This is probably old news to many of you by now, but I got highfiber.org up and running again 5 days ago (version double-zero, for anyone keeping count), after receiving an unexpected number of comments about it on my previous entry. I made the decision to bring it back, appropriately, on Friday the 13th and proceeded to throw it together over the weekend. By Tuesday we were online, although issues with our hosting prevented me from stabilizing the app until mid-Wednesday.
I really missed these binges. Every version of highfiber was written in a literal fit of coding fury, and there’s nothing quite like the adrenaline high of building something you love. I used to lock myself up for days (sometimes weeks) at a time to design and build, confident in the fact that when I emerged I’d have a wicked new site that’d blow everyone’s mind. This new version is a little more humble I guess; I’ve gotten older and a lot more aware of my shortcomings. At the same time, I’ve come to appreciate certain truths about online community websites, which is what has been guiding most of the architectural decisions for this version.
Here’s a couple of them, for anyone interested in this sort of thing.
1. Graphic design isn’t important. Lots of social networks fall all over themselves to be really snazzy and flashy and noisy, and for the most part, they just completely miss the point. With social networks, the primary focus should be the user-interactions, and making sure they occur in the smoothest way possible. What it looks like basically matters fuck-all; if their friends are there, people will come. (If anything, the graphic design should be as understated and unassuming as possible, because you want people to concentrate on the content.)
2. Being unique is important. Although Highfiber is a forum at its core, it has a couple of flourishes that make the HF experience singular online. The patting/poking system has gone a long way towards embodying what this site is all about, and the rules of what a user can or cannot do based on his current score is almost like an ongoing experiment in micro-scale socio-economics.
3. Being stable isn’t important; communicating with your members is. To a certain extent, users will forgive downtime as long as you explain what’s going on. None of my sites have ever been known for their stability, so the broadcast bar is usually one of the first pieces of admin functionality I write.
4. You won’t get it right the first few times. One of the hardest things to accept for me back when I was just starting out was why some of my ideas worked and some didn’t. A lot of the time I would rationalize it by saying that people just “didn’t get it” — it was their loss, not mine. Now I know a little better. The truth is, your ideas are only as good as your audience’s acceptance of them. (The corollary to this rule is: “Choosing the right audience for the right idea is pretty frickin’ important too.”)
These days, my strategy is really simple. I release a build that represents my best guess at solving a particular problem, then pay lots of attention to how people react to it. Sometimes you’ll get lucky and people will be very vocal about their feelings; other times, you have to infer it from behavioral patterns. Either way, you need to be flexible enough to move on your learnings.
Case in point: my current experiment on double-zero involves adding a cost for every comment. Essentially, a user is charged one point (or more) every time they post a comment, which decreases their overall score. I won’t expound on the reasons for this particular solution just yet, but it’ll be interesting to see how people maneuver with the new rule in place.
Lastly: thanks to everyone who left comments on my last blog entry. I’m amazed, impressed, humbled and a little scared of your fervent devotion to this community, and I hope double-zero eventually meets all of your expectations. Here’s to another 7 years of highfiber!