[webkit-dev] MathML layout refactor proposal

[dbarton at mathscribe.com]

2nd try - hopefully as raw text, not an html attachment this time:

Since I was the one who changed RenderMathMLBlock to derive from RenderFlexibleBox instead of RenderBlock back in 2012, I thought I should reply to the recent question on it and give my reasons for doing so.

1. Ideally, we want most webkit developers to not even have to think about MathML. If you make some change to the rendering and layout code and it doesn't break non-MathML layout features, including flexbox, then ideally that should almost always mean that MathML layout is still ok.

2. Conversely, it would be ideal if the MathML layout code didn't have to worry about a lot of non-math things - floats, columns, margin collapsing, internal RenderBlock voodoo, etc. We just want good fast MathML layout, that integrates well with HTML, CSS, and JavaScript.

3. MathML must introduce few or no bugs, and especially no security bugs.

Basically, I think deriving from RenderFlexibleBox helped with all 3 points. Besides helping directly with these points, I think it also allows the MathML implementation to be thousands of lines of code, instead of tens of thousands, which also helps with all 3 points. (I think that also will help with convincing other browser vendors to implement MathML, if and when they consider doing so.) At the end of 2012, the webkit MathML code was around 2,500 lines (admittedly it didn't implement all of MathML). In comparison, the MathPlayer plugin for Internet Explorer, MathML implementation inside Gecko, and the MathJax javascript library, were each about 20,000-40,000+ lines of code. Another comparison is that the webkit rendering/svg source code folder then contained about 17,700 lines. If we try to implement MathML from scratch, sort of like SVG, then I think it will take many developer-years to do, which frankly may never happen.

Gecko MathML doesn't use flexbox, but that's because flexbox didn't exist when Gecko's MathML code was orginally written. Another point is that webkit MathML's use of anonymous blocks predates its use of flexbox, and is somewhat independent of it. Finally, using flexbox (and also anonymous blocks) had nothing to do with Google's decision to turn off MathML, and eventually delete the code.

From http://www.w3.org/TR/css-flexbox-1/#overview we have:

Flex layout is superficially similar to block layout. It lacks many of the more complex text- or document-centric properties that can be used in block layout, such as floats and columns [and margin collapsing]. In return it gains simple and powerful tools for distributing space and aligning content in ways that webapps and complex web pages often need. The contents of a flex container:

can be laid out in any flow direction (leftwards, rightwards, downwards, or even upwards!)

can have their display order reversed or rearranged at the style layer (i.e., visual order can be independent of source and speech order)

can be laid out linearly along a single (main) axis or wrapped into multiple lines along a secondary (cross) axis

can “flex” their sizes to respond to the available space

can be aligned with respect to their container or each other

can be dynamically collapsed or uncollapsed along the main axis while preserving the container’s cross size

MathML may not absolutely need all these, but several (like changing the order for mroot) are very useful, and outlawing floats and columns and margin collapsing is great for avoiding bugs or unintended interactions with other code. So again by using flexbox, I think MathML can be thousands of lines instead of tens of thousands of lines, and will need a lot less new tricky lines to avoid buggy interactions with existing layout features and voodoo.

Finally, I'd like to say that I think it's great that you guys are asking for feedback on this. There are obviously webkit layout and rendering experts that know a lot more than I do, and I urge you very strongly to try to get one of them to help you with this design decision. Otherwise someone else in a few years may delete your code, as has happened to me. :)