search

WICG / import-maps

How to control the behavior of JavaScript imports
https://html.spec.whatwg.org/multipage/webappapis.html#import-maps
Other
2.69k stars 70 forks source link

Remove advisements about future Fetch-based version #230

Closed domenic closed 3 years ago

domenic commented 3 years ago

I think we should set aside that idea, and close #169. It has serious problems with establishing the appropriate base URL, as discussed in previous issues. So we can remove these paragraphs from the spec and avoid distracting readers with the idea that the spec might dramatically change in the future.

This also removes the indication that we modify "fetch a module worker script graph". It's not clear how we'll integrate in the future, but it seems unlikely it would be via this exact mechanism, since that would flip "acquring import maps" to false immediately, before any modules are ever resolved.

Preview | Diff

hiroshige-g commented 3 years ago

I think we should set aside that idea, and close #169. It has serious problems with establishing the appropriate base URL, as discussed in previous issues. So we can remove these paragraphs from the spec and avoid distracting readers with the idea that the spec might dramatically change in the future.

Sounds good to me, but just to clarify the context, are there any changes in our mind or technical updates that contributed to this more complete deprecation of fetch-based import maps? Or do we remove this now given no updates for a while?

This also removes the indication that we modify "fetch a module worker script graph". It's not clear how we'll integrate in the future, but it seems unlikely it would be via this exact mechanism

Sounds good, totally deferring to https://github.com/WICG/import-maps/issues/2.

since that would flip "acquring import maps" to false immediately, before any modules are ever resolved.

Flipping "acquring import maps" to false before any modules are ever resolved sounds working as intended to me (acquring import maps should be false at the time of specifier resolution). Anyway I agree removing the section.

domenic commented 3 years ago

Sounds good to me, but just to clarify the context, are there any changes in our mind or technical updates that contributed to this more complete deprecation of fetch-based import maps?

Nothing recent, but I think the known issues around establishing an appropriate base URL are pretty serious technical problems. It's possible they're surmountable, but I can't think of how. So currently fetch-based import maps are in my mental category of "have serious problems and would need a lot of work" instead of "natural extension of specifier-based maps".

Flipping "acquring import maps" to false before any modules are ever resolved sounds working as intended to me (acquring import maps should be false at the time of specifier resolution).

Hmm, yeah, I think I miswrote this commit message. I think I meant "before the worker script is even fetched".