I don't mean to belittle the challenge of designing a non-trivial, non-binary trap detection scheme, but this just feels wrong. Angband just wouldn't be the same without the exquisite agony of killing some important monster, walking over to pick up the loot, and falling through a trap door you forgot to detect.

edit- I'd really like to see the turret type traps in V. Maybe also a rolling boulder type. I also had a kind of crazy idea for a trap- a trap of transmutation. It would randomly change an item in your backpack into some other substance. If gold, you'd get a "gold ingot" in your inventory, heavy but valuable. If lead, just a worthless hunk. If you can identify the trap and have the use for gold, you can throw an item onto the trap and see what you get.

I should generally think making the trap situation robust and thematic is a good way to go about things---take a page from Brogue at the least.

Instead of funneling everybody towards a catchall tool/spell/stat, why not have each lead with their best to varying nuanced degrees? The Wise know enough to leave something alone, the Strong not a damn thing aside from smashing it as hard as they can so as to break the thing, etc. I also favor the notion of just having them be a Vault concern.

In general, I see Angband as being about a theme, a personality to it, moreso than Adventures in Mathletics and Checklists. Math can get things really screwed up when such precise wranglings lord over all----look what happened to Portralis in the not distant past where things go so out of hand with numbers, not intent/situations, a wholesale re-do of sorts had to get underway and is now slowly coming back in Alpha-form.

It is fine to want to support Angband as a historical notion, but I would like to imagine much of what makes Angband "Angband" isn't rooted so much in dogma and sacrosanct anachronisms of gameplay-user notions, like Nethack or some such, but rather core elements and good times beget by solo and shared camaraderie of the various exploits.

Venerate the Spirit, not the Body, or some such. Shockbolt's fine work alongside and/all UI enhancements should help to open many good avenues to grow the community and encourage good times to be had by all.

One the more general theme of Vanilla development, along with the questions in the OP, maybe the "traditionalists" here can discuss Vanilla Angband should be. In the whole scope of roguelikes, some of them much more complex in a variety of different ways (quests, maps and levels, character options, items), some of them much simpler, what niche does Angband carve out? Whatever it is, it should do that as well as possible.

Take it to the V4 forum!

[EDIT: or to be more verbose: this thread seems to suggest replacing a long-established V mechanic with a whole bunch of untested thought-up-on-the-spur-of-the-moment alternatives: I thought that was exactly what we were trying to get away from.]

A.

My problem with that skill is not the skill itself, just that it works only on adjacent squares. It should be LoS -thing (visual range thing). Then it would really matter.

[EDIT] typo corrected.

I totally agree with Timo's description of the problem, and I think several of the proposed solutions would work very well.

I would happily merge any of these strategies into V4 if someone implemented one of them. With strong consensus I'd merge any of them into V if they played well in V4 first.

I've been very busy this fall so I will make no promises, but I think it's an area where we could quickly see some drastic improvement.

As I said in my original post, remove the reliable method. Extend search as skill to LoS.

With poor searching skill you might still stumble upon some traps, but that would be acceptable price of that skill being poor one. Helmet of seeing and amulets of searching would have valuable flag then.

With light radius of three you have at least three turns before you step on one, which means that search skill of 50% should have quite good probability to detect the trap before you actually step on one. This also would mean that you probably step on traps in early levels quite a lot more.

This is for debating what should we do for the vanilla, not for a variant based on vanilla.

I have brought up one thing that could be fixed. I haven't yet seen any other posts about same thing.

Then if there is some sort of consensus that this is something that could/should be done, then test it in v4, I'm not against that.

That sounds a better process.

Hey, if we want searching to be an interesting thing, then we need there to be more pattern about where traps are found.

In the extreme, if every space is equally likely to conceal a trap, the player will end up searching at every move (boring) or never searching (unsatisfactory).

it would be better if you could look at the dungeon layout and guess where a trap is likely to be. This is already true for vaults, but not so much, I think, for rooms/corridors.

So perhaps the trap placement algorithm needs to be part of the solution.

A.

It seems to me that traps should be added to a level after everything else has been generated, and go in squares around doorways, staircases and floor objects. Traps should be protecting something, not just randomly stuck in the middle of the floor.