Nethack is the game I've played more than any other.. having played it literally for decades and decades. Yet I switched to Dungeon Crawl Stone Soup about 5 or 6 years ago, and really haven't gone back except to try Nethack again every now and then only to feel it too painful and slow to continue playing.
I've been spoiled by Crawl's autoexplore, its richness of gods and god-granted powers, its colors (in text mode, not tiles), and many other new features and innovations. That said, Nethack is still the deeper game in so many ways. Crawl's "don't do it like Nethack" approach has prevented it from taking all sorts of great ideas from Nethack and making Crawl in to a richer game -- things like making more advanced use of pets (like training them to steal from shops), more things to do in shops, use of altars beyond merely converting to a god, use of potions beyond merely drinking them, more uses for item inscriptions, all the ways to combine items, the list goes on and on.
All this makes Crawl seem like the shallower game, though it's still plenty of fun, and I still play it to the exclusion of most any other roguelike for now. There's a lot of room for improvement, though. I just wish they'd take more of the best parts of Nethack and make Crawl even more fun.
My main problem with nethack is the over-reliance on "spoilers" if you don't want to figure out every (sometimes rather non-obvious) mechanics in the game through trial-and-error. It's really hard to get into nethack if you're not willing to spoiler yourself, and once you start reading spoilers it's hard to know when to stop if you don't want to spoil the fun. On top of that trial-and-error in a game with perma-death can quickly become discouraging as you die stupidly because you didn't know that you couldn't levitate above a sink for some reason.
Dungeon Crawl on the other hand is a lot friendlier, in general the mechanics are well explained and you know what to expect and how things work. Contrast for instance the curse mechanic between nethack[1] and DCSS[2]. Nethack is obviously a lot more intricate but good luck figuring out all of that on your own, and good luck finishing the game if you haven't figured out most of that.
It comes down to opposing philosophies. Without trying to sound too derisive, Crawl is a game made for gamers. It focuses on gameplay to the exclusion of all else. It's like those sports TV programs where they use aggressive editing to condense a 3.5 hour baseball down to 30 minutes.
NetHack comes from the opposite angle. It's the D&D world your brother's friend spent all summer creating while you were away on vacation. It's best enjoyed on a rainy Saturday afternoon when you've got a clear schedule ahead. Gamers might call it tedious or repetitive, but then so is knitting or dicing vegetables by hand. It's just not meant for people in a hurry.
It's a meditative exercise that also happens to be pretty challenging if you aren't a Munchkin. In that sense it's kind of like correspondence chess for one: sure, nobody's stopping you from cheating (or using other meta-game exploits) but you're really robbing yourself of a lot by doing that. On the other hand, reading books on chess theory (or the NetHack wiki) is just fine.
Crawl consumed my spare time for some years, i even got good enough to streak 4 wins in a row. I disagree that it's a shallower game. Crawl is streamlined, yes, but the game mechanics are so finely tuned that at a tactical level its depth surpasses Nethack. That is, if you play an interesting combination and not MfBe or something :p The tactical complexity leads to strategic complexity at the margins when you're pushing for low turn count wins or streaks with difficult combos.
Ah, now I want to log back on and do a run :) I'm curious to see how the game has evolved since I last played.
> I just wish they'd take more of the best parts of Nethack and make Crawl even more fun.
Not that it's your (or anyone's) responsibility to do this, but quick reminder that both Crawl and Nethack are completely Open Source. So if anyone ever wanted to do a downstream fork that was pure Crawl + a few Nethack features... they could.
Just want to maybe stick the idea in someone's head.
How do you submit patches for inclusion in nethack proper? Just spent five minutes wandering nethack.org, and am still fairly convinced that it's a "source dump"-style project, not something that you can just submit a patch on their reviewboard.
It absolutely is a source dump game, or at least it was the last time I checked (several years ago).
Nethack is often brought up as a jewel of Open Source games, but its development style closely followed closed source methodologies. By comparison, most closed source early access games today have more open and transparent dev cycles than Nethack ever did. That's a side topic, but Open Source game design is a really interesting topic all by itself that I wish more people would talk about.
In any case, source dump models are in a lot of ways even better for downstream forking, because you don't have to regularly integrate new patches in. Whenever a new version of Nethack comes out you'd have to merge in the changes from that version, and then reapply whatever patches you made.
The wonderful thing about Open Source is that your patches don't always need to be included proper. :)
Apparently the patch for the bug I reported in 2004 (C343-20) didn't make it into this version, despite being fixed four years ago. The glacial timescales are kind of endearing at this point.
The issue is that if you die by having a cursed ball and chain fall onto your head while descending a staircase, the screen doesn't transition to the lower level so it's not clear on which level the bones file would be generated. (Per emails from the dev team, it does get correctly generated on the lower level.) Apparently fixing this issue somehow required rewriting huge portions of the graphics engine, which probably explains the snarky comment next to the bug in the issue tracker. :-)
:) To be fair, if I was writing any program and someone came to me with a bug like that, my first thought would be "#!@? I don't want to deal with this."
Personally, since I'm less interested in MS-DOS modes and fixed-length 8bit charset modes, I switched to the redesigned NetHack 4, which works fine with modern UTF-8 terminals.
NetHack 4 is geared towards progression of the game and staying up to date with the changing world rather than mothballing the old game. It's a separate project from NetHack 3.x, with different goals and development teams.
The continued existence of an MS-DOS port doesn't detract from versions for current operating systems in the slightest... Vim and Emacs even maintain current ports for MS-DOS and I doubt anyone is going to give them up just because a DOS port exists.
NetHack 4 is a radical shift for the game, and many of its ideas were incorporated into NetHack 3.6 (the upstream project), I think even its developer was placed onto the devteam for upstream NetHack while maintaining the fork simultaneously. It's a cool project, but NetHack 3.x is the conservative one, both in gameplay and in maintaining ports to operating systems of yesteryear (they clearly aren't against removing them, but they support them when there is at least one person to maintain it).
+1 here. The most significant change of Nethack 4 is to separate game logic into a common library named libnethack[0], so you can write your own interfaces and communicate with server using JSON.
Loosing MS-DOS support is kind of a shame, because there's really no reason the game can't support it, but also because software for MS-DOS is some of the most portable software you can write these days.
DOSBox has accidentally achieved write-once-run-anywhere.
> "Elbereth" now erodes based on attacks by the player, not monsters scared
That puts Elbereth back into useful territory. The way it was nerfed in 3.6 made it borderline useless when encountering a single foe that outmatched you, and entirely useless encountering a group.
This is possibly bad news for people who thought that maybe The Problem With Nethack™ was that it was too easy to get out of the early game, but probably good news for everyone else.
As of Nethack 3.6.0 Elbereth no longer works on a square that you're not standing on, so it can't be used to guard a stash. Now you need to use a scroll of scare monster for that purpose.
You shouldn't even need a web browser to view this announcement -- it was posted on 27 April 2018 to rec.games.roguelike.nethack. Assuming you already have a newsreader set up, you can just go ahead and read that message.
I just went ahead and logged in to a machine where I had slrn installed, then signed up for an account with http://www.eternal-september.org/ and configured my .slrnrc with
set force_authentication 1
nnrpaccess news.eternal-september.org [user] [password]
In keeping with history of the "Net" prefix of NetHack, this was posted to Usenet in the rec.games.roguelikes.nethack newsgroup, not a mailing list. Usenet isn't totally dead, it just seems that way from the trendier parts of the Internet. :) That newsgroup is still primary for NetHack.
But sure, there are still more traditional ways to read newsgroups than the Google Groups interface.
I could never get into any other roguelikes, but I LOVE Dungeon Crawl Stone Soup. I easily have over 100 hours in it, and just recently finished an actual game for the first time (15 runes, Gargoyle Hunter).
I find DCSS's development goals very interesting in contrast to Nethack. While Nethack is widely famed for "the dev's think of everything", DCSS has been on a streak of streamlining mechanics, such as the latest release collapsing all food items to rations instead of the pizza, royal jelly, and others. Though rip wand of healing.
Hey, I was a crawl dev from 2009 until whenever I pushed my last commit. Your reaction suggests that you're either unfamiliar with or misunderstand crawl's development philosophy.
Crawl development strives to eliminate grinding and eliminate tedium. Players should never be rewarded for performing repetitive actions. Every choice should be meaningful. This has been the case since the game's revival in 2006 and goes back to the introduction of the food clock in Rogue. A few years back I asked Ken Arnold about the food clock. He introduced it because a timer is a great way of making stupid, time consuming strategies useless. Crawl follows in this tradition.
We removed the requirement that you have a bladed weapon to butcher corpses. Why? Because it's a game and not a murder-hobo simulator. If you need a fix for mindless simulationism you might enjoy Dwarf Fortress, the premier dwarf terrarium.
If you get a chance, you might want to take a look at 5e; mechanically it's still streamlined, but lots of people who didn't like 3e/4e seem to like it.
I was drawn to roguelikes because of their preference complexity and systemic approach compared to rest of industry's obsession with GFX. On the other hand, I'm turned off by the fantasy setting, narrow corridors and spoiler-based gameplay. I respect them and their legacy, but Nethack and DCSS don't work for me.
For anyone else that feels the same, Cataclysm DDA is a great choice. It features open dynamic world with no hard-coded goals, lots of survival mechanics, vehicle and base building, different ways to advance (skills, equipment, mutations, implants) and countless ways to approach any situation.
The game can also be very atmospheric - my @ once got sick so I parked my RV out of danger's way, spent the day reading books, listening to rain hitting the rooftop and planing my next move. It's an experience unlike any other game.
I’ve never been able to bring myself to ascend a 3-rune game either. By the time I get to zot, I know I can win — why not go for the bonus round? I don’t think I’ve won (though I cloned my save file one time to try both 3 and 15 one time and did technically win the 3). My closest was a wipe on sentinel island at 13-14 runes.
Huh, this seems like a new way of thinking about Crawl to me as I've only ever won three and four runes, and I've never collected more than six in a single game. So it still feels like a huge gap from that to 15!
Haha thanks, it was definitely a great feeling. I feel like I "beat" DCSS and haven't had any real desire to play since, but it's for the best I think...
I had dozens of Minotaur Fighters and Gargoyle Fighters that made it up to 7+ runes before dying. The worst death by far was a MiFi that died to Hellion Island because I was playing in bed, shifted my body slightly, and then accidentally held down the space bar for a few seconds right as I encountered it for the first time ever. Oops!
I am surprised no one has expanded on the DCSS experience in regards to auto/manual yet. The ability to seamlessly switch between automated exploration/combat and precision controls is the killer feature that kept me hooked, I think.
Has anyone won this game honestly? I've patched the C code and only got so far. I tried savefile hacking and that didn't get me to the end. I was only able to win by compiling a version with debug symbols and playing inside the debugger. (eg. find the flag for the Wizard of Yendor's invisibility status and clear it.)
I won 14 times, back in high school when I had way too much free time. Once with every class, plus an illiterate barbarian.
One of the big thing is knowing enough to recognize when you're in danger, before you're in danger. Stuff like "That q is too tough for me to take right now", or "I'd be screwed if that n steals X", or "I don't have poison resistance yet, so I better not fight those o at range," or "It's dangerous to be in the mines with this AC and no source of light." This comes with experience.
Another big part of it is understanding the sequencing, and how to build an ascension kit. Do you want to try to get to minetown early to grab a magic lamp? When is it safe to go to mines end, and do you need magic resistance to do it? What's the earliest you can take on a unicorn?
There are lots of other tricks (identifying wands, potions, etc), clearing the big room safely, finding the portals on the elemental planes, stealing from shops, what gives resistances, maximizing wishes, etc.
Ultimately, Nethack got boring for me because the late game is too long, and doesn't differ much between different classes. Ascension kits tend to be very similar (GDSM + silver shield or SDSM + CoMR?). I really liked what Derek Ray did with Sporkhack, and I never managed to win that, although I died on astral a couple times.
I've been playing for almost 30 years, and I think I've ascended about 40-50 times total. My latest 3.6 install, I've ascended 11 times so far, as Wizard, Monk, Knight, Archeologist, and Samurai. Pretty sure I ascended in the high 20s on previous install w/ 3.4, and I know I have ascended at least once with every character in 3.4 (Tourist was easier than I thought it would be). Can't remember my install before that.
I concur with others. Elbereth was useless w/o a wand of fire in 3.6.0. Taught me to learn new survival habits, like staying much closer to my pet and making sure I can always go from burdened to unburdened by dropping only a sack.
Ascending is mostly luck and discipline, and the discipline comes from practice and focus. I find playing nethack calming, and it's a good way to measure how well I'm able to stay in the moment. If I die stupid, it's usually because I'm not all there.
If I'm doing something new, and I am uncertain about what my odds are, I'll read the source code.
I've only modified the source in 3.6.0 to create a general random number generator using a more robust (and more consistent) algorithm than is likely used in most libc variants. My interest there was mainly in perfect cross-platform state reproducability given the same key sequence.
There's only one person known to have won without spoilers (that is, without reading the source code or information derived from the source code). However, you can see lots of experts playing live on nethack.alt.org and they regularly win.
I've personally won four times (with Valkyries and Barbarians, two of the easiest classes) and once got a Priest to the Astral Plane and died there. :-( Recently I've switched to Crawl, also a great Roguelike with a very different philosophy.
Watching other people play on NAO and/or asking people for help or advice should really improve your NetHack skills!
That is quite a modern concept. Back in ye olden times, people just played local versions of games, and won, and didn't announce their achievements to the world. I am sure there is a plethora of people who won and just didn't brag about it.
Winning Nethack games is just like any other hard accomplishment; you have to plan, know what's needed to win, and have patience. It's an unforgiving game, but winnable with no cheating. I have ascended a couple of times as a Valk and Wizard. I stopped playing a decade ago because it's very hard to play casually; you just get so involved. I am in awe of people who do mind-boggling conduct ascensions like the one here :
Unless you require "honestly" to avoid having ever read the source code. This is one of the rare games where I think understanding the source actually improves the gameplay. There are soooo many interactions that you can't possibly factor it all in; you just sort of get a sense of the programmers' philosophy and that guides the rest of your play. You would have to spend your whole life playing NetHack to develop a thorough understanding of the game rules; you only have to spend 1/10 of your life reading the source to get an equivalent perspective.
It's the only game I can think of where spending 1/10 of your life developing a perspective on it is actually time well spent. The source code (and the comments!) are a joy to read.
It might have been 20 years ago, or even 10, but today there are a lot of popular roguelikes which are still challenging but have fewer unavoidable deaths and less reliance on spoilers. Dungeon Crawl is a good example, but there are a lot of roguelikes even gentler than that.
The linked post is from 2014, and specifically mentions DCSS being "lower headroom". I expect a significant issue here is different use of "forgiving"; IIUC proponents of DCSS would tend to stress YASDs as being "unforgiving", but the focus of the linked article is more on flexibility in play style - forgiving somewhat-less-than-optimal play, forgiving outright recklessness, and occasionally killing you regardless of your behavior are all three somewhat different things.
Perhaps for a very specific and arcane definition of "forgiving." (Or maybe you just mean "more forgiving than Rogue," which is not saying much.) NetHack is famously brutal, and the people who beat it tend to sink hundreds if not thousands of hours first.
> "The game's strategy should not be dependent on spoilers." This is a request to increase the headroom
Not necessarily. You can give players enough information to theorethically enable deduction of the correct choices, will still making the deduction hard. Mixing up my genres, Go is like this, where the rules are very simple and it's just that deriving good gameplay from them is a tricky problem that takes a long time.
Yes — won once, and died another time on the Plane of Fire. (I meant to zap down, not at myself! Damn. Learned to think before acting. NetHack is not as real-time as it feels when you're in a sticky situation.)
3.6.0 is hard thanks to Elbereth being significantly nerfed. I've ascended a few times, but it was much more luck-based due to the early game being much harder. 3.6.1 should be significantly easier due to its reboosting of Elbereth.
---
Read the excellent wiki, spam Elbereth as much as possible, don't get trapped in open rooms, establish good playing habits, and go slow. Much of Nethack's late game is dependent on good habits.
I've won, exactly once. It was the first time I got a knight into the quest (I can pretty regularly get barbarians and Valkyries near gehennom) and it turned into my first time to the wizard, my first invocation ritual, and ultimately my first ascention. I have since got the amulet once, but then had my brain eaten.
Not remotely recently, but my friends and I played a lot of NetHack back in the 90s. I Ascended once or twice with no cheating (and with low-to-no use of Elbereth), and had a couple of good runs that could plausibly have worked out but didn't due to either bad fortune or (in one memorable case) very poor life choices involving a Ring of Conflict.
IIRC, we didn't even really read up on anything spoilery until we were getting down to the castle occasionally. Learning to survive the early-game was a journey both fun and frustrating.
With 3.4, I got to the point where I could probably win at least 1 in 5 games (and did so with every class). 3.6 got more difficult but I'd guess I could win roughly 1 in 10.
(None of that came without playing a lot of games where I did poorly over months and months, along with some liberal use of explore mode, and reading a lot of spoilers. It's a time investment for sure, not necessarily a wise one.)
I've won lots of times in 3.4.3. It feels “easy” but formulaic once you get the hang of it, so you challenge yourself. Best conduct I ever did was just blindfolded through the whole game (from first turn to last).
Edit: I remember looking at BotHack at the time and being surprised at how extensively and elaborately hard-coded the strategy and evaluations were. This is definitely not machine learning!
(defn- read-book [game]
{:pre [(have game book #{:noncursed})]}
(if-not (invocation-complete? game)
(if-let [[slot _] (have game "Book of the Dead")]
(with-reason "reading the book"
(->Read slot)))))
Fun Fact: the original Deus Ex, which takes place in the 2050s, has an Easter Egg. If you hack into a certain computer in the game, there is an email with NetHack release notes (version in the double digits, it being the year 2050).
Being 13 at the time, that was my first time hearing about NetHack, and I decided to try it eventually.
I am big fan of this game, but I wish it had a more appealing user interface - I mean appealing to the general public. Whenever I show nethack to a friend, they're put off by it, and it's a shame because nethack really is a great game and it should be better known outside geek circles.
I'm wondering if these types of categorical errors like 'BUGFIX: kicking cockatrice meat while barefoot now stones you' can be eliminated by writing in a language with a more advanced type system that'll check for these things at compile time instead of relying on years of hard labor and the stone knives and bearskins of C.
Ascended my first tourist yesterday! It's actually a great class to play after the quest. You just have to be very careful (and perhaps lucky) in the early game.
This playlist with a new player being led by an experienced player was very helpful for me when I was learning to play. Still haven't ascended, but what an addictive game.
I'm a Dungeon Crawl Stone Soup player, which is a cousin to Nethack, but I imagine the process is the same: get an IRC client, connect to Freenode, join #nethack, ask questions, and learn to use the utterly invaluable resources provided by the in-channel bots (e.g. the learndb).
NetHack's learndb (Rodney) isn't quite as vital as Crawl's, mostly because NetHack is more spoiler-reliant and (as a result) has a much more detailed and accurate wiki, as well as other sources of spoilers.
No. At first glance, it seems easy, just make other players appear like monsters and add a chat feature. But if you consider more details, you arrive at their FAQ answer:
> We think you can't do that in a playable way without compromising the basic idea of being able to think as long as you want about what you're doing, but many people have made many different suggestions as to the one obvious way to handle things. If you still like the idea, you can try Crossfire, a multiplayer roguelike for UNIX/X11. Other games to check out are MAngband and Diablo
I disagree with the reason for not implementing it, adopting something like "simultaneous turns" mode from Civilization series could provide that time to think for everyone. But I'll check out Crossfire.
I don't think Civ's simultaneous turns would work for NetHack. In Civ, you make quite a few actions on a single turn (especially in the late game). In NetHack you are always taking a single action at a time, though with some actions taking no time at all (viewing your inventory, naming and calling objects, naming monsters, applying a stethoscope for the first time that turn, etc) and other actions taking multiple turns (running, travel command, moving while slow).
Essentially, NetHack has a different concept of turns than Civ. NetHack uses turns as a discrete time resource whereas Civ uses the more conventional boardgame style rotation. You could try to convert NetHack into the Civ model by giving the player a number of action points based on their speed but this would allow fast players to move around a corner, take an action, and retreat without the enemy getting a chance to notice them. This would fundamentally alter the balance of the game.
There are a whole lot of multiplayer, D&D-like online games with NetHack and other roguelikes in their inspirational ancestry (some of which are actually branded as D&D games.)
On further reflection, most of them are single player / multiple character, excluding ones that are very far from any roguelike influence. Actual multiplayer that still have much resemblance to roguelikes (especially the turn-based nature) are harder to find (IIRC, Neverwinter Nights might qualify, though largely through the influence of roguelikes back on D&D, since it was a fairly direct translation of then-current D&D.)
I remember compiling NetHack back in 1993 or so for x86. The binary was well over 900 kilobytes of machine code (according to the size utility). I was like, "Gack! That's sized like some bloated Unix kernel!".
I read the headline with excitement and thought, "Oh, God, is it the late 80's again and all of this was a dream? Let me get on USENET and confirm...."
I've been spoiled by Crawl's autoexplore, its richness of gods and god-granted powers, its colors (in text mode, not tiles), and many other new features and innovations. That said, Nethack is still the deeper game in so many ways. Crawl's "don't do it like Nethack" approach has prevented it from taking all sorts of great ideas from Nethack and making Crawl in to a richer game -- things like making more advanced use of pets (like training them to steal from shops), more things to do in shops, use of altars beyond merely converting to a god, use of potions beyond merely drinking them, more uses for item inscriptions, all the ways to combine items, the list goes on and on.
All this makes Crawl seem like the shallower game, though it's still plenty of fun, and I still play it to the exclusion of most any other roguelike for now. There's a lot of room for improvement, though. I just wish they'd take more of the best parts of Nethack and make Crawl even more fun.