WebUSB is dead - Hackdot

Re:WebUSB is dead by anderspitman on Monday February 03, 2020 @05:43PM (#22230172)

Everything is unrealistic until it becomes real. I naturally tend towards cynicism sometimes, but I try to practice dreaming a little. I find I'm happier when I try to imagine the world as it could be and what I can do to push it that direction.

True, general web layout is a hard problem. But there are a lot of knobs to turn. What if we removed CSS layout altogether, and only allowed linear pages, a la reader mode? That's a big tradeoff, but maybe it would work. If that's too much, maybe only allowing flexbox would be enough. As for editing, at that point you're making a web app.

[ Parent | on HN ]

Re:WebUSB is dead by roca on Monday February 03, 2020 @05:54PM (#22230280)

> As for editing, at that point you're making a web app.

Are you getting rid of Web forms now? Also, remember that at the very beginning Berners-Lee's WorldWideWeb browser supported HTML editing, so it's definitely part of some vision of the "good old days".

Yes, if you cut down or eliminate CSS, or just redesign it, you could build a faster renderer.

But these questions point to one of major defects of the "cut-down Web" vision: lots of people have a limited set of Web features that they like, but they tend not to agree on what that set is. You like flexbox. Some people really like grid. Some people really need ruby, or floats, or columns, or editing, or vertical breaking (pagination). Pretty soon you've got something nearly as complicated as real CSS.

[ Parent | on HN ]

Re:WebUSB is dead by jancsika on Monday February 03, 2020 @10:01PM (#22231941)

> But these questions point to one of major defects of the "cut-down Web" vision: lots of people have a limited set of Web features that they like, but they tend not to agree on what that set is.

Democracy to the rescue:

1. Start a FLOSS mailing list thread to get as much input as possible on the best starting point for this reduced set of functionality. Gopher, resurrecting Xanadu, revisiting Hypercard, and so on. Let the conversation flow.

2. Invite the people who chose Firefox Reader View to private gitlab instance and get to work.

[ Parent | on HN ]

Re:WebUSB is dead by anderspitman on Monday February 03, 2020 @07:03PM (#22230821)

I think getting rid of editability would be great. The whole point is a minimal browser. I don't really want to cut down the web. I want to start from scratch but in a manner compatible with our current JavaScript VMs.

[ Parent | on HN ]

Re:WebUSB is dead by imtringued on Tuesday February 04, 2020 @06:49AM (#22234533)

I agree but a lot of the pre-flexbox stuff in CSS is just plain crap. However, I don't think removing obsolete CSS constructs will make anything faster or better because you're not forced to use them.

[ Parent | on HN ]

Re:WebUSB is dead by roca on Monday February 03, 2020 @05:57PM (#22230298)

If you want to push the world in your direction, join a browser company. You may or may not be able to push them in your direction, but you'll at least understand the problem space a lot better to see where it's worth pushing.

[ Parent | on HN ]

Re:WebUSB is dead by anderspitman on Monday February 03, 2020 @07:02PM (#22230799)

You're talking about this from a value-proposition point of view. Whether something is worth doing doesn't depend only on the final value. It also depends on the amount of effort to do the thing. If the amount of effort is tiny, then it's probably best to do the thing and see what happens.

[ Parent | on HN ]