Why We Engage in FLOSS: Answers from Core Developers

#1

Post by eqcho » Sat, Jun 02, 2018, 11:09 AM UTC

Why We Engage in FLOSS: Answers from Core Developers
arxiv.org

Re: Why We Engage in FLOSS: Answers from Core Developers

#2

Post by bfirsh » Sat, Jun 02, 2018, 12:44 PM UTC

If you're on a phone, here's an HTML version of the article: https://www.arxiv-vanity.com/papers/1803.05741/

Re: Why We Engage in FLOSS: Answers from Core Developers

#3

Post by adamdennis » Sat, Jun 02, 2018, 1:04 PM UTC

bfirsh wrote: post #2

If you're on a phone, here's an HTML version of the article: https://www.arxiv-vanity.com/papers/1803.05741/

Thank you!

Re: Why We Engage in FLOSS: Answers from Core Developers

#4

Post by blattimwind » Sat, Jun 02, 2018, 1:19 PM UTC

I think the converse would be very interesting as well: Why do people dis-engage?

Re: Why We Engage in FLOSS: Answers from Core Developers

#5

Post by jgtrosh » Sat, Jun 02, 2018, 1:25 PM UTC

Figure 9.b seems to have a problem with sampling? Half numbers of core contributors are plotted even though they (thankfully) don't correspond to the samples.

Re: Why We Engage in FLOSS: Answers from Core Developers

#6

Post by blattimwind » Sat, Jun 02, 2018, 1:27 PM UTC

jgtrosh wrote: post #5

Figure 9.b seems to have a problem with sampling? Half numbers of core contributors are plotted even though they (thankfully) don't correspond to the samples.

No, the negative lobes just show how contributor numbers are whole numbers. Natural, even.

Re: Why We Engage in FLOSS: Answers from Core Developers

#7

Post by jgtrosh » Sat, Jun 02, 2018, 1:40 PM UTC

blattimwind wrote: post #6

jgtrosh wrote: post #5

Figure 9.b seems to have a problem with sampling? Half numbers of core contributors are plotted even though they (thankfully) don't correspond to the samples.

No, the negative lobes just show how contributor numbers are whole numbers. Natural, even.

Yes, but do we need that information? It could be the case of any integer graph. Either just plot the samples, or curve them. Don't curve them with intermediary values, that means even less than a trend curve.

Edit: okay maybe the violin plot as it is shown has the correct density information. I disregarded that aspect of the plot.

Re: Why We Engage in FLOSS: Answers from Core Developers

#8

Post by smaddox » Sat, Jun 02, 2018, 1:50 PM UTC

bfirsh wrote: post #2

If you're on a phone, here's an HTML version of the article: https://www.arxiv-vanity.com/papers/1803.05741/

Brilliant! I'm never reading an arxiv PDF again!

Re: Why We Engage in FLOSS: Answers from Core Developers

#9

Post by vjeux » Sat, Jun 02, 2018, 3:13 PM UTC

As a core contributor of some very popular github projects, I get a ton of emails from researchers wanting me to fill out surveys. It would be nice if it was somehow centralized instead of a lot of independent people trying to do the same work over and over.

Re: Why We Engage in FLOSS: Answers from Core Developers

#10

Post by KirinDave » Sat, Jun 02, 2018, 3:20 PM UTC

I've never understood it. I've only had horrific experiences with my open source efforts. From the simplest cases like being ignored on PRs and tickets to mockery in social circles for algorithmic mistakes to finding out my work was enabling bad people to finding out people were misrepresenting my employment status while using my work in violation of the license, I don't think I've ever once had a positive experience with open source as a producer. Hell, I never even knew my extensions to the clojure time library were even being used until another maintainer who had added some incremental stuff over my work disappeared.

I wish I could go back in time and tell my younger self it was all a waste of time and I shouldn't do it. Maybe then I could be more positive about it now that I'm better at writing software.