undefined | Better HN

0 pointsthrowanem1mo ago0 comments

Thank you for explaining the joke.

You're quite abjectly wrong, though. Text adventures were heavily advertised, in their illustrious and very brief moment of sunshine, as 'accepting English input' (cf. Maher, The Digital Antiquarian), which by definition constitutes NLP. They were just extremely bad at it, hence their accompaniment by a constant stream of excuses like the one you just made. (You must have had to dust it off first! That one is older than me.)

0 comments

kqr1mo ago

Our understanding of what makes for a fun game has evolved significantly since the 1980s. Designers of text adventures today generally agree that structured, non-natural input is a good thing and reduces frustration in the end. I can't think of any prominent text adventure designer who still pretends the parser understands English. There are also no widely used text adventure development systems that even strive to understand English in their parsers.[1]

I would understand your joke if it was made in the 1980s, but today it only shows a very old misunderstanding of the genre. (One might say you must have had to dust that misunderstanding off first!)

[1]: The systems that do strive to understand English – usually through LLMs – generally do not result in very satisfying games. They are primarily made by AI enthusiasts rather than text adventure designers.

throwanemOP1mo ago

Now you're defending games from the 70s and 80s on the basis of technology and design attitudes from today.

Your profile says you are a quantitative analyst. (I take this as reliable, since it does not also call you a rhetorician or humorist.) The fetish for logic puzzles thus checks out, but I admit I had thought the denumeration of time among the arts of number. Or had you mistaken Graham Nelson, Emily Short, and allies for Infocom alumnae?