I’m just now coming down from this years RailsRumble. This year competition was fierce. I don’t think I looked at one other app during the rumble itself (alright, Eric Davis’ Mealette, but at that time it was little more than a wheel). Browsing some of the completed entries after the fact was just crazy awesome (hopefully I don’t use that work too much).
The entries page for RailsRumble has some stats:
In 48 hours, 237 teams competed and 137 deployed a completed application within the time period.
Of those 137 only 22 made it to public voting, unlike last year when everyone made it. I’m bummed we don’t all make it to voting, not much more to say about that I guess.
This year instead of going solo I hooked up with a couple friends and together we formed Bit Twiddlers. It was an interesting group. Myself, Curtis McHale our design muscle, and Eric Krause who wanted to get a reintroduction to rails by fire.
Eric and I have worked together before, but a few years ago and in .NET stuff. Neither of us had worked with Curtis. On top of it all we’re a virtual team. Curtis is way up north on the other side of the continent, I was at the beach on a vacation that lasted through Sunday (the end of the competition) and Eric travelled from Newport, RI to DC after a late night at the office Friday.
The App
I think we had a good idea, the Idea Bit Bucket. It’s a trusted network sounding board for your ideas, helping you decide which is the next best choice to execute on. We’ll continue to work on it, we probably had 80% of the functionality I had planned done for the final build of the weekend. We need to clean up the network piece a bit more, add different bit types and get on the same page about what a dashboard is. It sounds like more than it is, we were pretty close.
Any of the data that’s in there will live on once we move it off the rumble servers and onto a more permanent home.
What we could have done better
Practice. Yep, practice. I spent a hour or so helping everyone getting their environment up to speed. In Curtis’ case we basically ripped out a hand compiled version of ruby (1.8.8), all the cruft associated with it (gem, rake, paths, etc). All this done with TeamViewer connecting to a Windows Vista machine (Curtis) from a MBP (me, on an aircard, at the beach) to tweak an Ubuntu image running in VirtualBox. No, it probably couldn’t have gotten any more convoluted. Just a little bit of practice would have helped, worked some kinks out, got everyone familiar with the workflow, used to github and git, etc.
Sketch a lot more. Not sure what the technique is called, but you have everyone on a team independently sketch a design, then present. Then you go back, individually, and tweak your design, stealing the best bits from all the ones you’ve seen. Cool idea and very effective. We could have gotten past some “vision” issues early on.
Whittle the idea down to it’s core. The core idea for me is what I said earlier: “Trusted network sounding board for my ideas”. We could have gotten by with 3 things:
- Ideas (Title and a big text area supporting markdown, textile, whatever)
- Comments (to actually get feedback)
- Network (solid way to add people to your network)
I got distracted by stupid things:
Adding bits to ideas (bits are things that describe an idea that aren’t just words, basically file attachments). So I had to play with paperclip, and then jquery and a plugin to get a lightbox to show the bits.
Voting. I thought it would be cool to have everything votable, so your network could +/- an idea, or certain bits of an idea. I also thought it would be cool for you to call for a vote (which turns out to be a poll with 2 answers, yes or no). You could ask things like: Could I effectively monetize the bit bucket by charging per pixel uploaded? Or some such.
Who knows, voting might rear it’s head again, but I think most of that would be taken care of in comments.
Set Expectations. I really think it would have helped to set out some expectations from the beginning. Things like:
- Don’t sleep with a broken build.
- Be available.
- Get some rest.
I’m not saying any of this applied to us. Hell, it wasn’t until Saturday afternoon that I actually remoted into Curtis and Eric’s machines to help them with their environments.
Most of this just goes back to us not having worked together as a team before, the 48 hours of the rumble is not necessarily the warm and cozy time needed for team building, especially when I’m building the team.
The future for Team Bit Twiddlers and Idea Bit Bucket
Idea Bit Bucket will live on in some capacity, probably back to at least the core three things from above (ideas, feedback, network). If that sounds obvious, it’s because it should be, but for some reason I didn’t pick up on it early enough.
I’m more than happy to keep working with Eric and Curtis on this. I’m trying to drag Eric kicking and screaming to the world of ruby, because for me it’s been my happy place. As for Curtis, my creative side has been cowering in a corner of my psyche since about 1980, so I’m looking for a one or two kick ass designers who can work with me (and vice versa) to build some kick ass products.
Wrapping it up (already)
It feels like I’ve crashed off my Rails Rumble high not making it to the final voting round, but such is life. There were a lot of apps that looked way better than ours. Some of the folks participating in the rumble are just awesome. So from that point of view, I’m honored to be among them.
At the same time you don’t sign up to build a web app in 48 hours hoping to lose, so it’s going to piss you off a little. That’s okay, I can’t wait for the next one.
