FHIR Chat · OpenAPS: 10th most forked · social

Stream: social

Topic: OpenAPS: 10th most forked

---

Dave deBronkart (Sep 20 2019 at 22:51):

Well HERE's a tweet that answers "What would patients DO with data, if they had it?"
https://twitter.com/sulka/status/1174412175106072577

Dave deBronkart (Sep 20 2019 at 22:52):

(I'm the first to admit I don't literally know the detailed significance of this, since I'm not a programmer. But the coders I know all pop their eyes and go "Whoa, awesome" at news like this, so ... enjoy?)

Dave deBronkart (Sep 20 2019 at 22:53):

Links for those who don't know the space:

• OpenAPS - home of the Open Artificial Pancreas System (open source)
• NightScout (Wikipedia) -

Nightscout is a free and open-source project, and associated social movement, that enables accessing and working with continuous glucose monitor data. Nightscout software aims to give users access to their real time blood sugar data by putting this data in the cloud.

• github.com/nightscout

Dave deBronkart (Sep 20 2019 at 23:04):

I've learned that you "fork" a repository when you grab a copy to make (or propose) a change.

And now somebody please explain to me why being a "most forked" is cool. Does this just indicate the hot, wild level of engagement people have with the subject? Like, a zillion people tweaking it? (Or 32,167 in this case?)

Grahame Grieve (Sep 21 2019 at 01:59):

Yes it’s an indication of people being genuinely involved. 10th most forked? - that’s astonishing!

Lloyd McKenzie (Sep 21 2019 at 02:51):

It means that many people had both the interest and capability to either contribute changes or to create their own adjusted version. Given that the 'capability' isn't super common, it suggests that the interest/use is quite large indeed.

Lloyd McKenzie (Sep 21 2019 at 03:07):

As a comparison, the FHIR Java reference library (which is the foundation of a lot of FHIR production implementations) only has 17 forks...

Dave deBronkart (Sep 21 2019 at 13:04):

Many observations can be made, I guess.

First, it seems to be a spectacular specimen of patient engagement: "Give us a chance, and you might be AMAZED what we might do." Note, this is engagement in the sense of patients being actively involved in their care - not engagement in the sense of "following the doc's instructions" or "buying their pills as instructed, eating them, and buying more."

Second, it's worth noting that when it's life-and-death, patients are infinitely more motivated, intrinsically, than any vendor. (Remember that the initiating hashtag was #WeAreNotWaiting, after decades of industry not producing.

Third, the above make me reflect on the legal responsibility aspects. If I want to sell a product to people with an illness, I need to be damned sure it won't kill anyone, which can be really hard to be certain about. But when the patients themselves do it, NOT to sell but for self-care, there's no commercial transaction - it's just self-help, shared freely.

This is the point in Eric von Hippel's work at Sloan School about "Free Innovation" in any industry, including his article this year When Patients Become Innovators. Dana is of course the lead specimen. Another is Sean Ahrens, Crohn's Disease patient who has created the Crohnology community. And they touch on patient-driven clinical trials.

Mikael Rinnetmäki (Sep 21 2019 at 18:53):

To be honest, that's partly due to how Nightscout was originally designed, and how the instructions for getting it to use were crafted. You needed to make your own copy (fork) of the Github repository and then deploy that copy.

It's not mandatory to do that anymore, but it still explains a majority of the number of forks.

So I'd say (and @Sulka Haro has said it often as well) that the number of Github forks a bit of a vanity metric. But a cool one. And the community is definitely great, even more awesome than that metric lets you understand. :)

Grahame Grieve (Sep 21 2019 at 19:10):

@Sulka Haro congrats then. you have an account here?

Mikael Rinnetmäki (Sep 21 2019 at 19:14):

(You have been 'introduced' in topic https://chat.fhir.org/#narrow/stream/179166-implementers/topic/Encoding.20artificial.20pancreas.20decision.20data/near/168116583...) ;-)

Grahame Grieve (Sep 21 2019 at 19:15):

ahah. I did not make the connection

Sulka Haro (Sep 21 2019 at 19:35):

Yup here! Feel free to ask if you have any questions. :)

Virginia Lorenzi (Sep 22 2019 at 01:42):

honored to meet you :)

Michael Lawley (Sep 22 2019 at 02:01):

I can say I know 2nd hand the significance of this (my wife has T1D and OpenAPS, Nightscout (& Loop & AndroidAPS)) has changed her life. And it's not just that managing blood glucose levels is now a tractable task, it's the associated mental health burden that's been lifted.

Dave deBronkart (Sep 29 2019 at 02:13):

This is a superbly important anecdote for the patient empowerment community in HL7 to note. No medical device vendor would set out to create a product whose goal was to eliminate a mental health burden. But when patients innovate on their own, they can pursue any goal they want.

This is autonomy when your whole life is at stake. I can't emphasize this enough.

Dave deBronkart (Oct 05 2019 at 13:35):

To be honest, that's partly due to how Nightscout was originally designed, and how the instructions for getting it to use were crafted. You needed to make your own copy (fork) of the Github repository and then deploy that copy.

I finally realized why that comment back on 21 Sept struck me both as "oh.... I guess that makes sense" and "Still that's AMAZING." Because ... how many other health management tools have had thousands of personalised tweaks by individual users??

Dave deBronkart (Oct 05 2019 at 13:38):

Another thing about OpenAPS that continues to erode the paradigm is that when you have n of 1, all of the research controversy about p values disappears from the equation: there is no longer any obligation to predict how well it will work for many future buyers, because there ain't no buyers AND because any tweak is solely for one user.

This is so vastly different from the business planning that today's device makers and drug makers use.

---

Last updated: Apr 12 2022 at 19:14 UTC