You type in your alt text under the image. For extra flair, use <small></small> tags.
I see Dannii has given the proper advice, but if you don’t want to mess with that just say “ABOVE IS A PICTURE OF AN ANGRY BEAR” 
1 Like
A test (there should be alt text in this image):
1 Like
A screenshot of the “last visit” forum indicator.
Yes! Apps and game sites are a great way to learn games that are especially bookkeeping-heavy. Or even for stuff I wanted to get better at but were tough to get to the table. I wound up doing that with Sekigahara, because it can be a three-hour game, but I found Yucata where I can play asynchronously with fantastic implementation. It’s funny how that changes your play too, because after I’d gotten lots of sessions under my belt and sat down to play the physical copy with my wife, she still trounced me because I felt rushed having to make decisions in real-time, haha.
I also love how board games kind of train your brain to understand how programming works (on a very basic level), since the rules are the code, and the players are the computer executing the code. Hopefully the “code” is well-written!
I think the highest I’ve seen on BGG is The Campaign for North Africa, a 1979 war game with a 4.75 complexity rating. The game infamously takes 40+ days to play, and is apparently a deep enough simulation that it has players tracking stuff like gasoline consumption based on the weight of individual vehicles’ cargo or something crazy like that. It’s cool that it exists, that’s about all I can say about it.
We’ve got Lisboa, which is a euro-style game rated at 4.58, and while it is indeed complex, I think that’s relative. It’s often a question of chunking the thing into its smaller interlocking systems, and over time I got much better at recognizing the smaller mechanisms at work (much like @HanonO’s point about IF systems). With something that complex, you just have to be okay with not knowing exactly what to “go for” your first couple games, and play them close together enough not to forget what you’ve learned between plays (that’s the trickiest, because adult playdates are tough). And Lisboa’s got a nice solo mode so altruistic rules-learners can study up without subjecting others to too much confusion. But it is a top candidate for my “list of board games that scared the bejeezus out of my left-brain when I opened the box.”
2 Likes
Lots of games require the player to make inputs quickly and accurately. There’s definitely an audience for games with keyboarding challenges!
1 Like
Right… I’d be up for it!
There was a real-time, kinda-arcadey parser game (but with graphics) featured on the front page of Itch for a while recently.
I also enjoy that some board/card games have an app component that isn’t gimmicky. I played a card game based on Werewolf which had an app that served as the “host” and timer so everyone could play, and it audibly announced when to close your eyes, who was to open their eyes and do what at certain times, and then kept track of the play time with updates “30 seconds left!”
A well-designed app takes the place of games that used to include electronics, and can reduce card and token clutter - and even provide more scenarios and categories in word and trivia games. I’m surprised if there doesn’t exist a version of Monopoly with QR codes printed on the board and deeds where the app is the Banker and you scan your property and “debit card” - with the option to use or not use paper money.
Do you have an unpopular opinion, hot take, or a view different from the norm when it comes to Interactive Fiction? Bring 'em on!
For decades the central promise of (parser) IF was exploration of virtual worlds via a natural language interface. But technology could not deliver on this promise and anybody who played an IF game expecting the parser to actually understand even the simplest of natural language commands was in for frustration and disappointment.
Now that natural language processing is perhaps the hottest and fastest-moving subfield of computer science and communicating with a computer in natural language actually works, the IF community has been surprisingly sluggish to embrace and innovate with it.
2 Likes
Yeah, if anything “Plundered Hearts is actually (one of) the best Infocom game(s)” is 21st century conventional wisdom.
7 Likes
I find that very interesting. Yes, natural language parsing is well-developed and nearly everywhere. For example there is a chatbot on my bank’s homepage. What technique is used for such chatbots? (Please note everybody: We are not talking about creating game content by AI or driving the world model by AI, instead about enhancing the parser.)
On thing might be: The system requirements for running a self-contained game + LLM are ridiculously high compared to running a traditional IF game, and who wants to depend on privacy-gobbling cloud services for a game?
2 Likes
Running a large language model locally might be more intensive then running zork, but it’s not a problem for a machine that can run games with graphics. (Running on the cloud is a bit of a red herring, so your privacy will remain ungobbled.)
The real holdup is that nobody has any idea what they’re doing. The field is unexplored territory, being mapped in real time by best guesses and helpless flailing.
Whether it’s natural language or something else, in an enthusiast community I think the question is usually “who has the time, resources, interest, and knowhow to do it.” I’m not sure the community would reject it philosophically (though it might, I don’t speak for anyone)
1 Like
Is that really true? As far as I can tell, the offline version of AI Dungeon still won’t run on my PC, which has no problem with recent games like Indiana Jones and the Great Circle. Even the official online version is pretty slow, sometimes taking several seconds to respond to input. More like playing Zork on a C-64 than on a modern computer.
1 Like
The trend is towards smaller specialised models and away from do-everything dinosaurs like ai dungeon.
(Or at least it was when I was messing around with the technology a year ago, which is a century ago in ai time.)
I guess in this situation I’m imagining tiny models that are specially trained to produce one game, rather than downloading 80% of the internet and telling it to only think about orcs. But I am only guessing and flailing.
4 Likes
The attempts I have seen to use large language models with parser IF, such as LampGPT, seem to be all about improvising new responses to player input, rather than going the more obvious route of only intervening when the standard parser fails to understand the input, translating it into something the parser understands.
1 Like
It doesn’t “actually work” for the sorts of stuff IF needs, however? At least in terms of “extract the details of natural language input and transform it into something relevant to the world model”. Using an LLM as a “parser” would seem to me to be a lot of work to provide an illusory convenience to players that would ultimately frustrate author and player alike. The closest you could get to useful, I imagine, would be having a model suggest alternative commands, perhaps.
5 Likes
it would be interesting to have a parser more easily able to understand stuff like “i need to pick up the lamp” or “open the letter and read what’s in it” or “hey bob tell me something actually useful” or something. but i don’t know if that’s worth the cost of AI.
7 Likes
I believe there’s been a handful of games that used Perplexity (Kidney Kwest and Baby on Board, in particular) to allow for natural language input. The general consensus seemed to be that as of 2022 the technology wasn’t there yet (or at least not sufficiently there yet to justify the long load times.) Is it now? I have no idea. It might be tech-wise but the loading times will need to be pretty low to get player buy-in.
7 Likes