Open matthewfeickert opened 2 months ago
I talked to Allison and we are eventually planning to add some data visualizations to the OSPO website. These would be live / dynamic, running from a database back end so they would update with whatever frequency we decide to re-scrape the data from the GitHub / GitLab apis.
One open / unresolved question is where to host the server code and database. I dumped it onto my own server since it didn't seem appropriate to put it on a CCI server or the maps.datascience.wisc.edu server.
I wish that we had a common server for DSI to put web content. It would be nice to have one common server with subdirectories for each of the DSI staff so you could just upload content to something like: http://datascience.wisc.edu/staff/mfeikert or perhaps at a subdomain like: http://portal.datascience.wisc.edu/mfeikert
I could potentially request a shared hosting server and subdomain from DoIT like I did for http://maps.datascience.wisc.edu if DSI would like to have something like this as a shared resource.
Another approach might be to set up a CCI server which would give us more control so we could set up protected user accounts for each of the DSI staff. One way to do this might be to set the web server document root to the users directory.
Thoughts?
@amegahed made some nice visualizations of the data using the API he created like https://datascience.sharedigm.com/open-source/charts/github.html.
These are nice, but this also requires knowing where these visualizations live in advance. Is it possible to make these discoverable from a central website or to have them all be listed? If the idea is not to have a collection of visualizations pre-done, but to have the API be something that users interact with, is it possible to make a web interface to perform the query and get a quick visualization back?