Re: Interop 2024

I plan on having this on the agenda for Thursday.

@Brian: is it at all helpful if we identify the problems caused by the lack
of interop? Or is a more positive "this is what it enables" tone better?

    Neil


On Wed, Sep 27, 2023 at 3:27 PM Deyan Ginev <deyan.ginev@gmail.com> wrote:

> Hi Brian,
>
> Yes, "interop" is a great mechanism. But browser vendors don't care what
> is "important to me" - they didn't in 2013, and they still don't today.
>
> The realistic question in my mind is what needs to be strategically
> accomplished/prepared by the WG for an interop request to succeed in
> September 2024 (or 2025, or 2026...).
> Securing funding and recruiting browser representatives to join the Math
> WG are indeed highly impactful (and high difficulty).
>
> To quote your own words (which I agree with):
> "We are very likely to face the same thing that we faced last year: that
> math is no one's priority."
>
> (minuted at
> https://github.com/w3c/mathml-core/issues/206#issuecomment-1736177722 )
>
> I would certainly support a group vote that selects a small MathML
> implementation issue, with very narrow technical scope, which we then
> collectively offer for "interop" consideration.
> I suspect it still won't be picked up, but at least we'll maximize our
> chances if we suggest something that looks really painless to fix.
>
> Deyan
>
>
> On Wed, Sep 27, 2023 at 4:41 PM Brian Kardell <bkardell@gmail.com> wrote:
>
>> Interop is the single best venue you have to make a case to all of the
>> vendors at once that something is important to you.  Not only that, but new
>> stuff + interop are taking priorities so getting something beyond those is
>> extra hard unless we find someone to fund the work - which, while we have
>> done it thanks to a few sources -  seems to have limits for math :)
>>
>>
>>
>> On Mon, Sep 18, 2023 at 10:51 AM Deyan Ginev <deyan.ginev@gmail.com>
>> wrote:
>>
>>> Hi Neil, all,
>>>
>>> Is the interop effort a good place for Math WG members to independently
>>> start filing new issues? I am a little hesitant, myself.
>>>
>>> Wouldn't it make more sense to first see if we have buy-in from the
>>> respective browser vendors (and meet any conditions to gain that)?
>>> Once we hear back a "soft yes" from the right vendors, we could file an
>>> interop issue to make things official.
>>> Experience seems to show that "cold outreach" requests don't move the
>>> needle too much in MathML browser land.
>>>
>>> I can certainly imagine making CSS support for MathML Core a public
>>> "implementation priority" for the Math WG, where we do enough liaison work
>>> to have backing for a small number of features to gain parity.
>>> At which point there may be a then-successful interop issue for 2025 (or
>>> 2026,...)
>>>
>>> Just thinking out loud,
>>> Deyan
>>>
>>>
>>> On Mon, Sep 18, 2023 at 1:23 AM Neil Soiffer <soiffer@alum.mit.edu>
>>> wrote:
>>>
>>>> To give a little context to Brian's message, the interop effort is an
>>>> effort to make browsers behave the same/have the same features so
>>>> developers can count on a feature working in all main browsers when they
>>>> use the feature.
>>>>
>>>> MathML core has some significant features missing from Webkit/Safari
>>>> and Gecko/Firefox. This means that you can't really use a number of
>>>> features in MathML core. For example, you can't use CSS with MathML in
>>>> Safari or Firefox. This is a major frustration for me as a MathML full
>>>> polyfill author because I can't do some of the polyfills without having to
>>>> target each browser separately. I know I've seen others complain about this
>>>> and other issues.
>>>>
>>>> This is your chance to make the case for why some of the top
>>>> implementers in the browser world should concentrate on some feature. As
>>>> Brian has said more than once, there are A LOT of things outside of math
>>>> that need attention. We need to make a little noise if we are ever going to
>>>> get some math features to rise to the level of even being considered. If
>>>> you have bumped your head into some cross-platform issue with MathML Core,
>>>> say something by filing a Focus Area Proposal issue. The odds of it getting
>>>> addressed are not high, but the odds are zero if you don't file an issue.
>>>>
>>>>     Neil
>>>>
>>>>
>>>> On Fri, Sep 15, 2023 at 11:11 PM Brian Kardell <bkardell@gmail.com>
>>>> wrote:
>>>>
>>>>> Is now open:
>>>>> https://github.com/web-platform-tests/interop/issues/new/choose
>>>>>
>>>>> The core group had requested i let them know when it was.
>>>>>
>>>>
>>
>> --
>> Brian Kardell :: @briankardell :: bkardell.com
>>
>