blitz VS taffy

Have you talked to the Dioxus people recently?

They're working on a project called Dioxus Blitz; from what I'm told, they're trying to implement a minimal browser target, that provides some basic DOM-like features and renders with wgpu.

It's not exactly what you're hoping for, but you might find common ground.

https://github.com/DioxusLabs/blitz

(Also, the Linebender project is working on Masonry, with FFI as a medium-term goal.)

Your proposal sounds very similar to what the Dioxus folks are doing on Blitz, which is using HTML/CSS as the languages but doing native rendering. WebF seems much further along, and I know they've been exploring other rendering engines, so it may be an interesting conversation for you to have with them.

Others have already mentioned inlyne and blitz, however given that you're interested in combining web UI with game/wgpu rendering you may be more interested in bevy_ui (and the higher level APIs like bevy_ui_dsl and belly that sit on top of this). Bevy UI is working on a HTML model and is specifically designed to integrate with Bevy's main game renderer and other infrastructure. It's worth noting that a 0.11 release of Bevy of due in next few days, and that UI is going to be a strong focus of the 0.12 release cycle: which should lead to significant development of bevy_ui, and either a doubling down on the "HTML-like UI" approach or a pivot to a new approach.

Here is an example: https://github.com/DioxusLabs/blitz/blob/master/examples/buttons.rs

> But no one is saying, "Hey I have a CSS 2.1 compliant rasterizer and compositor that you can use in your C++ or Rust environment!" are they?

There’s actually quite a lot of interesting work going on in that general space, has been in various forms for some years. A couple that immediately spring to mind:

• Azul <https://azul.rs/> builds on WebRender, as used in Firefox. I haven’t looked at it for a few years, but it looks to have grown quite interesting now.

• Blitz <https://github.com/DioxusLabs/blitz> is based on from-scratch implementations of CSS layout and rendering, and wgpu rendering. It’s not usable yet, but is a very interesting concept. If one happens to be familiar with React Native: it’s kinda like that, or React Native Web.

You may also be interested in https://github.com/DioxusLabs/blitz which aims to be a full HTML+CSS renderer on top of wgpu (but is currently not nearly as complete as inlyne).

If you are focused on performance we are working on a native renderer for Dioxus called Blitz. It is very incomplete right now, but in the future it will be much more performant than the current webview based approach.

There hasn't been too much progress on Blitz in the last few weeks (it will come), but there is now a new project Freya which is using a Dioxus frontend and rendering with Skia. That's currently using it's own layout system instead of Taffy though.

The main desktop renderer for Dioxus uses wry (the webview tauri uses internally), so the startup time will likely be similar. We are working on a native renderer called blitz which should improve this, but development is still early

I maintain a standalone web layout engine[0] (currently implementing Flexbox and CSS Grid) which has no scripting support. WPT layout tests using is a major blocker to us running WPT tests against our library. Yoga (used by React Native) is in a similar position.<p>Do you think the WPT would accept pull requests replacing such tests with equivalent tests that don't use <script> (perhaps using a build script to generate multiple tests instead - or simply writing out the tests longhand)?<p>I could run against only the ref-tests, but if I can't get full coverage then the WPT seems to provide little value over our own test suite.<p>[0]: <a href="https://github.com/DioxusLabs/taffy">https://github.com/DioxusLabs/taffy</a>

> Is there any easy to use/hack HTML layouting engine where I could experiment with custom CSS attributes and bridge that gap? Would anything from Servo be suitable?

Servo could be used for this. You'd want to add support for parsing the CSS properties themselves to the style crate in https://github.com/servo/stylo and then the layout implementation to the layout2020 crate in https://github.com/servo/servo. You do effectively get a whole browser though.

I'm currently working on building a lighter weight / hackable layout engine based on a combination of https://github.com/servo/stylo (for css parsing and selector resolution), https://github.com/DioxusLabs/taffy (for box-level layout) and https://github.com/pop-os/cosmic-text (for flow/inline layout). I expect to have something decent in around 6 months

Neither of these setups currently have any support for pagination though.

I maintain a web layout library that is designed to be integrated into other software:

https://github.com/DioxusLabs/taffy

It needs to be combined with a text layout engine (such as https://github.com/pop-os/cosmic-text), and it doesn't support everything yet (notable features that are currently missing: "float", "display: inline-block", "box-sizing: content-box", "position: static"). But we have Block, Flexbox and CSS Grid support with more on the way.

All of these projects have in common that they use Taffy (the project that I work on!) for box-level layout (which currently gives them block, flexbox, and grid layout) , and are either using or planning to use cosmic-text for text/inline layout. This gives you a decent first approximation of web layout, but it's not perfect and there are major features like float, display: inline-block, position: static, box-sizing: content-box missing. Not to mention that none of these implementations currently resolve CSS selectors, so you are effectively limited to inline styles (if you're interested in something in that direction then you may be interested in https://github.com/vizia/vizia).

While there are a lot of Rust UI frameworks, none of them are really recommended for production use yet. I suspect a few of the will die off and work will coalesce a few once things mature a bit.

Another nice feature of the Rust UI ecosystem is that lots of it is being built in a modular way. For example I maintain a layout engine [0] library which just does layout and can be easily integrated by anybody creating a UI library. And there a bunch of similar composable libraries covering rendering, text layout, accessibility, window creation, clipboard access, etc.

[0]: https://github.com/DioxusLabs/taffy

You might be interested in the combination of Taffy [0] which handles box-level browser layout (block, flexbox, grid, etc) and Cosmic Text [1] which handles text-level layout and basic text editing functionality.

Integrating them into browsers while retaining accessibility could be tricky. But in they're general they're relatively small standalone libraries implementing most of the layout algorithms that browsers implement (although there are currently a few key missing features like laying out "inline-block" items in line with text).

[0]: https://github.com/DioxusLabs/taffy

[1]: https://github.com/pop-os/cosmic-text

I maintain a library (https://github.com/DioxusLabs/taffy) that implements both Flexbox and CSS Grid, and is designed to be easily embedded (similar to Yoga, which Ink is using).

https://github.com/DioxusLabs/taffy (disclaimer: I work on this crate) which does CSS layout given CSS styles. This would probably be much more useful once we merge support for display: block (https://github.com/DioxusLabs/taffy/pull/474), and if in the future we support display: table. Taffy doesn't handle text layout but is designed to integrate nicely with external layout systems.

Ok, "1.4GB" made me look into this more. I hadn't realised that we were using a "superlinter" action that includes linters for over 10 languages. Switching to a different github action brought to time down to 3 seconds! https://github.com/DioxusLabs/taffy/pull/463