The Tavern

The {Hunger} property is tedious. It encourages jewelry micromanagement. Some time ago we changed {Vamp} to impose a cost-on-wield to combat exactly this problem.

Some ideas for Hunger include:

• Inflate hunger costs rather than metabolic rate.
• Impose a cost on removal (like {Contam})
• Impose a cost on use (this is problematic)

My inclination is a combination of 1 & 2.

Hunger is tedious whether it has braces around it or not.

I like proposal 2 of yours the best.

1 seems to encourage micromanagement too, so I don't like that option on its own. For 2, it seems like it would be better to just generate the item with Contam instead. So I'd say get rid of it as an artefact property (I doubt the crown of Dyrovepreva will become brokenly overpowered), and make this change to rings of hunger.

Some other ideas:

• Cap satiation level, a la ghouls.
• Decrease nutrition provided by food.
• {Hunger-} simply doesn't work when under the effect of {Hunger}. Not a very elegant solution, I admit.

In all cases, adding a penalty on unequip, and maybe on equip as well, would help to discourage swapping.

These changes just encourage eating chunks. If there's a removal cost, I'd like to oblige the player to be full, forcing non-saprovores to eat permafood (like {Vamp}).

Might I suggest that hunger multiplies or add onto non resting satiation costs. Increase food consumption in battle, not rest.

Something like:
- You have to be at least "full" to take it off
- It then costs a bunch of hunger to take it off
(like equipping a vampiric weapon)
Would probably be the best way to do it, I think. Or just make it exactly like vampiric I guess, having the vampiric brand correspond to the {Hunger} tag makes sense to me.


Is it? I ignore all food costs on 95% of my characters without ever running into problems. The only characters for whom food ever really matters are spriggans, trolls and healers, at least for me.

Personally, as annoying as a ring of hunger is and how important it is to get the damn thing off, it does make you play differently to cope with it (resting less and rushing into fights so you can chow down on more corpses). And that's nice.

Might be an idea for a fixedart there. Something that made you constantly ravenous, but gave you bonuses of some kind base on how sated you were. Maybe it makes you strictly carnivorous while worn, too. If it could somehow encourage feasting on corpses mid-battle, that would be great.

Since (cursed) ring of teleportation exist, and can very easily kill a character, I never equip an unidentified ring without remove curse scrolls. This makes ring of hunger (or any other cursed bad ring) pointless for me.

Also, why is the amulet of cekugob {hunger}? It's decent, but not amazing, and -tele is already a more interesting penalty than {hunger}.

I'd say go for these two.

And you could also ban chunk eating for non-carnivorous races, to make that removal penalty really matter.

{Hunger} is tedious, Brendan, but it is also the tip of the ice burg. — {Hunger} being tedious is just an exceptionally bad case of hunger being tedious. So I'm going to respond to what I think is the deeper problem (and I beg your pardon in advance if this isn't where you wanted the thread to end up).

It would be a good idea, I think, for the developers to have a serious discussion about what food and hunger currently accomplish in Crawl, and what they should accomplish. Satiation only matters perhaps 5% of the time in any given game, but you have this mechanism that is a constant bother 100% of the time. This seems to indicate clunky design—I think most of the devs feel this way, but think that the hunger/satiation "minigame" in Crawl is too deeply ingrained. Perhaps—but my suspicion is that satiation (like victory dancing) is actually fixable, and that it seems much more important for balance than it actually is.

I see, broadly, two general approaches to making hunger less tedious:

For either approach, keep corpse, hide, and chunk generation as it is now. But make hides auto-drop W% of the time from applicable enemies because avoiding spoilers and streamlining interface are (rightly) design goals of DCSS.

1.) Remove chunk eating for satiation purposes from all but the five weird species—see below—and amulet of gourmand. Everything else remains equal. Increase spawn rate, or else the amount of nutrition provided, or both, for all permafood. Probably both. An increase in the "elasticity" and range of nutrition here would keep balance while reducing tedium. In other words, expand the "capacity" off everyone's belly and expand the edges of "Engorged -- Very Full --" etc. down the line.

2.) Automate as much as possible. Make all food costs draw directly from the food stores that you have in your inventory already based on nutritional cost. The interface "Engorged -- very full -- full ------ hungry -- very hungry -- near starving -- starving" would in this system only indicate how much edible (for your species) food one has in inventory. Chunks and butchering corpses remains a mechanism important for satiation purposes. Keep all the different fruit tiles and flavor, but upon pick up those bananas and strawberries go to "fruit: XXXX" nutrition counter, displayed at bottom of your (i)nventory screen. Your fruit nutrition counter, plus bread rations and chokos and the like make up your "non-meat: YYYY" nutrition counter. Beef jerky and the like add to your "meat: ZZZZ" stock. Remove all weight considerations from food because that is perhaps the silliest thing about them. Fedhas diminishes a portion of nutrition from your fruit stock based upon what abilities you are using. Spriggans only draw nutrition from the "non-meat" counter. And so on.

In both cases food can be made an entirely strategic decision. Given a certain level of the stat intelligence, when you choose to train spellcasting to a certain level, you have a good sense of how frequently you can cast some set of spells in a manner that is sustainable. You have all the same breaks to abusive stuff like stacking wizardry to cast high level spells early on or spamming pacification at *everything*, but the tedium is reduced to a minimum by taking the uninteresting decisions out of players hands. Which is what Crawl should strive to do for everything that is not interesting.

A few species with a very different relationship to the satiation clock (Trolls, Ghouls, Vampires, Kobolds, Felids). These species should, and could, maintain their special relationship to food in either approach outlined above. You could still keep in groumand even, and have it give you Troll-like eating abilities but without the faster metabolism; this could actually serve (as an added bonus) to make gourmand more distinctive as an option, as its benefits could be tweaked and amplified in either approach. (Or you can just remove gourmand, there's always that, it won't really affect the game much.)

Tangentially related: In either situation outlined above, I'd suggest automating sacrifice of corpses for Okawaru, Trog, and Fedhas. A certain proportion of corpses are automatically given up in the name of your religion. These gods take their tithe from you—that's what you sign up for when you worship them—and don't wait for the approval of a mere mortal before claiming a certain proportion of your kills.

For Nemelex, I'd suggest that you let people set sacrificial proportional weight in the ^ screen ("None – Normal – Focused," as in skill menu) and then have Nem auto-sacrifice. A certain portion of all the items of the given type simply disappear. You don't even see them. Nemelex snatches them up—being the trickster god, he can make use of them and has no moral problems with stealing—and the more you allow him to take, the more he likes you. (Greater piety, more decks, etc.) This would also have the side benefit of maybe making Nemelex a little bit balanced—in addition to the primary benefit of making him a lot more fun and less tedious to play!

Well the thing with the hunger randart property is that you can probably fix it pretty easily, whereas fixing food altogether would probably involve a lot more design work. Not to mention more lines of code than there were lines of cocaine in Ozzy Osbourne's medical history.

Combination of 1 and 2 sounds cool.

Yes, good point.



I have extremely little experience with coding, but I assume one of these is less fun than the other.