Rails rumble 2012

What? A team? pffft!

This year, after asking on the developers’ mailing-list at AF83, and not getting any positive reply, I decided to participate to my first Rails Rumble. The timing of this year’s Rumble was terrible for France. Two important conferences are happening over there, namely OSDC.fr, and the Open World Forum. Two activities in which my French co- workers are heavily involved with. Ergo, no co-workies for me.

I could not let this event pass (again) without trying, plus there were like 500 slots for teams, I’d just snag one. So I went for it solo, using Rails Rumble as an excuse to boot a little project of mine.

Scratch your own itch

I’m a big user of RSS, and its atomic variants. Working remotely has made this habit worse. I consume news from France, friends, tech blogs, web comics, and work through RSS.

To stalk efficiently all my co-workers, I’m subscribed to a bunch of feeds generated by various apps we use at AF83 (Github, Gitweb, Trac, Dropbox, a bunch of trojans installed here and there). Typically, my morning routine starts by opening newsbeuter and mutt to have a look at what’s been done, checking also e-mail notifications from a few other Trello boards… Of course, if I plan to have a productive morning, I’ll just skip all this, but that’s another story. :)

The thing is, I’d rather have all the work-related news, and notifications, aggregated in a nice timeline: they would form a personal news stream. I could filter and search on this. I’d like the possibility to setup groups, and tags, and {mail,SMS,tweet} notifications when it matters. In short, I want all the work-stuff in one place, and be able to manage the data stream easily. That’s roughly put, my Rails Rumble project. I called it Rivus after 5 minutes of intense googling for names and synonyms of the concepts of timelines, streams, and… rivers. ^^;

All of that from scratch in 48 hours seemed a little ambitious to me. So instead, I planned my time to build the basis for this app, trying to ship a viable MVP… Which proved difficult, since what’s been shipped needed at least 3-4 more hours of work to be called an MVP – which it got, but after the deadline. :(

Concluding

I wondered if keeping a log of 48 hours would be useful, and so the only log I have so far, is Git’s. It’s a little early to think of good lessons from this, but a few ideas came to my mind.

• First, don’t go solo. It’s harder, and you’ll need quite a lot of
  self-motivation and discipline to keep going.

• Plan ahead, with paper and pen.

• Keep it small. Don’t plan to add 50% more features later if you’re aiming
  for glory.

• Learn how to setup a server stack, and deploy your application to it
  automatically, or have one person who knows it on your team. Starting from nothing, if you need more than an hour to setup your application online, then you’re really doing it wrong.

• Don’t listen to your {girl,boy}friend when s-he offers to slip out for
  a while to have a drink, or see friends. I fell for it, twice: on the basis that time’s
  precious, so I’d rather spend it with interesting people rather than my code editor. Not a wise plan if you want to finish on time though.

• But do plan a few hours of distraction, you’re supposed to be having fun. :)

All in all, I had a lot fun and learned a few things. For those who would be interested in hacking with me on this little (Rails) project, I plan to open- source the code in a few days on Github (once I overcome the shame of releasing code riddled with FIXMEs, and without a single line of test).

And that’s about it for this evening, plus my Humble Ebook Bundle’s waiting for me.