Shaun Inman on developing apps

Shaun Inman is in a great position right now. When his web stats app Mint took off four years ago, he was able to leave his career in client services behind. Today his projects make enough money to sustain him financially, which is pretty rare, especially given that he usually starts out building things purely for his own interest. His secret? People simply adore the stuff he comes up with.

Shaun Inman is in a great position right now. When his web stats app Mint took off four years ago, he was able to leave his career in client services behind. Today his projects make enough money to sustain him financially, which is pretty rare, especially given that he usually starts out building things purely for his own interest. His secret? People simply adore the stuff he comes up with.

The switch to producing apps happened accidentally – he was just tinkering in the early hours. In fact, Shaun says, everything he does revolves around tinkering: “It was surprising when Mint took off and demanded enough of my time to allow me to turn down the client work. It was mostly just me scratching my own itch, then other people [began] finding the things I was giving away at the time. I built up a base of people who were interested in the same things as me and they went to support my decision.”

Shaun’s latest major project, self-hosted recommendation-engine-come-feed-reader Fever, followed the same model. “I was dissatisfied with the available solutions out there and overwhelmed by the number of feeds I’d accumulated over the years by participating in the web standards community. I know I wanted to make a feed reader for myself in order to address the shortcomings [I’d experienced].”

Fever sold as many $30 licences in its first week as the original Mint had in a month. After just two months, more than 2,000 people were using it, 10 per cent of Mint’s total user base. The release was also accompanied by a largely favourable TechCrunch post, although it questioned the lack of a demo or trial version. “It’s for the same reason I don’t offer a demo trial version of Mint,” Shaun explains. “The apps are written in PHP and MySQL, so once you have the source code, anybody who is moderately familiar with PHP can go and rip out the activation. I could do a feature-limited version but then I have to maintain two code bases, and I probably wouldn’t use a hosted version.

“There are plenty of free, hosted feed readers out there but none of them has ever appealed to me,” Shaun continues. “One of the main reasons is that I don’t like a third party observing my behaviour and profiting off it by selling advertising and market analysis and all that junk. And also, I went to school as a graphic designer and I’m a self-taught programmer, so creating a scalable, multi-server MySQL database through an application doesn’t really appeal to me. I don’t have the technical background to pull it off. I like being able to touch all the bits myself and I’m not really interested in putting together a team to build something like that.”

So, for now, Shaun Inman will continue to support his customers personally. “If I paid somebody to do support, they wouldn’t have that personal attachment, that personal stake in making sure the customer is happy. I’d have to educate them. As a developer it’s better to have direct contact with your customers because you get first-hand experience of what problems they’re having and which you need to take care of first.”

That personal attachment is also the reason Shaun hasn’t sold off any of his applications. “These projects are part of my personal brand,” he says. “I’ve never set up a company name to operate under. I’m not sure I could part with that. It would be like selling off your child to another couple. I’m not motivated by money, and I’m not sure that retaining some sort of creative control would be enough or whether it would make it worse because I’d be stuck to maintaining it without having directorial control. Maybe you could try and pry them from my cold dead fingers …”

Shaun Inman tackles both the design and development single-handedly and reckons his time is spread 60/40 in favour of development time. Whenever he hits a dead end and gets tired by one side, he hops over to the other. The process involves a lot of trial and error and, as Shaun puts it, results in a fractured workflow, but he always makes it through in the end. The trick is to stay focused. “Knowing that at any time you can just hop over can get you into a trap where you actually don’t solve any problems. You just start hopping back and forth,” warns Shaun. “But there are also a lot of positives that come from doing both sides. When you’re designing something, you can anticipate implementation problems. And when you’re developing, you’re a little bit more sensitive to the nuances of the design that you’re eventually going to be producing on top of that code base.”

Recently, Shaun’s also started dabbling in iPhone app development. His first effort, Horror Vacui, is an 8-bit two-player strategy board game. It secured Shaun a grant from a local organisation in Chattanooga to work on the next one, a Metroid-Vania side-scroller. Since he was a kid, he always wanted to make video games and simply loves the idea of being able to get a game on a popular portable device.

For Shaun, the hardest challenge when developing for the iPhone was working with libraries and frameworks. “I don’t like magic,” he says. “I like knowing how things work. I don’t like taking things for granted because if magic breaks, you don’t know how to fix it, unless it’s your own trick. I’m sure the Apple Objective-C frameworks solve a lot of problems that developers had for years, but as a new developer I haven’t encountered those problems and I feel like I’m jumping through hoops that are actually there to protect me and save me trouble in the long run.”

When things become a bit hairy, Shaun, who’s also a musician, picks up the guitar. He’s been in talks with other web developers about putting together an album of covers produced in their own special way. And of course, Shaun also initiated Scalable Inman Flash Replacement (sIFR), a typography technique that enables you to replace text elements on screen with Flash equivalents. The method was later refined by Mike Davidson and Mark Wubben, who’s taken over its development. For a while, Shaun lost interest in the web typography issue but now that the space is heating up he’s getting back into it, switching his own site over to Cufn. “It came along and didn’t have the Flash requirement, which is great. And it even works on the iPhone because it uses Canvas. That’s pretty sweet. But I’m not a fan of those commercial font-hosting services. They’re interesting but I have the same reservations about them as I do about other third-party hosted solutions. One, they can analyse your traffic. Two, what happens if a couple of sites on a particular font licensing network get slashdotted or digg’ed, what happens to other sites on that service?”

Shaun’s latest project is a little content management system called Less, or the Less Broadcasting System, inspired by Twitter. It came to Shaun when he realised that he’d abandoned his personal site in favour of Twitter’s single little text area – that creating a blog post had suddenly become much too complicated. “But what if your content management system was a single text area, I asked myself. So I’m building something that uses a slightly modified version of John Gruber’s Markdown. It allows you to create posts and links from a single text area and is smart enough to parse out a title and tags or categories, like a custom slug for a URL for that particular person. The idea is to reduce the barrier to expression on your personal site down to a level that Twitter has achieved with their service.”

During the development, Shaun’s created a personal URL shortening service called Lessn that adheres to the same minimal aesthetic as Less, both in terms of UI and infrastructure. If the success of Mint and Fever is anything to go by, the CMS will be a smash.