Mobile Last?

https://jonathanstark.com/blog/mobile-last

by tacon • 12 years ago

44 46 12 years ago

I just want webview-based mobile apps to support basic platform navigation conventions like scrolling to top when the top of the screen is tapped. That one simple gesture immediately outs webview apps 90% of the time. I'll stop preferring native apps when webapps start offering a benefit to me, the user, instead of just the developer.

BenSS• 12 years ago

The other one on iOS is edge-dragging to go back/forward in the navigation stack. Neither of these things is -hard- to implement with current tech either!

softdev122• 12 years ago

the benefit to the user comes because the developer can support more platforms. imagine you are an ios user, but the developer only has the resources to build for android. with phonegap/cordova, you as an ios user would benefit by having access to the app (rather than not having it all) if a cross platform solution is used by the developer.

izolate2• 12 years ago

Desktop is dead? I'm not surprised a "mobile software consultant" would say something so ridiculous.

jonathanstark10• 12 years ago

Hi! I'm the post author. Desktop is dead in the way TV and newspapers are dead. Yes, they're still around but there's no growth. They certainly shouldn't be the main focus for a web developer building a personal site from scratch.

Delmania• 12 years ago

Really? Considering Apple sold 5.5 million Macs last quarter, I'm not sure your claim that there's no growth holds water. If anything, the trend seems to be people using laptops (which are desktops) and smartphones, while tablets are stagnating. The advice to not focus on the desktop is just bad. You need to build for both.

izolate2• 12 years ago

Maybe not for a personal site, but for an web application, definitely. I mean, just look at your own site. It was made "mobile first" and as a result, looks like a blown-up mobile site on desktop. Now imagine how horrid a fully-fledged web application would look if it followed similar design principles.

api3• 12 years ago

I think that's a poor analogy. Desktop isn't dead at all. It's very much live in the way farming and electricity generation are alive -- very important and critical but not growing rapidly. It's a "mature" market.

Part of why "desktop is dead" is a big meme in the valley comes from the fact that the valley is dominated by big-money venture capital. To VC, anything not growing at double digit rates simply does not exist.

BTW... I do wonder if we'll see another leg of desktop growth at some point in the future as the developing world becomes more developed and starts wanting to do more of the things people do on desktops...? We might indeed be in an era when "mobile first" describes the order of computing device adoption by a culture.

api3• 12 years ago

I visited a place with a "desktop is dead" / "mobile is the next everything" message a while back. Couldn't help but notice all the desktops. Nobody was doing their work on a mobile device.

The boat people are totally missing is this: sure, mobile is growing like crazy, but what we're seeing isn't the death of PC. What we're seeing is the filling out and explosive diversification of the computing ecosystem. The future isn't a "next platform," it's a whole ecosystem of devices serving specialized functions or function domains. For PC that's "personal computing" -- doing "big" stuff and "real work." For mobile that's mostly communications along with navigation, etc., the sort of thing you want an ultraportable micro-communicator for.

I own a tablet, a smart phone, and a laptop, and of the three the tablet is probably the one I could do without. The smart phone and the laptop are used constantly. But the tablet is nice for reading and note taking.

Joeri• 12 years ago

Web browsing outside of tech circles us predominately a phone-based activity or on the verge of becoming one (it depends on whether you consider app interaction like facebook to be a web browsing style activity or not). The point of the article is that you have to build for your market, and for most web designers that means not just mobile, but phone.

jonathanstark10• 12 years ago

Yes, there are certain work products (code, for example) that still can only reasonably be done on desktop/laptop computers. But the list of work products that require a desktop/laptop are dwindling, not growing.

Printers are a good analogy. Look around most any office and note that the number of printers has drastically decreased. This is a sign of work products changing - i.e., people don't need to print out as much stuff on paper to get our jobs done. The same is happening with desktop/laptops now. It's industry specific currently but the trend is clear.

gphil• 12 years ago

This is exactly my feeling as well. I would take a huge productivity hit without a phone, or without a laptop. I would mostly miss my tablet on plane rides. I would love to see more data on this though, because I wonder if most non-HNers feel the same way.

ErikRogneby• 12 years ago

I wonder if the desktop first attitude at the Rails Rumble had something to do with when Rails came on the scene? That when people started using Rails desktop was king and mobile was painful. (wml anyone?)

I would bet that a node.js or a MEAN stack hackathon would have a higher percentage of mobile friendly results.

jonathanstark10• 12 years ago

I too wondered that while writing the post, but I can't think of anything about Rails that inherently makes the framework desktop-first.

kevinSuttle• 12 years ago

Had a similar sentiment earlier this year.

http://kevinsuttle.com/posts/the-right-questions-to-ask-mobi...

PaulHoule• 12 years ago

In a short time frame I'd hate to commit to cross device testing.

That said I wonder what the root cause of the incompatibility. Are these people using a fashionable client side framework that doesn't work on mobile? Or is there a problen with ror?

I can say lately I have made sites using java on the back end and very simple front ends (as little is as possible, very simple html, CSS for formatting and without trying at all these work great on tablets and phones.

gabemart• 12 years ago

I don't think you actually need to do cross-device testing to qualify as "mobile-friendly" in the context of a 48 hour hackathon. You just need to put together a view @media rules such that your UI is usable on all sizes of browser window and insert a meta tag to indicate to mobile browsers they shouldn't try to emulate higher resolutions.

It's impossible to say how hard this would be without discussing a specific web app. But if the benchmark is simply "usable" rather than "elegant", I don't imagine it would be very hard for most web applications.

ch82302• 12 years ago

Why was 'mobile last' a takeaway from a hackathon? Most dev's probably focused their efforts on the breakpoint they would demo. It is possible that are indeed doing 'mobile last' but I wouldn't conclude that from a survey of hackathon projects.

jonathanstark10• 12 years ago

Good point about "the screensize they would demo" but in the case of Rails Rumble, the contestants don't demo their work. The judges browse through them on their own on whatever device they like.

ch82302• 12 years ago

thanks for clarifying that does open up the need for responsiveness.

softdev122• 12 years ago

First comment, those competitions shouldn't be entered only to "win". the goal should be to build something useful not to have a few judges pick the idea as a "win". The percent of successful companies that come from these types of competitions is absurdly small.

Second comment,Cordova/Phonegap make it really easy to build mobile friendly apps that can be extended to the desktop web, all in pretty much 1 codebase.

PSeitz• 12 years ago

I always build mobile-first websites ... and they suck on desktop because they have vastly different screen sizes. Proper multi device support ist not cheap or free.

forrestthewoods• 12 years ago

Totally insane to call mobile free. No, if you spent a whole bunch of time testing on a bunch of platforms and a whole bunch more time writing things very carefully to work on a bunch of platforms then you can totally make a great experience for all of them.

I mean I get it. I also think it's insane to not support mobile. Everyone should be supporting mobile. Let's just not lie and say it's easy. It's a bunch of work that everyone should be doing. But it's still work.

hipsterrific3• 12 years ago

People often forget that testing cross platform functionality is difficult and time consuming. It can easily tie down designing and add more constraints than necessary. This is why I advocate focusing on essentials like usability, experience, and functionality rather than some gung-ho attitude of "we MUST be mobile first regardless."

jonathanstark10• 12 years ago

I'm not saying everyone must ALWAYS design mobile first. I'm saying it should be our default position, like doing layout with css instead of tables.

If you have some reason to ignore mobile (and there are many good ones) then go ahead and make an exception.

My issue with desktop-first in RR is that because of the nature of the situation, there were no good reasons to make that exception.

jonathanstark10• 12 years ago

It doesn't have to be hard. Some visual designs can be hard to do cross device but that's conscious design choice. The implementation is not that hard if you start small and work your way up.

derengel• 12 years ago

When developing a web application for phone, tablet and desktop, is it a good principle to use the same HTML for the three and a separate CSS for each device?

hawleyal• 12 years ago

I disagree. Of course designing mobile-first is considerably easier than trying to do it later. However, I would also say that designing mobile-first is actually easier in general, even if you never plan to target mobile specifically.

The idea is that you are designing for responsiveness. You are removing the worry about what should be on the left or the right or how wide things should be. You are focusing on what things are related, the relative importance of different things, and the content/interaction.

al2o3cr• 12 years ago

Meh. I looked at the winners - many of them wouldn't make a damn bit of sense as "mobile apps" or even mobile websites.

http://railsrumble.com/entries/winners

jonathanstark10• 12 years ago

I strongly disagree.

hipsterrific3• 12 years ago

I personally make decisions based on what the project requires rather than some template. Most websites for me are informational, so mobile-first works well. Not a whole lot of functionality to test so we can focus on building the website and make it look on any size device.

However, apps are a different ball game. I'd rather focus on usability, experience, and functionality. I'm not going to try and hobble the app by trying to beat a round peg into a triangle, square, whatever shaped hole. I find that most web applications which have large swathes of functionality benefit from native code, so I usually skip mobile-first design and just do a native app (Xamarin or Titanium works). Why? Because I don't want to sacrifice usability, experience, and functionality. Most apps, I've found out, have different experience, usability, and functionality on difference devices.

armandososa2• 12 years ago

You can do great mobile experiences on web, but it does require extra time and extra thinking. I don't want to download a new app for every new service I want to try, so I will say that there's still space for really good mobile web apps.

hipsterrific3• 12 years ago

Try and not all web apps should have a mobile web app. But there's are instances where a productivity app will benefit from having a mobile app than just a responsive website. I've worked on a couple of web apps that made mobile-first their paradigm and it sucked. Testing sucked. Dev'ing it sucked. We had to sacrifice a lot on usability and experience which sucked a lot. A mobile app would have made more sense than a responsive site. Then again, I used trulia and having them ask me to download their app was just stupid as the app provided zero value to me (their responsive site was perfectly fine).

armandososa2• 12 years ago

(Disclaimer: I'm not a native english speaker)

Responsive design, at least when it comes to web applications, is just lazy. Sure, your layout will totally shrink, but that does not make it "mobile ready".

The set of UI conventions that work and have worked for a long time on desktop won't work on tiny touch-enabled screens.

And if you really care for the user experience of those 2 billion mobile users, you'll have to rethink and redesign every widget and every interaction for fat,lazy and impatient fingers.

So I have to disagree with the notion that supporting a wide rage of devices requires an extremely low additional effort. That's simply not the case if you care about user experience both on mobile and desktop.

I'm guessing that, while on a hackaton, you'll have to choose.

Daiz2• 12 years ago

When building a site from scratch, the additional effort required to support a wide range of devices is extremely low.

If all you're doing is designing a simple website (like a blog - not exactly that complicated), then yes, mobile might not be that hard.

If you're designing a web app though, that's a whole another story. If you want everything to work properly, from user interface to the actual features (one could think that cross-browser compatibility is really good these days, but in practice there's still a surprisingly large amount of places where things differ and you need to account for it), then that will mean a lot of testing across browsers and devices, which will increase your development time a lot. Hell, just getting cross-browser compatibility with common up-to-date desktop browsers is a considerable time investment with web apps.

So with that said, I must agree with the organizer quoted in the post - when all you have is 48 hours, you can't really blame anyone for not spending the time making their web app work well on mobile. Hell, I probably wouldn't spend that much time on cross-browser compatibility either (probably latest Chrome & Firefox compatibility at most - and even that can be surprisingly involving), since the time really is quite limited, and if you end up with something cool (incidentally, as the blog post says, they didn't win, and spent time making things mobile & desktop -friendly) then you'll have all the time in the world to worry about compatibility and making it truly "production-ready" after the hackathon is over.

EDIT: Not to mention that depending on what kind of web app you're building, it might not even make that much sense on mobile - generally speaking, you don't use mobile devices for serious productivity, so if you're building a serious productivity app, full mobile support could be a waste of time when no-one would even use it seriously on mobile. If you're storing data server-side with user accounts, though (as in, your app's not purely client-side), then you could probably get away with a limited mobile version that only gives read-only data or something. Of course, this all depends on what exactly you're building and should be evaluated on a case-by-case basis.

robertnealan• 12 years ago

While I think Jon made it seem a little easier than in reality, as someone who uses and designs for both mobile/desktop I have to agree that when done right it's not that much more of a time investment - generally ~10% extra on a full design/build project. There's a real argument that in a hackathon 10% extra time isn't worth it, but in the real world where you don't always have 48 hour time limits I've always been in the opinion that the initial time investment is far outweighed by the benefits and the additional cost that you would incur by making it mobile-ready later (which usually requires either rebuilding the frontend templates from scratch or hacking away at the existing codebase in a way that is far less organized/sustainable).

As far as the long standing argument of whether building mobile-first or desktop-first, I honestly think that's up to developer preference so long as the whole team is building with the same mindset. Building responsive, mobile-ready sites/applications isn't free, but a lot of web developers/startups seem to make it out to be much more difficult than it actually is when done consistently and right.

jonathanstark10• 12 years ago

@robertnealan Well said.

Having personally worked on responsive redesigns for ew.com, techcrunch.com, and timeinc.com, I know as well as anyone that designing a full scale responsive site can be hard. Really hard.

That said, the hard parts are not the technology or the implementation. Rather, the hard parts are organizational dysfunction, legacy CMS software, integration with non-responsive 3rd party services like ad networks, and so on.

The complete lack of those sorts of challenges in something like a hackathon means that devs are free to design in a way that maybe they can't do at work or for clients. My shock was that given such freedom, they choose to basically ignore mobile.

Daiz2• 12 years ago

There's a real argument that in a hackathon 10% extra time isn't worth it, but in the real world where you don't always have 48 hour time limits I've always been in the opinion that the initial time investment is far outweighed by the benefits and the additional cost that you would incur by making it mobile-ready later (which usually requires either rebuilding the frontend templates from scratch or hacking away at the existing codebase in a way that is far less organized/sustainable).

Oh, I completely agree with this. I'd say that even during hackathons, mobile is definitely something you should keep in mind when you're building your design, markup, etc. just like you normally would so that things will be easier in the long run. However, the key thing is that when you hit something that would be easy to deal with on desktop but hard on mobile, you should just go the easy desktop route and leave the mobile solution for later (post-hackathon). And when you're building web apps, those kinds of moments will pretty much inevitably come because mobile is hard - compared to desktop, you will generally speaking always have less screen real estate (and input elements need to be larger, making it even smaller), a lot less processing power, lack of precise pointer input and hover (touch vs mouse), etc etc.

Again, the point here is the difference between building websites and web apps. With websites, your biggest concerns are generally that the site scales nicely to various screen sizes and that links are comfortably clickable (and that it all looks nice). This wouldn't be all that hard to do even under time constraints. Web apps on the other hand generally feature a lot more (single-page) interactivity, and getting that right is a lot harder.

untog2• 12 years ago

Honestly, if you asked these Rumble devs how many of them even tested in Firefox I think the percentage would be low. Such is the nature of these events - you move fast. And as the post says, the aim is to make an "innovative web application", not a "web application that would be delivered to users in an optimal fashion". It was also a Ruby hackathon - given that it's a server-side technology I doubt the client presentation form was given much focus.

I've made this mistake plenty of times before myself. I've dwelled on the the actual real life issues my concept would face and come away at the end of the weekend with a well-considered concept. But what people want at these events is a catchy, flashy demo that wows attendees and that can be tweeted/Facebooked/etc.

This is part of the reason I don't bother sacrificing my weekend to hackathons these days, but still. You're optimising for the wrong use case.

jonathanstark10• 12 years ago

But what people want at these events is a catchy, flashy demo that wows attendees and that can be tweeted/Facebooked/etc.

And then what? Opened on a mobile phone where the demo doesn't work? Mobile is where people are reading facebook/twitter so if you want to optimize for catchy/flashy for sharing in social media, you should definitely focus on mobile-first.

untog2• 12 years ago

You're assuming the tweeted link would go to the app itself - it wouldn't. It would go to their roundup of the event. These events are not built around making long-lasting projects.

bjacobel• 12 years ago

As long as we're talking about mobile usability - why is this site's font huge on desktop, then get smaller as you reduce the viewport?

Mobile Last?

Related Stories