Ivan Čurić
@baxuz.bsky.social
Your resident hopeless idealist.
They still haven't pulled the "protect the children" argument
November 30, 2025 at 12:23 AM
They still haven't pulled the "protect the children" argument
I did manual lazy components, modular js, ajax and server rendering, and hybrid routing in vanilla js back in 2016.
Shit was fast.
I'm never doing that again
Shit was fast.
I'm never doing that again
November 30, 2025 at 12:16 AM
I did manual lazy components, modular js, ajax and server rendering, and hybrid routing in vanilla js back in 2016.
Shit was fast.
I'm never doing that again
Shit was fast.
I'm never doing that again
People go for frameworks because it's a pain in the ass to do your own rendering and state.
Which is why I'm sticking with solid and lit.
Another thing is components. Practically every one needs components that are only now starting to get some native counterparts.
Which is why I'm sticking with solid and lit.
Another thing is components. Practically every one needs components that are only now starting to get some native counterparts.
November 30, 2025 at 12:12 AM
People go for frameworks because it's a pain in the ass to do your own rendering and state.
Which is why I'm sticking with solid and lit.
Another thing is components. Practically every one needs components that are only now starting to get some native counterparts.
Which is why I'm sticking with solid and lit.
Another thing is components. Practically every one needs components that are only now starting to get some native counterparts.
Very little to none. Which is why I try to at least enjoy the craft.
November 29, 2025 at 7:54 PM
Very little to none. Which is why I try to at least enjoy the craft.
"not unintentionally lethal" is an extremely low bar for deciding how and if something should be done.
November 29, 2025 at 5:52 PM
"not unintentionally lethal" is an extremely low bar for deciding how and if something should be done.
So if we consider this to be only a presentational feature, and go for the interactive approach, it would necessarily need to repaint on every update (expensive on huge images, especially HDR ones), whereas a matrix transform just needs to composite.
November 28, 2025 at 3:26 PM
So if we consider this to be only a presentational feature, and go for the interactive approach, it would necessarily need to repaint on every update (expensive on huge images, especially HDR ones), whereas a matrix transform just needs to composite.
object-view-box: xywh(x y w h) strictly overrides intrinsic sizing, so you might as well have put the w/h on a containing element, and the y / h on the transformed image.
HTML would still parse the full image size, as would dragging the image or using it as a canvas source since CSS doesn't apply
HTML would still parse the full image size, as would dragging the image or using it as a canvas source since CSS doesn't apply
November 28, 2025 at 3:26 PM
object-view-box: xywh(x y w h) strictly overrides intrinsic sizing, so you might as well have put the w/h on a containing element, and the y / h on the transformed image.
HTML would still parse the full image size, as would dragging the image or using it as a canvas source since CSS doesn't apply
HTML would still parse the full image size, as would dragging the image or using it as a canvas source since CSS doesn't apply
I see this as follows: having the size as a layout role is one thing, the viewbox is another, and an interactive viewbox a third.
Image size/AR can be controlled by the intrinsic sizing or by w/h attributes in html-only mode + whatever CSS layers on top, including its own w/h.
Image size/AR can be controlled by the intrinsic sizing or by w/h attributes in html-only mode + whatever CSS layers on top, including its own w/h.
November 28, 2025 at 3:26 PM
I see this as follows: having the size as a layout role is one thing, the viewbox is another, and an interactive viewbox a third.
Image size/AR can be controlled by the intrinsic sizing or by w/h attributes in html-only mode + whatever CSS layers on top, including its own w/h.
Image size/AR can be controlled by the intrinsic sizing or by w/h attributes in html-only mode + whatever CSS layers on top, including its own w/h.
What benefits does this have compared to implementing a `contain: strict` with a transform inside?
November 28, 2025 at 2:50 PM
What benefits does this have compared to implementing a `contain: strict` with a transform inside?
I love the "may include mistakes" part.
Imagine monetizing anything that needs a disclaimer like "may not work correctly, please do the job yourself afterwards"
Imagine monetizing anything that needs a disclaimer like "may not work correctly, please do the job yourself afterwards"
November 28, 2025 at 12:13 PM
I love the "may include mistakes" part.
Imagine monetizing anything that needs a disclaimer like "may not work correctly, please do the job yourself afterwards"
Imagine monetizing anything that needs a disclaimer like "may not work correctly, please do the job yourself afterwards"
Yeah, WebKit isn't really something to brag about.
November 26, 2025 at 9:14 PM
Yeah, WebKit isn't really something to brag about.
Does overscroll-behaviour: contain work on iOS Safari?
MDN says it does but I can't trust that it isn't subtly broken T_T
MDN says it does but I can't trust that it isn't subtly broken T_T
November 26, 2025 at 5:27 PM
Does overscroll-behaviour: contain work on iOS Safari?
MDN says it does but I can't trust that it isn't subtly broken T_T
MDN says it does but I can't trust that it isn't subtly broken T_T
A jebemu je i Tomašević seronja većeg kalibra. Nakon onog derneka dobije dozvolu za još jedan, o čemu mi pričamo.
November 25, 2025 at 9:40 PM
A jebemu je i Tomašević seronja većeg kalibra. Nakon onog derneka dobije dozvolu za još jedan, o čemu mi pričamo.
I mean, I didn't like the 2nd one nearly as much as the first game.
November 25, 2025 at 9:38 AM
I mean, I didn't like the 2nd one nearly as much as the first game.
