This is an old post and is probably extremely cringe. Please understand that I have moved on from these ideas. Still, it may contain some nuggets that point to some continuity in my thinking over the years, which is why I decided to post it here.
Git as I go
Over the past few weeks I’ve been getting acquainted with git, a version control mechanism that allows users to share code, text or other editable media in a collaborative environment. I’m learning that Github is a great venue for hosting text-based datasets and disseminating preliminary results in a comfortable way.
I’ve always been an advocate for open access to data and I had considered various options for storing and disseminating the database that I am constructing for my Master’s thesis. I really like text-based solutions since they are easy to modify across virtually any platform, they are easily convertible to more complex file formats, and there is a general air of simplicity when working with them. (I now use TextEdit as my primary word processor, switching over to word only whenever I need to do some advanced formatting or track changes.) Now, while I am using FileMaker to enter my data and Excel as a file-conversion intermediary, I am working on transforming my dataset to JSON and its spinoff dialects, geoJSON and bibJSON in particular. JSON is readable by both humans and computers, enabling easy modifications and additions of new data, with versatile ways of displaying and disseminating the information being stored.
Github makes it very easy to read JSON and many other text-based file formats, including tabular data files. Github has even integrated the Leaflet mapping tool so geoJSON data can be perused cartographically. I also like Github because it generates stable links for easy sharing, my data can be used by others with similar or widely different goals, and the effort that I put into the collation of this data can be fully recognized, and even expanded upon. However, Github was never intended to be a database-sharing platform and there are some limitations for this kind of use. It would be great if bibliographic entries could be generated so others can cite my work appropriately. Additionally, an option for some form of storage designed for longevity would better suit the goals of academics. Maybe there will be a push in this direction now that PLOS has updated their open data policy.
I’ve also started to push some of my preliminary analyses to Github as I work on them. Working transparently definitely impacts my mindset since I am encouraged to produce more meaningful results, stay organized, and participate with others doing similar things. I get to show off a bit too, extending my visibility to people at a distance. However saying something meaningful about the results I obtain is the other side of the coin that I must not forget about.
In general, I think this is a good direction for me. It feels great to get my work out there, even in an incomplete state.