"This pull request is brought to you by..."

All of my grant funded projects involve development of some sort, and many benefit from existing open source projects.  Sometimes, I have the opportunity to contribute back to these projects by creating documentation, or submitting a pull request.

Whenever I get a chance to contribute back to an existing open source project in this way, I often feel like I should be acknowledging the grant(s) that I'm funded under which let me do that work.  It not only helps funding agencies see that the grant they supported made a bigger contribution outside of the aims of the project, but someone reading the pull request can realize "Oh cool, my tax dollars going to the NIH in turn helped support this bug fix."

How to do this appropriately is what I'm trying to figure out (and hoping to get feedback on).  I don't think we need to overdo it - there doesn't need to be great fanfare for the little one-liner documentation change I made.  Having a one paragraph stock grant acknowledgement text that's longer than the lines of code changed may seem overkill (and could clog the git history).  But maybe just putting a brief acknowledgment of the grant in the pull request comments is a non-intrusive way to raise awareness.  I'm sure lots of people will miss it, but if someone wants to find it, it's there.

On the other hand, even though I'm doing this change for a grant-funded project, it doesn't mean that my funding agency officially supports the open source project I'm contributing back to, and doesn't necessarily endorse that project or my work.  And what if the new feature I'm trying to submit has a bug in it?  It should make me look bad, but I wouldn't want my funding agency to look bad too.  We can cover that often with disclaimers "... this work not endorsed by [agency] ...", which could be included with a pull request comment.

The other aspect of this is impact/metrics.  We could go back and track the number of pull requests made to a project when reporting to our funding agency.  To me, doing this at the pull request level instead of the commit level makes more sense - often times there are >1 commit per pull request.  I think we realize that if we get graded on those metrics, they will be abused ("oooh, I'll just submit one PR per commit!"), so weighing that would need to be handled with care.

Similar discussions have been happening around data sets and ensuring that people get credit for those.  I'm not necessarily saying that pull requests should be counted at the same level of a data set - different amounts of work go into each - but this may be a reasonable option for software developers in the scientific field to track output and impact.

Comments

Post a Comment

Popular posts from this blog

My ideal resume

As I'm getting ready for a yearly performance review, I started thinking about when I applied for this and previous jobs and all of the effort I put into building the "perfect" resume. Tweaking it to have the right keywords for the job description, making sure I didn't overstate or understate my effort, but making sure whoever read this knew I was the best candidate. I started thinking about what that really translates to at my job, and what skills are practical and important each day. I came up with this as my new resume (no, not sending it out, very happy at my current job but thanks). Of course I realize sometimes you have to "play the game", but tell me this isn't what you're really looking for in an employee: I know I need to keep learning, and so I do.  I can (and do) listen better than I can speak.  I won't be a jerk to anyone in the organization, no matter their role.  I will come in to work and genuinely want to be there. I&#
Read more

Chicago Informatics Week

Happy Informatics Week , Chicago! I must say I'm particularly pleased that today through November 7th we will have the unique opportunity to have the field of informatics not only "officially" recognized by our mayor, but to have it done so in a way that will provide more exposure to what informatics is and what impact it has made in healthcare. For those of you in the Chicago area who may not be aware of the events, you can find a list at: http://chicagoinformaticsweek.org/events.html This, of course, is right around the time of the AMIA Annual Symposium, and following will be the CTSA Informatics meeting.  I'm excited to be heading an AMIA panel on the changing face of security and privacy with respect to the role that "honest brokers" play (session S69, Tuesday from 3:30-5:00), and a poster describing recent work to integrate a research registry recruitment system into a commercial EHR vendor's patient portal. I'm looking forward to catch
Read more

Brought to you by the letter "B"

While poor time planning may be mostly to blame, I have also been quite busy the past few months with the new addition of a baby boy (our first child) to our household. Walking through the medical system in the context of delivery, follow-up and now pediatric care has given me some insight into those parts of healthcare. Previously my only personal interactions were the occasional office visit for check-ups. With no real complaints, I was more amazed at the amount of data being collected in the time prior to delivery, and the realization that my son has his own medical record number! At home, my wife brought to light a potential area for growth in newborn care - driven more out of need than any formal proposal. Tracking feedings and diaper changes was something done at the hospital, and once we were home the feeding chart became a very valuable flowsheet of information. Currently it's been done on paper (don't worry, I have plans to develop an app!), but has given our sl
Read more