Out with the JavaScript, in with the HTML

recursivedoubts a day ago | parent | next [-]

the web platform is making slow and not-so-steady progress as a hypermedia system, but things do seem to be picking up a bit

we are working on a proposal to bring more general transclusion and a few other things here:

https://triptychproject.org

▲

zamadatix a day ago | parent | prev | next [-]

I feel like the option for simplicity lies between "web component" and "make 4 pages". Something near "the button changes the CSS variable controlling the size".

You lose out on pre-downscaled images but gain that the images look sharper for high DPI users and don't have to maintain the image sets or deliver multiple copies when the size changes.

▲

DLA 18 hours ago | parent | prev | next [-]

Love this approach and use it often, usually with a Go backend. Blazing fast, simple, respects the way the web is supposed to work. Really nice touch using the CSS transitions.

▲

paularmstrong a day ago | parent | prev [-]

This is a weirdly unpopular opinion here when it comes to HTML & JS, but there's a time and place for everything. This is a neat small example, but hardly worth the effort of changing something that was already working fine.

With the change, I now need another roundtrip network request to get new sizes of the same content on the current page that would have been able to be done in just a couple hundred bytes of JavaScript.

Edit: also there is still no view-transition support on Firefox.

▲

jakelazaroff a day ago | parent [-]

You would have needed the round-trip network request anyway to get the new images, no?

The lack of Firefox support isn't a big deal because this is a progressive enhancement. Firefox users will still be able to switch icon sizes; they just won't see the fancy transition.

	▲	zamadatix a day ago \| parent \| next [-]
		> You would have needed the round-trip network request anyway to get the new images, no? It'd be a shared round trip request for all images (so long as you aren't still using HTTP/1.1) in the 1st example vs a request for the immediate images and then a separate page load. Both have their upsides/downsides depending on the rest of the page and how users usually use it.
	▲	paularmstrong 18 hours ago \| parent \| prev [-]
		The images will always need a network request, but the non-JS version requires another request for the new HTML source. And with the previous implementation, all users would get the progressive enhancement, not just non-Firefox users.