Ask HN: What do you think about app native vs. portable look-and-feel?
12 comments
I think usability almost always suffers when the native UI isn’t used. It’s not that platform-neutral UIs aren’t usable, or good looking. However when native UIs were prevalent, there were standards built up over years of hard-earned experience: for a Windows app, you wanted to get your tab order right, and you knew the convention of getting the Cancel button mapped to the escape key etc etc. See also the Mac and the HIG, encouraging apps to look and work roughly the same.
There were always outliers and ugly UIs, but it always felt like there was a uniformity that made it easier to get around in an unfamiliar app. Whereas now, electron apps look and work very differently (comparing slack to Spotify to VSCode and so on).
That said, I think very few people care as much as I do about it, and cross-platform UIs save a ton of development work.
There were always outliers and ugly UIs, but it always felt like there was a uniformity that made it easier to get around in an unfamiliar app. Whereas now, electron apps look and work very differently (comparing slack to Spotify to VSCode and so on).
That said, I think very few people care as much as I do about it, and cross-platform UIs save a ton of development work.
Not only usability, but also raw capabilities. Most native toolkits handle UI and basic ux well and has good interoperability with system components. With most cross platform frameworks, the developer is often reimplementing these or trying to resolve platform incompatibilities (always badly). And the project quicly balloons in complexity.
They’re great to get started, and when what you need is already available. But they’re a pain for everything else.
They’re great to get started, and when what you need is already available. But they’re a pain for everything else.
Part of my curiosity was sparked by the article on HN the other day discussing the hamburger menu icon: it used to be confusing, but now is pretty widely recognized as long as some guidelines are followed. So while I used to agree with you re: usability, these days I'm not so sure anymore.
One idle thought I had: when computer interfaces were still new, using physical analogies like file cabinets was good practice for teaching new users. Maybe GUIs are now commonplace enough that people are able to speak the different "languages" without as much trouble.
One idle thought I had: when computer interfaces were still new, using physical analogies like file cabinets was good practice for teaching new users. Maybe GUIs are now commonplace enough that people are able to speak the different "languages" without as much trouble.
I totally get where you’re coming from but for me it’s the predictability that’s key. Hamburger menus are familiar but they’re also not always there. On a traditional native UI you’d always have a menu bar, you’d always have a File and Edit menu. These days you may have a hamburger menu but it also may just be a wrapper for a menu bar. It’s just all a bit……vague I suppose.
Tbh it doesn’t matter really. As I implied in the previous comment, I can’t imagine many people bemoan the fact they can’t press Alt-f to get an illustrative menu drop-down, even if I do.
Tbh it doesn’t matter really. As I implied in the previous comment, I can’t imagine many people bemoan the fact they can’t press Alt-f to get an illustrative menu drop-down, even if I do.
Personally I'm obsessed with efficiency and performance - and platform specific native toolkits are rarely beaten at that. I really don't like seeing wasting many megabytes of (V)RAM and long seconds of CPU time just to display a marquee...
True, but it goes both ways.
My bank's mobile app is 200 MB. The website is 10 MB, but most of that is wasted — it could realistically be built in less than a 1 MB. Start-up times and performance are similar.
Why should I lose/waste 200 MB for something that i have to be online to use anyway?
My bank's mobile app is 200 MB. The website is 10 MB, but most of that is wasted — it could realistically be built in less than a 1 MB. Start-up times and performance are similar.
Why should I lose/waste 200 MB for something that i have to be online to use anyway?
On one hand, yes, I agree: duplicating the full functionality of a website for the sake of having an app is far from being ideal.
On the other hand, I also believe that this 200 MB comes mainly from 2 sources: bad engineering from the application developer, and bad engineering from the platform developer. These are not toolkit dependent, however. Bad engineering will create bad products, regardless of the tools used, even when some tools are harder to misuse to create bad products.
On the other hand, I also believe that this 200 MB comes mainly from 2 sources: bad engineering from the application developer, and bad engineering from the platform developer. These are not toolkit dependent, however. Bad engineering will create bad products, regardless of the tools used, even when some tools are harder to misuse to create bad products.
I don't like the janky, laggy feel of multiplatform apps. I don't like it when I press back and the app quits instead of going to the previous screen. I don't like it when I go several pages deep into a SPA, like a comment, then press back a few times and find that I have not liked the comment until I refresh.
A good enough team can handle all of this with any platform but I think the issue is that people who respect the experience use native, and those who want to cut costs don't.
A good enough team can handle all of this with any platform but I think the issue is that people who respect the experience use native, and those who want to cut costs don't.
I hate non native apps (like Spotify, Slack, Teams, and to some extent browsers). That said, I think newer generations don't even know what native apps look like anymore. I can't think of any native desktop apps anymore except Office... and on mobile everything is different even when native.
They look and feel similar to the user, I still think native is better but React Native looks and feels pro so I'm cool with it.
If the app is useful, nobody cares.
If it isn’t useful, it doesn’t matter.
Sure some developers have strong opinions about this…and if you are selling to developers those opinions matter (most commonly, a job interview ).
But apps are gonna be shit for other reasons and the attraction of native-or-not is that fixing those reasons are much harder. Even when you choose to use theory to make GUI choice to add to complexity of development).
Or to put it another way, the simplest thing that might work means you stay working instead of thinking (thinking != working). Good luck.
If it isn’t useful, it doesn’t matter.
Sure some developers have strong opinions about this…and if you are selling to developers those opinions matter (most commonly, a job interview ).
But apps are gonna be shit for other reasons and the attraction of native-or-not is that fixing those reasons are much harder. Even when you choose to use theory to make GUI choice to add to complexity of development).
Or to put it another way, the simplest thing that might work means you stay working instead of thinking (thinking != working). Good luck.
I don't use a lot of apps. The ones I do are "not optional" (like my banking app etc.)
Frankly, I couldn't care less about the UI looks or whether it's native or whatever. I get in, perform a task, get out.
As long as I can do what I need to do, the buttons can be anything.
I think developers care deeply about this. I think if you're in a competitive space then it matters. But I don't think Joe Public cares in the slightest.
Performance yes. Usability yes. Ability to quickly do what I want, yes. Whether it uses native check boxes or owner-drawn check boxes? Nobody cares (for some definition of nobody. )
Put another way, Usability is important. Yes it should look "nice". But at this point I probably don't know what the 'native UI' looks like. Nor do I care.
Frankly, I couldn't care less about the UI looks or whether it's native or whatever. I get in, perform a task, get out.
As long as I can do what I need to do, the buttons can be anything.
I think developers care deeply about this. I think if you're in a competitive space then it matters. But I don't think Joe Public cares in the slightest.
Performance yes. Usability yes. Ability to quickly do what I want, yes. Whether it uses native check boxes or owner-drawn check boxes? Nobody cares (for some definition of nobody. )
Put another way, Usability is important. Yes it should look "nice". But at this point I probably don't know what the 'native UI' looks like. Nor do I care.
My own views on this have evolved in the last 10 or so years. I used to vastly prefer applications using the platform's native UI controls, but at some point recently I realized I no longer really care all that much, and I don't think it factors in to my purchase or usage decisions anymore. I was actually surprised when I realized this.
Curious of others' opinions or perspectives.