Hacker News new | past | comments | ask | show | jobs | submit login
The Myth of Mass Collaboration (staltz.com)
43 points by zdw on Aug 24, 2022 | hide | past | favorite | 17 comments



To me, this seems to be taking an unrsalistic definition of "collaborate" and then being surprised that nobody meets it.

Is there any other venue where the author's definition of "collaborate" rings true? If your definition of collaborate is everyone works equally on everything as if by committee - no large project works that way. Neither do most small projects.


His definition is what you would expect in most work settings.

For example, on a team of 5 individual contributors who were supposed to be peers, were paid the same, etc, if one of them was doing 80% of the work, most people would think something was wrong there.


20% of the people doing 80% of the work wouldn't actually be that surprising. The Pareto Principle applies to a lot of cases.


It would be surprising if someone that high-performing weren't promoted into a different role. My claim isn't that such differences don't exist among people... only that we don't say such a group is "collaborating", which (to me) typically connotes some roughly equal breakdown of work. We say that the group has a leader, or that it is "A's project, but X, Y, and Z are helping a little."

The OP's thesis is that the rhetoric around "tens of millions of people collaborating to produce OSS" is misleading, and I think his point is fair. Most people would be surprised to see how single-individual-heavy the actual production is.


I don't think anyone uses collaborating to mean "collaborating equally".

The idea that "tens of millions of people collaborating [exactly equally] to produce OSS" is so obviously not true, its hard to imagine anyone seriously believes that.


That doesn’t sound out of the ordinary. You’d expect them to get promoted, a large raise or switch to a better (paying) company but those kinds of differences in performance sound normal.


In a work setting not everyone is paid the same and not everyone provides the same value. In open source, people aren't paid, which isn't equivalent to being paid "equally". Some do more than others.

I feel like in a work setting, the equivalent would be saying it is not collaborative because senior employees are more productive than the interns.


In my prior job, I worked at a company that had a large monorepo with thousands of engineers contributing to it every week. I truly felt like I was part of a mass collaboration of people that were collectively defining what our product should be like and how it should operate.


"To collaborate on something is to work together with others towards a common goal."

Yes, but open source is a power-house of an opportunity to collaborate with 'no common goals'.

If you make really good 'indexing software' - well, someone else can make medical related AI much better.

Etc..


The point of the article was that this is not happening through collaboration. If I import a library you wrote in my project, I am not "collaborating" with you. And out of the hundreds of libraries that a typical project uses today, the project authors are unlikely to even report bugs in more than a handful. Contributing core features to even one of them (which would count as collaboration) is unlikely, and doing so for more than 2-3 is basically unheard of.

This is not an indictment of the actual FOSS model. It's just an indictment of the use of the word "mass collaboration" to refer to it.

The point is that the vast majority of projects are built by a small team, with very small contributions from a larger cloud of devs.

Exceptions are things like Linux or Kubernetes, where many teams from many places actually form a loose organization that actively collaborates on the project as a whole. Of course, even there it's not like a fully connected graph, but it's more like a typical software org, with different teams focusing on different areas of the code base, but team leads coordinating with other team leads at various levels, and all following a somewhat common vision.


It's easier to write code than to read and understand someone else's code. And you need to at least read some code to add to it.

It's hard to learn another language than use the language you know. (Both computer programming languages and real languages)

Mass collaboration could be our goal, if we can create software that is inherently understandable and malleable.

The problem is that every line of code has a maintenance and understandability burden.

If you can create a system where development is as easy as casual contribution such as editing a wiki page or writing a comment that would be awesome.


Nothing about this is specific to software; it happens everywhere. Wikipedia was mentioned in the article, as well as various other examples.


He is ignoring the black swans.

Before the invention of cities and companies, there was hardly any cooperation besides hunting together.

>(1) After playing this for 3 years, I’ve learned that you have to learn about your team mates styles, and this can only happen if you consistently play with the same folks.

There are TikTok algorithms to find content that people like. It should be equally possible to find teammates that are fun to play with.

The difference is that a teammate algorithm stops being useful once it is successful thus it is difficult to monetize.


If you’re not familiar with the author, they’re a creator of Manyverse, one of the most popular Secure Scuttlebutt clients.


I was hoping for some interesting analysis of how big projects like wikipedia and iNaturalist are often dominated by the contributions of surprisingly few individuals. Unfortunately this just seems to be thought piece about definitions washing up on the rocks of reality.


From the article.

>I’ve written about this before [1] and gave some examples from Wikipedia, YouTube, Mastodon, and Tor.

From that linked article.

>On YouTube, 2 billion MAUs but only 15 million (0.75%) active creators

>On Wikipedia, 39 million registered users but only 128 thousand (0.3%) active contributors

>On Mastodon, 1 million active users but only 2 thousand (0.2%) instances

>On Tor, 2.5 millions users but only 6 thousand (0.24%) relay servers

So unfortunately your comment just seems to be a thought piece from someone who didn't really read the article.

1. https://staltz.com/some-people-want-to-run-their-own-servers...


It also seems a bit unfair how the author dismisses people making small fixes. Its different work but still important work.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: