Those first few rings are crap

[CliffStamp]

I would describe the basic philosophy as "Buying new gear is boring, but finding new gear is exciting."

Yes it is obvious that if there is a recall then it isn't Ironman proper, however if the focus has moved from shopping or even simply minimize its influence on kit/gear to pure exploring proper then that is a move towards Ironman play.

If you remove the ability (or even frequency) of buying/trading for kit then unless you have drops character selective you have made drop less useful as the odds of finding useful items has decreased massively.

[fizzix]

Yes it is obvious that if there is a recall then it isn't Ironman proper, however if the focus has moved from shopping or even simply minimize its influence on kit/gear to pure exploring proper then that is a move towards Ironman play.

If you remove the ability (or even frequency) of buying/trading for kit then unless you have drops character selective you have made drop less useful as the odds of finding useful items has decreased massively.

I don't think the effect is nearly as large as Derakon is making out. The town serves a huge purpose for gear in a normal game, or at least it can if you're willing to use it. You can also choose to avoid it. Even in older versions you really only cared about the BM after some level, or at least that's what I remember.

Here's what I buy from town:

spell books
_teleport
_identify, ?identify
? treasure detection
? trap detection (before -dtrap), ?mapping
!CCW
!Hero
all unid'd scrolls/rings/amulets/wands/staves to learn flavors when I can afford them without issue.

Here are things I often wind up buying from the BM

?banish/mass banish/destruction - a couple a game
!speed - about 3-8 a game
?teleportation/teleport level - in 3.4 only a few a game
=FA I usually get one from the BM before the dungeon
=Resist Poison, about 20% chance I'll get this from the BM
-teleport other - really expensive now (15k) but I'll buy it if I can
!stat gain 2-3 a game
,vigor - 2-3 a game

I'll often occasionally buy gear from the BM like helms of seeing or gloves of FA, but usually dungeon gear is superior.

[Magnate]

Doesn't help. v4's problem isn't that you don't know what the item does -- its properties are readily discernable because v4 uses rune-based ID. The problem is that you can't readily say which combinations of properties are useful without looking at them first, so the game can't squelch them for you. You have to say "Okay, that's a 3d4 sword with good pluses, slay demon, a weak electrical brand, and bonuses to STR and CON...but I'd rather keep my current weapon which has a strong acid brand and bigger dice."

In short, by making the potential variation of items so much bigger, our previous categorization-based squelch system breaks down.

But the v4 system is at least clever enough to squelch anything where you have squelched all its properties: so if you've squelched slay demon, weak elec brand, STR and CON, the sword will be gone. This isn't quite as useful as it could be because we haven't yet implemented pval squelch, i.e. squelch STR < +3 but keep >= +4. Now that slays and brands are pvals, this just got more important. We also can't yet set a "do not squelch this weapon if it has extra dice" marker, which would be very handy.

Interesting discussion about both ID and shopping. The former is one of the few areas of general consensus but hard to implement well; the latter is trivial to implement but probably the single least-agreed-upon part of the game. Ho hum.

[Derakon]

But the v4 system is at least clever enough to squelch anything where you have squelched all its properties: so if you've squelched slay demon, weak elec brand, STR and CON, the sword will be gone. This isn't quite as useful as it could be because we haven't yet implemented pval squelch, i.e. squelch STR < +3 but keep >= +4. Now that slays and brands are pvals, this just got more important. We also can't yet set a "do not squelch this weapon if it has extra dice" marker, which would be very handy.

Certainly it's impossible to argue that a more powerful squelch system would not be better-equipped to handle the greater variety we now have in items. I'm concerned that it will end up being too much effort to set up the squelch parameters though, especially as what you consider to be squelchable varies as you play the game. Squelch isn't a universal answer, in other words -- and frankly, I think the most valuable things it does now (now that TMJ has been fairly well tackled for most of the game) are:

1) letting you squelch consumables and jewelry you no longer need (e.g. ?TrapDetect, !NeutralizePoison, !Charisma, =Protection), and
2) letting you hide items without destroying them, in case you find yourself unexpectedly needing them (e.g. !RestoreStrength, ?Recall)

Squelching of equipment is probably always going to be difficult from a user interface perspective, though I expect we could at least do "Squelch all weapons that are strictly worse than this one".

[CliffStamp]

... the latter is trivial to implement but probably the single least-agreed-upon part of the game. Ho hum.

In such cases why not just option them? There was a time when features which almost everyone uses by default were not and there are always people who find them central/key/critical. The idea for example of being able to lose artifacts, haggling, or the always a point of contention auto-scum.

Simply having these as options alloys both groups to play the way they want and privately conclude of course they are playing the real true Angband as the haggle manually without auto-scum or the auto-roller and with the risk of lost artifacts if you don't find them on the level.

You could explore an option on limited stores which vastly reduced probability of key items (or even removed them completely from the inventory) and again meet both groups of users as no matter what you do, aside from bug fixing there is always going to be contention. It isn't like many of the things almost everyone accepts now (not all classes have 6 max blows for example) were trivially accepted when first proposed/changed.

[fizzix]

i.e. squelch STR < +3 but keep >= +4.

I've often wondered if each pval should have its own rune. (you'd probably need to finally put a cap on speed rings, say +25 or so) Then squelching by pval also squelches by rune.

[Derakon]

I've often wondered if each pval should have its own rune. (you'd probably need to finally put a cap on speed rings, say +25 or so) Then squelching by pval also squelches by rune.

Ugh, what a hack.

CliffStamp: gameplay-affecting options are easy to justify from the player's perspective, but they do make the game harder for the devs to balance and maintain. That's one reason why options have been fairly aggressively pruned in the past. Autoscum is gone, for example; the attitude taken being "if the game isn't generating interesting levels, then we should fix that, instead of automatically regenerating levels until an interesting one shows up."

It's very difficult to say "The line is here, and all options on this side are worthy while the ones on the other side are disfavored." Someone will be unhappy no matter what you do. Of course you can readily claim that any one option introduces only a marginal increase in maintenance burden, but that way code bloat lies.

[CliffStamp]

...the attitude taken being "if the game isn't generating interesting levels, then we should fix that, instead of automatically regenerating levels until an interesting one shows up."

But that is the point, this isn`t such a trivial argument. If levels are always interesting then well they obviously always are not by definition. You can`t have everyone in a population be above average for the same reason.

It was made an option simply because people would bounce up and down stairs until they got the feeling they wanted, other people thought this was abusive and then it was changed so you had to explore to get a feeling. This was not appreciated by the power-munchkin players who would prefer all special feelings (or whatever) and hence it was simply optioned and there was again peace on usenet.

Not all styles of play benefit of course from always interesting levels and it certainly is not a central theme in the game itself. A high level feeling originally was a reward - source of excitement, especially of course the you have a special feeling one as you were always then looking for a 4d5. At times you also may want to play for dl drops instead of level feelings, around 1500 being an obvious point.

But yes there is a balance, in order to please everyone you would end up with an option list of several pages and either code or external lists to source to customize. At some point you would be basically creating the ability to high level make a variant and that would be a bit unreasonable. The most you could hope to achieve would be to retain the people who consistently enjoyed the game and ideally bring in new people to play.

Every hard coded change, aside from bug fixes (and there are people who argue for exploiting them even), tends to have its criticisms, and the only reason we even have the same we have now is that someone looked at a piece of code and saw potential for expansion and from Moria thus Angband came. The question the maintainers need to ask is are they really trying to maintain Angband or are they trying to evolve it as was done with Moria.

The new combat system for example is a pretty big shift to claim can be done and still call it Angband proper.

[d_m]

I've often wondered if each pval should have its own rune. (you'd probably need to finally put a cap on speed rings, say +25 or so) Then squelching by pval also squelches by rune.

I was interested in this. I remember talking to Magnate about it. Of course, once bonuses start going above a certain amount it stops being reasonable.

[Magnate]

I was interested in this. I remember talking to Magnate about it. Of course, once bonuses start going above a certain amount it stops being reasonable.

I don't think I understand the need for this - is it just to keep the squelch UI simpler? My vision is that where we currently have a "Yes" in the "Squelch" column (I'm talking about v4's ego knowledge menu here, where it lists each affix) those Yes markers would be replaced with "<= +0" ... "<= +25" for pval squelch. And instead of pressing 's' to toggle squelch, if it's an affix with a pval, it just asks you for a pval.

Or is this something to do with improving ID-by-use?

[Magnate]

Every hard coded change, aside from bug fixes (and there are people who argue for exploiting them even), tends to have its criticisms, and the only reason we even have the same we have now is that someone looked at a piece of code and saw potential for expansion and from Moria thus Angband came. The question the maintainers need to ask is are they really trying to maintain Angband or are they trying to evolve it as was done with Moria.

The new combat system for example is a pretty big shift to claim can be done and still call it Angband proper.

We spent almost three years arguing over this very issue - whether Angband should be "maintained" or "developed" (== your "evolve"). It started with a blog post by takkaria in Jan 2009 (which was discussed here on Oook). It concluded in October 2011 with the launch of v4, an experimental codebase where more radical ideas could be tried without alienating fans of Angband-as-it-has-always-been. The new combat system, for example, is one such radical idea that's being tried in v4 (the others currently being new ego item generation and new traps & trap detection).

I'd be interested to know what you think of this split approach. 3.4.0 will be the first version released since v4 was created, and as such should contain far fewer contentious changes than in previous 3.x releases. (The arguments began with the release of 3.1.1 in Jan 2009.)

P.S. I used to argue for options for every possible approach to gameplay, until I joined the devteam. Now I share every maintainer's antipathy towards them. I think an option for store contents would cause huge balance issues that nobody really wants to deal with.

[d_m]

I don't think I understand the need for this - is it just to keep the squelch UI simpler? My vision is that where we currently have a "Yes" in the "Squelch" column (I'm talking about v4's ego knowledge menu here, where it lists each affix) those Yes markers would be replaced with "<= +0" ... "<= +25" for pval squelch. And instead of pressing 's' to toggle squelch, if it's an affix with a pval, it just asks you for a pval.

Or is this something to do with improving ID-by-use?

It was mostly ideological: that you could create a visual string that summarized a weapon's enchantment. I.e.

"BYH DZ EN"

which might translate into

+2str/con +3speed +1light

Once you use a unicode character set this could start looking pretty cool. I don't think there were incredibly useful gameplay differences, although it could possibly create an interesting situation where you see =speed and aren't sure whether it's +3 or >+8.

Anyway, not a big issue in any case.

[Estie]

On the topic of early rings: since dog got its plusses removed, pretty much all of them get squelched as found. The first keeper is reckless attacks for non-casters, and maybe mouse for casters (for the stealth, if the melee damage doesnt matter anyway).

In theory, something like a branded katana can actually profit from a high dex mouse ring, but finding such a weapon early is so rare that I dont bother.

[fizzix]

On the topic of early rings: since dog got its plusses removed, pretty much all of them get squelched as found. The first keeper is reckless attacks for non-casters, and maybe mouse for casters (for the stealth, if the melee damage doesnt matter anyway).

In theory, something like a branded katana can actually profit from a high dex mouse ring, but finding such a weapon early is so rare that I dont bother.

I find dog useful for the rfear. Especially for warriors that have miserable stealth anyways. Mouse is useful for rangers or rogues/mages that are using ranged attacks for damage. Reckless attacks is useful if it pops up early, but it seems that lately I've been getting it after my first =accuracy... Don't overlook =protection. AC is much more useful in recent versions.

[d_m]

I find dog useful for the rfear. Especially for warriors that have miserable stealth anyways. Mouse is useful for rangers or rogues/mages that are using ranged attacks for damage. Reckless attacks is useful if it pops up early, but it seems that lately I've been getting it after my first =accuracy... Don't overlook =protection. AC is much more useful in recent versions.

Interesting.

I play a lot of warriors and I find ring of the dog to be really bad. Maybe it's just because I almost always carry !Heroism anyway, or because I value stealth too highly, but even when playing low stealth characters (i.e. dwarves) I find =dog to be more trouble than it's worth.