Game Project Status Report

OK, thinking about projects time, what I've done with my 3-year summer vacation (extended, 2022 edition).

  • Scheme for local problems, sysadmin tasks, just general dorking around on the computer: 💯, best decision I've made in some time. In case it's not clear, Chez Scheme and Thunderchez.

    Schemers in general are annoying but less annoying than LISPers, so if the LISP community pissed you off, Scheme's might be 50% less toxic. I still have to block some people in IRC because they won't STFU or tolerate anyone Doing Things in Unapproved Manner. You know what would be amazing? A language as technically awesome as Scheme, with Python's friendly community. Python the language is trash, tho.

  • Haunted Dungeon, Scheme roguelike. Needs at most weeks of work, and then I can take a day and grind out binaries for various platforms (UUUUuuuugh Windows & Linux suck so much to interact with; Mac does for different reasons), ship it. It ballooned past my original tiny roguelike design long ago, but it's still not that big.

  • Multiple small Scheme programs & games, once I do that ship day I may just make a bunch of binaries. None of these are amazing but some are nice. Don't ask a developer to praise their own software, you know?

  • New Perilar CRPG, also Scheme. I have this at like 60% functionality, map generation's beautiful, and fuck-all for story, it's fine, same shipping problems, so much I need to think about if I get it to a playable state. Or it may be a learning experience.

  • Little Atari 8-bit game. Dungeon crawler with no purpose but it's cute, was meant to be a ZX SpecNext game but that's still not shipped after 2 years so… It's now at like 30% done, but I have some vision for it.

  • Open world action-adventure game I've got a bit of design for (in my paper sketchpad! Not even on the computer!), it could either be Scheme or Atari 8-bit or whatever. I don't actually know of anything like this design, tho original Zelda & Metroid are the parents (I clearly don't understand genetics) of all such games. Needs so much mapping & writing before I even start, but all the technical side is easy for me.

    Currently Atari stuff's in TurboBASIC-XL which is less bad than you'd think, but still really sucks compared to having a modern language; both Pascals and C's I have access to are less useful. I've been borderline to making a new language that compiles to 6502 ASM, but I know I'm lazy at tools-to-make-tools support.

    One nice part with Atari 8-bits is, shipping is easy. Put it in an ATR file, with an Atari emulator as seen on archive.org. One click, any browser shows it. Down side is, can't really charge money for this. Beg for patreon support; which I need to be better about giving you goodies in return for your cash, my fine patrons. Shall I write thee a sonnet?

  • Update & reupload my iPhone stuff. Should I even bother with Castles? I liked the underlying game but the UI is unbelievably shit, I had no idea what I was doing and limited by iPhone 1 screen/UI constraints. But Perilar, and some utility stuff, and maybe patch Brigand to be paid-up-front instead of IAP and say "this is what you could've had!". And I have my 3D game which never got shipped, just shown as demos. Worth spending some time on this and then never looking at it again.

  • Tabletop RPGs. Fuck this third goddamned plague year, which has made playing & playtesting RPGs just a nightmare. Every online group I try flakes out so fast they may be composed entirely of microscopic black holes. So I have my "original dungeon game" retroclone, and my much nicer sword & planet game, and a couple tiny gamelets. And with Hasbro's sabre-rattling at the OGL and "One D&D", I'm inclined to just ship only the sword & planet game, and turn the rest into world books for it. But I'd like to test it more than once with other Humans and also not get infected with plague. So. State of that world is uncertain.

Haunted Halloween

I've added a new seasonal game to my Mystic Dungeon: Haunted Halloween.

A text-mode twin-stick shooter (well, except it's an emulated Atari 800 "text mode", and the sticks are WASD and IJKL, I haven't written joystick support yet).

Five different levels:

  • Pumpkin Patch: Collect pumpkins for ammo.
  • Dark Forest: Find the path through.
  • Graveyard: Easy, just dodge gravestones and monsters.
  • Corn Maze: Unless you're Ted Forth you won't have a problem with this maze.
  • Haunted House: Just run thru the spooky house full of ghosts, and other monsters crawling in from the woods, get candy, get out!

Three difficulty levels:

  • Treat is turn-based, but there's so few monsters & candies you won't get a high score.
  • Trick is real-time, but you can generally outrun monsters.
  • Nightmare has twice as many monsters & candies, so it's the best way to get a high score… or die quickly, completely overwhelmed.

You can also play it like a stealth game, H hides you from non-adjacent monsters, so you can just run in, hide, wait for them to move off.

You collect candies for score (and banishing monsters earns candy), but every time you move to the next level you eat 10 candy to recover a hit and get some free pumpkins. So it's usually better to stay and clear all monsters, pick up all candies, then move on. But if you have a bunch of witches and ghosts, might be worth running away early.

The interesting thing from development is how little code is required for this kind of game. Halloween is under 1000 LOC, and that includes some long text blobs! Portal Worlds was 3000 LOC, Dungeon's over 3200 LOC and not even close to "done".

I'm still working on Public Caves, moving from BASIC to web-tech requires a lot more infrastructure!

MysticDungeon.club Random Thursday Update

Redesigned the front as a software gallery, got Portal Worlds working with my common input system, adapted Amazing (the dumb maze game).

I might get Heist adapted this weekend. Cityscape needs either a custom character set, or I add sprite graphics to the retro screen, which is a better plan. I have a bunch more JS games and demos, most can be adapted pretty quick. Porting the Mystic Dungeon RPG from Python is harder, but on the list.

Still thinking about the forum idea, I haven't seen a lot of interest yet, but a place for smack-talking would be nice.

It's really quite nice just having an easy way to focus on the game design or UI mechanics, and not have to make infrastructure from scratch every time.

Everything should be usable without an account, but you can't post scores unless you do!

MysticDungeon.club

I've finally got my web games/tech demo site MysticDungeon fully running SSL, a proper Node & database server, and all the existing games ported to my common "Learn2JS" framework. High scores and hit counters work for all of them; I haven't set up a really stable migration tool yet but that's on the TODO list before anything more serious gets stored there.

If you run into any bugs, let me know here or on fediverse.

Upcoming will be getting a couple features in PortalWorlds finished, then the rest of the Umbral Adventure world, and some more tools in Grimoire, which will be a tabletop RPG journal/toolkit, more for Referees to use as a virtual screen/notebook than as a coop gaming tool, but you could screen-share it if you needed to. Proper user accounts instead of an unverified screen name will be part of that.

I'm still thinking about if I should replace the old BBS with a forum, or what. Rebuilding the Mystic Dungeon game is on the list, that's part of what the Umbral Atari-like screen is for; nice ATASCII line-drawing characters instead of the few ANSI chars it supported.

Project Status

Made a lot of progress on Perilar Dark Weaver map generation. Hopefully this week I'll get ruins finished.

Spent a little too much time on the new Mystic Dungeon, the TTMS-76 virtual retro-console started on tilde.town but is now replacing the BBS. It now has scoreboards for the games! I still need to finish Heist, which turned out to be bigger than I first thought. In a bit I'll get yet another damned login system done, and from there a forum. I wonder if anyone else would be interested in making games for it? I've got a trivially easy framework, so if you know any Javascript at all it's fun to work with. All BSD licensed.

Updated StupidComments.css to block some more inline "affiliate" blocks and Youtube spam segments.

I started making a console Pomodoro timer, and it works, but needs persistence and a teeny bit of task management before I can release it. Very soon.

RPG-wise, I wrote a bit more of my "survival D&D" game Delvers in Darkness (aka Dungeon Hell), which is looking to come in well under 16 pages for a full Holmes-type dungeon game; maybe 32 if I write more on the setting, which since I complain about that in everyone else's games, I should. Haven't looked at my light game in a bit, and don't know when I'll get back to that.

Perilar: Dark Weaver

Perilar was my favorite of my games, an original iPhone RPG (loosely inspired by Rogue, Ultima, & Zelda). Alas, the App Store is a pain in the ass to stay current in, and Perilar needed updates I wasn't willing to jump thru Apple's hoops to deliver, so it's been gone from the store for a few years. Does anyone even remember me?

Then I wrote & released Brigand, a stripped-down realtime roguelike, which apparently everyone got mad at either because it was too hard (it was, but I liked that), or cost $9.99 which is SO MUCH MONEY on the goddamned App Store. And again the App Store made it obsolete and it wasn't reasonable to update it.

So I got back to a new desktop version of Perilar. I wrote a JS prototype (Fallen Kingdom) that wasn't fast enough to be usable, but let me rapidly test a bunch of new things. Now in Scheme, after a very long time, I have a nice, fast, nearing-complete sequel to Perilar: Dark Weaver.

Where I am right now:

  • World has both hand-designed and random sectors, tho the latter need some smoothing out to look like real terrain. I can walk across the world, at least until I hit impassable sectors. Porting my complex map and dungeon generators in is the next task. I have spent the last 6 months building features in the Town of Amity, and I'm ready to move on!
  • Most of the tile art is from the DawnLike set, with a lot of customization and new art where needed, and I've stuck to the DawnLike palette, it really has a nice old-timey look, a little less garish than the Atari 800, brighter than C64.
  • Player art isn't going to be these sprites, but the paperdolls I have are 2-facing (East/West), and I'd prefer 4- or 8-facing (you can move diagonally!); so I still need to find or draw (oh no) those.
  • NPCs have dialogue trees, stores, and special abilities (like the Innkeeper healing you; they're not super-powered).
  • Combat, with multiple attack/defense options, works in my test area. I haven't spread monsters around the sectors yet, but they've been developed and tested in the JS prototype.
  • Loot is extensive, magical weapons and armor have all the complex options you'd expect. I'm being a hardass on encumbrance in this one, because you can drop loot anywhere and come back for it. (Not quite the hardest possible ass; gold doesn't count towards weight, which it does in tabletop RPGs!)
  • Spells beyond Magic Missile are not implemented at all yet; will probably ship with only the dozen basic spells from the original release, and advanced spells added in an update. You won't find anyone to teach those for a long time anyway. Despite that, Wizards are still useful with magic wands.
  • New bosses, boss arenas, deeper dungeons, main quest, and sidequests.
  • At least one sector will be user-modifiable, tho I don't know if it'll be in the first release. You can buy furniture and walls, and fix up your own town. There's useful things you'll be able to get from that. (The building mechanic half works now; gathering doesn't).
  • Currently tested on Mac, should be buildable with no or very few changes to Windows, Linux, BSD, etc., but I need to get proper test environments for all of those.
  • Will be for sale on itch.io sometime this year. Price TBD.

I feel super awkward about self-promotion, but I do have a Patreon, and for Gold level you'll get betas; I haven't explained this, but at any level, when you've paid up whatever the cover price of the game ends up being, you'll get a full release license for it, too.

darkweaver-2020-01-08-2
darkweaver-2020-01-08-3

TODO App

I'm yet again frustrated by the state of checklist TODO apps. I had a perfect one once, ToDo on the Palm Pilot; aesthetically it was of its time, but for usability on a stylus-based PDA it was perfect. Everything since then has been a compromise.

Must:

  1. Run locally on iPhone and Mac.
  2. Sync automagically, preferably with iCloud and/or Dropbox.
  3. Have multiple lists, though does not need nested folders.
  4. Show a list of items I can check on/off.
  5. Be able to show or hide checked items, delete checked items with a command.
  6. Be able to reorder by hand (or sort with a command).
  7. Start up instantly right back where I was.
  8. Be usable while drunk, stoned, tired, or hungry. UI cannot be a giant pile of fiddly little switches.

Nice to Have:

  1. Due dates/expire dates.
  2. Priority tags.
  3. Search screen to show "What's next?", showing date, priority, list, item in that order.
  4. Notifications. But local notifications require a recent app launch, so you might miss stuff; or interacts with calendar which many people find annoying; or uses push notifications, which costs real money past a small number of users.
  5. Emoji & color tags, photos, long notes, etc.

Must Not:

  1. Give my information to Google, Microsoft, Facebook, or other evil mega-corporation who will weaponize my shopping for killer drones or advertising, or both. I'm dubious of Apple & Amazon, but at worst they seem to be venal, not evil.
  2. Be subscription-based.

Don't Care:

  1. Other platforms. Android, Linux, Windows, I don't use these so they're not a pain point for me. If mulle-objc works out, I could think about that.
  2. Pomodoro, GTD, and other fetishistic rituals.

State of the Field

  • Clear was good on iOS, but it's been "dead"/rebooting at Impending for 2 years already, and they won't be making a new Mac version.
  • Apple's Reminders is awful on the Mac (try finding the "show all" button, and keeping your list in order).
  • Things, Trello, OmniFocus, etc. are too slow & heavy for a grocery list.
  • Text files in Editorial or Drafts don't have usable (finger-sized) checkboxes.
  • Wunderlist got bought by Microsoft and destroyed as per usual. They're getting to be as bad a product graveyard as Yahoo! was.
  • RememberTheMilk is not super useful until you pay $40/year, which is not what a dumb todo list should cost.

What Am I Doing?

Thinking about this. I could write the app I want, release it for $5 on each platform, use Marzipan to make it use the same codebase. In theory, this is pretty easy; I can write a database-backed table view thing in a week. Making it nice is a while longer. Marketing is my least favorite thing, who do I use to get it in front of millions of people?

But I'd have to pay Apple to get back into the sharecropping business, and deal with their shitty Xcode tools and the rotting corpse of developer.apple.com which got fucked over by Apple marketing so you can't get to the FUCKING DOCUMENTATION. @invalidname went to work for them and got sucked into the black hole.

LOWREZJAM 2017

Signed up for LOWREZJAM 2017

I've just started drawing some "art", adapting an old pixel font nerds of a certain age will recognize. Finally, something my 1980s pixel art skills are good for!

Figure I'll make a metroidvania dungeon game; I don't mind action, I'm just not good at dealing with real-time animation of giant art sets. This'll make a good test of my Electron framework, before I ship something big in it.

Update: Realized there's an old game I loved to match the art that I'd like to do (an update/homage, not a straight ripoff), and it's not a metroidvania. Making a ton of level maps is beyond my current spare time, and procedural metroidvania's a hard research problem.

Premium Subscription ★☆☆☆☆

Day One Goes Premium Subscription
and of course the mob outrage in App Store ratings is what you'd expect: MacDrifter.

And this is why I only do bare minimum maintenance of my App Store software now. I released Brigand as free with a $10 unlock, and got savaged for it, so I pulled it. If Nintendo can't make that work with Mario, Apple giving them the front page, and millions in advertising, I sure can't. I love Brigand, but unless I put in more work changing the business model, I can't sell it; sunk cost fallacy tells me not to do that.

Productivity software should cost more than a game, but very few on iOS are willing to pay up front every single new version.

Apple doesn't let you give old customers an upgrade price, and presumably never will; maybe an upgrade killed Phil Schiller's pet/child/Camaro in front of him, or something, given the 9 years he's heard developers request this feature and told us to pound sand. And Apple does nobody any favors by Sherlocking and undercutting developers with "free" or cheap productivity apps.

The older solution of releasing a new numbered version and abandoning the old one every year or so was completely user-hostile. I just refused to do it, and would always switch apps whenever someone tried, and often found a better app by doing this.

Maybe the subscription model is terrible, but it's less terrible than anything else going on.

Michael Tsai wonders if the hostile reviews are from prices going up, but they're just catching up to desktop/web service prices, usually because a subscription gets you cross-platform access now.

Long-term, I think the App Store will be seen as the worst-managed disaster in the history of software. It went from a nice slot machine for indie devs and gallery for a few professional companies, to a predatory flea market full of thieves and frauds. Trying to tell anyone you make real software and here's a reasonable price, in that environment, is a waste of time.