Make project roadmap

[pmer]

Driftwood should have a roadmap of features to be worked on over the next few months. Since the primary obstacle to getting Driftwood in use are features, it makes sense to focus on those for the roadmap.

The other big things that might optionally go on the roadmap could be:

• Asking people to try using Driftwood and recording & analyzing what they think/say/do
• Making a super simple game or two ourselves
• Working on recruiting additional developers
• Writing documentation for game makers

To go about doing this, we should clarify the goals for project Driftwood, determine the audience, and focus on them while composing this document.

[seisatsu]

Sounds good. Before we start promoting Driftwood, we should have at least the following things in the base engine:

• Pixel and roguelike movement modes a la Tsunagari.
• Mouse support (maybe.)
• Looping areas (also maybe, yes I'm obsessed with this, I run the fansite.)
• Support for virtual layers (you can place these above map layers for overlays. Currently we only support as many layers as are defined in the map.)
• Scriptable UIManager with text, button, menu, dialog box, and custom graphic support. This will involve some sdl_ttf and overall it shouldn't be difficult to code the base functions. Everything's just a graphic and an x,y,w,h.
• Support for scriptable graphics effects such as lighting, rumbling, and variable transparency.
• Make sure people can't easily encounter crash conditions.
• Make sure engine info messages, errors, and warnings are clear and concise.
• The engine is bugtested and all known bugs are fixed. We'll get further with promotion if we look somewhat polished when we begin.

We also want a standard library containing well documented functions for doing common tasks with the engine using existing API. Several forms of combat, such as on-map combat or final fantasy style combat could reasonably be implemented entirely in the standard library using API functions. Inventory and common menu systems could also be handled by the standard library.

Documentation is of course essential and should be worked on once we're reasonably close to solidifying the API.

Eventually (either during or immediately after beta,) we will want an editor suite. Modifying Tiled is still an option, but we might need to hire a QT professional to navigate the code. Otherwise extra tools could be made and standard Tiled could be left as an external dependency for making maps. Something which combines menu-driven JSON option setting, a script editor with custom Python highlighting, and world packaging would be nice.

[seisatsu]

We can use GitHub projects to make the roadmap.

[pmer]

Cool. Which big items have been tackled and which remain? We don't have issues for all of this good stuff.