In the previous AI threads here, some people argued that there wasn’t much point of that kind of more “understanding” parser, as experienced players won’t have any use for it, and new players will be less helped by it than you might think.
5 Likes
Thinking about it: having something make assumptions about verb synonyms might be helpful. Nouns, on the other hand, might not be so useful. At least on the Inform side of the house we are always talking about duplicates and disambiguation, even in cases where the number of possible nouns are few.
3 Likes
My experience of playing the small handful of games that have tried having NLP parsers is that, in addition to the issue of them being kind of slow and large, they’re not more responsive than a well-implemented traditional parser game, but are much more likely to give you weird and unhelpful error messages due to whatever processing they did to your command. Like, playing [i] doesn’t exist, I remember typing “get the [noun]” and being told I wasn’t close enough to the [completely different noun] to interact with it, when actually the problem was that I had typed “get” and the game only understood “take” (or maybe it was the other way round, but it definitely only understood one of the two).
One could argue that none of this is an innate problem with the current technology available to create NLP parsers and it’s just happenstance that the ones that have been made so far have been bad, but I’ll remain skeptical until someone comes out with an NLP parser game that isn’t somehow a more friction-ful experience than a traditional parser game.
Based on the reception of The English Restaurant, I’m going to say no.
10 Likes
Lukewarm Take: Just because IF is text-based doesn’t negate a need for an aesthetically pleasing interface. Most IF looks utterly boring, even when the games are amazing to play.
11 Likes
My C64 runs games with graphics.
4 Likes
@Mike_G
I checked out this video after you mentioned the C64 version of Gauntlet. This one goes through all the incarnations of classic Gauntlet, from the arcade version through the C64, Amstrad, ZX Spectrum, Master System, NES, Megadrive/Genesis versions. Apparently, you’re not insane; the C64 version was top notch!
1 Like
I don’t disagree, but I try not to judge too hard because this is a community of amateurs (and I would frankly be throwing stones from my glass house). I wish the default layouts for most commonly used choice programs weren’t so ugly (Harlowe, Sugarcube) or violating of accessibility standards (Ink), which would give people a stronger starting base to work from.
That said, @manonamora is doing yeoman’s work here and has put together collections of snazzy free templates for both Harlowe and Sugarcube that I think more people should be aware of! There’s also more templates out there on Itch if one goes looking. I’m not exactly sure how best to shepherd newbies towards them, but I think we should normalize using templates for the CSS-challenged among us.
5 Likes
This intersects with thoughts I’ve had while reading this thread, about how much both computer interfaces and the way people discover computer games have changed over the past 40+ years. A basic parser prompt was super-similar to the command line or BASIC prompt found on all the computers Of the Time, whereas now it’s a real stretch to say “and now you will type your intentions.”
But the thing is, people have been trying to prettify the text interface for basically all that time. Most of the attempts we got, like Interplay ADVENT 2.0 and the Legend interface and on and on were Not Great, and I’d argue actually counterproductive. Once you prettify it enough that it’s actually pretty, you get SCUMM and now oops it’s a different game style.
4 Likes
I think he missed the important ones, the Amiga and Atari ST. Being a 16-bit game, it was the 16-bit ports that were up to it (except the Apple IIGS one, which was weak) rather than the admittedly interesting ‘how well can we translate this to 8-bit hardware’ mission of the other mentioned ports except the Megadrive.
-Wade
1 Like
Agreed.
I’m also talking about text adventure interfaces as well. No IF format is safe from my ire. 
A while ago, there was a topic about what can be done to grow the IF community. I suggested that IF should look sexy (market itself, basically). I don’t think that’s unreasonable.
Not everybody has the ability to design interfaces well, so it’s a big ask. Most programmers are not designers. Most writers are not designers. Most users are not designers. Most designers are not programmers… hot take!
4 Likes
I’ve lost the OP, but I’m pretty sure this community doesn’t like AI when it’s a way to scam artists of various kinds of their art and use it for free and not against AI per se.
I use ChatGPT a lot when in need of a search of something. It is way faster and usually gets the right answer right at the beginning while Google (which is another form of AI btw) can take evenings. Ofc, it must be double checked as ChatGPT is sometimes not completely sincere 
My other hot take. The best Infocom title ever is Leather Goddesses of Phobos. Period.
1 Like
I’m a writer, a programmer (sort of) but especially a designer… and I know I would HATE whatever interface you IF people could make.
The same applies vice-versa so, no, I’m not volunteering to design a parser interface that nobody would like.
4 Likes
Ha! Love it! 
In my experience, a good interface designer is not an artist. They are like architects; draftsmen. They express themselves creatively within a technical confine.
The best interface designers are also competent editors to their own work. They do the most, with the least… and still make it attractive! That’s where the excitement and challenge lies in interface design. 
1 Like
Making games is hard.
14 Likes
I did enjoy Gauntlet on the C64. When I said I could play it all day I meant I could literally play it all day and never die. I have no idea what level I made it to…whatever you could reach in something like six hours of gameplay I guess.
2 Likes
Related, here’s my actual hot take: if people want IF to appeal to more people, they should be willing to add a little bit of graphics & a bit of a GUI. Not to the point of directly representing everything described in the text (which would forfeit the advantages of a text-based medium), but there needs to be a minimum attempt to look like “a game” rather than a wall of text or else the general public simply will not be interested. In other words, I directly disagree with mathbrush’s take here.
Obvious reference points for me are 80 Days and Reigns, which were both successful in mainstream game markets, despite (IMO) being pretty pure choice-based IF from a structural/gameplay perspective, because they have a minimum level of GUI/graphics-y-ness wrapped around the whole thing.
Personally, my Ink game Serum Run got featured on the front page of Itch and played by thousands of people, and I’m 99% certain that this wouldn’t have happened if it wasn’t for the (very minimal) graphics. I maintain that it’s still fundamentally “an Ink game” (in fact, the gameplay-complete version that I sent to alpha testers was literally just a vanilla Ink web export), but I guess maybe the community sentiment here is that the addition of graphics makes it no longer IF, despite the gameplay being literally identical?
9 Likes
I think there’s unexplored potential in trying to design a parser interface which literally nobody likes.
> GET LAMP
Not if you ask like that.
> PLEASE GET LAMP
I'm not listening to you until you start using proper grammar, you plebian.
> PLEASE GET THE LAMP
Capitalisation, punctuation ... do they teach you nothing in school these days?
> Please get the lamp.
You didn't say "Simon says."
> Simon says please get the lamp.
You can't see any "Simon" here.
21 Likes
I think it’s important to remember that a lot of people make IF in their free time and aren’t necessarily going to crank out something like 80 Days on the weekends. Saying most things are ugly is a little like saying most things are badly written. I know nobody means it that way, but I wouldn’t want any new authors to be discouraged.
A lot of people here make stuff for the sheer joy of making it, and a lot of us play that stuff because we enjoy it. I think that’s alright.
(Edited for clarity)
13 Likes
Obligatory link to the IF-Arcade version of Pac-Man for those who haven’t encountered it.
Checkers: 1.71/5
Chess: 3.65/5
Go: 3.93/5
BoardGameGeek doesn’t just list contemporary games.
1 Like
Fifteen years ago, maybe, but no, if you pay attention to most major IF events these days you’ll reliably see a couple games with at least that level of graphics involvement and no one complains. It remains a small minority of games simply because most of us are amateurs making games entirely by ourselves; doing art/graphics requires yet another unrelated skillset on top of the ones you already have to develop to be successful in IF (writing, programming, game design), and not everyone has the time and energy to hone all those separate skills—plus it adds more time and effort to the process of making each individual game, and I, for one, barely manage to finish my projects as it is.
(Or, of course, you can use free art assets someone else has posted for the general public to use, but just from looking for stuff to use for cover art I can say that looking for stuff that fits the content of your game, meshes into a cohesive aesthetic, and has the right license for you to use it isn’t something that has zero time-and-effort cost either.)
Edit: To be clear, I’m not saying everyone would have graphics if only it were easier, just that I think the percentage of games that did wouldn’t be quite so tiny. Mostly my point is that there are reasons the percentage is that tiny that aren’t “the community as a whole just hates the idea of graphics”, which I don’t feel is true anymore.
11 Likes