Re: MathML Core authoring tools?

I personally only used the word "remove" for the particular case of 
"Mozilla Gecko/Firefox" and "Apple WebKit". As I said, I don't think 
it's really relevant to have them listed since the point of MathML Core 
is exactly to have native support in all browsers (which hopefully will 
soon be the case).

My main concern is that many of these tools are probably obsolete and 
didn't do any effort to align with MathML Core (or even MathML 4) which 
can be confusing for users. Typically, I'm thinking of tools generating 
MathML markup that is not supported cross-browsers. I'm fine with having 
a separate section for tools targeting MathML Core (again with some 
flexibility regarding what it means), so at least it can be easily used 
as a reference (e.g. from MDN).

I don't have strong opinions about what to do with legacy tools, but 
definitely the page in the current form (a long list of MathML <= 3 
tools) can be improved. The introductory paragraph should probably also 
be rewritten (reference to 'many browsers', MathML3 etc).

On 06/12/2022 20:04, Neil Soiffer wrote:
> I agree the list should be updated. And if the list is targeted to 
> MathML Core, then Deyan's suggestion of only including projects with 
> active development makes sense for MathML generators. However, there 
> is the big question of who has the time to do it.
>
> Two suggested changes:
> 1. Although some small number of people (~2%) still use IE, it seems 
> like a dead end so I'd be in favor of removing MathPlayer since it 
> only works with IE.
> 2. I would add mathlive to the list. It likely belongs in several 
> categories given that it also has a computation engine associated with it.
>
> I think it makes sense to add an "assistive technology" section 
> listing AT that uses MathML. Of course, getting a good list requires 
> time, and that's hard to get.
>
>     Neil
>
>
> On Tue, Dec 6, 2022 at 10:04 AM Deyan Ginev <deyan.ginev@gmail.com> wrote:
>
>     Hi Bruce, all,
>
>     Is there anything problematic with removing seemingly outdated
>     entries from that page? It leads with:
>     "This page by the MathWG keeps a non-exhaustive list", and maybe
>     it can also clarify that it bookkeeps "tools with active maintenance".
>
>     I think we can delegate to wikipedia for the "exhaustive"
>     encyclopedic bookkeeping, as long as we had a MathML historian (or
>     two, or three) who wanted to collect all of that metadata.
>
>     For some examples, the wiki page on "formula editor" has a table
>     of available systems, 36 (thirty six) of which claim to have
>     MathML support:
>     https://en.wikipedia.org/wiki/Formula_editor

>
>     Just that table spans 2+ vertical screens on my large monitor.
>
>     Similarly, there was a list I sent around with 16 mathml-to-latex
>     formula conversion tools, and I have another list with 10 actively
>     maintained latex-to-HTML5+MathML document converters. I think the
>     tally is even longer for the single formula latex-to-mathml tools,
>     which should easily be over 20 across programming languages. Oh,
>     and KaTeX is missing from the polyfills, which is unfortunate.
>
>     All of that said as to illustrate that with so many possible
>     entries it may be useful to strip away old projects with no
>     ongoing maintenance. To pick an example at random, the SnuggleTeX
>     release notes indicate the last release was in 2010:
>     https://www2.ph.ed.ac.uk/snuggletex/documentation/release-notes.html

>
>     So at the least I think an email to the authors is in order,
>     asking if they anticipate future work updating their project to
>     MathML Core. A lot of ground can be covered with ~50 emails, if
>     someone had a snowed-in Saturday afternoon.
>
>     Greetings,
>     Deyan
>

-- 
Frédéric Wang