[iamben]

We've had this discussion on HN before and I'll make the same point, because this is very similar with responsive websites as well. All too often I see designs that don't think about odd size tablets, or (as here) phones less than 375 or 414, or whatever else.

It absolutely boggles my mind, especially when you can attribute a "this is how much these users spend" amount to it (an ecom store, for instance). Most of the time it's not much extra effort to make it work properly (or even if you remove features, just keep it tidy), and absolutely worth it it real money terms.

[kawsper]

> It absolutely boggles my mind, especially when you can attribute a "this is how much these users spend" amount to it (an ecom store, for instance).

It also happens the other way around, they break it for odd-sized devices and then they go look at their numbers: "It's less than 1% of our revenue", WELL OF COURSE IT IS! It's broken!

[mgkimsal]

EXACTLY! And for some businesses, you may never get me back to try a redesign. That can also be used as justification - "those people won't come back again anyway".

This was some of the same 'thinking' back in the 'designed for netscape / IE' wars of the late 90s. I would routinely hear that "we get more sales from IE users". And me saying "it's because we didn't test on netscape and the form submission is broken" didn't seem to register with anyone.

I actually worked with one client who 'got it', and they were very particular that we tested with all major browsers of the day - they had multiple millions of dollars riding on orders coming in every month, and even small delays meant big ramifications down the line. It was almost all Netscape/IE, but we had to deal with multiple versions, and test accordingly, which wasn't trivial because they were trying to also add new JS functionality. We had a lot of user-agent testing at that time. :)

[iamben]

Yeah, see that loads, very frustrating. "Desktop CVR is double that of a tablet, so we don't bother with tablets..."

[skohan]

> Most of the time it's not much extra effort to make it work properly

I strongly disagree - with smaller devices you might be talking about an entirely different layout, and maybe a separate set of assets. More time for designers, more time for developers, and a whole category of tests for QA to verify during regression tests. You're talking about widening the entire development pipeline.

[enlyth]

Most people on HN think UIs should be autogenerated from backend code and that peak UI was Windows 95. Anyone worth their salt doing some real UI work knows how difficult it is to accommodate every arbitrary resolution and aspect ratio.

[iamben]

To reply to you both - I'm specifically talking about the web. I have no experience building apps, so can't comment there. But for the web, I'm absolutely happy to argue it's not that much more effort.

[ryandrake]

Whether it's the web or an "app," it should work sensibly at any reasonable resolution. It doesn't have to look good, or look like the pixel-perfect, happy-path Photoshop mocks the UX designer made, but it should function. Desktop developers have been doing at least an OK job of this for decades.

Yet here we are, where it's a struggle to find an app that even sensibly supports landscape mode, or supports a user-configured font size.

[skohan]

Arguably most of the examples given in the post "work sensibly". They have some aesthetic issues but they should function

[greggturkington]

Developers that think we should support the iPhone 3 just to be nice have not worked on enterprise-level apps. They underestimate the cost of the things you describe--design, QA, support, etc. It's more than a media query.

[furr]

X