Rapid Development at a 162 Year Old Institution: What I Learned This Summer

 

OneWeek | OneTool Development Team Pod (Missing in the picture are Julie Meloni and Patrick Rashleigh), by Effie Kapsalis, 2010

I recently took a weeklong break to go to camp - developer camp. No mosquito bites or arts & crafts, but lots of late nights and good stories to tell. The camp was OneWeek | OneTool, “a unique summer institute, one that aims to teach participants how to build an open source digital tool for humanities scholarship by actually building a tool, from inception to launch, in a week.”  Hosted by the Center for History and New Media (CHNM) and funded the National Endowment for the Humanities, Office of Digital Humanities, I was sequestered in the CHNM offices with eleven other “digital humanists” to “build something useful,” our one of few directives from CHNM Managing Director, Tom Scheinfeldt.

Well, the result was summed up perfectly by Doug Knox, a fellow Team Outreach member:

“What 12-18 people can get done in a week with loose coordination and a common goal is remarkable.” 

And you can now see and experiment with what we built, Anthologize.  It’s not a tool with a big scope - it does one thing well (although it is an alpha, so there are known issues). It enables researchers, curators, writers - and bloggers in general - to compile, edit, and publish anything available through RSS feeds. From Anthologize, you can send out your compile work as an eBook, paper publication, or TEI (an open XML format for storage and exchange).

The website text and information architecture brainstorm whiteboard from Team Outreach, by Jana Remy, 2010

However, I’m not here to talk about what Anthologize is (even though it will have many useful applications in the library, archive, and museum world). I want to talk about how it was done and how we can incorporate some of the ideas behind rapid, interdisciplinary technology development into the Smithsonian’s process.

Here was a brief recap of the OneWeek schedule:

Day 1: Meet & Greet/Learn about Open Source Development/Brainstorm
Day 2: Decide on project/Divide into teams/Start Work
Days 3-5: Build/Refine/Check-in (over and over again)

And before I go on, I’ll dispel the OneWeek myth. We actually spent the few nights (and for some of us, days) after OneWeek ended, fixing and testing. So, One Long Week. There were some key ingredients that made it possible for us to build a meaningful, open-source tool in a week:

1. First, we were a group of “do-ers and schemers,” which apparently is par for the course for digital humanists, as I learned from Meagan Timney, one of our test users.
2. We had a common goal which put our end user at the center: Build Something Useful. And as Steve Ramsey on Team Dev pointed out, it didn’t hurt at all that most of us would eventually be end-users ourselves.
3. Trust. Period. Even though I have worked in User Experience (UX) and Information Architecture, my role on OneWeek was Outreach. I trusted Teams UX and Dev to do what they do well. We met twice daily as a group (for no more than a half hour!) to discuss what each team had accomplished and the next to-dos. We did give feedback, but final decisions were essentially up to the teams.
4. There were no milestones, timelines, technical requirements, and other technology project management-y things. We had a google group where we posted final products. When it was all said and done, we had documentation and FAQs for the Anthologize website. The development process was organic and largely "do what needs to be done.” This happens to be very much the management style of CHNM (no surprise here). Tom Scheinfeldt summarizes this approach in his insightful post about the project here.

 Voting on OneWeek | OneTool Ideas, by Jana Remy, 2010

Now I’m not proposing that we work day & night at break-neck speed. Life, relationships, and hygiene cannot be put on hold forever. But I am hoping, and proposing, that we can do more inter-disciplinary team-based development to "build something useful." I’m certainly going to try to do more of this at the Smithsonian Institution Archives.

Anyone familiar with technology development in a large federal government and cultural institution knows that the OneWeek picture is very different from our reality. There are things that would have to be adjusted in the OneWeek process to make it work here. However, there are many things we can do now (trust, small interdisciplinary teams, etc.) irregardless of where we are. We really can’t afford not to.

Effie Kapsalis recently joined the Smithsonian Institution Archives as the Head of Web and New Media.To read an overview of how Anthologize could be used at libraries, museums, and archives, visit THE BIGGER PICTURE, a blog about visual archives and the Smithsonian.

Comments

[Nina Simon]

Great post, Effie. Thanks for sharing your experience. Maybe once a year SI staff could do the same?

[digitaleffie]

Hi Nina,

I would love it. Especially if we could get web folks with researchers, educators, curators all together to hash it out!

Effie

[Vincentstinks]

A nice post, though as you say, I am not sure how well it would work in the long run, and whilst trust works over short periods with do-ers and schemers, structure and focus through ticketing and stand up meetings is a good way to help drive a more splintered and less ambitious team.

Your post proves a good point though, that with focus and direction and a small finite goal, a lot can be achieved in a short period of time. Less encumbered projects can quite clearly be achieved in a reasonable manner.

[digitaleffie]

Thanks for your thoughts, Vincent. I think in large institutions where people stay for years, there is considerably more caution when it comes to embracing new ways of doing things. Elaborate process, meetings, and committees become the knee-jerk reaction, even in situations where trust and rapid development could be applied.

[john shupe]

Through a collective intelligence there is many ways to create some great ideas to materalize.