack without resist except saving
>> throw.
>>
>> You either need to add resistance or one-turn cure for this effect.
>> Othervise it can be way too deadly.
>
> I would guess that most casters would be able to mostly avoid such
> creatures, and melee types wouldn't suffer as much.
You can't. At least not Sauron. And one of the ringwraiths has this too.
Also this is *common* spell for monsters after stat-gain and it can
appear way before stat-gain, which means you would need to avoid too
many monsters for it to be reasonable. Because it appears that soon *at
least* cure needs to be potion of CCW or CSW. Not healing.
Consider this: Priest that should not have any problems against orc
group. There is one of the uniques in that group so you enter the room.
Surprise! there is memory moss there and because it is not evil you
didn't detect it. Suddenly you don't have any weapons to handle those
orcs. You can't escape with portal and can't blink. Can't read recall or
teleport. You didn't have money for teleport staff or it did get burned.
Or this simply is too early for that to be in your gear. And that moss
keeps on blasting you with amnesia. Can't kill it because orcs are in
the way.
Or lets make this much worse. There is Ethereal Dragon nearby and you
know you you can't handle it so you should avoid it. You carefully try
to go around it because you see a longsword 4d5 near it. Surprise! There
is a memory moss. And that Ethereal Dragon wakes up. Now what do you do?
Press @ to confirm suicide by frustration -more- ?
Killed by memory moss. Now that is what should read in tombstone.
For spellcasters:
Omarax becomes too dangerous to handle
Thuringwethil becomes too dangerous to handle
Angel uniques become too dangerous to handle
Vecna and Feathwath become too dangerous to handle
And of course anything with S_MONSTERS become pretty much too dangerous
to handle.
This also makes jelly pits completely impossible to handle. (yet another
reason to remove them from game).
Timo Pietilä
R. Dan Henry
<pca...@nospam.phony.com> wrote:
>None of you suggested a chance to drain experience, which also is
>in-theme for amnesia and could cause spell forgetting or similar.
That would be redundant with other attack forms. I think the map loss is
fine; you don't have to explore the whole level, anyway. Maybe just
forget equipment, but not inventory, that hasn't been *ID*ed -- the
tactical component of amnesia is the loss of activations from non-*ID*ed
equipment and that's worth keeping. As for spells, amnesia might cause
you to randomly forget a single spell. That wouldn't be too bad.
--
R. Dan Henry
danh...@inreach.com
Paul is dead!
Nick
> Nick writes:
> > Weak: Like current V squelching. Items always appear, have to be manually
> > squelched.
>
> You mean future V squelch, which will require three keypresses.
> Current 3.0.7 squelch is automatic.
Yes, OK; I meant the new version.
> > Extreme: As for strong, plus after level generation (and acceptance if using
> > autoscum) squelched items are silently removed. Possibly the same thing happens
> > with monster drops - the drop is generated as usual, then silently has
> > squelching rules applied before the items actually appear.
>
> I am opposed to this too. If you squelch !restoreDex, then get dex drained
> by time hound even though you have sustains, when you unsquelch !restoreDex
> any potions you have been ignoring should appear on the screen.
>
> *** Squelch should be about ignoring, not about removing or destructing. ***
OK, so squelched items would be invisible unless you unsquelch them?
> I am sorry to clutter this thread with what promises to be a long contentious
> discussion of squelch, but I couldn't leave such wrongheaded suggestions of my
> opinions on these matters without response.
I don't think it's clutter at all - it's been one of the more controversial
issues for some time, and I think it merits discussion when there's been a
largish change like this. While my last suggestion was mostly a lame attempt at
humour, I was also trying to progress the whole ID/squelch issue, at least in my
own mind. I'm still not entirely happy with any implementation I've seen yet
(although on the whole I probably like the NPP one best).
magnate
> On 2007-06-25 07:22:44, Eddie Grove <eddiegr...@hotmail.com> wrote:
> > Nick writes:
> > > Weak: Like current V squelching. Items always appear, have to be manually
> > > squelched.
>
> > You mean future V squelch, which will require three keypresses.
> > Current 3.0.7 squelch is automatic.
>
> Yes, OK; I meant the new version.
>
> > > Extreme: As for strong, plus after level generation (and acceptance if using
> > > autoscum) squelched items are silently removed. Possibly the same thing happens
> > > with monster drops - the drop is generated as usual, then silently has
> > > squelching rules applied before the items actually appear.
>
> > I am opposed to this too. If you squelch !restoreDex, then get dex drained
> > by time hound even though you have sustains, when you unsquelch !restoreDex
> > any potions you have been ignoring should appear on the screen.
>
> > *** Squelch should be about ignoring, not about removing or destructing. ***
YES. This is the perfect summary.
> OK, so squelched items would be invisible unless you unsquelch them?
>
> > I am sorry to clutter this thread with what promises to be a long contentious
> > discussion of squelch, but I couldn't leave such wrongheaded suggestions of my
> > opinions on these matters without response.
>
> I don't think it's clutter at all - it's been one of the more controversial
> issues for some time, and I think it merits discussion when there's been a
> largish change like this. While my last suggestion was mostly a lame attempt at
> humour, I was also trying to progress the whole ID/squelch issue, at least in my
> own mind. I'm still not entirely happy with any implementation I've seen yet
> (although on the whole I probably like the NPP one best).
I agree almost completely with Eddie: squelching is about saving
keypresses, so the Ey implementation really doesn't achieve much,
because you still have to press two keys (k!) every time you want to
squelch stuff. I won't be using it either - and sadly the venerable
Blackston squelch patch is almost certainly not going to compile with
the new V, so I too won't be going back to V.
The only area where I disagree with Eddie is that I see *nothing*
wrong with purple dots. Purple dots are items I'm not interested in,
so I can ignore them (and if I walk over them they will get tidied up,
i.e. destroyed). If I need a squelched item (e.g. !restoreDex after
draining), I can look for it by examining the purple dots, or I can
unsquelch it and see if any of the dots turns into a potion.
I really don't see why having squelched items invisible instead of as
purple dots makes any difference at all. To me it actually makes
things worse - I don't want to pick up squelched items, but I want to
see where they are. Also I think invisible items are tantamount to
destroyed-on-generation, which is a no-no as per the above summary.
Must get back to work, but agree with Nick that this is important.
CC
pete m
> I am sorry to clutter this thread with what promises to be a long contentious
> discussion of squelch, but I couldn't leave such wrongheaded suggestions of my
> opinions on these matters without response.
>
I agree with you, although I have no problem with "purple dots"
myself. NPP squelch is more than adequate. I too was unpleasantly
surprised to see that squelch now requires two keystrokes where it
previously required none. I am hoping this gets backed out before the
final release. (It was thrown in very late in development, sometime in
the last two weeks.)
pete m
> Consider this: Priest that should not have any problems against orc
> group. There is one of the uniques in that group so you enter the room.
>
> Surprise! there is memory moss there and because it is not evil you
> didn't detect it. Suddenly you don't have any weapons to handle those
> orcs. You can't escape with portal and can't blink. Can't read recall or
> teleport. You didn't have money for teleport staff or it did get burned.
> Or this simply is too early for that to be in your gear. And that moss
> keeps on blasting you with amnesia. Can't kill it because orcs are in
> the way.
Timo -- I rarely suffer more than a handful of successful Amnesia
attacks in a game, and they are rarely associated with a dangerous
situation. Losing lock is a bit scary, but not nearly to the extent
you make it out, because Amnesia attacks are rare. (25 monsters,
including ~7 dangerous non-uniques and ~5 dangerous uniques. Compare
this with ~120 monsters with various confusion attacks--you are just
getting hammered with it all game long. You should be able to work
around amnesia; working around confusion is just not possible.)
And things like the orc problem you mention just aren't that
dangerous--just leave the room, and you are safe. Or shoot the moss
with an arrow--even a priest can't miss. As for Vecna/Sauron--I
generally carry !Healing when I am killing them, unless I am playing
Priest/Paladin. It won't make much difference to start carrying it
for them as well. And Priest generally has an extremely good (or
perfect) saving throw by the time I take on Vecna.
I like the idea of a rare but dangerous new attack mode.
Eddie Grove
> The only area where I disagree with Eddie is that I see *nothing*
> wrong with purple dots. Purple dots are items I'm not interested in,
> so I can ignore them (and if I walk over them they will get tidied up,
> i.e. destroyed).
Squelching should be the ultimate in "tidied up". There shouldn't be any more
to do.
I am opposed because when you are running by them, you stop.
I am opposed because when I use the 'l'ook command to check everything on the
screen, I don't want to have to space through junk items.
I used to squelch all armor/weapons late in the game, and run around stepping
on dots to see what was an artifact. I've decided this is a bad thing.
They are just plain yucky! That's an aesthetic observation.
> I don't want to pick up squelched items, but I want to
> see where they are.
For those kind of items, NPP style set to "do not pickup" works. I do that
with restore potions [since I cannot squelch and later unsquelch without fear
of destroying them by accident].
It's not the same thing as squelching scrolls of trap detection when I am
carrying 9 rods that do the job, or squelching 100% junk items like ruined
chests, or squelching non-ego torches when I have the Phial, or
MagicForBeginners with a warrior at 4900'. I do not want to see where those
items are.
Eddie
Timo Pietilä
> On 2007-06-24 22:18:06, Timo Pietilä <timo.p...@helsinki.fi> wrote:
>
>> Timo Pietilä wrote:
>>> Andrew Sidwell wrote:
>>>
>>>> - Amnesia (the monster attack) is now a timed effect (a la confusion),
>>>> and has makes you forget how to read scrolls/pray/cast spells one
>>>> time in two.
>>> You either need to add resistance or one-turn cure for this effect.
>>> Othervise it can be way too deadly.
>> Earlier than potion of healing, that is. I don't like the fact that
>> monster can bypass the lock. Blindness, confusion and sound should be
>> only ones that can disrupt your spellcasting. This breaks one of the
>> fundamentals of the game.
>>
>> Timo Pietilä
>
> Look at it as slightly changing one of the fundamentals of the game.
That is not slight change. The lock has been there from frog-knows.
Breaking it means removing the lock, not changing it.
> I think it
> is an appropriate change. Most monsters that have FORGET are late enough for
> people to have healing potions. Making healing potions more common would be
> nice to go along with this change, though.
You would need to make them *A LOT* more common then. Vampire, Mind
flayer, Angel and Ring mimic are before stat-gain monsters with it and
at stat-gain you get Memory moss, Shade, Spectre, Phantom, Greater mummy
and Master vampire. Later more. And I didn't count uniques.
Problem is that those are very common monsters. Some of them good source
of XP and items, some are very hard to avoid (PASS_WALL, invisibles).
Those also are common stuff to get summoned. Anything with S_UNDEAD can
summon one of those.
Timo Pietilä
Timo Pietilä
> Timo Pietilä wrote:
>
>> Consider this: Priest that should not have any problems against orc
>> group. There is one of the uniques in that group so you enter the room.
>>
>> Surprise! there is memory moss there and because it is not evil you
>> didn't detect it. Suddenly you don't have any weapons to handle those
>> orcs. You can't escape with portal and can't blink. Can't read recall or
>> teleport. You didn't have money for teleport staff or it did get burned.
>> Or this simply is too early for that to be in your gear. And that moss
>> keeps on blasting you with amnesia. Can't kill it because orcs are in
>> the way.
>
> Timo -- I rarely suffer more than a handful of successful Amnesia
> attacks in a game, and they are rarely associated with a dangerous
> situation.
Problem is that it is way too common form of attack. It can happen
pretty much anywhere anytime unless you have 100% saving throw.
Situation for amnesia attack this far has not been dangerous because
amnesia is not dangerous. You don't assosiate it to danger that happens
two turns later because you are not hindered by this suggested form of
amnesia.
Of course this is mostly not the case of warrior-type and because you
play divers you don't know what it can be for pure spellcasters.
> Compare
> this with ~120 monsters with various confusion attacks--you are just
> getting hammered with it all game long. You should be able to work
> around amnesia; working around confusion is just not possible.)
Confusion can be dealt with resistance and CCW which are workarounds for
confusion. And you are right, you *SHOULD* be able to work around
amnesia. How do you do that without perfect saving? That's right. You
don't. Not until you have lots and lots of healing potions.
> And things like the orc problem you mention just aren't that
> dangerous--just leave the room, and you are safe.
If you can. I hope you can. You can't use spells. That means that as
long as amnesia is active you are vulnerable. Very vulnerable.
> Or shoot the moss
> with an arrow--even a priest can't miss.
Except when there is monsters between that moss and you.
> I like the idea of a rare but dangerous new attack mode.
I don't. Not this kind of attack.
Timo Pietilä
Steve Clark
Erm. Isn't there a saving throw against amnesia? Surely that would
mitigate most of the effects for high level Spellcasters and I notice it
doesn't affect sticks so you can take out that pesky memory moss with a
wand of stinking cloud.
PS. I agree with Eddie that making squelching more complicated is not the
way to go.
--
Steve Clark
{sjc...@ormail.co.uk using a 129Mb SA 6.06 RiscPC}
Elsairon
news:4ef8a15a...@ormail.co.uk...
> In article <f5nlue$2jsa$1...@news.vol.cz>,
> will_asher <will_...@yahoo.com> wrote:
>> On 2007-06-24 22:18:06, Timo Pietilä <timo.p...@helsinki.fi> wrote:
>
>> > Timo Pietilä wrote:
>> > > Andrew Sidwell wrote:
>> > >
>> > >> - Amnesia (the monster attack) is now a timed effect (a la
>> > >> confusion), and has makes you forget how to read
>> > >> scrolls/pray/cast spells one time in two.
>> >
>> > > You either need to add resistance or one-turn cure for this effect.
>> > > Othervise it can be way too deadly.
I think this is the most sane solution, if we want to keep the effect.
>> > Earlier than potion of healing, that is. I don't like the fact that
>> > monster can bypass the lock. Blindness, confusion and sound should be
>> > only ones that can disrupt your spellcasting. This breaks one of the
>> > fundamentals of the game.
>> >
>> > Timo Pietilä
>
>> Look at it as slightly changing one of the fundamentals of the game. I
>> think it is an appropriate change. Most monsters that have FORGET are
>> late enough for people to have healing potions. Making healing potions
>> more common would be nice to go along with this change, though.
>
> Erm. Isn't there a saving throw against amnesia? Surely that would
> mitigate most of the effects for high level Spellcasters and I notice it
> doesn't affect sticks so you can take out that pesky memory moss with a
> wand of stinking cloud.
>
>
> PS. I agree with Eddie that making squelching more complicated is not the
> way to go.
>
Using k! is better than destroying each individual item, but
having all squelched items not visible/pick-upable is much
better for lessening keystrokes. (which is the whole point
of squelch, no?)
A toggle [see/not see] squelched might give both camps solution.
--
Elsairon
Timo Pietilä
> In article <f5nlue$2jsa$1...@news.vol.cz>,
> will_asher <will_...@yahoo.com> wrote:
>> On 2007-06-24 22:18:06, Timo Pietilä <timo.p...@helsinki.fi> wrote:
>
>>> Timo Pietilä wrote:
>>>> Andrew Sidwell wrote:
>>>>
>>>>> - Amnesia (the monster attack) is now a timed effect (a la
>>>>> confusion), and has makes you forget how to read
>>>>> scrolls/pray/cast spells one time in two.
>>>> You either need to add resistance or one-turn cure for this effect.
>>>> Othervise it can be way too deadly.
>>> Earlier than potion of healing, that is. I don't like the fact that
>>> monster can bypass the lock. Blindness, confusion and sound should be
>>> only ones that can disrupt your spellcasting. This breaks one of the
>>> fundamentals of the game.
>>>
>>> Timo Pietilä
>
>> Look at it as slightly changing one of the fundamentals of the game. I
>> think it is an appropriate change. Most monsters that have FORGET are
>> late enough for people to have healing potions. Making healing potions
>> more common would be nice to go along with this change, though.
>
> Erm. Isn't there a saving throw against amnesia? Surely that would
> mitigate most of the effects for high level Spellcasters
For priests yes. Mages OTOH can have surprisingly low saving throw very
long into game. 30+race+0.9/lvl+Wis-bonus.
> and I notice it
> doesn't affect sticks so you can take out that pesky memory moss with a
> wand of stinking cloud.
That would require carrying YAmore item in inventory. But if change is
done maybe you need to carry staffs of teleportation and some powerful
ball-spell rods/wands.
> PS. I agree with Eddie that making squelching more complicated is not the
> way to go.
I haven't tried it, but for sound of it that did sound like a bad thing.
NPP squelch was good one and it was in 3.0.7s3. I don't know why they
choose to change that.
There is one very bad thing mentioned:
"Worthless items and emptied chests are automatically inscribed with
"squelch", saving some micromanagement."
If I have *remove curse* and enchant weapon I might want to keep Morgul
bows. Those have See_inv which makes them possibly very good if you can
enchant them up.
Timo Pietilä
Eddie Grove
> There is one very bad thing mentioned:
>
> "Worthless items and emptied chests are automatically inscribed with
> "squelch", saving some micromanagement."
>
> If I have *remove curse* and enchant weapon I might want to keep Morgul
> bows. Those have See_inv which makes them possibly very good if you can
> enchant them up.
Actually, ordinary rCurse works on bows, no need for the * version.
You can get around this problem by auto-inscripting bows. I auto-inscript the
big 3 weapons with "no" [anything will do] so that they don't ever squelch.
OTOH, there really needs to be a setting to squelch good/avg/cursed items, but
not to squelch worthless.
Eddie
Nick
happy with :)
Some suggestions for those:
Amnesia - how about making it a possible side effect of chaos attacks, then
making RChaos cover it (maybe someone else suggested that - imitation is the
sincerest form of flattery...)? And I think healable by !CCW would be fair,
too. I agree with Timo that it's quite a common occurrence - Angels and
Vampires (sounds like a TV show) are the most frequent culprits.
Squelch - this is really just a UI issue, IMHO, and one which people are very
particular about. I would argue that this is one case where _more_ options are
needed. It should not be too hard to make everyone happy (except the people who
are religiously opposed to all squelching).
Antony Sidwell
> I haven't tried it, but for sound of it that did sound like a bad thing.
> NPP squelch was good one and it was in 3.0.7s3. I don't know why they
> choose to change that.
I'll quote this from the comment at the top of squelch.c, which sounds
like it's Andrew's reasoning for the change between versions, though I'm
sure he'll post himself before long:
"The squelch code has a long history. Originally it started out as a
simple sval-dependent item destroyer, but then ego-item and quality
squelch was added too, and then squelched items on the dungeon floor
were marked by purple dots, and by this time the code was quite
unmaintainable and pretty much impossible to work with."
So whether this is the final solution for squelch in V I'm not sure, it
might just be that this version is more amenable to being altered in the
future to something better.
For myself, I'd've been happy with no squelch at all, but if we're
having it, I find the current solution is fine for destroying things
you're carrying. I'd prefer Eddie's "just hide the stuff" approach for
stuff you've not picked up, though, and would hope for that in the
future if squelch is here to stay.
As he says, if you're doing it, do it properly, and make it so that
things you consider junk don't interfere with any part of the game, be
that running or even your view of the things that do interest you.
Ignoring them is what you want to do, so ignoring them is what should be
catered for. Otherwise it's like having using an /ignore command on IRC
because someone's chattering inanely, but still having the program tell
you when someone says something, just not what they've said. It's
pointless, annoying, and clutters up your view of things you do want to
read.
I don't see any advantage to having an actual auto-destroy for things
you aren't interested in, that just brings up the whole "should it take
a turn, should it not, does that give an advantage or not" sort of
question. Feel free to enlighten me as to the advantages though,
because I'm sure someone can see one (or more!)
The down side I can see would be that you would lose the best bit about
squelch, which is the idea that merely by treading on a pile of armour,
weaponry or whatever you erase it from existence with a satisfying
*squelch* sound.
Phil Cartwright
> It's not the same thing as squelching scrolls of trap detection when I am
> carrying 9 rods that do the job, or squelching 100% junk items like ruined
> chests, or squelching non-ego torches when I have the Phial, or
> MagicForBeginners with a warrior at 4900'. I do not want to see where those
> items are.
Well, then, isn't the fix simple?
- Dump the "destroy" command entirely.
- Squelch has two levels, "boring" and "junk". Spots with no interesting
items but some boring items become purple (or whatever) dots. Spots with
only "junk" look like normal floor. Neither disturbs your running, and
the latter don't attract the "l" command (but can be manually targeted).
If modern versions still do compacting, junk is compacted first, then
boring, before anything else.
- Stepping on purple dots or "junk" has no blatant effect and nothing is
destroyed. Autopickup rules and the like ignore the affected items
whether the spot also has interesting items or not. Any spot with items,
however it looks, allows manual pickups and manual examination of the
list of items.
- Changing object types among interesting, boring, and junk promptly
updates the display appropriately. Floor dots may turn purple or develop
an object symbol, or vice versa.
- Objects with no resale value can be automatically classified as junk,
and {good} weapons as boring once yours is {excellent}, though you might
want to override this. That not-enchanted-enough mace of disruption of
extra attacks, given a few enchant weapon scrolls, might be a contender.
In a band that has reforging or scrolls of artifact creation a plain
mace of disruption might be very interesting indeed even if it's (+0,+0)
and you're tooling around with a Bastard Sword of Westernesse (3d6)
(+23,+27) (+2) or Holy Avenger (3d6) (+23,+27) (+4).
Phil Cartwright
> - Stepping on purple dots or "junk" has no blatant effect and nothing is
> destroyed. Autopickup rules and the like ignore the affected items
> whether the spot also has interesting items or not. Any spot with items,
> however it looks, allows manual pickups and manual examination of the
> list of items.
A further suggestion: the list should color-code junk as grey, and shove
it to the end of the list, and color-code "boring" as purple and shove
it above the junk but after everything interesting. (Purple or whatever
-- make it match the dot color, in case you go with something other than
purple to differentiate it from the existing purple dot squelch.)
momo125
Auto kill unwanted items and non egos (with good Psuedo). I can set it to not
destroy restores and then just quaff them when I walk on them.
I don't like purple dots and not showing on the screen make treasure rooms look
smaller
Hugo Kornelis
>Angband is soon to have a new official release.
Hi Andrew,
Congratulation on -almost- releasing a new official version of Angband.
It's been too long, due to the confustion about maintainership, since
the last release.
>Changes
>=======
>(All changes from 3.0.6 are mentioned.)
(snip)
The list is impressive, and I like most of the changes. I don't have
time to test (heck, I barely have time to play anymore), but I am
looking forward to seeing most of the UI changes in practice.
To add my two cents to the squelching debate, I have played NPP and like
it's implementation of squelch. I haven't played Ey, but based on your
description I think I like it's style of squelching less. I plea for
changing the squelch implementation to NPP style.
For the rest, it looks like I'm going to have to do some work to update
the spoilers for 3.0.8, and LOTS of work to either move changes to my
"work in progress" Dutch version of Angband or retranslate all sources
(but I will definitely base Dutch Angband -if I ever get to finish this
project- on 3.0.8 rather than 3.0.6, if only for the mouse support).
Thanks for taking up the role of maintainer and providing the community
with a new version of Angband to play (and to grumble about <g>).
Best, Hugo
--
Angband spoilers: http://www.juti.nl/hugo/Angband/Spoiler/index.htm
Angband UI Patch: http://www.juti.nl/hugo/Angband/UIpatch/index.htm
--
Strike & Co.
Eddie Grove wrote:
>
> I used to squelch all armor/weapons late in the game, and run around stepping
> on dots to see what was an artifact. I've decided this is a bad thing.
Why?
Once I get decent kit, I'm used to breaking everything to find
artifacts; what's so bad about that?
-Strike & Co.
Eddie Grove
I consider it abuse, that's all. I think the 'k'ill command should just
delete the item from the game, artifact or not. If preserve on, unided, the
artifact can be regenerated. It is absurd to think that a starting hobbit
rogue could destroy platearmor in one player turn, which is the implicit
current situation. IMO 'k' is shorthand for drop and ignore permanently.
Of course, I think staves of perception should perform strong pseudo on
everything in LOS, or there should be some equivalent to that.
You can be sure I hate wading through junk more than most people.
Eddie
Phil Cartwright
> It is absurd to think that a starting hobbit
> rogue could destroy platearmor in one player turn, which is the implicit
> current situation.
Sure he can -- he has a torch or lantern, right? :)
I basically agree though -- and said elsewhere "remove the destroy
command and make a true ignoring-stuff-squelch".
Hallvard B Furuseth
> That is not slight change. The lock has been there from
> frog-knows. Breaking it means removing the lock, not changing it.
What's the lock?
--
Regards,
Hallvard
camlost
> Timo Pietilä writes:
>> That is not slight change. The lock has been there from
>> frog-knows. Breaking it means removing the lock, not changing it.
>
> What's the lock?
>
If you have free action and resistance to sound, confusion, and
blindness, I don't think there's anything that can increase your fail
rates on spells or prevent you from casting spells or prevent you from
reading scrolls. Except for melee stun attacks? Not sure about the
last bit.
Joshua
Timo Pietilä
That's correct. And that melee-stunning used to be covered by sound resist.
Timo Pietilä
magnate
In some variants fail rates increase with very low mana. Not sure if
this happens in V - but you can of course get to zero mana and not
cast any spells. The real point about the lock is that you can
*always* read a teleport scroll - unless you're KO'd.
CC
Phil Cartwright
> In some variants fail rates increase with very low mana. Not sure if
> this happens in V - but you can of course get to zero mana and not
> cast any spells. The real point about the lock is that you can
> *always* read a teleport scroll - unless you're KO'd.
For that, you don't need resist sound, although it might help to avoid
getting KO'd by plasma or gravity hounds.
Timo Pietilä
> On Jun 26, 9:22 pm, Timo Pietilä <timo.piet...@helsinki.fi> wrote:
>> camlost wrote:
>>> Hallvard B Furuseth wrote:
>>>> Timo Pietilä writes:
>>>>> That is not slight change. The lock has been there from
>>>>> frog-knows. Breaking it means removing the lock, not changing it.
>>>> What's the lock?
>>> If you have free action and resistance to sound, confusion, and
>>> blindness, I don't think there's anything that can increase your fail
>>> rates on spells or prevent you from casting spells or prevent you from
>>> reading scrolls. Except for melee stun attacks? Not sure about the
>>> last bit.
>> That's correct. And that melee-stunning used to be covered by sound resist.
>
> In some variants fail rates increase with very low mana. Not sure if
> this happens in V
It doesn't. You have increased fail rate if you don't have mana to cast
the spell. Casting spell without mana can hurt your CON and cause short
paralyzation (through FA).
> - but you can of course get to zero mana and not
> cast any spells.
That is bad playing. You can try to cast a spell without any mana, but
risk for failing is very high.
> The real point about the lock is that you can
> *always* read a teleport scroll - unless you're KO'd.
I used to play priests without any scrolls, healing potions or or
wands/staves once I had the lock. I collected potions of restore mana
and carried those. Now because melee-stunning I carry CCW. I also
started to carry small amount of teleport scrolls "just in case", but I
hardly ever use any.
Timo Pietilä
Andrew Sidwell
> Angband is soon to have a new official release. Before that, though, I
> want to put out a prerelease version for testing purposes. I believe
> this version to be mostly bug-free and fairly consistent, but you may
> find otherwise, and if you do, I want to know! If no problems are
> found, I'll release 3.0.8 final without any modifications on 1st July.
>
>
> Download at:
>
> Source code: http://rephial.org/downloads/angband-3.0.8pr1.tar.gz
> Windows: http://rephial.org/downloads/angband-3.0.8pr1-win.zip
> RISC OS: http://rephial.org/downloads/angband-3.0.8pr1-ros.zip
Mac OS X port now available at:
http://rephial.org/downloads/angband-3.0.8pr1-osx.dmg
My apologies for the slowness here. :) I see that since only two
changes proved massively controversial, doing a prerelease can't have
been that bad an idea. The great thing about being maintainer is that
when you threaten to make a release, people remind you (sometimes
forcefully) of all the things that slipped one's mind. :)
Thanks to all the support-ful people who sent mail or replies.
Looks like CCW curing amnesia will be fine for everyone. I'm going with
that, along with another look at the monsters who have amnesia.
Squelch
-------
The squelch code started out as the NPP verison, which is basically the
Blackston patch, plus ego-item and quality squelch, plus the purple dot
(a horrible practice), plus always-pickup and never-pickup options, plus
autoinscribe. No-one really took the time to fix the code so it was in
even so much as the same style, and certainly no-one who ever cared
about interface looked at it (or if they did, they got frightened away).
My changes basically involved rewriting it so I could actually edit it.
I nuked:
* purple dotting, because it's a hack
* ego-item squelching, since I don't like it
* seperate categories for different kinds of armour, weapons, etc,
because I would just have to move along the list and alter them one at a
time
* "never pickup", because frankly, it seems redundant
* individual item kind squelch for weapons, armour, etc., all of which
are covered fine by quality squelch for my purposes
* auto-squelch, since the way it was done was icky
So basically, I did the above for usability, using my own experience as
the use-case. When people say "I want NPP squelch", which particular
aspect do you miss? If it's the purple dot or ego-items, it's not
coming back. :) It seems that by "NPP squelch", people actually mean
"destroying items in the backpack automatically", and I can do that.
It's a five-line job.
"magnate" wrote that squelch is about saving keypresses, but for me it
isn't, it's about reducing mental burden of looking through the
inventory list for what I want to destroy. I did stuff the Ey way just
because for my purposes, "k!" is fine. I understand now that
keypress-optimisers don't like it. :)
Eddie Grove wrote:
> If you squelch !restoreDex, then get dex drained by time hound even
> though you have sustains, when you unsquelch !restoreDex any potions
> you have been ignoring should appear on the screen.
>
> *** Squelch should be about ignoring, not about removing or
> destructing. ***
Destruction, for me, has always meant "I will never care about this
object again". If you want to write a patch to change the semantics of
"destroy", then I'm actually quite interested in it, regardless of
whether it works on V or NPP.
Antony (Sidwell) has written a patch which ignores items on the floor
that would be marked as squelch, which should keep you happy along with
auto-squelch. Note: monsters will still give messages when picking up
junk, but that's tiny.
There have been some quite complex approaches to squelch. I think that
really, the problem lies not there but rather with broken object
handling (with worthless items, generation, and the ID system). As far
as I'm concerned, squelch should be only an interim solution; having
automatic systems to combat boredom introduced by other automatic
systems misses the issue of "why have the boredom introduced in the
first place?".
I will be proposing a new ID system after 3.0.8 is out of the door which
should solve a bunch of issues with the current one, but I'm not sure
how people will react.
Two other notes: I particularly like making destroying artifacts
possible, and have them just turn up again later. I also plan to make
"destruction" a 0-energy action.
That was too long and a couple hours have passed since I started
writing. Apologies for spelling mistakes. :)
--
Andrew Sidwell
http://rephial.org/ -- the home of Angband
My email address changes monthly, and is the first three letters of the
month (in English), followed by the last two digits of the current year,
@entai.co.uk.
Andrew Sidwell
> For the rest, it looks like I'm going to have to do some work to update
> the spoilers for 3.0.8, and LOTS of work to either move changes to my
> "work in progress" Dutch version of Angband or retranslate all sources
> (but I will definitely base Dutch Angband -if I ever get to finish this
> project- on 3.0.8 rather than 3.0.6, if only for the mouse support).
Well, I'd like to move as much text out of the game as I plausibly can,
which will make your job easier, I think. Be warned that 3.0.9 (early
2008 release) will be making fairly big changes to the insides of the
game too, so you might want to hold off til then...
On a related note, I feel a twinge of guilt when making sweeping
interface changes because of the borg... My apologies to APW! :)
Tagore Smith
> > > Oh well, no future V for me.
I've felt the same way recently, but over different issues. I think
I've made my peace with junk, rightly or wrongly. But I'm glad you
posted something this blunt before I did. At any rate, I'm glad there
is an active maintainer now, even if our visions *cough* differ. At a
certain point a ship must have a captain. I will say that I think it
is a mistake to make ook, rather than r.g.r.a, the main place for
discussing changes to V- many people on r.g.r.a will then be presented
with firm decisions that they would rather have had the chance to
weigh in on.
> This is almost diametrically opposed to what I want. While I am opposed to
> the identify spell, I believe in learning by testing, not by generating
> identified.
This sounds like it would be interesting for a few games. I have to
wonder if it would not be even more tedious than the current system a
few games after that- I do test a lot of things early because of lack
of ID, and honestly, I find that tedious. I'm aware of the problem,
and have no solution, but I am not sure this is the right one either.
Never let it be said that I lit a candle when I could have cursed the
darkness.
>In fact, I don't think you should
> see damage dice until you pick up a weapon.
This makes some sense. In V, you often know exactly what things are
after detecting, as odd-dice artifacts seem much more common than odd
dice non-artifacts. In NPP you can at least be pretty sure they are
ego, and that matters quite a bit in the early game (recent warrior
went to get a 2d5 branded Main Gauche that he probably would have
otherwise ignored on an early level- was good enough to keep till
2000').
> > There is an automated "player recall" system, something like monster recall,
> > which decides at what point in the game the player will want to squelch a given
> > item and immediately silently set it to squelch (this will apply from everything
> > from Potions of Slowness right up to whether that Mace of Disruption of Extra
> > Attacks is enchanted enough to be a good swap). After level generation, all
> > monsters which would be trivial to kill will automatically and silently commit
> > suicide, leaving any unsquelched drop in a neat pile on the ground in a
> > player-configured order.
>
> All wrong.
I agree, without knowing more about this- what yardstick do you
measure this with? Equipment decisions are notoriously hard to make in
some cases... hard to see this kind of system getting it right in
general.
> I am sorry to clutter this thread with what promises to be a long contentious
> discussion of squelch, but I couldn't leave such wrongheaded suggestions of my
> opinions on these matters without response.
I think it may be one of many contentious discussions. This is why I
think the move to ook as the main place to discuss changes to V is
problematic. R.g.r.a has been the center of Angband, or V at least,
for a long time, and the move to a private system controlled by one
individual doesn't sit well with me, even though I am registered at
ook. No disrespect intended to pav, the site is a great resource. I
just don't think it should be turned into the main venue for
discussion of Angband by fiat, even if by the sort of benevolent
dictator I approve of.
Andrew Sidwell
> On Jun 25, 1:22 am, Eddie Grove <eddiegr...@hotmail.com> wrote:
>
>>>> Oh well, no future V for me.
>
> I've felt the same way recently, but over different issues. I think
> I've made my peace with junk, rightly or wrongly. But I'm glad you
> posted something this blunt before I did.
I'm not sure bluntness is ever really required. Making guarded comments
like "but over different issues" and not elucidating on them isn't so
helpful either -- just because if you actually said what the issues you
have are, then I might be able to listen to what you have to say about
them. :)
> At any rate, I'm glad there
> is an active maintainer now, even if our visions *cough* differ. At a
> certain point a ship must have a captain. I will say that I think it
> is a mistake to make ook, rather than r.g.r.a, the main place for
> discussing changes to V- many people on r.g.r.a will then be presented
> with firm decisions that they would rather have had the chance to
> weigh in on.
I ask everything I ask both here and on the forums. I reply to people
wherever they ask questions. Given this, I don't know where your
complaint lies...
Tagore Smith
> Look at it as slightly changing one of the fundamentals of the game. I think it
> is an appropriate change. Most monsters that have FORGET are late enough for
> people to have healing potions. Making healing potions more common would be
> nice to go along with this change, though.
Timo is quite correct about the consequences of this. The only way you
can consider it minor is to imagine that it will be very uncommon in
critical situations. I didn't comment on ook because I knew Timo would
when it was announced here, and his objections carry more weight than
mine, as they should.
Angband makes certain promises, and skilled play, particularly for low
HD classes, depends on those promises. That is to say, if you satisfy
these requirements, your 2000 point heal will always work. Your
tele_level will always work. Etc. You are right that it changes a
fundamental of the game, but you are wrong to think it is a minor
change.
I don't mean to be rude at all here, but it is hard to say this sort
of thing online without sounding overly harsh, so I will just say it,
and I'd like you to imagine all the ameliorating body language I would
use if we had this conversation in person. Timo is one of the better
Angband players around. having won many times, and many times with
ridiculous challenge characters. I am not as good a player as Timo,
but I still have dozens of wins to my credit. I believe the sub-2000'
part of the game is still virgin territory to you. No offense, but I'd
like to suggest that you take Timo's response seriously as he is in a
better position to know how a change like this will actually change
game balance. I'm not suggesting that you bow completely to his
authority- I just think you ought to consider that he knows what he's
talking about, and that his arguments should not be airily dismissed.
I've been commenting (as aeneas) on your dumps on ook, so I know you
have the desire to modify the game, even maybe maintain a variant, and
I encourage you to do that. It's not uncommon that variant authors are
not the best players (though of course you might become a very good
player in the future), but the best maintainers listen closely to the
best players of their variant. They have to. The best players are the
best players precisely because they do understand the game balance-
they know how to exploit the strengths of their characters and the
weaknesses of the game.
Your point about the lateness of the attack and the availability of !
Heal is incorrect and not germane. !CCW is available from the
beginning, but pure spellcasters try mightily to get a lock, even
though !CCW cures the effects prevented by the lock. Why? Because V
makes certain promises, among them that if you have 0% fail and the
lock you will never be unable to cast the spell, excepting a few
avoidable circumstances (KO is the first that comes to mind, but that
cannot usually be done from a distance in one turn- I have not meleed
a grand master mystic in some time), in _one_ turn. So it doesn't
matter how many !Healing you have- you still have to use at least two
player turns to cast the spell. That is a huge breach of the promise,
and it radically changes the balance for pure spellcasters who do not
yet have a perfect save.
What bothers me about this is that it is a change thrown out in
brainstorming on a message board and immediately adopted with what
seemed like little consideration, but it is in fact a big change. A
lot of the new stuff in V seems to have got there in this way. I still
play mainly V because it balances being balanced with being fun to
play better than any variant, IMHO. Part of the reason is that
previous maintainers have been ultra-conservative about V (JLE patch
excepted). I think that changes to V ought to be approached
conservatively.
As I said in another post, I am very glad that V has an active
maintainer now, and I appreciate that Andrew is willing to put in the
work. My time is pretty much my own these days, but I know I would
still be unwilling to commit to doing a good job of maintaining V. So
all props to Andrew for doing a thankless job that benefits us all.
That said... I have mixed feelings when I look at the change log. V is
starting to look like another variant to me.
A while ago I suggested, half-jokingly, that V ought to be co-
maintained, with Timo a balance/gameplay consultant. The reason it was
a joke was that I assumed that most big changes would pass through
r.g.r.a, and be commented on by Timo. This is why I do not like the
fact that the big changes are discussed on ook now. Only when they are
a fait accompli do r.g.r.a members who do not feel like registering
for ook get to see them- and by that time they are already in the next
release.
Tagore Smith
> That would require carrying YAmore item in inventory. But if change is
> done maybe you need to carry staffs of teleportation and some powerful
> ball-spell rods/wands.
Hmm- it surprises me that you say that... I am more concerend about
the effects of pure spellcasters by this (mages more than Priests as
their saves can fail to be perfect for longer). They already have OK
ball spells. The problem is that they now need two turns to cast a
spell they would have in one before.
Andrew Sidwell
> What bothers me about this is that it is a change thrown out in
> brainstorming on a message board and immediately adopted with what
> seemed like little consideration, but it is in fact a big change.
I hate amnesia myself, Pete Mack had mentioned several times since
3.0.7s that he wanted rid of amnesia, and Andrew Doull's suggestion to
alter it seemed OK. It wasn't quite as casual as it may have seemed,
but yes, it wasn't passed through many people before it was implemented.
It was just done, just like the recharging behaviour changes which I
seem to recall caused much outrage at the time. It's better to try
something and revert it than to not try it at all.
> A lot of the new stuff in V seems to have got there in this way. I
> still play mainly V because it balances being balanced with being
> fun to play better than any variant, IMHO. Part of the reason is that
> previous maintainers have been ultra-conservative about V (JLE patch
> excepted). I think that changes to V ought to be approached
> conservatively.
Previous maintainers to V have been nothing of the sort. :) The authors
of Angband were anything but conservative, 2.5.x had plenty of gameplay
changes, and Ben came along and rewrote a fair bit, not to mention the
JLE patch or the spell changes under Robert.
If you want to stay static, keep on playing 3.0.6 -- I don't have an
issue with that, but you won't get any of the other improvements, either.
> As I said in another post, I am very glad that V has an active
> maintainer now, and I appreciate that Andrew is willing to put in the
> work. My time is pretty much my own these days, but I know I would
> still be unwilling to commit to doing a good job of maintaining V. So
> all props to Andrew for doing a thankless job that benefits us all.
> That said... I have mixed feelings when I look at the change log. V is
> starting to look like another variant to me.
It's far from thankless, actually. :) It sounds like you have
misgivings, but I haven't picked up any specific objections.
> A while ago I suggested, half-jokingly, that V ought to be co-
> maintained, with Timo a balance/gameplay consultant. The reason it was
> a joke was that I assumed that most big changes would pass through
> r.g.r.a, and be commented on by Timo.
As far as I can remember, big changes get announced by maintainers with
the release of a new version and people object only after it's been
released. :) Though I do mean to ask Timo for a list of gameplay
issues, at the suggestion of various people, so I can tootle off and fix
them. I'm not a great player; I know that, and I'm happy to take as
much advice as possible from people who are.
> This is why I do not like the
> fact that the big changes are discussed on ook now. Only when they are
> a fait accompli do r.g.r.a members who do not feel like registering
> for ook get to see them- and by that time they are already in the next
> release.
Are you talking of the forums or the development mailing list? I think
you might be misperceiving what's going on -- if someone suggests
something over there, I reply over there, and that's all there is to it.
"Big changes" are no more discussed on oook than they are on r.g.r.a
(and I have a set of files which summarise the massive debate that went
on when I took over as maintainer, ready for future use and tweakage),
on the dev list, or with Antony (my brother) in real life, or on IRC.
If you want to keep track of the changes to the game, you want to keep
an eye on the trac timeline:
http://dev.rephial.org/trac/timeline
So please, don't believe I'm trying to cut out r.g.r.a. It's just not
true. :)
Eddie Grove
> Destruction, for me, has always meant "I will never care about this
> object again". If you want to write a patch to change the semantics of
> "destroy", then I'm actually quite interested in it, regardless of
> whether it works on V or NPP.
The only differences I have proposed are that destroy should work on artifacts
and take zero energy. I can code it if you like, but I think email overhead
is more than your doing it yourself.
> There have been some quite complex approaches to squelch. I think that
> really, the problem lies not there but rather with broken object
> handling (with worthless items, generation, and the ID system). As far
> as I'm concerned, squelch should be only an interim solution; having
> automatic systems to combat boredom introduced by other automatic
> systems misses the issue of "why have the boredom introduced in the
> first place?".
People have been saying that since Blackston wrote it. Guess what -- the
problem hasn't been addressed at all. I believe that *every single issue*
should be approached from a "try to avoid creating junk" perspective, but even
for a 100% no-brainer like removing the priestly pointy penalty, I couldn't
even get a majority to post that they agreed with me. I am forced to conclude
that people *want* junk, which they refer to as different objects for
different classes. OK, but in that case I require squelch.
Eddie
Christer Nyfalt
> There have been some quite complex approaches to squelch. I think that
> really, the problem lies not there but rather with broken object
> handling (with worthless items, generation, and the ID system). As far
> as I'm concerned, squelch should be only an interim solution; having
> automatic systems to combat boredom introduced by other automatic
> systems misses the issue of "why have the boredom introduced in the
> first place?".
>
> I will be proposing a new ID system after 3.0.8 is out of the door which
> should solve a bunch of issues with the current one, but I'm not sure
> how people will react.
It might be worthwhile to compare Angband with the item system of Diablo II,
which doesn't really seemed to have the too much junk problem even
if it has a smaller inventory.
And consider point-by-point what we really gain from being different.
In Diablo II:
- Only one level of Id. (This is on the TODO list for 3.0.9, right?)
- Items are automatically pseudo-ided, and you can see their quality
with mouse-over. (At least make weak pseudo-id detect normal items.)
- No worthless, cursed or harmful items. (It's this kind of crap I want
squelch for. I mean, come on, only a newbie falls for them.)
- Only two scrolltypes, a handful of potions, no food and no devices.
(Less is more. KISS. I especially dislike that a warrior need to carry around
devices.)
Let these points serve as a basis of further discussion about object
handling.
--
Christer Nyfalt
Timo Pietilä
> Tagore Smith wrote:
>> certain point a ship must have a captain. I will say that I think it
>> is a mistake to make ook, rather than r.g.r.a, the main place for
>> discussing changes to V- many people on r.g.r.a will then be presented
>> with firm decisions that they would rather have had the chance to
>> weigh in on.
>
> I ask everything I ask both here and on the forums. I reply to people
> wherever they ask questions. Given this, I don't know where your
> complaint lies...
For example this:
"New features for Angband from variants" by andrewdoull
in ook.
This is discussion I didn't even know existed. And because web-interface
still sucks big time I won't bother participating.
Problem is not you. It is dividing angband community.
Timo Pietilä
Timo Pietilä
Except that you might not get that turn. Unless you kill that memory
moss by other means, that is. You need to carry that potion of _healing_
if suggested change goes thru *before* stat-gain to get that two-turn
spellcasting. And it might still go wrong if that memory moss succeeds
again at the same turn you used that potion.
How many potions of healing do you have before stat-gain?
I agree that this is not very big problem late in game, but early this
is very very very bad.
Timo Pietilä
Timo Pietilä
> released. :) Though I do mean to ask Timo for a list of gameplay
> issues, at the suggestion of various people, so I can tootle off and fix
> them. I'm not a great player; I know that, and I'm happy to take as
> much advice as possible from people who are.
Just to notice. I'm going on four week vacation after three weeks. So if
you plan to ask do it quickly or leave it after my vacation.
Here is short list of things what I wan't to happen in vanilla that
should not be too radical to do in very near future:
Restore chaos resist giving immunity to confusion effect (but not damage).
Related to that: remove confusion or chaos resist on several artifacts
giving both. Replace resist confusion with resist blindness in Amulet of
Magi.
Resist blindness should be ability, not resist. It doesn't give you
protection against any form of damage like all other resistances do. It
is closer to free action than any resistance. This would also allow it
to be in magi, lothlorien, blessed, *slay evil* etc.
Remove branding ammo. Or make it random slay instead. Brands are too
effective on too wide scale of monsters.
PDSM are currently mini-Bladeturners. They need to be toned down. Not
much but still.
Replace ball-spells in elven rings with variety of healings (they were
made for heal and growth and preservation, not for war).
Ranger and Rogue spells should be reversed IMO. Rogue should be the one
with more attack spells, and ranger more warrior-like with bow as main
weapon. Keep the spells that are "in theme" for both classes (like
detection spells with rogue).
Timo Pietilä
Timo Pietilä
> Andrew Sidwell <ju...@entai.co.uk> writes:
>> There have been some quite complex approaches to squelch. I think that
>> really, the problem lies not there but rather with broken object
>> handling (with worthless items, generation, and the ID system). As far
>> as I'm concerned, squelch should be only an interim solution; having
>> automatic systems to combat boredom introduced by other automatic
>> systems misses the issue of "why have the boredom introduced in the
>> first place?".
>
> People have been saying that since Blackston wrote it. Guess what -- the
> problem hasn't been addressed at all.
Problem with this is not that nobody is not willing to do that. It is
mainly that nobody knows how to do that.
Sangband has done quite a good job for it. Reduced drops, less total junk.
Timo Pietilä
Eddie Grove
> Squelch
> I nuked:
> * ego-item squelching, since I don't like it
Once you have a shield of elvenkind, you never want to look at another shield
of resist fire. I guess if you figure you already went to the trouble to id
it, how much more trouble is it to kill it afterward? I can't argue so much
with that even though I hate it. But in an improved id system where you don't
have to use a separate ?id on each individual excellent item [as in NPP with
mass id], we really see a huge gain from ego squelching to wipe out the
clutter. I've squelched a dozen ego items with a single use of mass id.
If you ever move in another direction I propose, treating egos somewhat akin
to flavors in that once you learn a particular ego you recognize it the next
time without the need for id, ego squelch would again improve gameplay a lot.
Squelching weapons is an exceedingly intricate problem. I think I basically
know how to do it, but I've been thinking about it seemingly forever and
haven't made up my mind yet exactly how it should be done.
Eddie
magnate
> Looks like CCW curing amnesia will be fine for everyone. I'm going with
> that, along with another look at the monsters who have amnesia.
By "curing amnesia", do you just mean restoring your spell failure
rates (and other "stun"-type effects) to normal? Or do you mean that
all the ID'd stuff that amnesia caused you to forget will suddenly be
ID'd again?
> Squelch
> -------
>
> The squelch code started out as the NPP verison, which is basically the
> Blackston patch, plus ego-item and quality squelch, plus the purple dot
> (a horrible practice), plus always-pickup and never-pickup options, plus
> autoinscribe. No-one really took the time to fix the code so it was in
> even so much as the same style, and certainly no-one who ever cared
> about interface looked at it (or if they did, they got frightened away).
> My changes basically involved rewriting it so I could actually edit it.
>
> I nuked:
> * purple dotting, because it's a hack
What is wrong with purple dots? You yourself say "for me ... it's
about reducing mental burden of looking through the inventory list for
what I want to destroy" - substitute "visible objects" for "inventory
list" and you should see why purple dots make perfect sense. They're
markers which tell you there is an item there which you are not
*currently* interested in. Some would have them invisible, but others
like to see the dots because you can at least examine them and see
what's there (should you be looking for a squelched restore stat
potion, for example).
> * ego-item squelching, since I don't like it
Erm ... with respect I think you have to give a little more
justification than that, maintainer or no. As Eddie has pointed out,
one you are wearing Foo of Resistance, you will never want Foo of
Resist Element. The same goes for Weapon of Westernesse over Weapon of
Slay Orc. For these reasons ego-squelching makes perfect sense, and I
would be interested to hear counter-arguments. IMO it is the single
biggest improvement NPP made to the Blackston squelch implementation.
> * seperate categories for different kinds of armour, weapons, etc,
> because I would just have to move along the list and alter them one at a
> time
??? Again, with respect, I'm not convinced by your reasoning. I think
most players agree that they would squelch {good} daggers without
squelching {good} lochaber axes, let alone blades of chaos. Ditto
squelching {average} body armour of most types but not squelching
{average} mithril chain.
> * "never pickup", because frankly, it seems redundant
Agreed.
> * individual item kind squelch for weapons, armour, etc., all of which
> are covered fine by quality squelch for my purposes
Erm. I think my comments above apply here. I agree that you don't need
*both* category-squelch and individual item type squelch (which we
might call sval-squelch) - but I maintain that you need one or the
other. You seem to be getting rid of both.
> * auto-squelch, since the way it was done was icky
Why, because it didn't take a turn? I can't think of anything else
icky about it ...
> So basically, I did the above for usability, using my own experience as
> the use-case.
Usability is debatable. Your own experience and preferences differ
greatly from mine and Eddie's, at least. I think dropping ego-squelch
and dropping both category and sval-squelch are major mistakes.
> When people say "I want NPP squelch", which particular
> aspect do you miss? If it's the purple dot or ego-items, it's not
> coming back. :) It seems that by "NPP squelch", people actually mean
> "destroying items in the backpack automatically", and I can do that.
> It's a five-line job.
It's more than that. It's the ability to define what items interest
you, and get rid of those which don't with a minimum of fuss
(keypresses, mental energy etc.).
> Eddie Grove wrote:
> > If you squelch !restoreDex, then get dex drained by time hound even
> > though you have sustains, when you unsquelch !restoreDex any potions
> > you have been ignoring should appear on the screen.
>
> > *** Squelch should be about ignoring, not about removing or
> > destructing. ***
This is precisely why purple dots are useful.
> Antony (Sidwell) has written a patch which ignores items on the floor
> that would be marked as squelch, which should keep you happy along with
> auto-squelch. Note: monsters will still give messages when picking up
> junk, but that's tiny.
Er ... the original Blackston implementation already had this feature.
Items "that would be marked as squelch" appeared as purple dots, and
there was a toggle to determine what happened when you walked on them:
ignore or destroy.
> There have been some quite complex approaches to squelch. I think that
> really, the problem lies not there but rather with broken object
> handling (with worthless items, generation, and the ID system). As far
> as I'm concerned, squelch should be only an interim solution; having
> automatic systems to combat boredom introduced by other automatic
> systems misses the issue of "why have the boredom introduced in the
> first place?".
>
> I will be proposing a new ID system after 3.0.8 is out of the door which
> should solve a bunch of issues with the current one, but I'm not sure
> how people will react.
Well, I have long been on your side here - TMJ needs fixing and
squelch is indeed only an interim solution. But as Timo pointed out,
it's an interim solution that's been needed for many years, and it's
worth making it friendly to as many users as possible. I will be very
interested in your proposals for revising item generation and ID. I
wrote quite a long post a while back in one of Eddie's TMJ threads
about how item generation should be much more specific to the current
character (e.g. instead of generating Raal's or WoG randomly, generate
the relevant one) - I'll see if I can find it.
> Two other notes: I particularly like making destroying artifacts
> possible, and have them just turn up again later. I also plan to make
> "destruction" a 0-energy action.
Ugh. I will plant my flag in the ground here and say that I violently
disagree with both you and Eddie on the destruction of artifacts. This
is a roguelike, not a quantum physics lesson. An artifact is still an
artifact whether it's IDd or not - it has already been generated - so
it should remain indestructible. The fact that you two think it's
cheesy that you can ID an artifact by trying to destroy it is ...
irrelevant. If you don't like that, come up with indestructible non-
artifacts to confuse people! But don't get all Schrodinger on us.
> That was too long and a couple hours have passed since I started
> writing. Apologies for spelling mistakes. :)
No problem at all - thanks for all your efforts. You're doing a grand
job and I'm sorry I disagree so much with your chosen approach to
squelching. I hope I've managed to explain why.
CC
magnate
>
> Well, I have long been on your side here - TMJ needs fixing and
> squelch is indeed only an interim solution. But as Timo pointed out,
> it's an interim solution that's been needed for many years, and it's
> worth making it friendly to as many users as possible. I will be very
> interested in your proposals for revising item generation and ID. I
> wrote quite a long post a while back in one of Eddie's TMJ threads
> about how item generation should be much more specific to the current
> character (e.g. instead of generating Raal's or WoG randomly, generate
> the relevant one) - I'll see if I can find it.
Hmm, the long post turned out to be about spreading out stat potions
to get rid of stat gain - I assume you have your own ideas about that,
connected with solving TMJ. The flash of insight on TMJ was provided
by RDH:
> Letting stat potions count as "good" (and possibly even "great") so long
> as they're for an unmaxed stat (yes, that's hackish, but sometimes play
> has to come before ideally beautiful code) might also be helpful, as
> getting "good" drops would help, not hinder stat gain.
I replied: "Dan, this is a stroke of utter genius. It's the answer to
TMJ! Wrath
of God is a DROP_GREAT if you're playing a priest or paladin, and not
otherwise. Similarly blessed weapons. You could even extend this idea
to include current status: a char without ESP would consider any item
giving ESP a good or great drop, where a char with ESP already might
not (though it is handy to have alternatives).
Yes, rewriting the drop code to take stuff like class, stats and kit
into account would be a major change - but people have always said
that the solution to TMJ had to be radically different to just
tinkering with drop rates and rarities and depths. I think this might
be it."
Not sure if it will feature in your plans, but it seems to me that
adding pieces of code to parse the player's current stats and kit and
create a temporary categorisation of possible drops would enable you
to solve TMJ without having to do much to change the actual item
generation code. All the stuff about depth checking and so on could
stay as is, along with the generation of good or great items etc. -
you simply add a routine that will redefine what constitutes good and
great.
Must do some work,
CC
Antony Sidwell
>> Eddie Grove wrote:
>>> If you squelch !restoreDex, then get dex drained by time hound even
>>> though you have sustains, when you unsquelch !restoreDex any potions
>>> you have been ignoring should appear on the screen.
>>> *** Squelch should be about ignoring, not about removing or
>>> destructing. ***
>
> This is precisely why purple dots are useful.
This is *exactly* what the current implementation of squelch in svn lets
you do. If you have hide_squelchable turned on so that such items are
invisible, then you can go to the squelch options, mark that you don't
want [whatever] squelched any more, and it will no longer be hidden from
you. So you don't have to check through purple dots, just choose the
thing you've changed your mind about from the menu and it will now be
visible again.
>> Antony (Sidwell) has written a patch which ignores items on the floor
>> that would be marked as squelch, which should keep you happy along with
>> auto-squelch. Note: monsters will still give messages when picking up
>> junk, but that's tiny.
And I fixed the "monsters giving messages" bit before it went into svn,
so I *think* ignored stuff shouldn't impinge on the game at all if you
don't want it to.
--
Antony Sidwell
Eddie Grove
> Eddie Grove wrote:
> > Andrew Sidwell <ju...@entai.co.uk> writes:
>
> >> There have been some quite complex approaches to squelch. I think that
> >> really, the problem lies not there but rather with broken object
> >> handling (with worthless items, generation, and the ID system). As far
> >> as I'm concerned, squelch should be only an interim solution; having
> >> automatic systems to combat boredom introduced by other automatic
> >> systems misses the issue of "why have the boredom introduced in the
> >> first place?".
> > People have been saying that since Blackston wrote it. Guess what -- the
> > problem hasn't been addressed at all.
>
> Problem with this is not that nobody is not willing to do that. It is mainly
> that nobody knows how to do that.
I know how to do it. I propose things to address TMJ every so often.
The group never agrees with me. Perhaps junk is a good thing.
It's been a couple months, so here are a few more I don't remember posting:
(1) Combine INT and WIS into a single stat(*). Currently wis items are junk
if you use mage spells, and int items are junk if you use priest spells. This
exists solely to create junk. There is no gameplay benefit to the current
design, unless you wish to produce items that are class-specific, i.e. junk to
other classes.
(2) Combine the mage and priest spell books into a single set. Currently mage
books are junk if you use priest spells and vice versa. Thirty-plus spells
are enough for each class, and that leaves another 12 slots [more when you
allow for overlapping spells] for multiple copies of spells if you truly
insist that a spell like id should be in a town book for one class and a
dungeon book for the other.
(3) Pick a constant like 4 for GOODTOGREAT. Whenever a monster with DROP_GOOD
produces a drop of N items according to current code, instead produce
N/GOODTOGREAT with DROP_GREAT and N%GOODTOGREAT with DROP_GOOD.
> Sangband has done quite a good job for it. Reduced drops, less total junk.
Well, if you want to go S style, start by reducing pits from 5x19 to 3x7.
Also, reduce the sizes of escorts and summonses. One of the contributions to
junk is that there are too many monsters.
(*) Actually, as posted before, my preference is to remove both, tying mana to
CON and spell failure to DEX, but I know that's far too radical for most.
Eddie
Twisted
> Resist blindness should be ability, not resist. It doesn't give you
> protection against any form of damage like all other resistances do. It
> is closer to free action than any resistance. This would also allow it
> to be in magi, lothlorien, blessed, *slay evil* etc.
This is true of resist fear too -- though I favor putting in a few
actually damaging fear attacks. (Great Wyrms of Terror, say, breathing
fear for 400+ damage. Not to mention undead with damaging fear-
inducing spells or melee hits.)
> Remove branding ammo. Or make it random slay instead. Brands are too
> effective on too wide scale of monsters.
Branded ammo damage from rangers is simply sick. Seconded.
> PDSM are currently mini-Bladeturners. They need to be toned down. Not
> much but still.
PDSM are *rare*, damned rare. I've seen two. Both were dropped by
Sauron (in separate games, both of which I won).
> Ranger and Rogue spells should be reversed IMO. Rogue should be the one
> with more attack spells, and ranger more warrior-like with bow as main
> weapon. Keep the spells that are "in theme" for both classes (like
> detection spells with rogue).
Rogue should have thief-abilities. "Spells" to steal from monsters
(one random item generated according to the monster's drop, and the
monster drops one fewer item if killed; risks waking the monster if
sleeping and hasting it if already awake) and a short-range controlled
teleport that can land inside vaults (fairly high level and high mana
cost; with some randomness to arrival point. Would not always be as
useful as it sounds, e.g. with a GCV some cells of which are too far
inside for the spell's range and all of which are normally occupied.
Cells with a ghost or other wall-walker or wall-eater could be cherry-
picked, but you'd spend a round adjacent to the would-be occupant!
Also useful for non-vault-raiding purposes, such as tactical combat
and evasion. Having the only controlled tport in the game would
differentiate rogue and make the class not as weak compared to the
others. Suggest letting you target any spot within around 10 squares,
similar to phase door's range, and actual arrival spot is a random
unoccupied square, cave_dirty or not, from a radius of 2 or 3 about
the target. A single vacant GCV cell could be hit exactly but more
usually there'd be some uncertainty as to exact landing situation. If
no suitable landing site, spell leaves you unmoved but you still use a
turn and the mana for casting it.)
If you make bigger changes to the id system, rogues might be left as
the only class with an identify spell. As things stand, they might be
given a fairly low level instant-pseudo-id spell called "Appraise" or
some such and a high level dungeon book spell "Lore" that provides
*ID* capabilities. An in-theme new detect would detect all, but only,
monsters that have drops of gold or "good" or better items. This
rogues might get instead of a general detect monsters spell, with a
general detection depending on a rod like with warriors or a deeper
spell like with priests.
Other in-theme stuff could be monster traps (as a spell perhaps) or
temporary poison branding of melee attacks (not especially powerful
except in the fairly early game).
Combat differences might be made and some of the combat system
screwiness in V addressed without resorting to (ugh) O combat. For
example, the value of weapon weight and STR in getting criticals
changed so criticals are a bigger deal in melee. Warriors getting a
boost to criticals instead of lots of blows, but roughly the same
damage output. Rogues should get lots and lots of blows with light
weapons but do terribly with big weapons. If they get an early temp-
poison-brand spell, they might get a late temp-1-extra-attack spell
(cumulative with weapon extra-attack bonuses). If they are modified to
get six blows (only with the lightest weapons mind you) this gives
them potentially nine with a Dagger of Extra Attacks (and a 3d4 dagger
at that, if facing a non-poison-resistant foe). That's nothing to
sneeze at, particularly if the dagger is also heavily enchanted.
Another thing could be ... speed. A short boost of more than +10 (say
+15, or +5 but cumulative with haste self), in a relatively deep
spellbook. Earliest and lowest-mana access to haste self (with mages
getting it, from the same book, at a somewhat higher level and mana
cost).
Twisted
> Erm ... with respect I think you have to give a little more
> justification than that, maintainer or no. As Eddie has pointed out,
> one you are wearing Foo of Resistance, you will never want Foo of
> Resist Element. The same goes for Weapon of Westernesse over Weapon of
> Slay Orc.
Even a Dagger of Westernesse (1d4) (+11,+13) (+1) over a Scythe of
Slicing of Slay Orc (8d4) (+17,+20)? ;)
Twisted
> For example this:
>
> "New features for Angband from variants" by andrewdoull
>
> in ook.
>
> This is discussion I didn't even know existed. And because web-interface
> still sucks big time I won't bother participating.
>
> Problem is not you. It is dividing angband community.
Seconded. (Re: web interfaces sucking, re: merits of newsgroups over
web boards, and re: dividing community.)
But you already knew my opinions on such matters as I've been quite
vociferous about them in the past.
Speaking of terrible web interfaces, WTF is with google groups these
days? Following up to multiple posts in a big thread like this is a
real pain now because for some reason, if you submit one followup and
then go to compose another one you get a read-only text area to
compose it in! Yep -- read-only, as in no blinking cursor (though it
does let you select text). This didn't used to happen. Hitting back
and forward works to fix it, but here's another even wackier problem
-- you have to hit back *twice*, then forwards once. Also have to hit
back twice generally to back out of a thread after posting something
to it (and otherwise only once). First back button click simply has no
effect, but the second works as expected. Strange.
If I could only find a free news server. No ISP seems to provide one
for customer use any more, and the only one that I've found with
research that doesn't have any serious "real names and unmunged emails
so we can spam you" type restrictions is that aeio one, or whatever it
was. And that one has a tiny limit on posts per day that put me off as
I recall.
I wonder if there's been a decision from on high to kill usenet by
some sort of powers-that-be. ISPs are all big telcos these days, and
there tend to only be one or two ISPs in any given service area, often
only a handful of players nationally. I find the dropping of NNTP by
nearly all major ISPs in a short period of time to be suspicious, the
same as if they all raised prices a similar amount at the same time
which would make me suspect collusion. Between the constant push
toward web forums (of those wanting to have discussion groups, and of
individual newsgroup users once a parallel web forum's been created)
and ISPs apparently all getting "drop NNTP" marching orders from some
big shot (and who could order all of the big telcos about and be
obeyed unquestioned? Not even the president ... someone in the shadows
methinks, probably in one of the huge media conglomerates that
actually seem to be the present risk of world domination...) it looks
like there's a concerted and organized attempt to kill the biggest
experiment in free speech ever, orchestrated by an obscure,
unaccountable and unelected, and evidently very powerful individual or
agency...one that clearly doesn't like any decentralized, uncensorable
forum existing, one who wants all speech to have to go through
centralized servers where it can be more readily subjected to
censorship.
Well, to whomever is apparently trying to strangle usenet, a hearty
FUCK YOU! (One more reason to hate this unsubtle shove towards using
webboards -- you can't use the proper words for some occasions without
at best getting it changed to stars or something like that by some
unaccountable so-called moderator. On any of them. So much for free
speech.)
Eddie Grove
> Speaking of terrible web interfaces, WTF is with google groups these
> days?
I can't even figure out how to show a thread in tree format any more.
Someone is working hard to make it annoying.
Eddie
Twisted
You can show a thread in tree format on GG? :P
I've taken to using a public nntp server to read (which doesn't post
or have binaries, naturally) and just find the ones I want to respond
to in GG to post. It got to be too big a PITA to read news with GG's
interface, especially when it keeps changing all by itself.
Between GG's shoddiness and everchangingness, Microsoft's greed, and
the extreme potential for lockin and data being held hostage, I find
it simply unbelievable that anyone is still seriously pushing the idea
of "software as a service". Yeah -- it's just what every computer user
wants. You'd get the same crappy software they're used to, only you'd
have to pay for it every month or it stops working, instead of just
once or whenever there's something useful in the way of new features
in a new major version. And the interface is suckier. And your
documents are who the Christ knows where, readable by God only knows
how many other people with the devil knows what motives and aims,
rather than being yours to keep confidential, delete, encrypt, back
up, restore, keep after you stop paying, and generally do with as you
please. And of course the interface, the bugs, the workarounds to the
bugs, the security holes, the ways to mitigate the security risks, and
so forth all change gratuitously and unavoidably from time to time (a
problem that also plagues anything with automatic updates you can't
turn off or do without, such as antivirus apps). And of course you
have to have a good net connection. If you want to keep your data
locally (and have the choice!) doing so entails giving some remote
host real-time ability to control your computer and access its disk
drives at dangerous privilege levels -- they could snoop all your
files or delete stuff with the access for all you know, for as long as
you're still connected.
It only works even for news posting (and web search) because you've
got no real data to want to keep, removing the main threat of lock-in,
confidentiality breach, data loss, or other insecurity. The worst case
from switching vendor or the vendor screwing things up is losing your
read/unread post info. The worst case from the vendor being a peeping
Tom is your search patterns, newsgroup habits, and read/unread posts
info being seen by third parties. Of those I'd only be worried about
the search patterns; the newsgroup stuff can be guessed by anyone who
sees your postings and does a search or two anyway. Well, except for
the fact that you regularly lurk, but don't post, in that latex-and-
tentacles fetish newsgroup ...
NNTP reading-behavior privacy seems to require using Google Groups
without logging in and through an anonymizing proxy; using an NNTP
server obviously exposes your interests to the server operator. Then
again, anything that isn't encrypted is seen by the proxy, too. What's
needed is the ability to use one anonymous proxy to do a key exchange
with a site, and then another to surf the site with an encrypted
stream; only if the proxies collude can a third party both read the
stream and know your IP address, and you'd pick both proxies.
As for search privacy, it would be nice if privacy-conscious folks
could run their own search spiders. Program it to retrieve each
successive URL through a different, anonymizing proxy and none of the
proxy operators has a clue as to the pattern of retrievals to guess
your search term or even know it's part of a search. Visit any site
via a proxy manually and you've later got deniability by saying your
search robot must've went there for some reason while looking for
something perfectly innocuous. The problem being web site masters seem
to be moving in the general direction of restricting access to
anything to a) humans and b) Googlebot (and letting Googlebot index
stuff they later refuse to show humans without a login! Talk about
cheating...) On top of which, joe random can't be expected to somehow
make or obtain a working spider, let alone build a Google-sized index
of the web on their own. I could probably program a working spider to
start from some start page and follow links looking for a given search
term, and then try to follow links to sites with more and more use of
that term, or something like that, without it trying to build an
index, but I don't know if my ISP would hate that and find some excuse
to block it, or sites block it, or ... :P Besides, I'm not joe random,
who wouldn't know a compiler from a toenail clipping or a debugger
from a deerfly ... there's also the issue of a start page. The obvious
start page would be the google results for the particular search
query, which renders the whole idea somewhat superfluous...
Phil Cartwright
> On Jun 28, 5:41 pm, Eddie Grove <eddiegr...@hotmail.com> wrote:
>
>>Twisted <twisted...@gmail.com> writes:
>>
>>>Speaking of terrible web interfaces, WTF is with google groups these
>>>days?
>>
>>I can't even figure out how to show a thread in tree format any more.
>>Someone is working hard to make it annoying.
>
>
> You can show a thread in tree format on GG? :P
>
> I've taken to using a public nntp server to read (which doesn't post
> or have binaries, naturally) and just find the ones I want to respond
> to in GG to post. It got to be too big a PITA to read news with GG's
> interface, especially when it keeps changing all by itself.
aioe.org has a public newsserver that permits posting. I don't think it
has binaries though. Or is that the "aeio, or whatever it was" whose
posting limits you don't like? :P
[trimming a monster rant from out of left field]
> ... that latex-and-tentacles fetish newsgroup ...
Oooh, kinky. I'd never have guessed either, if you hadn't blurted it out
in a usenet post!
[trim ... aren't you glad I'm not going to quote the remaining six megs
of rant?]
Andrew Sidwell
> Andrew Sidwell <ju...@entai.co.uk> writes:
>
>> Squelch
>
>> I nuked:
>
>> * ego-item squelching, since I don't like it
>
> Once you have a shield of elvenkind, you never want to look at another shield
> of resist fire. I guess if you figure you already went to the trouble to id
> it, how much more trouble is it to kill it afterward? I can't argue so much
> with that even though I hate it. But in an improved id system where you don't
> have to use a separate ?id on each individual excellent item [as in NPP with
> mass id], we really see a huge gain from ego squelching to wipe out the
> clutter. I've squelched a dozen ego items with a single use of mass id.
>
> If you ever move in another direction I propose, treating egos somewhat akin
> to flavors in that once you learn a particular ego you recognize it the next
> time without the need for id, ego squelch would again improve gameplay a lot.
Well, I might as well link to my current thoughts on a new ID system
here: <http://rephial.org/wiki/NewID>. Nothing's set in stone.
Basically: one level of ID, auto-pseudo-ID of sorts, chance of
automatically knowing item charges or ego item types.
Along with that, worthless items would be removed, cursed items would
always be mixed-blessing and thus worthwhile. I think it might work,
but it may be far too radical for some. I intend to release a version
as a "testing" version with that system implemented, seperate from the
main bit of development, just to see how useable it is. If everyone
hates it, well, I might start my variant instead.
People's feedback here seems very aggressive over the 3.0.8
*prerelease*. This isn't the final version. Unless something's
mentioned in the changelog, it's not intentional. I (or any of the
other developers) are not trying to screw everyone over by stopping
macros from working, removing the 'n' action, or whatnot. They're just
mistakes, so bugreports are welcome, but threats of "no more V for me"
just aren't useful. Perhaps QA hasn't been great this time round, but
no need to get aggressive. Thanks. :)
Andrew Sidwell
> On Jun 28, 1:52 am, Andrew Sidwell <j...@entai.co.uk> wrote:
>> Looks like CCW curing amnesia will be fine for everyone. I'm going with
>> that, along with another look at the monsters who have amnesia.
>
> By "curing amnesia", do you just mean restoring your spell failure
> rates (and other "stun"-type effects) to normal? Or do you mean that
> all the ID'd stuff that amnesia caused you to forget will suddenly be
> ID'd again?
Amnesia doesn't cause you to forget ID'd stuff anymore. Amnesia now
just has the scroll and spell failure effect.
>> Squelch
>> -------
>>
>> The squelch code started out as the NPP verison, which is basically the
>> Blackston patch, plus ego-item and quality squelch, plus the purple dot
>> (a horrible practice), plus always-pickup and never-pickup options, plus
>> autoinscribe. No-one really took the time to fix the code so it was in
>> even so much as the same style, and certainly no-one who ever cared
>> about interface looked at it (or if they did, they got frightened away).
>> My changes basically involved rewriting it so I could actually edit it.
>>
>> I nuked:
>> * purple dotting, because it's a hack
>
> What is wrong with purple dots?
I seem to recall seeing the code and thinking "that's a hack", so it
never made it into V. It wasn't even in 3.0.7s.
Either you want to see items or you don't; purple dots are saying "I
can't make up my mind".
>> * ego-item squelching, since I don't like it
>
> Erm ... with respect I think you have to give a little more
> justification than that, maintainer or no. As Eddie has pointed out,
> one you are wearing Foo of Resistance, you will never want Foo of
> Resist Element. The same goes for Weapon of Westernesse over Weapon of
> Slay Orc. For these reasons ego-squelching makes perfect sense, and I
> would be interested to hear counter-arguments. IMO it is the single
> biggest improvement NPP made to the Blackston squelch implementation.
Ego-squelching looks like it should be one of the easier TMJ problems to
fix, by adjusting or rewriting the probabilities of ego-items at
different depths, combined with generating fewer items generally. I
can't do that for this release, but I plan to at least have a go at this
for 3.0.9. (I need to do some data visualisation with the current
ego-item file to get a picture of what to do, though, and become a
better player than I am right now.)
Another reason I don't like it is because it goes against the entire
grain of what ego-items are. They're meant to be excellent, quite
special items. Automatic squelching feels wrong.
The third reason is that I didn't like the UI for ego-item squelch.
It's very hard to get that right, I think, and I don't really want to
try. Nor do I want to release something with UI I'm not happy with.
And the last reason I offer is the slippery slope argument. If I put
ego-item squelching in now, I (or someone else) will get complained at a
lot for taking it out at any future point.
As Eddie said, if you've had to manually ID an item, then you can spend
the time to manually destroy it, at least for now. This might not be
optimal
>> * seperate categories for different kinds of armour, weapons, etc,
>> because I would just have to move along the list and alter them one at a
>> time
>
> ??? Again, with respect, I'm not convinced by your reasoning. I think
> most players agree that they would squelch {good} daggers without
> squelching {good} lochaber axes, let alone blades of chaos. Ditto
> squelching {average} body armour of most types but not squelching
> {average} mithril chain.
The old system didn't let anyone squelch {good} daggers while not
squelching {good} lochaber axes, so this one doesn't either.
>> * individual item kind squelch for weapons, armour, etc., all of which
>> are covered fine by quality squelch for my purposes
>
> Erm. I think my comments above apply here. I agree that you don't need
> *both* category-squelch and individual item type squelch (which we
> might call sval-squelch) - but I maintain that you need one or the
> other. You seem to be getting rid of both.
Have you downloaded it and had a look at the menus? I have the feeling
I did a poor job of describing it.
You can do sval-squelch for staffs, wands, rods, scrolls, potions,
rings, amulets, and food. You can do quality-squelch for melee weapons,
missile weapons, ammunition, armour, jewelry, and diggers.
>> * auto-squelch, since the way it was done was icky
>
> Why, because it didn't take a turn? I can't think of anything else
> icky about it ...
No, because the code was icky. It scattered the code to squelch all
over the place. Adding it back in was 8 lines' work. So this is a
non-issue.
>> So basically, I did the above for usability, using my own experience as
>> the use-case.
>
> Usability is debatable. Your own experience and preferences differ
> greatly from mine and Eddie's, at least. I think dropping ego-squelch
> and dropping both category and sval-squelch are major mistakes.
>
>> When people say "I want NPP squelch", which particular
>> aspect do you miss? If it's the purple dot or ego-items, it's not
>> coming back. :) It seems that by "NPP squelch", people actually mean
>> "destroying items in the backpack automatically", and I can do that.
>> It's a five-line job.
>
> It's more than that. It's the ability to define what items interest
> you, and get rid of those which don't with a minimum of fuss
> (keypresses, mental energy etc.).
>
>> Eddie Grove wrote:
>>> If you squelch !restoreDex, then get dex drained by time hound even
>>> though you have sustains, when you unsquelch !restoreDex any potions
>>> you have been ignoring should appear on the screen.
>>> *** Squelch should be about ignoring, not about removing or
>>> destructing. ***
>
> This is precisely why purple dots are useful.
If squelching is about ignoring, why do you want to be told about items
you want to squelch?
>> Antony (Sidwell) has written a patch which ignores items on the floor
>> that would be marked as squelch, which should keep you happy along with
>> auto-squelch. Note: monsters will still give messages when picking up
>> junk, but that's tiny.
>
> Er ... the original Blackston implementation already had this feature.
> Items "that would be marked as squelch" appeared as purple dots, and
> there was a toggle to determine what happened when you walked on them:
> ignore or destroy.
The original patch didn't ignore them, it kept them in menus, and it
kept them visible. This one hides them. As in, you'll never even
become aware of their existence. The original patch did *not* do that.
>> There have been some quite complex approaches to squelch. I think that
>> really, the problem lies not there but rather with broken object
>> handling (with worthless items, generation, and the ID system). As far
>> as I'm concerned, squelch should be only an interim solution; having
>> automatic systems to combat boredom introduced by other automatic
>> systems misses the issue of "why have the boredom introduced in the
>> first place?".
>>
>> I will be proposing a new ID system after 3.0.8 is out of the door which
>> should solve a bunch of issues with the current one, but I'm not sure
>> how people will react.
>
> Well, I have long been on your side here - TMJ needs fixing and
> squelch is indeed only an interim solution. But as Timo pointed out,
> it's an interim solution that's been needed for many years, and it's
> worth making it friendly to as many users as possible. I will be very
> interested in your proposals for revising item generation and ID. I
> wrote quite a long post a while back in one of Eddie's TMJ threads
> about how item generation should be much more specific to the current
> character (e.g. instead of generating Raal's or WoG randomly, generate
> the relevant one) - I'll see if I can find it.
I'm not so sure that item generation should pay that much attention to
the current character; I think a massive step up would just be removing
worthless items, making fewer items occur, and make those that do occur
be of higher value. It's easy to say that, but I haven't done the
requisite source-diving to make it happen yet.
>> Two other notes: I particularly like making destroying artifacts
>> possible, and have them just turn up again later. I also plan to make
>> "destruction" a 0-energy action.
>
> Ugh. I will plant my flag in the ground here and say that I violently
> disagree with both you and Eddie on the destruction of artifacts. This
> is a roguelike, not a quantum physics lesson. An artifact is still an
> artifact whether it's IDd or not - it has already been generated - so
> it should remain indestructible. The fact that you two think it's
> cheesy that you can ID an artifact by trying to destroy it is ...
> irrelevant. If you don't like that, come up with indestructible non-
> artifacts to confuse people! But don't get all Schrodinger on us.
I don't get the connection between quantum physics and Angband here, I
have to say, but I don't think that's really what you were getting at,
somehow...
Eddie Grove
> Another reason I don't like it is because it goes against the entire
> grain of what ego-items are. They're meant to be excellent, quite
> special items. Automatic squelching feels wrong.
Heck, half the artifacts aren't special. Most excellent items are useless.
E.g., you cannot have both hats "of wisdom" and hats "of intelligence" be
useful to the same character. I am sure to want to squelch one immediately,
and both as soon as I've found a non-cursed artifact hat. Do you plan to
guarantee not to generate them any more after an artifact hat is identified?
Let me go into gory detail. I made a char dump every time I returned to town
in a winning V priest game I may post some day. I just wrote a perl script to
extract every excellent item I was ever wielding in any such dump. 47 trips.
Here's the full list.
Hard Studded Leather of Resist Cold (-1
Mithril Chain Mail of Resist Lightning (-1
a Jewel Encrusted Crown of Lordliness
a Katana (Defender
a Long Bow of Accuracy (x3
a Long Sword (Holy Avenger
a Morning Star of *Slay Undead* (2d6
a Pair of Hard Leather Boots of Stealth
a Quarterstaff of *Slay Undead* (1d9
a Small Leather Shield of Resist Fire
a Small Leather Shield of Resist Lightning
A grand total of 11 ego items wielded just before recalling from town in a
full winning game. That's it, barring bugs in my script or my scanning the
output. Also 2 dragonscale armors, but those are different IMO.
The overwhelming majority of "excellent" wieldables in my game were
effectively junk. Most don't compare to artifacts, nor to better egos.
> The third reason is that I didn't like the UI for ego-item squelch.
> It's very hard to get that right, I think, and I don't really want to
> try. Nor do I want to release something with UI I'm not happy with.
That's an excellent reason. I might make the same stand in your place.
Let me suggest a solution where you don't have to associate your name to a
shoddy UI. Make the only way to squelch egos be through the 'k'ill command
answering 'E' to the confirmation, and make it be permanent. Then you can
remove the monstrosity that is the ego squelch menu. The only reason I ever
use the ego squelch menu is to fix a mistake, in which case tough nougies.
If anyone cares, they can write a decent UI for you.
> And the last reason I offer is the slippery slope argument. If I put
> ego-item squelching in now, I (or someone else) will get complained at a
> lot for taking it out at any future point.
It's a bit late for that sentiment. Those of us who have been begging for
squelch for years have been spoiled by 3.0.7s. Rightly or wrongly, we
consider that to be the current state of V.
Eddie
Timo Pietilä
You need to open some message. Then there is "options" in grey
background. There is settings for "fixed width font" and "tree view".
It still cuts out everything between [] in subject line when you reply.
Timo Pietilä
Timo Pietilä
> magnate wrote:
>> On Jun 28, 1:52 am, Andrew Sidwell <j...@entai.co.uk> wrote:
>>> * ego-item squelching, since I don't like it
>> Erm ... with respect I think you have to give a little more
>> justification than that, maintainer or no. As Eddie has pointed out,
>> one you are wearing Foo of Resistance, you will never want Foo of
>> Resist Element. The same goes for Weapon of Westernesse over Weapon of
>> Slay Orc. For these reasons ego-squelching makes perfect sense, and I
>> would be interested to hear counter-arguments. IMO it is the single
>> biggest improvement NPP made to the Blackston squelch implementation.
>
> Ego-squelching looks like it should be one of the easier TMJ problems to
> fix, by adjusting or rewriting the probabilities of ego-items at
> different depths, combined with generating fewer items generally.
I think that if you can do that then you can solve entire TMJ problem.
Main problem with TMJ is just crappy egos. Second problem is autoID junk
consumables. I think those are much easier to solve, because they are
junk pretty much to all at certain point, when egos can vary a lot.
Every game is different in equipment point of view. What is in inventory
stays pretty much same game after game.
But you *still* need squelch for egos if you want crappy egos stay away.
Current quality-squelch implementation actually is pretty much useless,
because you might not want to squelch any big-dice weapons, nazgul bows
or dragon armors. There should be some way to say "squelch everything
except these".
Timo Pietilä
Eddie Grove
> Eddie Grove wrote:
> > Twisted <twist...@gmail.com> writes:
> >
> >> Speaking of terrible web interfaces, WTF is with google groups these
> >> days?
> > I can't even figure out how to show a thread in tree format any more.
> > Someone is working hard to make it annoying.
>
> You need to open some message. Then there is "options" in grey
> background. There is settings for "fixed width font" and "tree view".
Thank you.
It is invisible to me, white text on white background, but knowing to search
for "options" I managed to highlight it and click on it.
I'd never have figured it out on my own.
Thanks again.
Eddie
magnate
> magnate wrote:
> > On Jun 28, 1:52 am, Andrew Sidwell <j...@entai.co.uk> wrote:
> >> I nuked:
> >> * purple dotting, because it's a hack
>
> > What is wrong with purple dots?
>
> I seem to recall seeing the code and thinking "that's a hack", so it
> never made it into V. It wasn't even in 3.0.7s.
Well, I'm not a coder so I'll defer to your judgement there.
> Either you want to see items or you don't; purple dots are saying "I
> can't make up my mind".
I don't understand why I seem to be alone in this. Purple dots are not
saying "I can't make up my mind", they're saying "I want to
rationalise a whole disparate set of items I currently consider
useless so that I can ignore them while still knowing they are there,
should my situation change and I actually then want one of them".
But hey, I'll leave it there. I'll try to play a version which
squelches without using purple dots, and see if I get annoyed by not
being able to recover un-squelched items. Then I'll get back to you.
> >> * ego-item squelching, since I don't like it
>
> Ego-squelching looks like it should be one of the easier TMJ problems to
> fix, by adjusting or rewriting the probabilities of ego-items at
> different depths, combined with generating fewer items generally. I
> can't do that for this release, but I plan to at least have a go at this
> for 3.0.9. (I need to do some data visualisation with the current
> ego-item file to get a picture of what to do, though, and become a
> better player than I am right now.)
This is a noble aim and one I agree with. If ego items become
sufficiently rare and good that squelching them isn't needed (because
the few undesirable ones are not too painful to destroy/drop
manually), that would be fine ...
> Another reason I don't like it is because it goes against the entire
> grain of what ego-items are. They're meant to be excellent, quite
> special items. Automatic squelching feels wrong.
... but you do still have a 100-level dungeon. It's quite likely that
many items which are truly awesome at 500' will be useless at 4500',
whatever you do to improve item generation.
Sure, it would be nice if ego-items were all "yay" moments, but the
reality is that for three-quarters of the game, most of them are
"meh". Please don't underestimate the difficulty of changing that
situation by tweaking item generation.
> The third reason is that I didn't like the UI for ego-item squelch.
> It's very hard to get that right, I think, and I don't really want to
> try. Nor do I want to release something with UI I'm not happy with.
Damn I wish I could code. I have in my head an excellent UI for
squelching, indexing item sval by ego type, but I can't create it.
> And the last reason I offer is the slippery slope argument. If I put
> ego-item squelching in now, I (or someone else) will get complained at a
> lot for taking it out at any future point.
You can blame Jeff and Diego for that ...
> As Eddie said, if you've had to manually ID an item, then you can spend
> the time to manually destroy it, at least for now. This might not be
> optimal
"might not be optimal" is a delightful understatement. That was not
one of Eddie's wiser sayings.
> The old system didn't let anyone squelch {good} daggers while not
> squelching {good} lochaber axes, so this one doesn't either.
Your premise is right, but I don't think it justifies the conclusion!
> >> * individual item kind squelch for weapons, armour, etc., all of which
> >> are covered fine by quality squelch for my purposes
>
> > Erm. I think my comments above apply here. I agree that you don't need
> > *both* category-squelch and individual item type squelch (which we
> > might call sval-squelch) - but I maintain that you need one or the
> > other. You seem to be getting rid of both.
>
> Have you downloaded it and had a look at the menus? I have the feeling
> I did a poor job of describing it.
No, I've not downloaded it yet, sorry. If you are still keen to get
more feedback I'll have a look.
> You can do sval-squelch for staffs, wands, rods, scrolls, potions,
> rings, amulets, and food. You can do quality-squelch for melee weapons,
> missile weapons, ammunition, armour, jewelry, and diggers.
This inconsistency seems strange to me - an hangover from Blackston,
perhaps. What would be so bad about sval-squelch for the latter group?
> >> Eddie Grove wrote:
> >>> *** Squelch should be about ignoring, not about removing or
> >>> destructing. ***
>
> > This is precisely why purple dots are useful.
>
> If squelching is about ignoring, why do you want to be told about items
> you want to squelch?
In case I need one of them in future. Like I said, I don't understand
why this doesn't seem obvious to other people.
> >> Antony (Sidwell) has written a patch which ignores items on the floor
> >> that would be marked as squelch, which should keep you happy along with
> >> auto-squelch. Note: monsters will still give messages when picking up
> >> junk, but that's tiny.
>
> > Er ... the original Blackston implementation already had this feature.
> > Items "that would be marked as squelch" appeared as purple dots, and
> > there was a toggle to determine what happened when you walked on them:
> > ignore or destroy.
>
> The original patch didn't ignore them, it kept them in menus, and it
> kept them visible. This one hides them. As in, you'll never even
> become aware of their existence. The original patch did *not* do that.
Apologies - I understood a less comprehensive meaning of "ignore".
Yes, they did still appear in menus. As for their being invisible, I
don't actually support that type of squelch, as per the purple dots
debate. But as I said, I'll try it and see.
> >> I will be proposing a new ID system after 3.0.8 is out of the door which
> >> should solve a bunch of issues with the current one, but I'm not sure
> >> how people will react.
>
> > Well, I have long been on your side here - TMJ needs fixing and
> > squelch is indeed only an interim solution. But as Timo pointed out,
> > it's an interim solution that's been needed for many years, and it's
> > worth making it friendly to as many users as possible. I will be very
> > interested in your proposals for revising item generation and ID. I
> > wrote quite a long post a while back in one of Eddie's TMJ threads
> > about how item generation should be much more specific to the current
> > character (e.g. instead of generating Raal's or WoG randomly, generate
> > the relevant one) - I'll see if I can find it.
>
> I'm not so sure that item generation should pay that much attention to
> the current character; I think a massive step up would just be removing
> worthless items, making fewer items occur, and make those that do occur
> be of higher value. It's easy to say that, but I haven't done the
> requisite source-diving to make it happen yet.
Here Angband is hamstrung by its own breadth. As I said above, your
aim is noble, but with nearly a dozen classes and hundreds of race-
class combinations and 100 dungeon levels, I don't think you can
reduce TMJ to the point where squelch is unnecessary simply by making
fewer better items. There will *always* be items which are junk to one
class or another, but not all. Introducing a bias towards the current
character is one way to address that problem. Another way is to reduce
the variety of the game: fewer races, classes, stats, spells, etc.
There are probably others.
> >> Two other notes: I particularly like making destroying artifacts
> >> possible, and have them just turn up again later. I also plan to make
> >> "destruction" a 0-energy action.
>
> > Ugh. I will plant my flag in the ground here and say that I violently
> > disagree with both you and Eddie on the destruction of artifacts. This
> > is a roguelike, not a quantum physics lesson. An artifact is still an
> > artifact whether it's IDd or not - it has already been generated - so
> > it should remain indestructible. The fact that you two think it's
> > cheesy that you can ID an artifact by trying to destroy it is ...
> > irrelevant. If you don't like that, come up with indestructible non-
> > artifacts to confuse people! But don't get all Schrodinger on us.
>
> I don't get the connection between quantum physics and Angband here, I
> have to say, but I don't think that's really what you were getting at,
> somehow...
http://en.wikipedia.org/wiki/Schrodinger%27s_Cat
The basic idea is that you can't look into the box to see if the cat
is alive without interfering in the probability of the outcome.
Similarly, your idea allows the user (by destroying) to affect the
outcome of whether s/he has found an artifact or not. My position is
that if s/he has found an artifact, it is therefore indestructible,
IDd or not. If you allow artifacts to be destructible when unIDd and
indestructible once IDd, you are materially altering their properties
solely on the basis of IDing them (looking into the box). This is fine
if they are subatomic particles, but not if they are physical items.
CC
andrewdoull
> On Jun 29, 1:42 am, Andrew Sidwell wrote:
> > magnate wrote:
> > > On Jun 28, 1:52 am, Andrew Sidwell wrote:
> > >> I nuked:
> > >> * ego-item squelching, since I don't like it
> >
> > Ego-squelching looks like it should be one of the easier TMJ problems to
> > fix, by adjusting or rewriting the probabilities of ego-items at
> > different depths, combined with generating fewer items generally. I
> > can't do that for this release, but I plan to at least have a go at this
> > for 3.0.9. (I need to do some data visualisation with the current
> > ego-item file to get a picture of what to do, though, and become a
> > better player than I am right now.)
>
> This is a noble aim and one I agree with. If ego items become
> sufficiently rare and good that squelching them isn't needed (because
> the few undesirable ones are not too painful to destroy/drop
> manually), that would be fine ...
Take the artifact power code. Use it to compute the relative 'power' of
ego-items. Use this relative power to boost the to-hit, to-dam, to-ac or damage
dice of egos that are too 'weak' for the level, assuming some kind of minimum
power curve. Decrease the relative frequency of all weapons and armour,
particularly at depth, excluding bows and ammunition.
Allow stacks of potions and scrolls to appear for all items out of depth
(excluding some deep ones like Healing and Life). Some weak potions/scrolls
should have their effects combined or replaced (Replace Boldness and Slow Poison
with scroll of Benediction which does both).
Allow 'dangerous' potions and scrolls to be useful, either by allowing them to
be set in traps, thrown at enemies or coating weapons, or even boost spell
damage (stinking cloud + potion of poison = higher damage stinking cloud).
Potions of Water should do damage to fire-based monsters and vampires. Potions
of Slime Mold Juice should do damage to jellies/slimes/molds.
Make wands of healing damage undead. Make wands of haste monster damage golems.
Make other wands replaced by upgrades deeper in the dungeon. Replace some staffs
with rods which have identical effects (Treasure Detection, Object Detection,
Detect Doors/Stairs).
Have all high-books treated as 'good' until one copy is found, but increase
overall rarity otherwise. Have all stat potions treated as 'good' if the stat is
less than maximum (maybe reduce chance of 'goodness' by a percentage based on
how close stat is to maximum, but this could be exploitable).
Reduce drops by approximately 33%, 66% for 'good' drops. Allow 'good' treasure
drops which have 10 * the currency value of a non-good treasure drop.
Add Slime Mold, Animated Torch, Animated Dagger monsters which drop slime molds,
torches and daggers respectively. Have archer monsters (and other shooters) drop
arrows. This allows some 'essentials' to be more common at all levels in the
dungeon, regardless of the decrease in drops.
That should cover about 90% of the TMJ problem.
Andrew
--
The Roflwtfzomgbbq Quylthulg summons L33t Paladins -more-
http://unangband.berlios.de http://roguelikedeveloper.blogspot.com
andrewdoull
> Have all high-books treated as 'good' until one copy is found, but increase
> overall rarity otherwise. Have all stat potions treated as 'good' if the stat is
> less than maximum (maybe reduce chance of 'goodness' by a percentage based on
> how close stat is to maximum, but this could be exploitable).
Have a limited number of potions (Restore Mana, Healing, *Healing*, Life),
scrolls, wands, staffs e.g. the Morgoth essentials, treated as good.
pete m
> On Jun 29, 1:42 am, Andrew Sidwell <j...@entai.co.uk> wrote:
> > The third reason is that I didn't like the UI for ego-item squelch.
> > It's very hard to get that right, I think, and I don't really want to
> > try. Nor do I want to release something with UI I'm not happy with.
>
> Damn I wish I could code. I have in my head an excellent UI for
> squelching, indexing item sval by ego type, but I can't create it.
Give it a try--with the new menu code, fixing UI should be a lot
easier for inexperienced coders.
If you need help, you can get it easily via the angband dev mailing
list, or by the IRC group.
Andrew Sidwell
> On Jun 29, 1:42 am, Andrew Sidwell <j...@entai.co.uk> wrote:
>> Another reason I don't like it is because it goes against the entire
>> grain of what ego-items are. They're meant to be excellent, quite
>> special items. Automatic squelching feels wrong.
... this wasn't one of my finest arguments ever ...
> ... but you do still have a 100-level dungeon. It's quite likely that
> many items which are truly awesome at 500' will be useless at 4500',
> whatever you do to improve item generation.
Of course. So you just don't generate the same items at 500' as you do
at 4500'.
>> The third reason is that I didn't like the UI for ego-item squelch.
>> It's very hard to get that right, I think, and I don't really want to
>> try. Nor do I want to release something with UI I'm not happy with.
>
> Damn I wish I could code. I have in my head an excellent UI for
> squelching, indexing item sval by ego type, but I can't create it.
Please do some UI mockups in a text editor or something, and send me
them, or post them here. That's the nearest alternative you'll likely get.
>> The old system didn't let anyone squelch {good} daggers while not
>> squelching {good} lochaber axes, so this one doesn't either.
>
> Your premise is right, but I don't think it justifies the conclusion!
I think you misconstrue what I said. If you want it in the form of a
syllogism:
1. NPP squelch didn't have this feature
2. I was simplifying NPP squelch for V
3. V squelch doesn't have this feature
You want that feature? Fair enough, but adding that feature was never
anything I intended to do, nor anything anyone had asked for up to now.
>> Have you downloaded it and had a look at the menus? I have the feeling
>> I did a poor job of describing it.
>
> No, I've not downloaded it yet, sorry. If you are still keen to get
> more feedback I'll have a look.
It's just that looking at the thing in question is much better and
saying what you think is much better than looking at a short textual
description of it and saying what you think. :)
>> You can do sval-squelch for staffs, wands, rods, scrolls, potions,
>> rings, amulets, and food. You can do quality-squelch for melee weapons,
>> missile weapons, ammunition, armour, jewelry, and diggers.
>
> This inconsistency seems strange to me - an hangover from Blackston,
> perhaps. What would be so bad about sval-squelch for the latter group?
Well, there's nothing that bad about it. It's just that I don't see
there's that much point in it when you have quality squelch for those
item types. Timo's suggestion in the "Squelch suggestion" thread would
make it worthwhile (setting never-squelch flags), but as it is now, I
just figured that it was an extra six/seven long menus which would just
get annoying to page through.
>> I'm not so sure that item generation should pay that much attention to
>> the current character; I think a massive step up would just be removing
>> worthless items, making fewer items occur, and make those that do occur
>> be of higher value. It's easy to say that, but I haven't done the
>> requisite source-diving to make it happen yet.
>
> Here Angband is hamstrung by its own breadth. As I said above, your
> aim is noble, but with nearly a dozen classes and hundreds of race-
> class combinations and 100 dungeon levels, I don't think you can
> reduce TMJ to the point where squelch is unnecessary simply by making
> fewer better items. There will *always* be items which are junk to one
> class or another, but not all. Introducing a bias towards the current
> character is one way to address that problem. Another way is to reduce
> the variety of the game: fewer races, classes, stats, spells, etc.
> There are probably others.
I've not played that many computer games comparable to Angband, really
-- mainly things like Theme Hospital and Wesnoth -- but I think the junk
problem can be reduced to the point where squelch is not mandatory.
I think the "scope" of the game does need to be reduced. Reduce the
number of dungeon levels a bit. Reduce the current 51 different types
of melee weapons down a bit and nuke the priest weapon restriction.
Increase variation in damage dice for weapons. Nuke pointless curses,
make ID simpler, amongst other things. However, I'll leave that to when
I have time to reply to the other thread.
>> I don't get the connection between quantum physics and Angband here, I
>> have to say, but I don't think that's really what you were getting at,
>> somehow...
>
> http://en.wikipedia.org/wiki/Schrodinger%27s_Cat
>
> The basic idea is that you can't look into the box to see if the cat
> is alive without interfering in the probability of the outcome.
> Similarly, your idea allows the user (by destroying) to affect the
> outcome of whether s/he has found an artifact or not. My position is
> that if s/he has found an artifact, it is therefore indestructible,
> IDd or not. If you allow artifacts to be destructible when unIDd and
> indestructible once IDd, you are materially altering their properties
> solely on the basis of IDing them (looking into the box). This is fine
> if they are subatomic particles, but not if they are physical items.
Heh. I understand the basics of quantum physics easily enough, I just
didn't see the connection. I thought to make artifacts destroyable even
when ID'd, though, which is probably why I didn't see the link. :)
Eddie Grove
> magnate wrote:
> > On Jun 29, 1:42 am, Andrew Sidwell <j...@entai.co.uk> wrote:
> >> The old system didn't let anyone squelch {good} daggers while not
> >> squelching {good} lochaber axes, so this one doesn't either.
> >
> > Your premise is right, but I don't think it justifies the conclusion!
>
> I think you misconstrue what I said. If you want it in the form of a
> syllogism:
> 1. NPP squelch didn't have this feature
> 2. I was simplifying NPP squelch for V
> 3. V squelch doesn't have this feature
>
> You want that feature? Fair enough, but adding that feature was never
> anything I intended to do, nor anything anyone had asked for up to now.
Oh, we want it all right, but we understand how hard it is to do correctly.
A good 3d5 maingauche with +15 damage is better than many good lochabers.
When you have a longsword with firebrand, and maybe want to squelch fire
daggers but not cold daggers, things get much more complicated. If anyone is
interested in discussing the nitty-gritty, send me mail and we can try to get
it right.
It would be a little easier if criticals did not correlate to weapon weight.
I like the idea of looking at (hit roll number / AC), maybe 2+ or 3+ should
mean a critical, but that's not so important.
Eddie
Brian Cramer
news:87ps3df...@hotmail.com...
Umm, weird argument, because I checked and double-checked in 3.0.7s3, and
you CAN, after a fashion. In the Quality squelch menu, you can squelch
weapon catagories, and even though you can't distinguish between daggers and
swords, you can distinguish between daggers (Sword catagory) and lochaber
axes (Is a lochaber axe a Hafted Weapon or a Polearm?). Thus, you can set
daggers to squelch Good quality and below, and squelch lochaber axes Average
and below. If you can do it by weapon catagory, it should be possible
(although tougher) to do it by exact weapon type. Since I don't know how
the weapon catagories are handled in the code, I'll leave it to the
programmers to deal with, but the example which you were arguing about was
possible in at least one version, so claiming both old and new versions
don't do it was a bit... false.
Nick
> Andrew Sidwell writes:
> > > On Jun 29, 1:42 am, Andrew Sidwell wrote:
> > >> The old system didn't let anyone squelch {good} daggers while not
> > >> squelching {good} lochaber axes, so this one doesn't either.
> > You want that feature? Fair enough, but adding that feature was never
> > anything I intended to do, nor anything anyone had asked for up to now.
>
> Oh, we want it all right, but we understand how hard it is to do correctly.
> A good 3d5 maingauche with +15 damage is better than many good lochabers.
> When you have a longsword with firebrand, and maybe want to squelch fire
> daggers but not cold daggers, things get much more complicated. If anyone is
> interested in discussing the nitty-gritty, send me mail and we can try to get
> it right.
>
> It would be a little easier if criticals did not correlate to weapon weight.
> I like the idea of looking at (hit roll number / AC), maybe 2+ or 3+ should
> mean a critical, but that's not so important.
ToME and FA have average weapon damage calculations as part of the 'I'nspect
command; these take into account slays, brands and criticals. That could be
part of quality squelch.
Nick.
--
"There is no safety, and there is no end. The word must be heard in silence;
there must be darkness to see the stars. The dance is always danced above the
hollow place, above the terrible abyss."
- The Farthest Shore, Ursula Le Guin
topazg
> On 2007-06-30 20:31:33, Eddie Grove wrote:
>
> >
> > Oh, we want it all right, but we understand how hard it is to do correctly.
> > A good 3d5 maingauche with +15 damage is better than many good lochabers.
> > When you have a longsword with firebrand, and maybe want to squelch fire
> > daggers but not cold daggers, things get much more complicated. If anyone is
> > interested in discussing the nitty-gritty, send me mail and we can try to get
> > it right.
> >
> > It would be a little easier if criticals did not correlate to weapon weight.
> > I like the idea of looking at (hit roll number / AC), maybe 2+ or 3+ should
> > mean a critical, but that's not so important.
>
> ToME and FA have average weapon damage calculations as part of the 'I'nspect
> command; these take into account slays, brands and criticals. That could be
> part of quality squelch.
>
> Nick.
This does have inherent risks though. For example a reasonable dagger may score
up as more valuable than an excellent executioners sword if you get 3 blows with
1 and 1 with the other -- future potential of damage needs to be incorporated to
do this safely.
The other issue of course is slays and brands. Cold brand is great in the first
1500' ft, and fairly pointless in the last 1500' as most things that would be of
concern are resistant. Slay dragon is pretty useless up to 1000' (baby dragons
are simply too easy to make it useful), and very helpful between about 3000' and
4000'.
Again, somehow this would be of great benefit when trying to weigh up quality of
items, but then you risk some crazy and controversial evaluation algorithms that
I would not be brave enough to attempt.
--
Take Care,
Graham
Pos(0.3.0a2) Alpha "Natar" XX L:1 DL:50' !A R--- !Sp w:Short Sword +0,+0
Pos(V/T//NPP) W H- D+ c-- f PV+ s- TT? d P++ M+
C-- S+ I- So B ac GHB- SQ+ RQ+ V+ F:Better monster AI (Acting like decent
players without automatically knowing where the player is - randomly roaming the
dungeon etc...)
Andrew Sidwell
There's a point at which you do genuinely have to say "we can't automate
this fully, human judgement is required". I think it was just hit.
Andrew Sidwell
http://tinyurl.com/33a4j3
It contains, TTBOMK, all of the bugs reported for 3.0.8 (barring any
major reworking of squelch), with some bugs grouped together into one
bug (like macro issues, UI niggles).
Fixed bugs are under the "Resolution: fixed" heading. This should give
people some idea of what's reported and what isn't; if I've missed
something in collecting feedback, please reply *to this post* so I can
catch up on it.
Not much is fixed right now; seems I chose a bad time to release since I
don't have regular, sensible-hours access to a computer at the moment.
Also, please note that TMJ is not going to get fixed for 3.0.8.
Timo Pietilä
> There is now a realtime-updated buglist for 3.0.8pr1 here:
> http://tinyurl.com/33a4j3
>
> It contains, TTBOMK, all of the bugs reported for 3.0.8 (barring any
> major reworking of squelch), with some bugs grouped together into one
> bug (like macro issues, UI niggles).
>
> Fixed bugs are under the "Resolution: fixed" heading. This should give
> people some idea of what's reported and what isn't; if I've missed
> something in collecting feedback, please reply *to this post* so I can
> catch up on it.
- JohnCW said in "Re: 3.0.8 V : My thoughts so far" that casting
resistance and casting resist poison crashes the game.
- 12x20 font lacks block char in Windows XP
- Identifying item onto floor doesn't give selection if there is
multiple, it just identifies first item in stack.
- Inscription for pickup doesn't work (=g).
Those I didn't found in buglist.
Timo Pietilä
magnate
> magnate wrote:
> > On Jun 29, 1:42 am, Andrew Sidwell <j...@entai.co.uk> wrote:
> >> The third reason is that I didn't like the UI for ego-item squelch.
> >> It's very hard to get that right, I think, and I don't really want to
> >> try. Nor do I want to release something with UI I'm not happy with.
>
> > Damn I wish I could code. I have in my head an excellent UI for
> > squelching, indexing item sval by ego type, but I can't create it.
>
> Please do some UI mockups in a text editor or something, and send me
> them, or post them here. That's the nearest alternative you'll likely get.
Ok, will do. I'm off on hols for the next two weeks, but I guess
you're not in too much of a hurry!
> >> The old system didn't let anyone squelch {good} daggers while not
> >> squelching {good} lochaber axes, so this one doesn't either.
>
> > Your premise is right, but I don't think it justifies the conclusion!
>
> I think you misconstrue what I said. If you want it in the form of a
> syllogism:
> 1. NPP squelch didn't have this feature
> 2. I was simplifying NPP squelch for V
> 3. V squelch doesn't have this feature
Ah, apologies - I thought you were saying the feature was not
necessary, but it was merely that the feature was not in the code you
were importing. Fair enough.
> You want that feature? Fair enough, but adding that feature was never
> anything I intended to do, nor anything anyone had asked for up to now.
If I start from the UI, that will show what kind of squelch I think is
required. For instance, I want a "do not squelch any weapon with extra
dice" switch. Extra dice are sufficiently rare for this to be a catch-
all, rather than a per-weapon switch.
> >> Have you downloaded it and had a look at the menus? I have the feeling
> >> I did a poor job of describing it.
>
> > No, I've not downloaded it yet, sorry. If you are still keen to get
> > more feedback I'll have a look.
>
> It's just that looking at the thing in question is much better and
> saying what you think is much better than looking at a short textual
> description of it and saying what you think. :)
Agreed. Will look at it before trying to come up with a UI.
> >> You can do sval-squelch for staffs, wands, rods, scrolls, potions,
> >> rings, amulets, and food. You can do quality-squelch for melee weapons,
> >> missile weapons, ammunition, armour, jewelry, and diggers.
>
> > This inconsistency seems strange to me - an hangover from Blackston,
> > perhaps. What would be so bad about sval-squelch for the latter group?
>
> Well, there's nothing that bad about it. It's just that I don't see
> there's that much point in it when you have quality squelch for those
> item types. Timo's suggestion in the "Squelch suggestion" thread would
> make it worthwhile (setting never-squelch flags), but as it is now, I
> just figured that it was an extra six/seven long menus which would just
> get annoying to page through.
True - I think indexing sval by ego type would only work if you could
keep it onto a minimum number of screens.
> I think the "scope" of the game does need to be reduced. Reduce the
> number of dungeon levels a bit. Reduce the current 51 different types
> of melee weapons down a bit and nuke the priest weapon restriction.
> Increase variation in damage dice for weapons. Nuke pointless curses,
> make ID simpler, amongst other things. However, I'll leave that to when
> I have time to reply to the other thread.
Thank you - this 'reduction in scope' I think is crucial to solving
TMJ. I wish you luck getting it past the "less is terrible" crowd.
> > IDd or not. If you allow artifacts to be destructible when unIDd and
> > indestructible once IDd, you are materially altering their properties
> > solely on the basis of IDing them (looking into the box). This is fine
> > if they are subatomic particles, but not if they are physical items.
>
> Heh. I understand the basics of quantum physics easily enough, I just
> didn't see the connection. I thought to make artifacts destroyable even
> when ID'd, though, which is probably why I didn't see the link. :)
Ah, sorry - I had no idea you intended to allow artifacts to be
destroyable once ID'd. I don't really have too much of a problem with
that. Inscriptions are there to avoid accidents.
CC
Andrew Sidwell
> Andrew Sidwell wrote:
>> There is now a realtime-updated buglist for 3.0.8pr1 here:
>> http://tinyurl.com/33a4j3
>>
>> It contains, TTBOMK, all of the bugs reported for 3.0.8 (barring any
>> major reworking of squelch), with some bugs grouped together into one
>> bug (like macro issues, UI niggles).
>>
>> Fixed bugs are under the "Resolution: fixed" heading. This should give
>> people some idea of what's reported and what isn't; if I've missed
>> something in collecting feedback, please reply *to this post* so I can
>> catch up on it.
>
> - JohnCW said in "Re: 3.0.8 V : My thoughts so far" that casting
> resistance and casting resist poison crashes the game.
I couldn't reproduce this, so it must have been fixed by some other
change, but I've filed it anyway for completeness.
> - 12x20 font lacks block char in Windows XP
Filed.
> - Identifying item onto floor doesn't give selection if there is
> multiple, it just identifies first item in stack.
>
> - Inscription for pickup doesn't work (=g).
Both of those are under "Fix pickup system".
Thanks,
Igenlode
> Angband is soon to have a new official release. Before that, though, I
> want to put out a prerelease version for testing purposes. I believe
> this version to be mostly bug-free and fairly consistent, but you may
> find otherwise, and if you do, I want to know!
Here are my observations on the only part of the game I am qualified to
judge, namely the town and shallowest levels :-)
I've rolled up a Dwarf Priest and played her down to 1100', and
encountered the following issues /en route/:
Squelch: apparently I'm the only player who actually agrees with the
maintainer that it ought to be automating routine manual destruction,
rather than trying to fix item distribution by pretending that dozens of
items were never created in the first place :-p
I get great satisfaction from stepping on a wand and seeing "You squelch
a Wand of Heal Monster" -- I find it downright distressing to approach
an object in order to find out what it is (I tend to identify objects by
stepping on them; not a good survival trait, I'm sure) and find that it
has vanished out from under me without explanation...
If the number of keypresses is an issue, how about using 'K' for
'destroy squelched'?
If I try to do a squelch while wearing a cursed item, the error message
is 'no squelched items to destroy' -- which isn't actually true! Should
probably be 'The <x> you are wearing appears to be cursed'
Even after I uncurse it, squelch still doesn't work until I take it off.
Character generation: in the initial selection screen, if you press ?
for information, then return to race/class selection, the previous
selection is forgotten, and the highlight goes back to the top item. I
assume this is a bug. At any rate, if you go to look up the special
features of a Dwarf and think "Right, I'll have one of those" and press
Escape followed by the return key, you end up with a Human instead!
The hitpoint/window/delay menu entries should be triggered by lowercase
h/w/d as well as H/W/D -- I think someone else mentioned this.
The Help menu instructs us to read the README file by pressing '1', but
this doesn't actually work. Only letters are accepted as input here.
Mouse support: it's a clever idea, but it's not possible to play with
mouse only as yet (no means to exit from shops, for a start), and as
Timo observed, clicking back into the main window -- after, for example,
typing an email -- is a potential killer.
Shift-'q' to quit seems to have been remapped to Control-c instead,
without notice.
Shops: the keypad movement keys don't work in the shops, which I assume
is a side-effect of the new-style shop menus which can extend down below
'z'. It's a pain to have to switch from using up/down movement keys to
up/down arrow keys in order to move the highlight here -- I assume we're
only supposed to be selecting things via their letters now?
-more-: the shop display shows a -more- prompt even when all the stock
is in fact on view, if the list reaches the bottom of the screen (i.e.
'a' to 'z')
Trying to sell multiple items (i.e. several weapons) to a shop involves
what seem like unnecessary extra keypresses before the second and
subsequent 's' commands. Typical example: you sell an item and get the
appropriate message displayed, then press 's' to sell the next item and
get a -more- prompt to clear instead, before pressing 's' a second time.
Maybe the shop messages aren't clearing properly?
You can't sell unwanted Torches, Lanterns and Rations back to the
General Stores -- you have to junk them. Not that vital cash-wise, but
it does seem a bit queer.
Neither the arrow keys nor the movement keys work to browse spells.
Mouse click doesn't seem to work either. Presumably here again we are
intended to select items directly via the appropriate letters rather
than scrolling to neighbouring items...
The delay factor doesn't seem to affect running speed -- I always
assumed it did, but I could be wrong. At any rate, it now seems to be
possible to zip around at such a speed that you can't actually see how
you got 'here' from 'there' -- it's more like teleport..!
Monster AI: groups of missile/spell-attacking monsters keep shooting
each other in the back, with fatal results at low levels. This makes
life easier for the player, but doesn't seem very intelligent...
With 'flow by sound' and 'flow by smell' set, uniques now seem to have
difficulty tracing the player when he blinks away to a nearby location;
they just sit there looking puzzled. This actually makes life *more*
difficult, ironically, since you have to go back within direct line of
sight before the monster you've just blinked away from will move towards
you -- in a twisty corridor with a unique who's faster than you, it
makes 'shoot & scoot' pretty useless.
Problems with prayer books: the prayers in both "Words of Wisdom" and
"Chants and Blessings" reverted to 'unknown' after they had successfully
been read (from 'untried'). They eventually displayed as expected, but
only once all the rest of the prayers in the books had been learned and
cast.
Pick-up problems: with 'always pick up' switched *off*, the ',' command
no longer seems to pick up the object you are standing on. The '-'
command sometimes picks up the object moved onto, and sometimes displays
"Press Return to pick up this object" instead: I couldn't work out why.
(In fact, I've even seen it do first the latter and then the former
during two attempts to pick up the same object!)
I had 'no stacking' enabled, which made the game's apparent insistence
on treating all floor items as stacks of one object even more
annoying... When attempting to pick up/use/fire an object from the
floor, if there is only one object present the code should surely
default to that object instead of asking you to confirm that you want to
use object 'a'. I tried switching 'query_floor' on and off, but didn't
find a satisfactory combination; at present, the game seems to be set up
to work chiefly on the assumption that you will be automatically trying
to pick up everything you step on.
When you look at objects, this 'stack mentality' produces weird messages
such as "It is on a Jet Ring" (for a Jet Ring sitting by itself on the
floor).
Quaint observations:
I liked the Scroll of Blessing entitled 'lo prius patus' -- very holy!
Second-hand clothes:
You bought a Pair of Hard Leather Boots [3,+0] for 14 gold.
You bought a Hard Leather Cap [2,+0] for 14 gold.
You drop a Metal Cap [3,-1] (r)
You drop a Pair of Soft Leather Boots [2,-1] (q)
The Pitiful-looking beggar picks up a Pair of Soft Leather Boots [2,-1].
The Pitiful-looking beggar picks up a Metal Cap [3,-1].
<grin>
--
<Igenlod...@nym.alias.net>
We live in a culture in which being well-spoken is considered
proof of insincerity.
Elsairon
news:200707020110...@remailer-debian.panta-rhei.eu.org...
> On 24 Jun 2007 Andrew Sidwell wrote:
>
>> Angband is soon to have a new official release. Before that, though, I
>> want to put out a prerelease version for testing purposes. I believe
>> this version to be mostly bug-free and fairly consistent, but you may
>> find otherwise, and if you do, I want to know!
>
> Here are my observations on the only part of the game I am qualified to
> judge, namely the town and shallowest levels :-)
>
> I've rolled up a Dwarf Priest and played her down to 1100', and
> encountered the following issues /en route/:
<snip many observations>
> You can't sell unwanted Torches, Lanterns and Rations back to the
> General Stores -- you have to junk them. Not that vital cash-wise, but
> it does seem a bit queer.
I think shops should at least buy what they sell. Maybe for a minimum
of 1 gold, or something. Many times I have stocked up on lanterns and
what not in the dungeon.
For me grabbing loot and selling it in the town is a staple of the genre.
At level 1, an ego-lantern might be just the awesome loot sellable to
buy that nice stack of rare scrolls in the black market or something.
I'm not sure what the non-sellability of some items adds to the game
other than increasing junk.
But if it is junk, I vote for sellable junk, rather than complete, totally
worthless (squelch bloat) junk.
<snip remaining observations>
> --
> <Igenlod...@nym.alias.net>
>
> We live in a culture in which being well-spoken is considered
> proof of insincerity.
--
Elsairon
Nick
wrote:
> Mouse support: it's a clever idea, but it's not possible to play with
> mouse only as yet (no means to exit from shops, for a start), and as
> Timo observed, clicking back into the main window -- after, for example,
> typing an email -- is a potential killer.
Mouse-only play is scheduled for 3.0.9. I'll be posting an outline of how it
might work (<plug> for a sneak preview, try FAangband</plug>) and asking for
feedback before too long. The clicking back into the main window is a good
point.
> Shops: the keypad movement keys don't work in the shops, which I assume
> is a side-effect of the new-style shop menus which can extend down below
> 'z'. It's a pain to have to switch from using up/down movement keys to
> up/down arrow keys in order to move the highlight here -- I assume we're
> only supposed to be selecting things via their letters now?
>
> -more-: the shop display shows a -more- prompt even when all the stock
> is in fact on view, if the list reaches the bottom of the screen (i.e.
> 'a' to 'z')
>
> Trying to sell multiple items (i.e. several weapons) to a shop involves
> what seem like unnecessary extra keypresses before the second and
> subsequent 's' commands. Typical example: you sell an item and get the
> appropriate message displayed, then press 's' to sell the next item and
> get a -more- prompt to clear instead, before pressing 's' a second time.
> Maybe the shop messages aren't clearing properly?
>
> You can't sell unwanted Torches, Lanterns and Rations back to the
> General Stores -- you have to junk them. Not that vital cash-wise, but
> it does seem a bit queer.
Shop interface issues have been discussed, um, somewhere (I suspect the forums)
- they are a work in progress.
> Neither the arrow keys nor the movement keys work to browse spells.
> Mouse click doesn't seem to work either. Presumably here again we are
> intended to select items directly via the appropriate letters rather
> than scrolling to neighbouring items...
As for shops. The new menu interface is a massive improvement, but as it was a
complete change it takes time to get all the i's dotted and t's crossed.
> Pick-up problems: with 'always pick up' switched *off*, the ',' command
> no longer seems to pick up the object you are standing on. The '-'
> command sometimes picks up the object moved onto, and sometimes displays
> "Press Return to pick up this object" instead: I couldn't work out why.
> (In fact, I've even seen it do first the latter and then the former
> during two attempts to pick up the same object!)
Again, issues being worked on (not by me, I should add :)).
> I had 'no stacking' enabled, which made the game's apparent insistence
> on treating all floor items as stacks of one object even more
> annoying... When attempting to pick up/use/fire an object from the
> floor, if there is only one object present the code should surely
> default to that object instead of asking you to confirm that you want to
> use object 'a'. I tried switching 'query_floor' on and off, but didn't
> find a satisfactory combination; at present, the game seems to be set up
> to work chiefly on the assumption that you will be automatically trying
> to pick up everything you step on.
And again.
This is all useful info - have you reported these things on the bugtracker?
Twisted
> This is all useful info - have you reported these things on the bugtracker?
He posted to rgra -- he just DID report these things on the
bugtracker. ;)
Twisted
Header@[127.1]> wrote:
> I liked the Scroll of Blessing entitled 'lo prius patus' -- very holy!
Indeed. They do say "cleanliness is next to godliness", and a Prius is
definitely quite a clean little low-emissions vehicle...
Timo Pietilä
Actually it is not. It uses around 5 liters/100km. I can give you
examples of several diesel cars that consume (and pollute) less *and*
beat Prius in every way for performance. And few gasoline cars too.
Only way to make Prius clean is to use it as pure EV, which in turn
makes it slooooww.
If you want *clean* car take a look at http://www.teslamotors.com
But this is OT so I don't rant any more.
Timo Pietilä
Andrew Sidwell
> On 24 Jun 2007 Andrew Sidwell wrote:
>
>> Angband is soon to have a new official release. Before that, though, I
>> want to put out a prerelease version for testing purposes. I believe
>> this version to be mostly bug-free and fairly consistent, but you may
>> find otherwise, and if you do, I want to know!
>
> Here are my observations on the only part of the game I am qualified to
> judge, namely the town and shallowest levels :-)
<snip a lot of useful reports, which are noted for later fixing>
> Shift-'q' to quit seems to have been remapped to Control-c instead,
> without notice.
What port are you using?
> Monster AI: groups of missile/spell-attacking monsters keep shooting
> each other in the back, with fatal results at low levels. This makes
> life easier for the player, but doesn't seem very intelligent...
>
> With 'flow by sound' and 'flow by smell' set, uniques now seem to have
> difficulty tracing the player when he blinks away to a nearby location;
> they just sit there looking puzzled. This actually makes life *more*
> difficult, ironically, since you have to go back within direct line of
> sight before the monster you've just blinked away from will move towards
> you -- in a twisty corridor with a unique who's faster than you, it
> makes 'shoot & scoot' pretty useless.
Unchanged from previous versions here.
> Problems with prayer books: the prayers in both "Words of Wisdom" and
> "Chants and Blessings" reverted to 'unknown' after they had successfully
> been read (from 'untried'). They eventually displayed as expected, but
> only once all the rest of the prayers in the books had been learned and
> cast.
An odd one, given I've not touched that code.
> Pick-up problems: with 'always pick up' switched *off*, the ',' command
> no longer seems to pick up the object you are standing on. The '-'
> command sometimes picks up the object moved onto, and sometimes displays
> "Press Return to pick up this object" instead: I couldn't work out why.
> (In fact, I've even seen it do first the latter and then the former
> during two attempts to pick up the same object!)
The latest development version has changed the options somewhat.
Particularly, I nuked the "jump (flip pickup)" command -- though if it's
widely used, it can return easily enough, I think -- and now with
always_pickup off, 'g' is the key you want to press to pick up items.
> Quaint observations:
>
> I liked the Scroll of Blessing entitled 'lo prius patus' -- very holy!
That's what you get for generating scroll names using Latin source texts. :)
> Second-hand clothes:
>
> You bought a Pair of Hard Leather Boots [3,+0] for 14 gold.
> You bought a Hard Leather Cap [2,+0] for 14 gold.
>
> You drop a Metal Cap [3,-1] (r)
> You drop a Pair of Soft Leather Boots [2,-1] (q)
>
> The Pitiful-looking beggar picks up a Pair of Soft Leather Boots [2,-1].
> The Pitiful-looking beggar picks up a Metal Cap [3,-1].
I must admit I'm totally flummoxed by this, too.
Timo Pietilä
> Igenlode wrote:
>> With 'flow by sound' and 'flow by smell' set, uniques now seem to have
>> difficulty tracing the player when he blinks away to a nearby location;
>> they just sit there looking puzzled. This actually makes life *more*
>> difficult, ironically, since you have to go back within direct line of
>> sight before the monster you've just blinked away from will move towards
>> you -- in a twisty corridor with a unique who's faster than you, it
>> makes 'shoot & scoot' pretty useless.
>
> Unchanged from previous versions here.
I have noticed this too time to time. Monsters seem to lost their
ability to walk around corners every now and then. Mostly I notice it
when I use teleport other. Once smart monster is now stuck in some
corner and don't know how to reach player even that it there is obvious
route to player.
It is rare, but it happens.
>> Second-hand clothes:
>>
>> You bought a Pair of Hard Leather Boots [3,+0] for 14 gold.
>> You bought a Hard Leather Cap [2,+0] for 14 gold.
>>
>> You drop a Metal Cap [3,-1] (r)
>> You drop a Pair of Soft Leather Boots [2,-1] (q)
>>
>> The Pitiful-looking beggar picks up a Pair of Soft Leather Boots [2,-1].
>> The Pitiful-looking beggar picks up a Metal Cap [3,-1].
>
> I must admit I'm totally flummoxed by this, too.
Why? All humans in town have TAKE_ITEM flag.
Or do you mean that those messages are created in shop?
Timo Pietilä
Andrew Sidwell
I think I may have misunderstood. I thought it was a bug report that
when dropping items, they get disenchanted...
Andrew Sidwell
> Igenlode wrote:
>> no longer seems to pick up the object you are standing on. The '-'
>> command sometimes picks up the object moved onto, and sometimes displays
>> "Press Return to pick up this object" instead: I couldn't work out why.
>> (In fact, I've even seen it do first the latter and then the former
>> during two attempts to pick up the same object!)
>
> The latest development version has changed the options somewhat.
> Particularly, I nuked the "jump (flip pickup)" command -- though if it's
> widely used, it can return easily enough, I think -- and now with
> always_pickup off, 'g' is the key you want to press to pick up items.
Forgot to add this: I'd appreciate it if when the second prerelease is
out, you could fiddle with the setup then and see if the system works
better for you.
Timo Pietilä
> There is now a realtime-updated buglist for 3.0.8pr1 here:
> http://tinyurl.com/33a4j3
>
> It contains, TTBOMK, all of the bugs reported for 3.0.8 (barring any
> major reworking of squelch), with some bugs grouped together into one
> bug (like macro issues, UI niggles).
>
> Fixed bugs are under the "Resolution: fixed" heading. This should give
> people some idea of what's reported and what isn't; if I've missed
> something in collecting feedback, please reply *to this post* so I can
> catch up on it.
This is not actually 3.0.8 issue, but a lot older. In limits.txt there
is entry
# Maximum number of objects on the level
M:O:512
That is too small. For example if you happen to have immunity to some
element and wipe out entire dragon pit of corresponding color you get
item compression. I have increased that to double 1024 in my copies and
haven't noticed any weirdness yet. If there is no coding limit which
forces that limit to be that small I would like to get it increased.
Also _when_ compression happens I would like to have it start from gold.
By the time you can get enough items that this can happen gold is
meaningless, and anyway items you collect are probably much much
valuable than gold you can get from floor. Unless gold has separate
memory space that is.
Timo Pietilä
magnate
> Timo Pietilä wrote:
> > Andrew Sidwell wrote:
> >> Igenlode wrote:
> >>> Second-hand clothes:
>
> >>> You bought a Pair of Hard Leather Boots [3,+0] for 14 gold.
> >>> You bought a Hard Leather Cap [2,+0] for 14 gold.
>
> >>> You drop a Metal Cap [3,-1] (r)
> >>> You drop a Pair of Soft Leather Boots [2,-1] (q)
>
> >>> The Pitiful-looking beggar picks up a Pair of Soft Leather Boots [2,-1].
> >>> The Pitiful-looking beggar picks up a Metal Cap [3,-1].
>
> >> I must admit I'm totally flummoxed by this, too.
>
> > Why? All humans in town have TAKE_ITEM flag.
>
> > Or do you mean that those messages are created in shop?
>
> I think I may have misunderstood. I thought it was a bug report that
> when dropping items, they get disenchanted...
The joke was about the beggar picking up his old knackered armour
(presumably after a pasting from an acidic cytoplasm or something)
after he bought the new stuff.
Not a bug report, I think.
CC
Eddie Grove
> Andrew Sidwell wrote:
> > There is now a realtime-updated buglist for 3.0.8pr1 here:
> > http://tinyurl.com/33a4j3
> > It contains, TTBOMK, all of the bugs reported for 3.0.8 (barring any
> > major reworking of squelch), with some bugs grouped together into one
> > bug (like macro issues, UI niggles).
> > Fixed bugs are under the "Resolution: fixed" heading. This should give
> > people some idea of what's reported and what isn't; if I've missed
> > something in collecting feedback, please reply *to this post* so I can
> > catch up on it.
>
> This is not actually 3.0.8 issue, but a lot older. In limits.txt there is entry
>
> # Maximum number of objects on the level
> M:O:512
Is there any point in it being less than 10,000?
> That is too small. For example if you happen to have immunity to some element
> and wipe out entire dragon pit of corresponding color you get item
> compression. I have increased that to double 1024 in my copies and haven't
> noticed any weirdness yet. If there is no coding limit which forces that limit
> to be that small I would like to get it increased.
>
> Also _when_ compression happens I would like to have it start from gold. By
> the time you can get enough items that this can happen gold is meaningless,
> and anyway items you collect are probably much much valuable than gold you can
> get from floor. Unless gold has separate memory space that is.
>
> Timo Pietilä
I seem to recall that there is some silliness about compacting squelched items
first. That sounds smart, but then it compacted everything else anyway! If
compacting squelched items [and gold] drops the number of items down to less
than half the limit, compaction should stop. I could be wrong about this,
since I didn't really understand the compaction code, but I remember being
annoyed.
OTOH, we are all mostly using supercomputers [exporting anything in use today
to China 10 years ago would have been a felony]. I don't see why the game
doesn't simply malloc a new table of twice the size when the old table fills
up.
Eddie
Eddie Grove
> That is too small. For example if you happen to have immunity to some element
> and wipe out entire dragon pit of corresponding color you get item
> compression. I have increased that to double 1024 in my copies and haven't
> noticed any weirdness yet. If there is no coding limit which forces that limit
> to be that small I would like to get it increased.
>
> Also _when_ compression happens I would like to have it start from gold. By
> the time you can get enough items that this can happen gold is meaningless,
> and anyway items you collect are probably much much valuable than gold you can
> get from floor. Unless gold has separate memory space that is.
One more thing -- I prefer to allow different types of money to combine,
so that you have far fewer money drops to pick up when you clear a deep
dragon pit. Just be careful about overflow. Users get so picky about
negative pvals when they pick up money. :)
Eddie
Otto Martin
>Make wands of healing damage undead.
IMO only undead with nether/life-draining spells & breaths.
(I'm not entirely comfortable with the positive and negative energy
concept used in D&D...)
>Make wands of haste monster damage golems.
This I don't really understand...
>Make other wands replaced by upgrades deeper in the dungeon.
Now this is a really good idea. In fact, a good deal of the junk problem
could be reduced by creating a "powerful" or "deep" flag, which could be
given to many consumables. The deep versions would give a much more
powerful effect, and at the same time, we could reduce the number of
items.
Otto Martin - much of the snipped text had good ideas as well, IMO
--
"I'm a dragon. I don't need tact."
Arzosah, The Golden Falcon, Katharine Kerr