7DRL version postmortem / devlog

BRIEF POSTMORTEM / INTERMORTUM

I'm slowly working on a 1.0 release of the game, but I wanted to post my original 7DRL dev log and some brief thoughts.

I wasn't explicitly planning on participating in the 7DRL challenge this year, but in the weeks leading up to it I had an idea that I liked too much to ignore.  It was the same basic premise that a few other games (eg. Cultist Simulator, Stacklands) have explored:  What if you took the combining-two-elements aspect of Doodle God or Little Alchemy or whatever that AI-powered on is, but added some more traditional gameplay and narrative content to it.  I think when you try to do this, you wind up quickly realizing that the combinatorial exposion becomes unmanageable much sooner than it seems like it would.  Both CS and Stacklands added timers to card interactions, and only a fairly small fraction of combinations actually yield any results.  This is great, and allows them to have a long arc of content, but it wasn't what I wanted to do.  I wanted to make a game where:

• the only thing you do is combine one card with another, and
• every combination does something

I decided I'd write custom interaction text for each combo, because... well, that's just the kind of thing I do.  This allowed combinations to "do something" even if they don't yield a novel new card or a meaningful change to the game state -- you get a little story bit, or a joke, or an adjective added to one of the cards.  To further manage the number of possible combinations, I split the cards into:

• stat and object cards that the player can drag around, and
• dungeon/hazard cards that the player cannot

This made it so I didn't have to support dungeon cards being combined with other dungeon cards, and dramatically reduced the combo burden.  There were still 1400+ custom-written combos in the initial release (1900+ as I write this,) so it didn't reduce it to a reasonable number, but still.

I'm really, really happy with the way this game turned out, and I'm still having a really good time working on the next version of it.  It did pretty well in the Temple of the Roguelike rankings, and it's the second most-played of anything I've published here on itch, which confirms my earlier suspicions that I was onto something special.

THE LOG I KEPT DURING DEVELOPMENT

prior to jam

• added support to my basic game library code for dragging, dropping and combining cards
• wrote some backend database tools for handling the combinatorial explosion I have unwisely decided to create
• (sometimes all you get is a joke or an adjective added to one of the cards, but you at least get that.)
• did some brainstorming about basic mechanics, character classes, locations, interactions
• fully outlined/themed/named a reasonably-scoped (I hope) set of levels,

day 1a)

• very short, didn't start until late afternoon when the jam officially began, and spent a lot of the evening on an airplane
• adapted the card library test project to work the way the game needs to work
• wrote data export from the database and made sure I could easily paste it into a JS library file for the game to use
• made the full set of location cards and interactions, content complete except for icon selection, which was a hassle to do on the plane

day 1b)

• another very short day, at my in-laws' house, managed to work for (1.5ish hours) starting at 4am before everybody else got up.
• got the first tier of classes spindled and combos handled and written.  six explodes way more than four does.  this is gonna be a lot of work!

day 2)

• second day at in-laws' house brief period in morning
• added third tier of classes
• filled out the matrix of basic stat combos
• added HP/sanity system and readout, though no fail state (or even time passage system) yet
• continued backend tool improvement
• on the flight home:
• worked on filling in pile of missing combos for the initial basic set of player cards
• finished that!  Okay, so now adding a new dungeon monster or hazard or shrine is a pretty well-defined task, and won't take a ton of time.
• adding a new player card is going to be more work, and then it'll have to be filled in for every dungeon card
• these are manageable.  this is going to work.

day 3)

• early start day at office, locked in, let's do this.
• dungeon cards for "graves" type
• added "tool" player card
• added support for basic containers
• added time passage system, injury/horror, game-loss state
• added quality tier display to cards, WIP
• added support for doors unlocking new branches
• (these can be arbitrarily complex/custom, which I was worried wouldn't be the case)
• added a couple of new character unlocks
• added stat decks for every class
• (still need to do other perks)
• updated game flow, added treasure -> store -> restart loop
• added basic system for NPC stores in town
• beggar and statue done so far

day 4)

• woke up at 3am thinking about it, yes ha ha ha, yes.
• finished some missing class combo writing
• started on combos for forest cards
• finished the three monsters, spider web door type, mushroom ring
• did some visual polish to procrastinate those combos
• forest level gen and village unlock
• finished level, church unlock, most NPC stores in town

day 5)

• woke up not hung over and fairly enthusiastic, despite intense and mentally-taxing trivia experience last night
• I have to get most of the rest of the levels done today if I'm gonna have a day to polish.  Luckily I think most of the dungeon cards that I need are in place.
• finished saloon, upgrades
• improved stat card draw VFX, feels nicer now
• made darkness, approach, locked container dungeon cards
• made spectre and caster enemies, only eight cards left to make (apart from endgame)
• made barrow level type
• wolf enemy
• made debug unlock card for easier testing of stuff
• made banditcamp, dungeon, darkforest level types
• finished all of the dungeon cards on the list, orc, statue, trapped hallway, golem
• finished all second tier level gen

day 6)

• woke up with a pretty good idea of how the whole game progression works
• good ending = ring both bells, unlock everybody for easy defeat of dark lord as Demigod
• from TODO list: create last 3 player cards and combos, this is maybe a good early morning project
• from TODO list: figure out what this TODO was supposed to be?  third-tier locations is a good guess I bet
• from TODO list: publish playtest build on asymmetric, try to get at least a placeholder of the final fight 
• made magicweapon and charm and magicartifact player cards.
• finished battlefield (w/ soldier unlock), taintedforest, bell tech, necropolis
• everything else on the TODO list outside of polish.  and with like ten minutes to spare to go pick up my kid.  pshew.

day 7)

• fixed some bugs from playtest
• music system (Allegro by Emmit Fenn)
• tons of playtesting, fixing some very disruptive bugs in level gen
• fixed some infinite combos Kevin found
• added particle effects to title and interstitial game screens