(Before reading your post, I had uploaded a new version with updated font.)
I now changed the "outside" to mean "West" there. Do you also mean adding "look" action to each exit to tell where they lead? (Changing the directions to location names would be a bit difficult.) Or that I tell that in the text box below?
I have now submitted the current version to MSXdev'20. If someone is interested, I wrote a blog post on the why and how. There would still be room for some more content in the ROM image, but the game's enough as it is.
(Of course, now I see I didn't tell well enough what happened in one part...)
"How much of the engine" would you like? Overall the scripts are definitely not mature (say, the Python preprocessing script doesn't check that the variable indices are within range, that all the referenced scripts/texts/... are present, some variables need to be adjusted by hand, verbs and directions are hardcoded into the Python script rather than the CFG, some hardcoded paths, ...).
I use only tniASM, so I don't know how much work it'd take to convert the original code + the generator to work with other compilers.
I'm thinking that the memory mapper support might be easiest done simply by forcing the game be split into episodes, and each episode taking one 16KB page with their own initialisation. The game I made takes at least 20KB in the data, but I didn't really try to optimize the number of tiles used towards the end.
The ROM on the MSXdev site was updated a while back (very minor changes -- no game-crashing bugs or such, just control instructions on the title screen, minor gfx tweaks and a scripting fix to the texts on the rocky beach).
I've read both of the blog posts and I do wonder why you're saying that you don't like the things that's in there.
There's a lot that I want to comment but let's start with a few points.
1. There's nothing wrong with empty rooms just because there's nothing there. Not every room needs to have a puzzle, a character, random events and many exits. Those rooms would actually be boring and confusing, too. Empty rooms have their uses and are fine.
2. MSXDev doesn't accept 128k ROMs, so you shouldn't probably be working with those if you were posting them there.
3. An adventure game with your kind of mechanics has been done in games like Knight Tyme. It's a great game, maybe you can play it too.
I always tell people that their first game doesn't need to be perfect, and this one doesn't need to be perfect, either. Your game is prefectly fine.
@Timmy: I had only heard the name of Knight Tyme, but had no idea what type of game it was, let alone that it had similarities with Jäästä. I'll need to look into it.
Yes, empty rooms etc. can work to give the sense of transition between places and pacing, something I felt Shadowgate missed. But if the memory limits mean choosing between a 15-minute game with such transitions and a 30-minute game without transitions, I'd choose the latter, especially since the game has no replayability. I'd rather cover the transitions with a text blurb like the ones when leaving the iceberg.
tfh already addressed the ROM size, but admittedly, next time the rules may change back to only smaller ROM sizes. But regardless, using memory mappers in a game would be my first time, and I think that's reason enough to try doing it.
I need to check if this is what "humble bragging" is. I hate promoting myself and feeling proud, so maybe blogs like those are a roundabout type of saying "look what I made".
