Add notes for WG-03-11 (#519)

* Add notes for WG-03-11

* Update 2020/WG-03-11.md

Co-Authored-By: Ms2ger <[email protected]>

Co-authored-by: Ms2ger <[email protected]>

Author: binji

2020/WG-03-11.md

@@ -48,4 +48,73 @@ None.
 
 ## Meeting Notes
 
-Added after meeting.
+### Introduction of attendees
+
+* Ben Smith
+* Ms2ger
+* Luke Wagner
+
+### POLL: [standardize phase 4 proposals](https://github.com/webassembly/proposals#phase-4---standardize-the-feature-wg)
+
+1. [Non-trapping float to int](https://github.com/WebAssembly/nontrapping-float-to-int-conversions)
+
+1. [Sign-extension operators](https://github.com/WebAssembly/sign-extension-ops)
+
+1. [Multi-value](https://github.com/WebAssembly/multi-value)
+
+We discussed how best to advance from phase 4 to phase 5. Because of the low
+attendance of this meeting, we decided to send out an email to the WG mailing
+list instead of polling the group.
+
+We discussed that we should probably make a set of PRs for each proposal before
+sending the email out, so the group can see what they'd be agreeing to merge,
+and if there are any merge issues. It also requires us to linearize the set of
+proposals, which may be desirable.
+
+AI(binji): Make a PR against the spec for each proposal.
+
+AI(binji): Send out email to WG asking for approval to merge PRs
+
+### Discussion: new spec document design (Ben Smith + Eric Prud'hommeaux)
+
+Ben presented to the group. The current core spec document is generated in a
+convoluted way (Restructured text, to single-page HTML, extracting MathJax and
+converting to Katex, python scripts, Bikeshed). This is not a maintainable
+solution. Ben Smith, Eric Prud'hommeaux, and Philippe Le Hégaret had a meeting
+to discuss improvements to this process a month ago. For the most part this
+process was done to ensure that the mathematics snippets in the document were
+renderable without JavaScript (which MathJax requires). However, since we can
+already produce a PDF from the spec document source, this would fit the
+requirement of having a non-JavaScript version of the document. We can then use
+the standard multi-page HTML w/ MathJax document as the main spec document.
+
+This would require some changes to both documents to include the required W3C
+layout and front matter, but would be significantly less work than we are doing
+currently.
+
+In general, the group seemed to think this was a good direction to go, though
+Ms2ger did mention that it was annoying that the document reflowed when the
+MathJax finally loaded. The group agreed. It also seems that perhaps the
+feature detection for MathJax is not quite right, since it doesn't detect that
+Firefox can render MathML, and says that MathML is not available if you try to
+choose it in the settings.
+
+Ben also mentioned that we were considering moving to using the W3C's new
+living document process. It would be a better fit for WebAssembly in many ways,
+since it doesn't require going through the entire REC process. However, as far
+as we know, no other group has adopted this process yet, so there was some
+hesitation from the group. We discussed that perhaps if someone from the W3 was
+able to help us through the process, it might make it easier for us to make
+this decision.
+
+AI(binji): Follow up on living document
+
+### Discussion: go offline for most meetings
+
+Because of low attendance of this meeting, we discussed having more decisions
+made asynchronously over email. We weren't exactly sure the correct email
+address for the WG mailing list.
+
+AI(binji): Find email for WG mailing list
+
+## Closure