next »

[Radix]

I just gave Brogue a shot myself until I got bored and coincidentally (or not!) that happened to also be level 8 in my case. It's pretty cool, I like that it's very Rogue-ish and pretty much just adds a lot of new elements to the basic game, but it is indeed way too easy on those early boards. I generally like the restricted diagonal movement but in this case with the more winding organic layouts it just makes the floors a pain to navigate and as a result they take two or three times as long to complete as they would in Rogue, compounding the issue.

Auto-explore... that's interesting. Technically cool but to me it seems like if the first seven floors of your roguelike are too simplistic for players to want to navigate themselves without it being a chore then you really ought to be starting the player on level eight.

I can see how it might be a good introductory game though.

[Waltorious]

Auto-explore... that's interesting. Technically cool but to me it seems like if the first seven floors of your roguelike are too simplistic for players to want to navigate themselves without it being a chore then you really ought to be starting the player on level eight.

Dungeon Crawl Stone Soup also has auto-explore, and it's one of my favorite features. Coupled with auto-travel (i.e. automatically go down the nearest stairs, or automatically take me to level 5 from level 2, etc.) it makes moving around so much easier that pretty much any roguelike that makes me navigate manually seems annoying now.

The auto-explore in Brogue is nice but not quite as good as Dungeon Crawl's.  It seems to take less optimal routes sometimes and it often seems to ignore items as well.  So I find myself using a mix of manual travel and auto-explore.  I agree that the limited diagonal movement can be hard to get used to.

One other complaint about Brogue is the inability to examine items. I identified a ring as a "ring of awareness" but that's all the info I got.  I didn't actually know what it DID.  I eventually figured it out, I think, but I prefer when identifying an item actually tells you its function.

[Hayden Scott-Baron]

Why are the user interface for roguelikes so poor? Is it legacy, as some sort of assumption for basic keyboard shortcuts for common actions?

I'm trying out Dungeon Crawl Stone Soup right now, and I'm astonished that even the mouse interface often requires me to shift or ctrl-click on things, or use the keyboard to confirm my intent.

I'm a big fan of Shiren the Wanderer, and I've had some good fun with Powder, but other than that I'm finding this genre to be so poorly presented that it's hard to overcome. 

Help me tiggers, there must be some games with a decent user interface!

[Oddball]

That's my main issue with roguelikes. It's like the makers are deliberately holding on to arcane control schemes to foster some kind of mystique about the genre. The first thing I did when I started making Hack, Slash, Loot was throw out the archaic roguelike controls and replace them with context sensitive mouse clicks. Everything is controlled by a single mouse button, no hotkeys or key combos to learn whatsoever.

[tametick]

Why are the user interface for roguelikes so poor? Is it legacy, as some sort of assumption for basic keyboard shortcuts for common actions?

I think a lot of it is due to the fact that hard-core players actually *want* their dozens of keyboard shortcuts, and that's the demographic roguelike developer cater for.

[Hayden Scott-Baron]

I'm almost much at the point of 'fuck these guys' with roguelikes. It's like a revoltingly insular coder's only game of soggy biscuit.

Much like visual novels, FPS or even 2D platformers (double jump anyone?), they have a ton of potential and succeed in so many areas, but can't resist pandering to the established conventions and handbook of interface design.

[Radix]

I agree that the limited diagonal movement can be hard to get used to.

It's not hard, it's not the only game to have that, it's actually a feature I like. However it isn't suited to levels with complex cave-like shapes.

And yes, auto-explore is a cool feature. That wasn't what I was objecting to.

[s0]

I think it's great that roguelike developers don't let themselves get bogged down by "accessibility"  concerns and focus on the substance of their games. Also, all roguelikes use more or less the same set of commands, so once you've learned the controls for one of them, you've essentially learned them all.

[Hayden Scott-Baron]

I think it's great that roguelike developers don't let themselves get bogged down by "accessibility"  concerns and focus on the substance of their games. Also, all roguelikes use more or less the same set of commands, so once you've learned the controls for one of them, you've essentially learned them all.

You mean, much like fangames and community made map packs? Those are great too.

Accessibility isn't a dirty word. It can also lead to accessing a new type of game design rather than tiptoeing around the exact same ruleset.

[PaleFox]

There are accessible roguelikes too. POWDER is famously made to be played on a Gameboy Advance; DoomRL is simplified and yet very enjoyable; there are others as well. It's just that the complex interface has become an expected trope, similar to tropes in other genres. Why doesn't the main character in an FPS ever use more than one pistol? Or tape a flashlight to a gun? That might be helpful in some situations, surely.

Why do so many indie platformers have lowfi pixel art?

[s0]

Yup, there are plenty of accessible, simplified "coffeebreak" roguelikes, most of which have already  been mentioned in this thread. And while a lot of them are great games, they're not the same as the "traditional" hardcore roguelikes. There's a reason why most of them have complicated interfaces. Stone Soup and Nethack Vulture's Eye are pretty much as far as you can go simplifying a roguelike's interface without sacrificing depth (both could still be pushed a bit further I guess).

About context sensitivity: While a lot of the interactions in a game like Nethack, like equipping a food ration as a weapon for instance, are meaningless, simply the fact they're even possible is part of the game's charm. There are also a lot of possibilities for emergent gameplay, coming up with wacky and unusual solutions to problems, etc. because of that as well. It's an openness you don't see in often in games and one that I think couldn't be conveyed effectively with a simplified interface like DoomRL's.

And if you're looking for "innovation", I'd recommend checking out some 7DRL compo entries, though I think that there's merit refining a formula over time and not "pandering to the established conventions" of mainstream interface design and "accessibility" maxims as well.

[PaleFox]

There's a certain amount of silliness available in nethack and such not available in other things. For example, it's often a tradition in nethack to kill Vlad (a notoriously easy boss encounter) with some nonsense item, such as an amulet or pies, to show the player's contempt -- but that's only possible because the game allows you to beat people to death with a dead chicken if you so desire.

You can dip things into liquids, engrave the ground or objects, name them, write on them, eat them, whatever. It's part of the charm and appeal I think.

[Hayden Scott-Baron]

None of the functionality or complexity behind ASCII based roguelikes needs to be lost in improving the interface. There's nothing at all that needs to be stripped out in order to make the controls and presentation better.
None of the abstract visuals need to be lost in making it less messy.

It's relatively pointless to bring up, however. Plenty of people are content, (or "content" if you "prefer") with nethack derivatives, so there's no reason to rock the boat. It's just disappointing to see so many games paint themselves into using walls and floors made of ##### and .....

[s0]

It's just disappointing to see so many games paint themselves into using walls and floors made of ##### and .....

Wait, I thought this was about interface complexity, not visuals?

[Hayden Scott-Baron]

Visuals are part of the interface. An arcane series of unrelated character symbols is part of the interface. A lack of visual feedback is part of the interface. 

[s0]

I don't think roguelike ASCII visuals are all that "arcane" TBH. I dunno, I've just never had any problems with them. And there's text feedback, i.e. "you hit the dragon for 5 damage, the dragon is severely wounded." I mean, what's the issue with accepting that a "#" represents a wall and a "." represents a floor anyway? Besides, there are tilesets.

The thing is, roguelike devs are indies just like you or me. They don't have infinite budgets or man hours (in fact they often have no budgets at all). And as I'm sure you know, polished graphics and detailed visual feedback take a lot of time to make, and especially in the case of roguelikes which often evolve constantly over long periods of time, having to come up with some sort of graphical representation for everything while keeping things consistent can slow down development considerably.

Basically, what the developers of traditional roguelikes do is freeing up time they would otherwise spend on graphics, sound and accessibility to improve the aspects of their game they think are important. That's what makes hardcore roguelikes unique, that they're willing to sacrifice everything for depth and complexity. Also, seeing a roguelike gradually evolve is a great thing and it's what makes the genre pretty much exclusive to indies. Working on a game for 20 years would be commercial suicide in the mainstream industry.

[Radix]

I'd argue that the fixtures of the Rogue genre are a shorthand that helps define a game and sets player expectations in a useful way. Kind of like how I've occasionally used Mario (SMB 1-1 in particular) for stupid little kotmk things where I want to present a setup that begins in a completely familiar way so that the deviations are more striking.

To put it another way, you're using common presentation to go along with a common basic ruleset and common control conventions to make a particular audience comfortable and building on those rather than reinventing any wheels, deliberately, conceding that the conventions may not be exactly optimal.

For that hard core of roguelike fans the conventions are in some way part of the appeal. Nethack has a graphical tile mode and things like Falcon's Eye, but I'd wager most players play in textmode regardless. I'm not going to talk about benefits although a case could be made for certain things, but there's a definite preference among that core.

Could more accessible games open the genre up to far more players than would be put off by any changes? Of course, and there are people doing that sort of thing, and some really interesting, valuable projects have come out of that approach. But for the most part roguelike development is a labour of love, just like the people who still work on classic adventure games or classic IF or other dead love-letter genres that are intended as much for their own communities as for players in general.

tropes

Just FYI, "trope" doesn't mean what tvtropes implies that it means and it's a little uncomfortable to watch somebody misuse the word when you know exactly where they got their false definition from.

[rszrama]

Good thoughts above. And I'll just add that using a mouse to walk around a dungeon just seems stupid to me. That's what arrow keys are for. Having to click on everything I want to use seems equally stupid to me. That's what hotkeys are for. Sure, perhaps you can provide combined input formats, but some of the things you're castigating the genre for are time savers and "accessible" gameplay items for fans of the genre.

Ditto what was said above about graphics. I have two options: I can use ASCII graphics or I can use Oryx's tileset. I'm not an artist, and there's no way I can create distinctive art for everything in a game. But if everyone in the RL world knows an @ is a player, I'll use that. And if no one would EVER be able to recognize any dragon sprite I could make as a dragon, I'll use a D for that.

One interesting point of comparison may be with interactive fiction. The way the argument went above, it seems like IF should've died out as soon as Maniac Mansion came out. No more having to memorize interaction verbs, figuring out what to look at, having to keep maps. You can just walk around, point, and click.   

In other words... the way you interact with the world is part of the experience. But of course, no one's stopping innovation, and various games have already been mentioned that "break the mold".  I don't see why every other developer should change their games and intended experience just to keep up with the "easier to play" games.

[Mipe]

Heh. Some people, like my mom for example, believe that having to use keyboard for games is stupid. They want to move by clicking on the destination, not by repeatedly jamming the arrow keys. They also want to be able to click items instead of going through a chain of obscure keybinds.

[Hayden Scott-Baron]

I can use ASCII graphics or I can use Oryx's tileset.

Honestly, I'd find it interesting if there were a middle ground. Plainer graphics with walls and surfaces represented. I love cute squat characters, but I appreciate it doesn't fit the tone of every game, nor the sense of imagination.  A radar with blips can be pretty engrossing if the context is understood, as movies and television have shown us.  

One interesting point of comparison may be with interactive fiction. The way the argument went above, it seems like IF should've died out as soon as Maniac Mansion came out. No more having to memorize interaction verbs, figuring out what to look at, having to keep maps. You can just walk around, point, and click.

This is a pretty incomplete comparison. I do get where you've coming from, but the visual presentation of IF is akin to what human beings are used to. We're also used to typing into our computers, now moreso than ever.   It's true that the verb based interface is a barrier to entry, but it's one of the only barriers to entry.  In theory an IF game could be made entirely via SMS or IM (even Facebook IM!) and the user wouldn't ever know there was a human on the other side. Far fetched, of course.  

I agree that IF has much in common though, and it hasn't been able to bridge the interface problem. Most IF fans have little interest in addressing this problem, and that's perfectly fine too.

Representing a top down map with ### and @ is by no means akin to how people are used to seeing visual map data. I'm not suggesting that it's practical to be able to implement something better, but I think it's blind to imply that ASCII is some sort of optimal representation of tile based map data.



As a quick and sloppy test, here's a university map, inverted. We have a key to symbols, we have labels, we have colours. People are used to looking at map data, be it google maps, building plans, or the GPS on their car, but they're not often told that & is a tree when it could very simply be rendered as a universally accepted tree shape.  ASCII characters aren't even square.

This is not a proposal as a solution to ASCII visuals, I just brought up as a reference to similar data in other formats.

I agree that dragons et al are difficult to represent, but even a large ominous green circle or square might be better than the letter D. It's also important to pay attention to the grid for movement, and to other necessary data, but I don't think ASCII does any of these better.

Obviously graphical tilesets aren't the solution, especially when they're offered as an option. It's almost always a sloppy layer to the interface rather than an integral part of the game design. As for contextual usage, I expect some RL games do this, but they could offer a list of interactions and put the common ones at the top, with the facility to scroll to the more peculiar behaviours.  

Often it's the Application interface that sucks as much as anything. Obviously this will vary from game to game, but for obvious reasons they feel like DOS or Windows 3.1 products at best, with little regard for modern application interface. This isn't entirely about mouse input either, there have been a great many developments in how to present information to users.