As Twttr (sic) celebrates its 4th birthday, I figure it’s as good of time as any to blog about something I’ve been thinking for a while. No, don’t worry, I’m not going to quit tweeting, but I will quit competing. Which sadly, is probably what they want.

Twitter engineer Alex Payne sent out a prophetic tweet last month. In this message to the Twittersphere, he basically says that is going to be so badass and feature-rich that you’ll soon rethink your need for 3rd party Twitter clients. This caused an uproar in the developer community as many (over-reacting) people took his comments to mean Twitter was going to try & kill off the alternative clients. @Al3x and the rest of Twitter HQ went into damage control mode to explain that Twitter wasn’t attacking alternative clients and that they were still supportive of the developer community. Hugs all around, right? No. I think most people saw the writing on the wall at that point.

I know I have. So, after 2 years of developing my own Twitter clients, I’ve decided that I’m finally throwing in the towel. Twitter has built a great web app, so there’s little need for me to continue. There’s part of me that is sad, but mostly I’m really happy for Twitter. Also, I’m relieved as I can now focus on something else.

A little background… It wasn’t up until recently that Twitter’s own web client ( lacked most of the features that I wanted, so I was forced to build them on my own. I began building Tweenky almost 2 years ago and the goal was simple… create a Twitter web client that had the following features:
A) A friendly Ajax interface
B) Integrated searching
C) Groups
D) Saved searches
E) Fixed the @reply problem where replies were not visible to your replies feed unless it started with “@username”
F) Had other basic shortcut features (like retweet links)

When it was ready in the summer of ‘08, I released it to the wild with the help of TechCrunch and other tech blogs, who all praised its set of features. I’m not going to claim I was the only one working on such features. Most of them were just obvious extensions to how people really wanted to use the Twitter service. They would have been implemented by Twitter themselves had the service been stable enough to add feature development resources. It’s funny to think that between 2006 and 2009 remained largely unchanged. Why? Because they were generating too many failwhales and fixing those was the #1 priority.

By 2009, the engineering team had rebuilt Twitter into a stable platform and they were finally able to let the front-end developers loose and start working on features. First came some ajaxy goodness, then integrated searching the replies/mentions fix. Later in the year they added Lists and the Retweet feature. At that point, I noticed Tweenky started to become less & less useful. Others did too and the userbase started to decrease.

Enter 2010… The front-end team is beginning to crank out features & tweaks at a fast pace. So far this year we’ve seen hovercards, location dectection, and integrated maps. It’s finally at the point where the speed of innovative features is out-pacing what the developer community will be able to keep up with. There are still some major clients, such as Tweetie (on the desktop) that haven’t even integrated Lists yet. I’m not going to attempt to work on the hovercards or integrated maps, not because I can’t do them, but because what’s the point? I’ve actually begun using the web client more than my own client because it simply lacks essential features. Sure I can add them, but once I’ve completed that, the larger-than-1-person-front-end team at Twitter will have rolled out a couple more slick features, and I will always be playing catch-up.

So here’s the point of this post… I’m done. From here on out I suspect the majority of my Twitter time will be spent on the web client. Don’t take this the wrong way, I’m actually really happy for Twitter and the awesome front-end/UX team they’ve assembled (which includes a number of ex-Yahoo’s =D ). They’ve implemented most of the “must-have” features that 3rd party developers have been working on for years. This is a good thing because those features are now available to the majority of the Twitter userbase instead of a small portion. I suspect over the course of 2010 and beyond, the pace that we see new features will continue to increase, and with every new release, more & more 3rd party developers will cease working on their own clients. This will be a bitter pill for some in the developer community to swallow but the side-effect is they’ll be spending less time on simple, basic features that should have, and instead hopefully on innovative non-client apps or things completly unrelated to Twitter.

I’m mostly happy with this direction. The main reason I’ve developed Twitter clients is to geek around and gain experience in areas I feel my knowledge is lacking. I’ve never approached my client development as “OMG, I have to get as many people as possible to use this thing so I can make money and/or sell it!” I’ve never attempted to monetize my work. I’ve just approached it as there’s a certain user experience I want to have with Twitter, and if anyone else wants to join the fun, cool. No? That’s cool too. Work hard and good things will come. Having converted the original Tweenky client from mostly PHP to all JavaScript, I’ve been able to gain valuable experience with jQuery, YUI3, & JS in general. To me, that is satisfying enough. All the JS, REST API, and scaling knowledge I gained through this process is one of the reasons I now have a job at Yahoo.

So what’s next? I dunno. If I’m spending X less hours per week trying to replace, I can now spend X hours working on something else. I’ll most certainly work on some non-client Twitter apps, but I’m hoping to spend the majority of my time on non-related Twitter projects. Maybe some much needed Node.js hacking? Maybe some WebOS apps? Hmmm… Stay tuned.

P.S. Tweenky has always been an open-source project. You can find the source code on GitHub. You can also find Tweenky’s cousin “Tweetanium” (a YUI3 rewrite) on GitHub as well.