It’s been nearly two months since BarCampMilwaukee, and I still haven’t finished blogging it. Some journalist I am, huh? Just when I became convinced it was no use continuing, Christian Wilcox prodded me for details on the mini-mashup that happened Saturday-night-into-Sunday-morning as September turned into October. So, for Christian, Pete and anyone else patiently waiting for this (yeah, right), here’s a Thanksgiving reminiscence.
The BarCampMilwaukee Mini-Mashup
As you may recall, the mini-mashup session was scheduled for the late-night period (everyone knows programmers work best in all-night sessions, no?). Pete Prodoehl led the session and several others appeared. Pete kindly recorded the rollcall: “(Here’s the folks sitting in right now: Victor Nwankwo, Blake Hall, Kevin Ciesielaski, Christian Wilcox, Scott Reynen, Mike McCallister and Justin Lowry…).” Of these 8 folks, 2 didn’t have laptops and/or weren’t programmers — me and someone else. What was moderately startling at the time is that everyone in the room who came equipped for coding was carrying a PowerBook! Macs are allegedly just for graphics geeks, not coders, right? I no longer believe this–thanks, guys!
Pete offered an introduction to the mashup concept, and several folks shared sites they’d either worked on or simply liked. Victor showed off some of his work with music and event sites in Chicago (and elsewhere), which was pretty cool, if not exactly mashups. Right around midnight, Pete asked whether folks had the interest or energy to try this now. After a little bit of hemming and hawing in the crowd, Victor took the lead in affirming that, yes indeed, he’d like to give it a go. When no one left the room after the declaration, the project began. I was reasonably certain at that time that if no one had declared in favor, we might have all gone to bed. It is perhaps no accident that Victor had just arrived at BarCamp, but without him the mashup would never have happened.
That first half-hour was spent brainstorming ideas. I wasn’t taking notes at this point (my notebook was still upstairs at the time, but I later copped a pad of paper from the registration table — thanks folks at C2!), so any brilliant ideas that went by the wayside will go unrecorded. Then someone mentioned 10×10, an interesting and curious site that generates photos related to current world news tagged from Reuters, the NY Times and BBC. The tags are then ranked according to frequency, 1-100. Click a tag or a photo and you get a clickable list of news stories related to the tag. This site is updated hourly. Here’s a more detailed explanation.
There is a history function associated with the site, but it’s clunky. People also wondered if there was a way to track the rise and fall of a particular tag during a particular day/week/month/etc. That was the “ah-ha” moment — this could be done! Thus was born “10x10xTime.” Suddenly, the room was abuzz with activity. Everyone took an assignment, a host was found (with just a little difficulty), and the domain registered (http://tenbytenbytime.org/). By 1:30 AM central time, things were well on the way.
Did I mention the difficulty finding a host? Memory might be playing tricks, but as I recall: Victor went to his regular hosting firm, but had some kind of trouble getting things set up. Tech support was less than supportive (probably woke somebody up), but a backup host was found (or the original got things fixed; Victor’s phone rang a couple times during this process). This was fun to watch!
Meanwhile, it was decided to pull all the site files into a database to make them searchable, and some debate went on over what tables to create. They finally settled on Date, Time, Word, and Frequency tables. A basic interface for the site was built — all in that first hour. Then, before 2 AM, the database was created, functions were selected, and it was all uploaded (without data, I think) to the host. There was more discussion about what the search page would look like (a PHP page was agreed to, in part because you could drop it into Drupal if you wanted to), and what sort of search queries should be allowed (my note reads: “Search #5 for each day/hr” as an example).
By 2 AM, I was starting to wilt, and fewer discussions were happening (by gosh, people had coding to do) to keep me awake. Actually, others were also talking about setting a bedtime, but I get ahead of myself. More decisions were made: results would be sortable by frequency. People could search based on a date range, a single date, and/or a word. At 3 AM the PHP front page was ready (If you go to the site now, there isn’t much else to look at, which is the main reason I didn’t formally link). This was pretty much done when I tanked out; I learned that everyone except Victor crashed not much later. We found Victor asleep in the room when I woke up. Can’t say exactly what he did in that time, but what did get done (at least conceptually) was mighty impressive — to me, at least.
This would be a good time to (again) thank Pete, Tegan, James, the Waldron clan, and all the other folks who put BarCampMilwaukee 2006 together. We learned quite a bit, had some fun, and hope to expand the connections made. Let’s see if we can do it all again in Madison in February, and on an even bigger scale in Jefferson County next summer!
Happy Thanksgiving to everyone. Be sure to take some time off from your celebration to get your BitTorrent client working on openSUSE RC1 — I’m depending on you!