An interview with Tweetie creator Loren Brichter
Atebits software released Tweetie for Mac this week and Aulia Masna, executive editor of Macworld Indonesia, scored an interview with author Loren Brichter in between his sleeps. That is, if he slept at all. Given the popularity of Tweetie for the iPhone and the run up to the Mac version's release, it's a wonder the man gets any rest. In this interview, he shares some insight into how the app came to be and the thought process behind its (not so?) radical design.
Tweetie for the iPhone is hugely popular. Why do you think people like it so much?
I think it struck the perfect balance between power and simplicity. Before Tweetie, you had a choice between clients that had a simple interface but were lacking in features, or clients that had every feature imaginable but an interface where couldn't figure out how to use half of them. Tweetie brought to the table some really powerful features in a really intuitive package. Something appropriate for software designed for an Apple device.
It took a while before you released Tweetie for the Mac. What prompted that? Did you always plan to release the Mac version second or was it because the iPhone version was so popular?
I suppose the world moves pretty fast. Tweetie for iPhone had only been out for five months before Tweetie for Mac launched. The popularity of the iPhone app certainly helped trigger the Mac version, but not for the reasons you might think. As the iPhone version grew more popular, I spent more and more time trying to handle support requests via Twitter. I tried a few clients, but ones with multiple-account implementations were rare and poorly thought through. I needed something that I could use myself. It had to be fast, full-featured without being complicated, and powerful enough where I didn't have to go to Twitter.com to get things done. And AIR apps are like modern day Java applets... sure, they run on every platform. But they also suck on every platform.
What do you think of the current market for Twitter applications? Previously the only native Mac app was Iconfactory’s Twitterrific but now there’s a whole range of apps that people can choose from including both native and those based on Adobe’s AIR platform. Are we at or close to a point of saturation?
I really don't know. One of the fantastic things about Twitter clients is how easy it is for users to jump from one to another. Just type in a username and password and off you go. It's possible for anyone to write a Twitter client nowadays and have the opportunity to completely blow everyone else out of the water. It's very exciting. Very democratic. And it certainly seems like everyone and their mother is trying to do just that. I'm just happy to be part of it, I know the developers of other clients and I can say definitively that competition is making all of us write better apps.
You once said that you’re not a big fan of ads, but now there are two versions of Tweetie for Mac, an ad supported version and a paid version ($14.95 until May 4). Has your position changed or is it due to the economics of a Mac app?
The economics were certainly a factor. If the App Store got one thing right, it's the ease with which you can buy an app for your iPhone. It's two taps. There's more of a hump with Mac apps, and people are used to try-before-you-buy. The other factor was Fusion Ads. (http://fusionads.net/). They serve some of the most beautiful ads around. Many people mentioned that they registered Tweetie but then opted to keep the ads on anyway (there's a checkbox in the Preferences).
When you put the app into private beta stage, it seemed that you cast quite a wide net based on the number of people talking about it on Twitter. Was it larger than you expected and how much did the early users affect the final product?
I had a pretty clear vision of what I wanted to build from the beginning, but the feedback was invaluable. Everybody uses Twitter a little bit differently, and I tried to find those few options which when tweaked in the right combination, could make Tweetie work exactly how you wanted. There's still a ton of work to do in this regard, and I'm happy to get feedback from everyone.
How did you approach designing Tweetie for the Mac? It almost looks like a radical departure from the standard Mac app.
Radical, yet not. It's an evolution of UI concepts, many borrowed and extended from the iPhone. It fits in as a Mac app, looks and feels like one, but the navigation and functionality is next-generation. The drill-down is inspired by iPhone, it was important to be able to make it easy to delve into tons of information without requiring the app to sprawl across your entire monitor. The sidebar was particularly interesting to design. I needed a UI that worked well with exactly one account but also scaled to many accounts. Old school Mac implementations might have had a double set of tabs, or a Mail-style sidebar, or a tab bar with an account dropdown box. All of these are flawed. Some take up too much screen real estate, some aren't good enough at providing at-a-glance badge notifications on a per-account-subsection basis. The sidebar design in Tweetie for Mac solves these problems in an extremely elegant and scalable way. It's new and different, which is hard for some people to swallow, but somebody needs to push the envelope and try new things or we're all going to rot in UI hell.
TweetDeck is a very popular app due to its ability to create groups based on a person’s preference but you are against the notion of client side grouping. Will your position stand, given that you’re willing to relent with respect to ads?
To be clear, I'm not against the notion of grouping, I just think it's a hack to do it client side. It's possible (and common) for someone who is following lots of people to find gaps in the timeline of subgroups. Having said that, my initial feeling towards groups came from the perspective of a developer of a mobile application. Mobile apps by definition are run in brief spurts. This makes proper grouping next to impossible. Desktop apps are a different story - because they are running all the time, and can constantly poll your Twitter stream there is a much smaller likelihood of gaps. All I can say right now is that I'm flexible. I would like to take the time to point out (yet again) that there is a really amazingly robust solution to groups where the work is done server side and there won't be any gaps in your stream. Plus you can use it in Tweetie for Mac and Tweetie for iPhone today in a really graceful manner. It's called multiple accounts. The idea is to create another "read only" account where you follow some subset of people. You use that account only for reading, whenever you tweet you can always switch to a different account for posting (Tweetie on both platforms makes this ridiculously easy to do).
You also mentioned that the core of Mac Tweetie will become the new base of the iPhone version. How much of a change are we talking about here internally, visually, and functionally?
That's all still pretty secret, but I can say that the things that I got right with Tweetie for iPhone won't change, and the things that I got wrong will be vastly improved. The internal changes are going to be major. Very major. The new core powering Tweetie for Mac is going to make a lot of interesting stuff possible.
Before Tweetie you had Scribbles, a different kind of drawing app. Has Scribbles seen an uptake since people discovered Tweetie?
Sadly not that I can tell. But the few people who do use it absolutely love it.
Are we going to see something resembling an advanced version of Scribbles?
It's on the back burner at the moment, but I've been mulling over a much improved version for a while.
What are your plans after this? Sleep, presumably?
What's sleep? But seriously, I'm throwing my weight behind Tweetie for Mac and iPhone. I have a fantastic new foundation with which to build a true next generation Twitter client. People have just seen the tip of the iceberg.
Thank you for taking the time for this interview, Tweetie for Mac looks great, we’ll look forward to more from atebits.