Make /digging temporary

[and into]

Currently, wands of digging permanently alter dungeon layout, which enables one to make "kill holes" that one then lures enemies into. It also encourages one to dig out tunnels in advance of the orb run to minimize time traveling between ascending stairs/hatches.

My proposal is to make /digging temporary: Starting at the furthest tile that has been dug out of the wall, the tunnel "caves in" if the tile is unoccupied. Only the furthest tile is susceptible to caving in (so you and enemies cannot get trapped).

I don't think this is a significant nerf to reasonable uses of /digging, but it prevents strong (but not very fun) abuses and makes the number of charges on the wand actually matter more so than now, so (at least theoretically) one has to be more judicious in using it. I believe this is superior to current design.

(In cases where one has dug out multiple tunnels, there would need to be some additional coding to ensure that a passage back to the dungeon remains open. Not sure how hard that would be to implement, but if there is a relatively easy way to do so, I think digging should be made temporary.)

[elmdor]

To make a proper killhole you usually only need to dig two squares, one for the enemies to flood into and one for you to stand in. If this change were implemented, killholeing would still be possible, you just couldn't go back into the hole after you left it. So players would just dig a second killhole. There are usually plenty of extra digging charges in every game so all this would do is add tedium.

Another problem with this is that right now the 'tab' key makes you attack or advance toward enemies, so if an enemy failed to enter the nearest square, you could leave your current killhole accidentally just by tabbing, and it could close up behind you when you didn't intend to leave just yet.

I also don't see how this would work with tunnels designed to connect disconnected levels, such as Orc:1.

Finally, this change would also remove much of the interest from digging enemies such as earth elementals, troll shamans and cacodemons (as one could just retreat and wait out the effects of their terrain changes) and it would also be a big inconvenience for Formicids.

It's true that digging isn't much of a scarce resource right now, but it seems like the way to fix that would be to reduce the average number of charges in the dig and disint wands.

[Hurkyl]

I also don't see how this would work with tunnels designed to connect disconnected levels, such as Orc:1.

Or some other annoying dungeon layouts with really long paths between points of interest that are amenable to digging major shortcuts.

[Quazifuji]

I think making digging temporary has a good chance of working as a solution to some of the issues with digging, but I feel like before we try anything drastic and complicated, we should try just increasing the rarity and lowering the charges of wands of digging. That's been suggested before, I assume it's incredibly easy to implement, and it has enough potential to work that it seems worth trying.

[dpeg]

I was in favour of removing wands of disintegration, and I still am. Moving digging to be top tier (which gives it less charges and makes recharging apply fewer charges) are also good.

[Sprucery]

Yeah, combine wands of digging and disintegration and make them as rare as the big 3 and give them max 9 charges and we're set.

[swoonis]

Can we just remove digging entirely? It's useless bloat really.

[Quazifuji]

Can we just remove digging entirely? It's useless bloat really.

I'm not sure what you mean by "useless" here, but I can't think of a way you could mean it that I wouldn't disagree with. If you're talking about gameplay, then no, digging is extremely powerful in many cases and not at all useless. If you're talking about game design, then no, digging has interesting uses that add to the game in my opinion.

I think there are reasonable arguments for removing digging, but not because it's useless. Rather, it has some degenerate uses (pre-digging tunnels before orb run, excessive luring with killholes) and one could argue that the problems digging creates outweigh the good things it adds to the game. But I think making it rare and/or temporary has a high enough chance of reducing the problems while keeping the interesting parts that I think they should be tried first.

I do agree with dpeg about Disintegrate, though. If a damage wand that's countered by MR is interesting, remove the digging part. If a damage wand that's countered by MR is not interesting, remove it entirely. Having a one-tile alternative to digging's bolt does nothing but occasionally make kill holes easier to set up, and that's not a good thing.

[Shard1697]

I'm more in favor of the idea from earlier where once you leave a floor, all walls revert to normal.

[Hands]

I don't think pre-digging the orb run is optimal. I think the benefit is so marginal that it is outweighed by other small factors like piety loss and use of recharging. I did pre-dig the orb run for my first few wins. And I enjoyed doing it too because it felt like I was meticulously preparing for something epic. As soon as I realised that this was not the case I stopped doing it. I often use digging on the fly on the orb run though, sometimes in ways that feel quite tactically satisfying. Kill-holes are another issue, but I don't think (what I'm assuming to be) a tiny proportion of players routinely doing something they find tedious in the mistaken belief that it increases their chance of winning should be a factor in the decision process.

[goodcoolguy]

I am 100% sure this notion that predigging for the orb run is optimal is nonsense. You don't know what the best route to the exit is a priori because it depends on how and where monsters spawn on the way out. If it were always possible to dig narrow, highly angular tunnels that reduce your exposure to normally generated dungeon to zero most of the time, it might be, but I think that's almost never possible. As soon as your best move is to abandon your prechosen route, the piety and in some games charges are wasted. Also worth noting that the shortest path is not always continuous. On average, you'll often save time by doing a cold teleport off the downstairs because you're at one the farthest points on the level from upstairs and most other places are substantially closer to stairs.

On the other hand, you stand to save maybe twenty turns of using digging wands over the course of about seven or eight hundred turns of orbrun. Instances in which those turns actually matter tend to intersect strongly with times where you would abandon your pre-chosen route for one that involves a new tunnel or a teleport.

If you really want to predig the orbrun properly, you're going to be doing a hell of a lot digging and spending a lot of time doing it. It's not just a single path you need to do and you're saving about 20 orbrun turns by doing it. It is not optimal play.

As to the proposal in the OP generally, the main uses of digging are killholes and short cuts on the orbrun. The proposal does little to address either of those in my opinion, though I don't think the latter is actually problem.

[Shard1697]

You don't know what the best route to the exit is a priori because it depends on how and where monsters spawn on the way out.

It's still better to have a shorter route. Less tiles to walk is virtually always better.

Also the amount of piety lost from taking the time to predig is entirely meaningless and I'm surprised anyone is mentioning it like it's something to seriously consider.

[goodcoolguy]

?????

You don't actually get a shorter route.

And it takes thousands of turns to predig, thousands of turns that are entirely wasted in exchange for 20 turns.

[Shard1697]

?????

You don't actually get a shorter route.

what on earth are you talking about

Here's a floor, chosen at random.

Standing on the stairs I just went up to get here in this screenshot. You can clearly see by looking at it that from any of these pink entry points, the route to blue exit point could be optimized by digging, but let's go ahead and make sure, just to make a point.

First, let's travel from where I am to the closest stairs up.


Without a better route dug, that took 41 turns to reach the next stairs.
Now let's dig a better path and see how much faster it is.


Only a few charges required to remove all obstructions from my path. And this new route only takes...


22 turns. With digging, the route becomes almost twice as fast.
This means that dangerous enemies like fiends, pan lords, seraphs, will only get around half as many chances to spawn and be a problem.

And it takes thousands of turns to predig, thousands of turns that are entirely wasted in exchange for 20 turns.

If you're not walking back and forth like I was taking screenshots it only takes like 50-70 turns per floor, which really isn't very much, and I don't see how you can say those turns are "wasted" when for no real cost you get to spend less turns doing the orb run, giving hostile demons less chances to spawn. What's the problem the player is going to run into here? They're not gonna run out of permafood unless they seriously fucked up already. They're not gonna lose a significant amount of piety. If they don't have enough charges to do it all, they can still optimize some of it.

[goodcoolguy]

You are deeply confused, friend. Being on the orbrun does not prevent you from digging.

[Shard1697]

Ok, but digging also takes a few turns, so again, it is a shorter route. Also I think it's better to plan it out before being in the thick of it, even if that means once in a while a charge or two ends up wasted because you change your route.

It's not something I make a habit of doing because I think it's boring, but I think it's untrue to say it's not optimal. Looking at that example again-when you're dealing with 20-40 turncount per floor, 5 turns is not insignificant.

[goodcoolguy]

Ok, but digging also takes a few turns, so again, it is a shorter route. Also I think it's better to plan it out before being in the thick of it, even if that means once in a while a charge or two ends up wasted because you change your route.

It's not something I make a habit of doing because I think it's boring, but I think it's untrue to say it's not optimal. Looking at that example again-when you're dealing with 20-40 turncount per floor, 5 turns is not insignificant.

No, you don't get a shorter route than digging while on the orbrun, which is obviously what I was saying. You cannot actually plan an entire orbrun route, unless you intend to fight every monster you can't just walk away from.

Did you actually read my original post?

[Hands]

You won't get the quickest route just by looking on the map and cutting a route to the nearest staircase anyway. At any given down-stair you don't know the up-stair that lies on the minimal path (unless there's an interface feature I don't know about, which is quite possible). Accounting for potential digging would be more difficult.

[Shard1697]

No, you don't get a shorter route than digging while on the orbrun, which is obviously what I was saying. You cannot actually plan an entire orbrun route, unless you intend to fight every monster you can't just walk away from.

Did you actually read my original post?

Yes, I did. Yes, you cannot know where all monsters will spawn. Yes, it is still worth it to dig beforehand, because it still saves you turns compared to digging during the run, which means it's a shorter route. Not shorter in terms of distance traveled but in turns taken during the run.

You won't get the quickest route just by looking on the map and cutting a route to the nearest staircase anyway. At any given down-stair you don't know the up-stair that lies on the minimal path (unless there's an interface feature I don't know about, which is quite possible). Accounting for potential digging would be more difficult.

This is true, but digging without planning minimal paths is shorter than not digging without planning minimal paths(and digging with planning minimal paths will be shorter than not digging with planning minimal paths).

[Hands]

No, you don't get a shorter route than digging while on the orbrun, which is obviously what I was saying. You cannot actually plan an entire orbrun route, unless you intend to fight every monster you can't just walk away from.

Did you actually read my original post?

Yes, I did. Yes, you cannot know where all monsters will spawn. Yes, it is still worth it to dig beforehand, because it still saves you turns compared to digging during the run, which means it's a shorter route. Not shorter in terms of distance traveled but in turns taken during the run.

It's a tiny, tiny benefit. In my opinion it's a smaller benefit than the tiny, tiny benefit of having a little extra piety and some more potential blasts on your HW wand.

[goodcoolguy]

No, you don't get a shorter route than digging while on the orbrun, which is obviously what I was saying. You cannot actually plan an entire orbrun route, unless you intend to fight every monster you can't just walk away from.

Did you actually read my original post?

Yes, I did. Yes, you cannot know where all monsters will spawn. Yes, it is still worth it to dig beforehand, because it still saves you turns compared to digging during the run, which means it's a shorter route. Not shorter in terms of distance traveled but in turns taken during the run.

Do you disagree with my estimate that at best, it [that is: predigging versus digging while orbrunning] will save you approximately 20 turns [of zapping digging wands] on the orbrun, an average of less than one turn per floor? And do you agree that if you're forced off your favorite route during the run, that many of those saved turns will be lost when you dig new tunnels to get to stairs?

According to your own estimates, digging every floor from zot to the exit takes over 1500 turns, by the way. In my first post, though, I mention that this is wildly optimistic. To do it right, you have to dig a lot of different paths per floor to account for teleportation and rerouting, not just the piece of your single favorite path that lives on each level. This would involve easily over 100 digs and take many thousands of turns.

[Sprucery]

Autotravel finds the optimal route to the exit. Personally, after I pick up the Orb and teleport away, I set the exit from D:1 as the autotravel target and then just use ctrl-g to go there after every interruption on the way.

It is also useful to use all upward hatches during the game so that autotravel can use those too.

And no, I never dig routes beforehand (and very rarely use digging during the ascension anyway). I would find it boring.

[dpeg]

The problem that too much digging is availbe is not related whether you predig the orbrun or not.

[4Hooves2Appendages]

If digging was extremely rare then pre-digging the orbrun might not be a choice in every game.

[Leszczynek]

You are really putting too much thought into digging the orb run path. The orb run is already easy even without any digging, so why is it even any issue to you? Because for whatever reason you feel compelled to do it just because you can? Then it's probably your, not the game's, problem. Now, if the orb run was sufficiently hard to make it a necessary tactic, then there'd be something to discuss; with the way things are, it's a non-issue in my eyes.

As for the killholes, just give many more enemies the ability to destroy or, if you don't want layouts to become too empty, shift walls. Deep elf elementalists are already good at messing up the killhole strategy, deep elf troll earth mages and cacodemons are also doing fine. You could also make digging each square take 3-5 turns and emit a lot of noise. You would still be able to create really useful shortcuts while less optimal ones like that shown some posts above would no longer be faster. While you'd probably whine about digging before orb run even more, that would push killholes into a rarely useful strategy.

The only thing I agree with is removing wands of disintegration, I wouldn't cry if they were gone.

[njvack]

Only the furthest tile is susceptible to caving in (so you and enemies cannot get trapped).

I'm pretty sure this still lets you trap things if you dig intersecting tunnels:

  Code:


XX4X
1111
X24X
X233


If you first dig tunnel 1 (E-W), then 2 (N-S), then 3 (W-E), then 4 (S-N) and stand on the 4 in the third row, I think you'd ultimately wind up like:

  Code:

XXXX
XXXX
XX@X
XXXX


I suspect there might be a marginally-useful exploit of this — dig a "track" and pillardance monsters until it's not accessible from the outside, then tele out.

[and into]

njvack: yeah I tried to address that in the OP at the end (though perhaps my wording was not clear).

Avoiding tedious but optimal practices is a good design goal even if few players actually perform those practices. At any rate, it is an explicit design goal of DCSS, regardless.

About pre-digging the orb run: It is not an all or nothing proposition. Using digging plus the method Sprucery outlines to chart and dig a course through dungeon ahead of time is almost certainly overkill. But pre-digging some tunnels on a couple of levels with really unfortunate layout is probably optimal, even if the benefit usually doesn't matter. Triply so on Z1, U1, and D1, where there is only one ascending staircase!

That being said, I like dpeg's solution, which unlike OP doesn't introduce a somewhat clunky new feature. Besides, digging should probably be made rarer just for balance; it seems silly that in many games you have access to almost as many charges of digging as you do of throw flame/ frost....

Yes disintegration can be removed; we already have plenty of "wands of mess up low MR dudes" and one-tile digging is not different enough to justify its own wand.

[elmdor]

I'd be cool with disint removal if you could zap a digging wand for less than the max distance, similar to how you can target arrows to fall short with .

[Hands]

Autotravel finds the optimal route to the exit. Personally, after I pick up the Orb and teleport away, I set the exit from D:1 as the autotravel target and then just use ctrl-g to go there after every interruption on the way.

It is also useful to use all upward hatches during the game so that autotravel can use those too.

And no, I never dig routes beforehand (and very rarely use digging during the ascension anyway). I would find it boring.

Autotravel computes the shortest path, but it doesn't as far as I'm aware present the information to you till after you've traveled.

[Rast]

I was in favour of removing wands of disintegration, and I still am. Moving digging to be top tier (which gives it less charges and makes recharging apply fewer charges) are also good.

FR combine digging and disintegration, make the disintegration effect ignore MR.

[dowan]

Let's not all forget about LRD or shatter in our excitement about removing disint and lowering availability of digging! Of course... there's still always the option of dragging a troll around to dig for you... but I'd have to see someone actually pull it off to say it's really even possible, let alone a problem...

[Sprucery]

Autotravel computes the shortest path, but it doesn't as far as I'm aware present the information to you till after you've traveled.

Yeah, I didn't mean to imply that it did.

[CanOfWorms]

Let's not all forget about LRD or shatter in our excitement about removing disint and lowering availability of digging! Of course... there's still always the option of dragging a troll around to dig for you... but I'd have to see someone actually pull it off to say it's really even possible, let alone a problem...

!lm perunasaurus rune=slimy 42 -tv:<T46800:>T47899:x4

[dowan]

Yeah, good point, if you can do it to get the slime rune, I suppose you could do it to drill paths for the orb run. Now, if you think pre-digging with a wand is tedious...

[tabstorm]

Almost no one predigs the orb run. You do not need to pre dig the orb run. I have died zero times on the orb run in years of playing, and I never pre-dig, and most of the time I don't even bother to zap wands of digging on the orb run, I just hit GD1 until I have reached D:1. The probability of pre-digging being the difference between life or death is vanishingly small. There are hypothetical optimal things that are annoying and have a nontrivial effect on your chances to win (see: the times when glowing weapons weren't identified, or the days of item destruction) and then there is stupid stuff like predigging the orb run.

Also, how would this work if a player digs a tunnel to camp in like so:

  Code:

###.##
##.###
#.####
.#.###


Wouldn't the filling-in of the tunnel trap the player in this situation?

[removeelyvilon]

Yeah I agree on killing 3 birds with one stone and merging (a powerful version of) disint with dig, making it a top-tier which is basically the player version of energy bolt; that also incorporates the idea of an "ultimate" attack wand that was floating around.

[Laraso]

At any given down-stair you don't know the up-stair that lies on the minimal path (unless there's an interface feature I don't know about, which is quite possible).

Press X. (Has to be uppercase, lowercase x will not work)
Hover over an upwards staircase.
Press [
The game will now display the map of the floor above, focused on the staircase that you will arrive on if you go up that specific set of stairs.

This works for downstairs as well. You just have to press ] instead.

[Hands]

At any given down-stair you don't know the up-stair that lies on the minimal path (unless there's an interface feature I don't know about, which is quite possible).

Press X. (Has to be uppercase, lowercase x will not work)
Hover over an upwards staircase.
Press [
The game will now display the map of the floor above, focused on the staircase that you will arrive on if you go up that specific set of stairs.

This works for downstairs as well. You just have to press ] instead.

Right but this doesn't tell you the minimal path from where you are to the exit. The greedy algorithm won't work in general.

[KoboldLord]

It seems much more efficient just to remove the orb run in general. If the game needs a victory lap for some reason, there's no reason to recycle old maps for it when literally infinite new maps could be generated that would not allow for tricks like pre-digging or screen-shotting finished levels. Once the player grabs the orb, perma-mark the player, have a steady stream of orb run monsters spawn from all upstairs on every level, and generate a new Zot 4 through Zot 1 for the player to retreat through. Because Zot is all chaotic from the start, and is in even more disarray because the orb is being stolen. If you want, throw a Ziggurat on the new Zot 1 that the winning player can do for extra (worthless) points on their score, but no reason to force every player to walk through 22 more mostly-vacant levels populated only by a handful of chaff to cut through or a couple pan lords to run past.

[and into]

@tabstorm: Already addressed the zig-zag tunnel problem in the OP, though there would still potentially be awkward situations in terms of how exactly "cave-in" would work. The other points were addressed subsequently in the thread. I have already conceded that making digging charges more scarce is a good change that could obviate the problem because, aside from general rebalancing that could arguably merit such a change, non-tactical uses of /digging outside combat would have a more substantial cost attached them.

(The merits and demerits of the orb run itself are something of a separate issue, but I think it would be fine to discuss that in this thread as well.)

[duvessa]

@tabstorm: Already addressed the zig-zag tunnel problem in the OP

Only in the sense that you addressed that you have no idea how to solve it...

[and into]

@tabstorm: Already addressed the zig-zag tunnel problem in the OP

Only in the sense that you addressed that you have no idea how to solve it...

Unless I am misunderstanding what tabstorm was getting at, I did mention in the OP how one could avoid that particular problem. There's any number of ways to make digging non-permanent, the simplest probably being what Shard1697 mentioned upthread, but I'm not going to flesh out a proposal that I have already been convinced is unnecessary.

[Arrhythmia]

It seems much more efficient just to remove the orb run in general. If the game needs a victory lap for some reason, there's no reason to recycle old maps for it when literally infinite new maps could be generated that would not allow for tricks like pre-digging or screen-shotting finished levels. Once the player grabs the orb, perma-mark the player, have a steady stream of orb run monsters spawn from all upstairs on every level, and generate a new Zot 4 through Zot 1 for the player to retreat through. Because Zot is all chaotic from the start, and is in even more disarray because the orb is being stolen. If you want, throw a Ziggurat on the new Zot 1 that the winning player can do for extra (worthless) points on their score, but no reason to force every player to walk through 22 more mostly-vacant levels populated only by a handful of chaff to cut through or a couple pan lords to run past.

I've posted about this again and again, but I think it'd be very good if as you're running up, the dungeon is "falling apart" (i.e., tending towards bad terrain). It'd keep the "victory lap", i.e., you'd have all the old landmarks from it (oh yes I remember this fucking vault), but the maps would actually be different, in that like, lava is bursting out of the ground while walls are falling down around you. Think the "Escape From the Cave of Wonders" sequence from Aladdin. Plus, pre-digging would be useless.

[Sprucery]

I've posted about this again and again, but I think it'd be very good if as you're running up, the dungeon is "falling apart"

I'd be fine with that as long as autotravel would still work.

[dpeg]

I think it'd be very good if as you're running up, the dungeon is "falling apart" (i.e., tending towards bad terrain).

Yes, sounds intriguing. I believe that it should be possible to invent a system that's both cool and gameplay-wise interesting.

I'm a sucker for dynamic visuals anyway, but the only thing I could achieve was the Slime:$ vault walls losing colour when you kill the Royal Jelly. Would be nice to have more like this.

[tabstorm]

@tabstorm: Already addressed the zig-zag tunnel problem in the OP

Only in the sense that you addressed that you have no idea how to solve it...

Unless I am misunderstanding what tabstorm was getting at, I did mention in the OP how one could avoid that particular problem. There's any number of ways to make digging non-permanent, the simplest probably being what Shard1697 mentioned upthread, but I'm not going to flesh out a proposal that I have already been convinced is unnecessary.

I still don't get it. If tunnels like the one I posted start caving in from the farthest tile dug out, the minimal length tunnel that wouldn't trap the player while they were in it would be the following:

  Code:

######
######
#.####
.#@###
######


You only need 3 tiles for a functional tunnel to prevent enemies from spamming summons or smite-targetted attacks.

So, unless you're going to forcibly push the player out, I don't see how this is resolved by saying "There will be some additional code". People will just camp in the tunnel and zap lightning wands to attract enemies. Having tunnels disappear when you leave the floor will not solve this "problem". I don't think digging tunnels to get a better position is necessarily bad, but it appears to be considered enough of a problem to add a new monster with a complex mechanic to Elf:3 to combat player tunneling.

[genericpseudonym]

Almost no one predigs the orb run. You do not need to pre dig the orb run. I have died zero times on the orb run in years of playing, and I never pre-dig, and most of the time I don't even bother to zap wands of digging on the orb run, I just hit GD1 until I have reached D:1. The probability of pre-digging being the difference between life or death is vanishingly small. There are hypothetical optimal things that are annoying and have a nontrivial effect on your chances to win (see: the times when glowing weapons weren't identified, or the days of item destruction) and then there is stupid stuff like predigging the orb run.

Also, how would this work if a player digs a tunnel to camp in like so:

  Code:

###.##
##.###
#.####
.#.###


Wouldn't the filling-in of the tunnel trap the player in this situation?

This potential problem comes from giving a separate timer for each new zap of the wand. But it seems like the more elegant and problem-free solution would just be to look at all the dug tiles together, as a whole. You can run some kind of pathfinding algorithm and then have only the tile that is the most steps from a non-dug tile collapse. This would make it impossible for the player to be trapped by cave-ins, at least in normal situations. You'd still run into problems with disconnected levels like Orc and Slime.

[duvessa]

problem-free, except for the problem that now players have to know how the algorithm works in order to get the most out of their dig charges

resetting the walls upon leaving the level has its own problems but not nearly as many as this overcomplicated idea

[andreas]

I've posted about this again and again, but I think it'd be very good if as you're running up, the dungeon is "falling apart"

This sounds flavorful and fun to me!

Digression:

Autotravel finds the optimal route to the exit.

I'm sort of skeptical of this. There are other desiderata besides getting to d1 in minimum turns that autotravel ignores. For example, I'd rather take a slightly longer route if it has me caught out in the open much less. I've also heard that it's important to keep low turns spent on each floor; it sounded like a slightly longer overall path might be worth it if it avoids a long time spent on particular floors (but it's hearsay, I don't really understand how orbrun spawning works). Autotravel also ignores when you could easily improve your overall time by digging a shortcut (if you didn't predig).

[goodcoolguy]

The optimal orbrun route is unknowable a priori and obviously autotravel knows nothing about it. It doesn't even try to dig.

If you want to do temporary digging without trapping players or much worse monsters, you probably need to displace things when the walls recover. The easiest way to do this would be random teleportation, but you still get the problem that trapping monsters will often help the player.