Flare alpha vs. beta
Flare’s engine is close to the feature-completeness line. The features we worked on for v0.18 affected flare-game in a few fundamental ways:
- Changed from 4 to 10 equipment slots
- Changed from highly random items to a smaller, more memorable item list
- Adding more armor slots means more absorb, so damage and health has to increase across the board
- With items no longer being mostly random, items need to be specifically placed on enemies, in loot containers, or on vendors.
Some of the core decisions about a game’s flow depend on these changes. Right now flare-game’s alpha demo is obsolete. It might be possible to rework those old maps and retrofit all the new data into them, but it would be a significant amount of work for that alpha demo’s final release.
Meanwhile the engine code for v0.18 is essentially ready. It could use more testing with real content/data but the major features are implemented. v0.18 is also the last scheduled Alpha Release. v0.19 will include the remaining features we want for Beta and for 1.0.
If v0.18 is the last alpha release, it does make sense if we update the alpha_demo one final time. Here’s what we’d need to do:
- Set new damage and absorb on all new items
- Set new damage and health on all creatures
- Place all new items in the game (on creatures, containers, or vendors)
- Update all containers to open-once logic
- Add item descriptions to many new items (if possible)
- Playtest a lot. All these new numbers won’t magically balance, they’ll need several iterations.
The alternative would be to just create new content, which would later be candidates for final release content. One reason this may make more sense is that the flow of the game content is much more important when it’s not driven by random items. As an example: we’d want early quests to reward items that are useful to low level players (magic swords/staffs/bows). We’d want to place interesting optional items in side quest areas. It’s easier to do that deliberately when creating maps instead of trying to retrofit them into the old maps.
Either way a full game release on the v0.18 engine is a ways away due to the amount of content work needed. Here’s one decision I’m not sure about: should I tag the engine as v0.18 so that we can keep working on the next features? I’d say yes, except there are specific expectations from downstream repos. Many flavors of linux are packaging flare-engine and flare-data separately. If we just update flare-engine it actually won’t work well with the old data. If we don’t really want the v0.18 code to be released yet, it makes sense if we hold off on tagging it.
What I may do is move the goalpost slightly for v0.18, so the code can continue while we work on release content. The main feature left for Flare 1.0 is cutscene support, and I think it’s something we can implement now. I’ve commissioned Flare artist Justin Nichol to work on some digital paintings that will serve as intro cutscenes for the Flare world. The cutscene system will likely also be used for displaying in-game credits.
So I have a lot of content to hammer out no matter what route I choose. Should I create the beginning of a new adventure, with maps designed for the new itemization? Or should I clean up the old demo and get something out a bit sooner, but perhaps messier?