- From: Neil Soiffer <soiffer@alum.mit.edu>
- Date: Thu, 5 Nov 2020 12:17:03 -0800
- To: public-mathml4@w3.org
- Message-ID: <CAESRWkDOnefe_+1cZq+hhw-qCP6VMg-2XHe_yeGNPcL1sFxB2Q@mail.gmail.com>
*The meeting was recorded: https://benetech.zoom.us/rec/share/Jd-szRHtTdxRYvd0ROt4rO6iSZyAj8R2qlPjMoY2SW5QdFCWpCEvNoepEKOgt8V4.ehxmoV9sZw1GJXBv <https://benetech.zoom.us/rec/share/Jd-szRHtTdxRYvd0ROt4rO6iSZyAj8R2qlPjMoY2SW5QdFCWpCEvNoepEKOgt8V4.ehxmoV9sZw1GJXBv> Passcode: Fo+T1$P? Starts about 11 minutes in.Attendees:Neil SoifferLouis MaherMoritz SchubotzBruce MillerDavid CarlisleDeyan GinevSteve NobleMurray SargentDavid FarmerThanks to Louis Maher for taking the minutes. It was a very lively discussion with lots of technical points about a screen share that Deyan made. So the minutes and recording may be harder to follow than usual.1. Move to WG -- making sure there is consensus to do this. Last chance to make comments on charter.Draft charter is at https://mathml-refresh.github.io/charter-drafts/math-2020.html <https://mathml-refresh.github.io/charter-drafts/math-2020.html>. PI: has editorial changes. NS: Consensus agreement on moving forward to form WGAction Item: NS will move the charter to the working group.NS: Hopefully when we are a WG, we will get more participation from the vendors.2. Continue discussion based on Deyan’s (live demo) ideas: https://dginev.github.io/tiny-mathml-a11y-demo/ <https://dginev.github.io/tiny-mathml-a11y-demo/>DG: The core formula gets buried in the general syntax.You want to reduce the markup.NS: Can you use a nested approach to simplify the syntax? Something where the table points to the rows and the rows concatenate the contents.DG: is concerned with where to break the line, and the types of meanings that are implied with the line-break (several top-level formulas OR e.g. single right-hand-side expression broken across several table rows, in the same index column - e.g. column 3 in the demo examples).DC: Do not use a table markup at all for top-level formula if RHS is single expression, just put the RHS alone in a table.DG: Note that real documents have expressions aligned so that they can stack vertically and horizontally (e.g. row of 12 columns, col 1-3 contains one expression, col 5-7 second, col 9-12 third. Each of those expressions can be line-broken on multiple rows).NS: Mark it up at the top level. MS: In OfficeMath with the math paragraph, we haveThis also gets the binary plus instead of a unary plus at the line break because the layout program knows that a manual line break is inserted.Moritz Did not understand how the equals sign was being shown. What is the difference between $op and $ number.DG: OP = operatorNot much difference between $op and $number, short-term implementation conveniences. They could all be numbers. Details to be decided later (mostly cosmetic).Moritz: What is the advantage of DG’s system and the standard implementation.DG: pmml+a11y avoids doubling the tree size.NS: Concerned that you have to keep two trees linked. NS: UUsing Intent you would only have to add small amount of syntax. Progressive refinement. If the Intent was missing, then the math would still be shown.Moritz: Inventing a new technology would have the same complications as the current systems.BM: You still have the problem of having the presentation and content mathML synchronized.Moritz: Will demonstrate a content MathML example next week.DC: With core, content MathML is not valid.NS: Believes content MathML is still valid. The parser will still accept it.Browsers do not do anything with content MathML.DG: There are more problems with content MathML than we have discussed w.r.t adoption.NS: SD: and MS: will do demonstrations next week.We will meet again next week.DG: Post-meeting note, here is a reference to my use of “sigil”:https://en.wikipedia.org/wiki/Sigil_(computer_programming) <https://en.wikipedia.org/wiki/Sigil_(computer_programming)>*
Received on Thursday, 5 November 2020 20:17:30 UTC