Re: [dispatch] draft-ietf-dispatch-javascript-mjs-02: goal parameter

[Bradley Meck <[e-mail]>]

Sorry for the late reply, I am new to some of this process.

I'd refer https://github.com/bmeck/I-D/issues/1#issuecomment-322554737 ,
and the comments leading up to it.

The original intent of this draft was to introduce things related to JS
Modules and the syntactic goal related to them. The addition of the MIMEs
to reflect reality was done according to your comments in the issue
tracker. This draft could be split, but I would not personally submit one
with just the new MIME registrations as that was not the original intent of
this draft.

There is a problem of not wanting to implement things before being
standardized as Adam Roach points out. I don't think that reflecting the
reality of implementations is possible if implementations such as Node
won't ship the parameter without it being standardized. I also don't think
separating to match reality makes much sense for additional features like
new file extensions when it comes to file extensions as WHATWG explicitly
states to not use file extensions during MIME sniffing.

Per the changes related to JS Modules:

It seems that Allen and I think it is useful.

* We are both members of TC39, which is in charge of the JS language
specification and there is an open issue tracking this MIME update.
* That tracking issue does not have any disagreement with the current
changes.

I am also a Node.js contributor with heavy involvement in the upcoming
Module work, and am seeking to use MIME as a means to disambiguate the
intended interpretation of source texts.

Adam of IETF remarks that it makes sense to land it preemptively if it
would be useful later.

Per the comment of lack of interest from WHATWG:

* It seems Firefox is interested.
* Node has interest but no say in WHATWG, if it did there would have the
required 2 party interest if I understand things.
* No outstanding disagreement on the feature seems to be found on a quick
search. Seems like just a lack of interest, not a statement that the
parameter should not exist.

The lack of Node as a relevant figure in WHATWG I think shows how this MIME
can be useful outside of browsers and strengthens that lack of interest
from WHATWG does not mean lack of utility for other technologies.

I don't think a change to this update is needed, but am going to reach out
to these parties again and see if they would like to reply here as well.

Cheers,
Bradley


On Fri, May 18, 2018 at 10:03 AM, Anne van Kesteren <[e-mail]>
wrote:

> I'm not convinced adding a goal parameter at this point is useful
> since not all JavaScript hosts seem keen on adopting it. In
> particular, a change proposal to the HTML Standard has not gotten
> sufficient implementer interest:
> https://github.com/whatwg/html/pull/3205.
>
> I'd much prefer it if the registration was first updated to "match
> reality" and additional features are only added if there's actually
> sufficient implementers behind them. (Otherwise you have another
> Cookie2.)
>
>
> --
> https://annevankesteren.nl/
>
> _______________________________________________
> dispatch mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/dispatch
>