[treyd]

I always thought it was bizarre that there's no serious selfhostable FOSS virtual tabletop software [1]. The overlap between the two communities is very large and passionate. Foundry is selfhostable but it's a clunky proprietary node.js application and there's a ton of janky UI issues. Roll20 is far worse. I presume that this Diceright project is better since it's actually using a canvas. But why doesn't someone just use a proper game engine that can target the web (with wasm/emscripten/etc)? You're making a game after all, and it would mean that people who are happy downloading and running a native application locally for better UX would be able to.

[1] edit: That I was aware of when writing this comment initially.

[tomlagier]

Nit: Foundry uses Pixi.js for rendering, so it's using more or less standard JS gamedev tools to draw on a WebGL context in a canvas.

Edit: See https://foundryvtt.wiki/en/development/guides/pixi

Edit2: For an example of why gamedev toolkits don't necessarily produce performant, highly usable software, check out Dungeondraft (https://dungeondraft.net/). It's built with Godot and gets the job done, but as an application it's a total mess. I'm working on an alternative but (surprise!) it's a challenge.

[Capricorn2481]

> I always thought it was bizarre that there's no serious selfhostable FOSS virtual tabletop software

> Foundry is selfhostable but it's a clunky proprietary node.js application and there's a ton of janky UI issues.

I don't think it's that bizarre when a genuinely good product gets this kinda reception. It's a thankless job

[Aeolun]

Foundry is amazing. But it’s only amazing compared to the competition. I’ve ran a campaign on there for the first time yesterday, and the UI is just…

[manbash]

Criticism for Foundry aside, it's an extordinary effort to implement a mature self-hosted alternative, especially as FOSS implies it's going to be a side project for most if not all of its developers.

Foundry's upside is that it's extensible (modules), which increases traction and nurtures a community.

Most of the FOSS attempts will probably die if they won't take that approach (extensibility that is).

[treyd]

Yeah that's how I see how it would work. Just build a framework for the players to interact in according to the basic mechanics of tabletop games and then expose a rich module API and hand it off to users. If you're clever about it I could imagine you could machine-translate the art asset formats made for Foundry, or design it to use it directly.

[delecti]

I think Owlbear Rodeo fits the bill, somewhat.

They recently released a more feature-rich cloud-based V2, but simultaneously released the source and self-hostable images for their V1.

Edit: I meant to add this initially but just forgot: https://blog.owlbear.rodeo/owlbear-rodeo-legacy-edition/

[bee_rider]

People who are into D&D and people who are into open source programming already both have time consuming, creative, indoor and sedentary hobbies. I think they mostly look for additional hobbies that scratch less-overlapping itches.