Unlike the world of art + design, creative science is rarely a one-man show. Sure, you could lock yourself away in a room and commit to work diligently until hypothesis X is validated or disproven. But a scientist rarely possesses all the skills needed to throughly tackle the suite of complex problems at hand. Sometimes, they could theoretically complete all the work by themselves, but parts of the process would be so frustrating and painful that the effort just isn't worth it (that's what Ph.D. programs are for--training, perseverance, and struggle. Once you pass that milestone, collaboration and delegation make life so much easier). When a group of scientists target a specific problem, their collective effort can be seriously efficient and time-saving. Your problem might require someone else's skill set. You might learn about an easier way to get things done.
The #Phylotastic hackathon rolls on. I can honestly say that the second day was even more exhausting than the first. Yesterday we went through an iterative process of discussions, allowing each subgroup to define and prioritize the tasks they would complete before the end of the week. I was surprised that we didn't dive straight into programming (to me, the term "hackathon" implied an all-hours coding bender), but clarifying the goals was important to get everyone on the same page. With such a short timeframe to produce Phylotastic tools, we want to avoid wasted or duplicated effort.
So, each group broke off in the morning. We made design decisions, noted our challenges, and identified the outcomes and deliverables we plan on producing. I've ended up in the "Shiny" group. We're tasked with building the user interface of Phylotastic - a sleekly designed website (as well as non-web based modules) which shows off what you can do with the Phylotastic toolkit. There's no way we have enough time to get everything bug-free and fully functional by the end of the week, however, so we're focusing on a few demo "test cases" which showcase the long-term vision for Phylotastic.
Team Shiny is inherently dependent on all of the other groups - we need to know how to interact with the other computational pipelines and APIs that will be developed this week. After defining our own goals, we needed to hear about the other group goals--then talk with them and align our work to be fully complementary. After each breakout session the entire hackathon reconvened and disseminated our musings to all participants. Some of the discussions got very technical. Sometimes we went off on passionate tangents (taxonomic name resolution has become a running joke here). But many of the exchanges helped us to crystallize the path forward--what programming language to use, how many APIs (and what their exact functions should be), and what group would be responsible for what products.
By 5:30pm, I was beat. It was time for another night of beer at the divine Geer Street Garden.
No comments:
Post a Comment