"It's not like people see our $50,000 a year and think 'Hey, I want a piece of that pie.' They'd much rather look towards things like Minecraft, where there are hundreds of millions of dollars."
The brothers have this massive, massive blindspot regarding the importance of the UI. That $50,000 cap they perceive is barely scratching the surface of what would be possible for them if they made the damn thing a bit more accessible to the average person. It's great that their small niche base loves ASCII art and doesn't mind fumbling their way through a hostile interface in order to play the game, but that is the primary reason the game stays confined within a tiny niche corner of the gaming market.
They have the real possibility that if they did that seed work, even just on exposing an API into the game, as opposed to venturing into a full-on UI overhaul, that enough new people would be introduced to the game that they could hire a third developer to focus specifically on the API and UI aspects of the game, leaving them free to keep working on the bits they love.
But do you know much about dwarf fortress? The brothers have responded to questions about this a number of times (regarding UI) but more importantly than that, it appears that they don't particuarly care about making the big bucks.
They are totally happy just pottering along with a loyal hardcore following, earning a very middle class salary, doing something they are massively passionate about. In fact, it seems that what they have is all they ever want in life. They are in their nirvana.
It's not the ASCII interface. I love ASCII art, and they managed to pull a high information density. You have graphic tile sets too. Although fan-made, these have a great quality.
The problem is the inconsistent input system. The same task uses different shortcuts in each menu, which makes it hard to remember. Even selecting rectangular regions of the world is done in a particular way depending on what you are doing.
(Just having fun analogy time as I wake up, once I got rollin it just kinda got outta control)
I don't actually use C++, not since about 2005 anyway, but like many on HN that won't stop expressing strong opinions on the UI. The semicolons dripping everywhere are a massive UI problem. Makes it look like a Perl obfuscation competition. A modern language like Intercal does not use semicolons, well, not as much. They need to get with the program and use UTF-8 so I can program in my native Klingon. I'm sure this Stroustrup character likes his language very much, thank you, but he could make a heck of lot more money, for a lot longer, if he replaced the UI with something a little more modern and polished, like a mobile iphone app where you can end your statement lines with pink hearts and smiley faces. And C++ needs social networking and multiplayer.
C++ also has a horribly inconsistent UI, look at these two ways to add one:
blah = blah + 1;
vs
blah++;
How is the average noob off the street supposed to handle inconsistency? There should only be one way to do it; just like sex. And don't get me started on those Perl heathens.
Its just a hostile UI and C++ would be more accessible to the average person with a better UI. Because the only people who matter are the average people, F those outliers.
That's nice that some people have buckled down and learned all the peculiarities of the existing C++ UI, but we'll just have to abandon them with our "improvements". I have Stroustrup's latest edition of C++ and its bigger than my metro area phone book, which is far too long compared to the copy of K+R I got in 1983 or so which was so small it was mistaken for a copy of readers digest. Heck even "Modern Perl" sitting on my desk right now is only 280 pages in dead tree form.
Niches are inherently bad and we should all want to either conform or dominate those who conform, so its inherently, of itself, bad, if only a fraction of the population "gets" what they're doing. Just like art and music! Quantity has a quality all its own. Can't they just watch football like normal guys? I mean ten trillion horseflies can't all be wrong, or however that goes.
>The brothers have this massive, massive blindspot regarding the importance of the UI.
No they don't. They keep saying over and over again that they know the UI is bad, and they know it drives away potential players. They said it right in this very article even. I am always amazed at how many people are so desperate to tell others how to run their affairs that they invent false ignorance in others so they can step in and tell them how to do things.
If the history of Dwarf Fortress interests you, I recommend you check out the history of roguelike games (probably the nerdiest of all my passions.) Dungeon Crawl: Stone Soup, for example, is a beautiful illustration of applying open source software development tenets to a massive game, as dozens of contributors have continuously evolved and enhanced it over time (and its massively fun to boot.)
Speaking of the history of roguelikes, I've been pretty impressed by Nicolas Casalini, aka Darkgod. I'm not sure exactly how long he's been involved in the scene, but he's been developing since at least Pernband back in the late 90s.
His Tome2, in the early part of the last decade, was arguably the pinnacle of old-school band style roguelikes (as opposed to the hack style popularized by nethack, or more modern, user-friendly ones that have been cropping up the past couple years). The current Tales of Maj'Eyal is one of the most polished, engaging roguelikes I've come across.
I like in this particular article, how the interviewer delves into how it is possible to sustain interest in coding a single game for so long (and potentially for decades). Anyone who has coded games as a single dev probably knows how bogged down you get as the game becomes more complex. Generating graphics yourself is a huge time suck which they have neatly avoided. What's intriguing is that the game design appears to be flexible enough to spend time coding mineralogy and then jumping to say horticulture while still being manageable. It would be interesting to know more about the design to see why it is so extensible.
I've been reading articles about this fascinating game for several days now, but haven't had the guts to invest the time into it given its steep learning curve. Kudos to these guys for doing exactly what they want to do and (yet) still making a living by coding.
Ever since I read the NY Times article posted here a couple years ago (see below), I have had a similar response. I love reading about Dwarf Fortress, but I'm hesitant to actually try it.
It's actually a lot more accessible than it seems.
For anyone interested in learning to play DF, I highly recommend YouTube user 51ppycup's "Dwarf Fortress for Dummies" series - he gets the basics across really well and you'll be striking the earth in no time.
If you are just curious about the game, there are many "Let's Play Dwarf Fortress" videos on YouTube. Expert players talk through the game features and progression.
I'm biased as a long time player but with a graphics tile set I don't think the graphics are any worse than your average DOS or console game from the early 90s.
The only problem is the input system but anyone who has mastered Unix can easily master DF.
> "We've just had our mitts on this one for a long time without passing it along," he says, "but that's not going to be a forever thing, obviously. It's only going to stay afloat as long as people keep it afloat, right?"
> "I think the game has actually gotten easier compared to what it was in 2006," Adams reasons. "It's not really the mechanics that matter so much, since a lot of the mechanics in DF are under the table. A lot of the updates don't matter either -- I mean, I spent a month on beekeeping, but you're not confronted with beekeeping, and you don't need to learn how to do it, but if you want to make wax crafts and honey, then it's an avenue you can explore."
Perhaps it's the language weenie in me, but I can't help but think they're doing this wrong. If you're going to work on something for 30 years (with the Moore's law that implies), wouldn't it make a heck of a lot of sense to use the highest-level language you can, start building up DSLs for specific things like crafts, and other strategies so you don't spend a month hacking on a sprawling C++ codebase just to add in beekeeping?
[+] [-] axefrog|12 years ago|reply
The brothers have this massive, massive blindspot regarding the importance of the UI. That $50,000 cap they perceive is barely scratching the surface of what would be possible for them if they made the damn thing a bit more accessible to the average person. It's great that their small niche base loves ASCII art and doesn't mind fumbling their way through a hostile interface in order to play the game, but that is the primary reason the game stays confined within a tiny niche corner of the gaming market.
They have the real possibility that if they did that seed work, even just on exposing an API into the game, as opposed to venturing into a full-on UI overhaul, that enough new people would be introduced to the game that they could hire a third developer to focus specifically on the API and UI aspects of the game, leaving them free to keep working on the bits they love.
[+] [-] robbiep|12 years ago|reply
But do you know much about dwarf fortress? The brothers have responded to questions about this a number of times (regarding UI) but more importantly than that, it appears that they don't particuarly care about making the big bucks.
They are totally happy just pottering along with a loyal hardcore following, earning a very middle class salary, doing something they are massively passionate about. In fact, it seems that what they have is all they ever want in life. They are in their nirvana.
Good for them!
[+] [-] kaoD|12 years ago|reply
The problem is the inconsistent input system. The same task uses different shortcuts in each menu, which makes it hard to remember. Even selecting rectangular regions of the world is done in a particular way depending on what you are doing.
Inconsistency is awful for users.
[+] [-] VLM|12 years ago|reply
I don't actually use C++, not since about 2005 anyway, but like many on HN that won't stop expressing strong opinions on the UI. The semicolons dripping everywhere are a massive UI problem. Makes it look like a Perl obfuscation competition. A modern language like Intercal does not use semicolons, well, not as much. They need to get with the program and use UTF-8 so I can program in my native Klingon. I'm sure this Stroustrup character likes his language very much, thank you, but he could make a heck of lot more money, for a lot longer, if he replaced the UI with something a little more modern and polished, like a mobile iphone app where you can end your statement lines with pink hearts and smiley faces. And C++ needs social networking and multiplayer.
C++ also has a horribly inconsistent UI, look at these two ways to add one:
blah = blah + 1;
vs
blah++;
How is the average noob off the street supposed to handle inconsistency? There should only be one way to do it; just like sex. And don't get me started on those Perl heathens.
Its just a hostile UI and C++ would be more accessible to the average person with a better UI. Because the only people who matter are the average people, F those outliers.
That's nice that some people have buckled down and learned all the peculiarities of the existing C++ UI, but we'll just have to abandon them with our "improvements". I have Stroustrup's latest edition of C++ and its bigger than my metro area phone book, which is far too long compared to the copy of K+R I got in 1983 or so which was so small it was mistaken for a copy of readers digest. Heck even "Modern Perl" sitting on my desk right now is only 280 pages in dead tree form.
Niches are inherently bad and we should all want to either conform or dominate those who conform, so its inherently, of itself, bad, if only a fraction of the population "gets" what they're doing. Just like art and music! Quantity has a quality all its own. Can't they just watch football like normal guys? I mean ten trillion horseflies can't all be wrong, or however that goes.
[+] [-] papsosouid|12 years ago|reply
No they don't. They keep saying over and over again that they know the UI is bad, and they know it drives away potential players. They said it right in this very article even. I am always amazed at how many people are so desperate to tell others how to run their affairs that they invent false ignorance in others so they can step in and tell them how to do things.
[+] [-] alex_c|12 years ago|reply
http://lparchive.org/Dwarf-Fortress-Boatmurdered/
It's a fairly long read, but as always when Dwarf Fortress is involved, it's well worth the effort.
[+] [-] fleshgolem|12 years ago|reply
Still a great read though
[+] [-] jmduke|12 years ago|reply
[+] [-] ricree|12 years ago|reply
His Tome2, in the early part of the last decade, was arguably the pinnacle of old-school band style roguelikes (as opposed to the hack style popularized by nethack, or more modern, user-friendly ones that have been cropping up the past couple years). The current Tales of Maj'Eyal is one of the most polished, engaging roguelikes I've come across.
[+] [-] boneheadmed|12 years ago|reply
[+] [-] boneheadmed|12 years ago|reply
[+] [-] sage_joch|12 years ago|reply
http://www.nytimes.com/2011/07/24/magazine/the-brilliance-of...
[+] [-] ecdavis|12 years ago|reply
For anyone interested in learning to play DF, I highly recommend YouTube user 51ppycup's "Dwarf Fortress for Dummies" series - he gets the basics across really well and you'll be striking the earth in no time.
Dwarf Fortress for Dummies 2012: http://www.youtube.com/playlist?list=PLD1A3FE72C0DCAC66
[+] [-] _pmf_|12 years ago|reply
And by steep, you mean vertical.
[+] [-] thejosh|12 years ago|reply
[+] [-] cpeterso|12 years ago|reply
[+] [-] INTPenis|12 years ago|reply
The only problem is the input system but anyone who has mastered Unix can easily master DF.
[+] [-] gwern|12 years ago|reply
> "I think the game has actually gotten easier compared to what it was in 2006," Adams reasons. "It's not really the mechanics that matter so much, since a lot of the mechanics in DF are under the table. A lot of the updates don't matter either -- I mean, I spent a month on beekeeping, but you're not confronted with beekeeping, and you don't need to learn how to do it, but if you want to make wax crafts and honey, then it's an avenue you can explore."
Perhaps it's the language weenie in me, but I can't help but think they're doing this wrong. If you're going to work on something for 30 years (with the Moore's law that implies), wouldn't it make a heck of a lot of sense to use the highest-level language you can, start building up DSLs for specific things like crafts, and other strategies so you don't spend a month hacking on a sprawling C++ codebase just to add in beekeeping?
[+] [-] idoescomputers|12 years ago|reply