For the past year, my time spent on W3C work has been quite completely subsumed by my having joined the Data on the Web Best Practices Working Group. It’s been an amazing experience, and I wanted to blog a little about what it’s been like.
The 2014 Technical Plenary/Advisory Committee (TPAC) meeting was held in Santa Clara, and I attended with the intention of identifying a working group to join. There were several on my list of candidate groups, so I began popping into the sessions for them, to see which would strike a chord with me. The second working group meeting I attended was the Data on the Web Best Practices one, and I found myself so intrigued by the discussion there that I forgot to go to the next one on my list. The meetings are generally set up so that the working group members are seated at tables in the center and observers sit in chairs against the wall. The working group is expected to conduct its business and the observers are expected to simply observe. The DWBP group was interested in who the observers were, and they asked observers to introduce themselves. When I told them that I was from LBNL, they immediately encouraged me to take part in the proceedings. They were hoping to get more input from the scientific world about their specs. I was asked to take a seat at the table right then. I ended up joining the group that day and have been involved heavily in development of their (now our) primary deliverable, the Best Practices document.
The Best Practices document will be a W3C Recommendation (which means it’s an official W3C-endorsed specification) when complete. We are collecting best practices for publishing data on the web, attempting to address issues that prevent data sharing. The working group has members from around the world, but a large number are from Brazil. Another large subset is from Europe. Only two of us are from the U.S., so we end up doing early-morning conference calls, every week, year-round to move the project forward. We wake up just before 6:00 am, call in to the MIT WebEx service, open an IRC client, and begin conversing with friends from across the world. It’s still somewhat thrilling to me to have that opportunity, despite the inconvenience of the hour. The working group also organizes face-to-face meetings twice a year. I was able to attend one in Austin, but had to skip the ones in Sao Paolo and Zagreb. Still, the face-to-face meetings always incorporate WebEx and IRC, so I’ve been able to participate in all of them. (IRC is used in an interesting way, to scribe notes about what is said and to manage a queue for speaking.)
The DWBP group is also working on two vocabulary documents, one that is the Dataset Usage Vocabulary, and one that is the Data Quality Vocabulary. The former is meant to create a means to describe instances of using a dataset, to help in tracking provenance and enabling dialog between dataset publishers and consumers. The latter vocabulary is for describing the quality of a dataset in the many dimensions involved. Neither the DUV folks nor the DQV folks are charged with creating new technology, but their work lays the foundation for future browser capabilities.
Over all, I’ve been most impressed by the openness of the process. Comments are always taken seriously, and working group members strive to address every comment made, as resolving each one is part of the W3C Process (yes, it has a capital P). At times the work can seem too mired in the process itself, but there is a history of good reasons for the method of working, and it ensures that all of us are heard. Much of the work is done via email between calls, and those emails are public, as are their archives. It’s an impressively inclusive organization peopled by techies who care. Thus far, I’ve felt fortunate to be among them. I’m happy to talk more about my experience to anyone at the Lab who is interested in participating. It’s a lot of work, but I’ve found it to be highly rewarding, and we’re not even done yet.
And if you’re so inclined, take a look at the Best Practices doc, http://w3c.github.io/dwbp/bp.html. Comments welcome!