There are currently two mature @TC39 proposals for the low-level functionality that would make it possible to declare *any* file type as a dependency:
1. Import text (Stage 3 as of last week!): https://github.com/tc39/proposal-import-text
2. Import bytes (Stage 2.7): https://github.com/tc39/proposal-import-bytes
Perhaps we should push for these to get implemented *first* instead of ad hoc types for every possible type of dependency (CSS modules, HTML modules, etc), which doesn’t scale.
@westbrook When designing any product/API/language a common tension is:
1. Cover all use cases at 80% (low-level primitive)
vs
2. Cover only the common cases at 100% (high-level primitive).
The right call depends on the specifics, including:
1. What % of use cases would 2 leave out?
2. How painful are those use cases now?
If the answers are "high" and "very", 1 tends to be the right call, and buys you time to do 2 well, because now it's about reducing friction, not making things *possible*.
@westbrook Of course, here CSS modules have already shipped in 2/3 so it's low-hanging fruit to support them. I'm not suggesting doing Import Text *instead*, unless we figure out time travel meanwhile.
But we should definitely focus on these primitives before asking for any other types IMO or this situation will keep repeating.
@leaverou Thanks for clarifying! 😅 Because, I like the idea of "getting everything so right", but I am glad that that doesn't (in this case) superceed getting CSS import attributes into #webkit 🤞
I also like text, especially with so much text prompting going around, and agree it'll be a good next stop towards normalizing the super powers that import attributes provide at large!
Lea Verou, PhD
westbrook