UXP Development Priorities

[dbsoft]

So I have a couple things still hanging in the air, some performance API fixes and changes and the Ion enhancements. However I was thinking maybe since the Ion changes aren't really necessary for functionality, before working on that full time I should tackle one (or more) of these other issues. I've kind of made a list of the issues I think should be prioritized going forward. Does anyone have thoughts on if I should try to take any of these down before getting lost in Ion? Or general thoughts on what should be prioritized in UXP?

Future priorities based on issues reported:

Implement min()/max()/clamp() CSS functions:
https://repo.palemoon.org/MoonchildProd ... ssues/2230

Support calc() in more places:
https://repo.palemoon.org/MoonchildProd ... ssues/1827
https://repo.palemoon.org/MoonchildProd ... ssues/1826
https://repo.palemoon.org/MoonchildProd ... ssues/1765

Fetch/Streams API:
https://repo.palemoon.org/MoonchildProd ... ssues/1442

Add support for <link rel=preload> meta (FranklinDM seems to be working on this already):
https://repo.palemoon.org/MoonchildProd ... ssues/2158

Implement CSS4 "revert" keyword:
https://repo.palemoon.org/MoonchildProd ... ssues/2045

ES2022:

Top-level await:
https://repo.palemoon.org/MoonchildProd ... ssues/2229

Regexp match indices (this doesn't seem to have an issue yet?):
https://developer.mozilla.org/en-US/doc ... hasIndices

[FranklinDM]

Implement min()/max()/clamp() CSS functions:
...

I was planning to work on this once I'm done removing Stylo from UXP, but had problems due to lack of free time, unfortunately.

[dbsoft]

I was planning to work on this once I'm done removing Stylo from UXP, but had problems due to lack of free time, unfortunately.

Oh I completely understand, but if you have eyes on that I'll save that for you!

[Moonchild]

We probably need to get some of those CSS things in soon. They will be more important for web compatibility now we've tackled a lot of the JS roadblocks in the last few releases (my unending thanks to everyone helping out with that!). I'd say the min/max/clamp and revert ones are probably highest priority in that area.

[RealityRipple]

Luckily most the even newer stuff being churned out is... let's say "not going to be implemented". Did you know someone wants to add full MIDI device support?

[Moonchild]

most the even newer stuff being churned out is... let's say "not going to be implemented"

Would be nice if they would stop piling on more and more to the already obscene web specifications. In fact, it would be nice if they would remove at least half of the web spec. In the current era of just about every website being made with front-end tools that can translate to whatever basic spec, it doesn't matter if all of the syntactic sugar is implemented or not hardly anyone will factually use it. I think we were in a good place with ES5, maybe ES6. CSS2/3, and sicking with the trifecta of html/css/js in balance.
But heh I guess we don't live in that universe anymore
Still... let sites be generated. Use compilers/transpilers. It's being done for all other languages, so why not for the web? offloading all that work on just-in-time processing in browsers is madness, IMHO.

Anyway /rant

[Mæstro]

Still... let sites be generated. Use compilers/transpilers. It's being done for all other languages, so why not for the web? offloading all that work on just-in-time processing in browsers is madness, IMHO.

Off-topic:
My favourite way of handling this has been to mirror troublesome sites in Archive Today: a JIT mess is converted at once into static HTML. Unfortunately, I cannot do this when the site requires a login or uses a form (though scriptless forms have existed for decades), and some of my VPN servers are blacklisted and I am too lazy to change server. Ghost Archive is similar, but inferior.