@andy OMG! YES!
I am sick of websites thinking I'm on a phone just because I have a vertical monitor.
https://shkspr.mobi/blog/2021/02/just-because-i-have-a-vertical-screen-doesnt-mean-im-on-a-phone/

@andy Excellent stuff, Andy, thanks :)

@andy @simon What a great resource, and wonderful website 👏👏

@andy
The 2nd most popular viewport's height is 666?! ⛧🤘

BTW, the text on the orange and purple blocks have contrast below 4.5.

@andy

Much needed work, thank you!

Only the other day, I encountered a popup which covered all the other words, wouldn't move, and had its "X to close" off the right hand edge of the screen.

In order to get to the X and get rid of it, I had to switch on auto-rotate and tip my phone on its side.

Vexing!

@andy beautiful website, and excellent domain name 😁

@andy I started offering different images depending on viewport size a week ago. I felt smart.

Ah and of *course* this is built with @eleventy

@andy Thanks for this. I'm generally not a CSS person but it's interesting to try to follow the design patterns to help provide maintainable structure. I've got some reading up to do on CUBE before I forward this along.

PS: On desktop Firefox(?) viewports the inline view of buildexcellentwebsites shrinks as the page is scrolled making it hard to read any of the content.

@andy SUPER useful, ty!

@andy you and your team have done an excellent job again, thanks for sharing 💪💪

@andy I seem to have put this on the front page of HN https://news.ycombinator.com/item?id=37208211

@andy Hey, this looks really interesting, and I agree with the takeaway. But I wonder if there's a missed opportunity to get more out of the data … e.g. if you hadn't treated each pixel difference as a separate viewport, maybe there would appear some clusters which could offer some guidance for more sensible default breakpoints? Unfortunately the "all the viewports" visualization doesn't really help there since the widths are not visualized really … am I missing something?

@andy This is *excellent*, Andy. Thank you.

@andy That's interesting. I visited the site in the timeframe with multiple configurations and the iPhone 13 mini still wasn't there once?

I wonder if an AdBlocker has prevented the transmission. 🤔

@andy This is so cool! Thanks for making it, love the design of the site too ☺️

@andy this was lovely (once again)

@andy Very neat - there's a small typo (repeated twice) - "a much difference picture".

@andy This is excellently done, thank you!

@andy Very cool article.

And the attached data finaly gave me something to procrastinate with.

I tried to visualize the view ports in relation to one another. (The screenshot is zoomed in and the pixels are not to scale)

But you can defiantly make out mobile phones.

@andy great article! However I still feel there are so many aspects left to explore to improve our design and testing workflows. Are there larger clusters of similar viewport sizes that we can pay extra attention to? What percentage of users use viewports within those clusters? If you plan to test specifically for those clusters, is it better to use the average size or the minimal size...