B
Story
CSS-only infinite scrolling carousel animation
Have you tried this in a mobile browser? I’m seeing the headline cut off because the page ends up being wider than my viewport because of the position of the later carousel elements.

It otherwise looks great but the side scrolling is too much of a compromise for me.

I’m using Safari on iPhone.

Over a decade on from “mobile first” I’m always surprised at how few developers actually test their work on mobile devices.
It’s fairly common for these sorts of things to be iPhone-specific, and you can’t test on Apple devices without owning one, or paying altogether too much.

I’ve found Epiphany (a.k.a. GNU Web), which is based on WebKit, to be decent proxy for testing on Safari—in my last project of this sort, where I was developing on Firefox and could check in Chromium, I was able to reproduce each of the three or four Safari issues that were raised, in Epiphany.

It’s fairly common for these sorts of things to be iPhone-specific

Having done a bunch of mobile web work recently, I don't think that's actually true -- I hit weird issues on both Safari (on iOS) and Chrome (on Android). (And Firefox for that matter)

Safari does have its weird glitches but it's not inherently glitchier than other browsers.

There are some APIs it doesn't implement, but you don't need real hardware to catch that, MDN / caniuse has all the compatibility data and there are tools like ts-browser-compat for automated checks.

you can’t test on Apple devices without owning one, or paying altogether too much

Right, that's the real problem. But again I don't think it's Apple-specific. If you're a Mac user, in principle you could install Android Studio and test your app there, but a) that's a huge hassle, and b) an emulated phone on a desktop isn't close enough to a real phone to be useful for QA.

You really need to test on both iOS and Android hardware, there's no two ways about it.

I guess it would be good if you could install Safari on Android, and/or real Chrome on iOS. Then you'd only need to buy one phone. And that is on Apple, as they're the one blocking it in each case.

> It’s fairly common for these sorts of things to be iPhone-specific, and you can’t test on Apple devices without owning one, or paying altogether too much.

That's correct, but Apple phone users are customers with money used to paying.

> You can’t test on Apple devices without owning one, or paying altogether too much.

Yep! And it sucks, but this is the #1 reason why most web developers are using macs. If you want to test across all major platforms, it's the only way to do it from one machine (testing windows on a mac is a bit iffy on Apple Silicon macs, but it can be done).

>but this is the #1 reason why most web developers are using macs

Have you got a link to that? I'd be interested in seeing the stats and breakdown.

I don't have any stats, but this has been my professional experience in the UK. It's entirely possible that this doesn't hold up worldwide, or outside of specialist web development agencies.
  • Kiro
  • ·
  • 3 days ago
  • ·
  • [ - ]
The issues differ wildly between Android and iOS and often a fix for one device, OS and browser breaks something in another. It's not easy. Also, not everyone has access to an iPhone or a Mac with a simulator, and even the simulator has its own irregularities with how things are behaving that are not consistent with an actual device.
Which is a shame because one of the upsides of building mobile-first is that you generally end up with simpler/easier to reason about results.

(not a comment on this site implementation per se, but a general remark)

  • Hasu
  • ·
  • 3 days ago
  • ·
  • [ - ]
It works fine on both Chrome and Firefox on my Android (Pixel 8). Maybe the developer doesn't have an iPhone?
thank you for pointing out. it turns out that a `overflow: hidden` was missing. should be ok now.
  • Lorin
  • ·
  • 3 days ago
  • ·
  • [ - ]
I did this back in 2018 with full mobile support and nice edge fading: https://codepen.io/Lorin/pen/OEJGrM
It's the same with a sufficiently narrow browser window on my laptop.

Firefox: https://i.ibb.co/NW69hK2/carousel.png

  • ·
  • 3 days ago
  • ·
  • [ - ]
Same with Brave on android, text is breaking out and overflowing beyond the vp
  • kalev
  • ·
  • 3 days ago
  • ·
  • [ - ]
It’s <marquee> with extra steps, except for the pause which i’m not sure is possible
some how it's deprecated: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/ma...
Marquee, blink, i, b... Older people like me remember the old "innocent" days of the internet. It's probably nostalgia, but the modern web doesn't seem exciting anymore...
<marquee> quickly got considered a bad design as it prevents revealing the whole information at once and requires a lot of focus and patience. However here I guess in this form it can be used for presenting other products as an additional row
I think you could probably cause it to move on click by using the target selector, like in this old article here that is probably a bit out of date nowadays https://blog.logrocket.com/dynamism-and-css-calculations/

>But there is a pseudo-class in the location category that I don’t often see used: the target pseudo-class, which allows us to style a part of the page that is the currently active target. In other words, if we have a link that leads to a part of the page, we can style that part of the page when the link is clicked.

thank you for your suggestion. will explore the target selector to see what can be done.
duplicate items are not good for SEO.
A. Prove it.

B. Who cares? Anything in a carousel is content that doesn't matter.

It's also not "CSS only".
is it not? Because of the @keyframes rule?

https://developer.mozilla.org/en-US/docs/Web/CSS/@keyframes

I think they're saying because the technique requires adding a second copy of the content in HTML, it's not "CSS-only."

For the pedants, they could change the title to "No-JS infinite scrolling carousel animation."

  • ·
  • 3 days ago
  • ·
  • [ - ]