What happened to Matchmaker?

Haskell Foundation
#1

I’ve been reading the Haskell Interlude interview with @jmct and I’ve found the following passage somewhat surprising:

And actually, one of the things I’ve been surprised about is how many emails I get about that. So it’s people who are excited about Haskell, they love the language, they learned it in a class or whatever, and they’re like, “How do I help? What are things I can volunteer for?” And so part of my time is helping match them with projects that might be able to use that help

This reminded me of an early project of the Haskell Foundation, called Matchmaker, see also Seeking a Project Lead for Matchmaker. This was: a project of the Haskell Foundation to help open-source maintainers and contributors find each other, and provide a smoother experience for people wishing to invest themselves in the open-source Haskell ecosystem.

It looks like the project is dead (inactive for 3 years). The temporary substitute is also inactive (and a similar, independent initiative shared the same fate).

I wonder what happened to this project. Why didn’t it work out? Did somebody decide this is not the way to go? If so, why? Or maybe there wasn’t enough manpower to finish it? Maybe nobody wanted to be the project lead? Can anybody that was involved (@Kleidukos perhaps?) share some details?

5 Likes
#2

Hi, thanks for your interest. Ultimately yes it was a matter of workforce. Since I was volunteering on this, I had to disengage when life threw other priorities my way. I’d be more than happy to take up the project again with other people, however, because I know for a fact that volunteers looking for projects is still a relevant subject today.

1 Like
#3

It is my perception that this project went overboard with a technical implementation of an idea.

It could start much simpler with a github project and an appropriate process that allows people to ask for help and look for work (e.g. simply on the issue tracker).

5 Likes
#4

I guess that was this:

But perhaps it was lacking people with enough time to run the process.

3 Likes
#5

Or maybe the process wasn’t appropriate? I’m not saying that it wasn’t, but it’s an option.

Or maybe it lacked promotion/marketing?

#6

Thanks for the explanation! I was actually somewhat excited about the project when I first learned about it a few years ago.

I have not looked at the internals, but at first glance this project looks rather simple. It should be more accessible to newcomers than, say, Cabal or HLS. I guess that people with some web development experience that are at the “advanced beginner” level in Haskell could make substantial contributions. If a more experienced Haskeller or two could do some code review and guide the contributors a bit, maybe enough workforce can be obtained by directing the stream of volunteers towards this project?

But this needs a rather clear vision of the final product and some kind of a roadmap. Is any of these available? If so, are they written down somewhere?

Also, there is no guarantee that this will actually help the HF, the maintainers of various projects and the potential volunteers. I wonder if the idea behind this project and the requirements were ever discussed in public? I think that for the project to succeed it would have to satisfy the needs of these three groups (and possibly some others).

#7

I’d recommend to look at similar projects (are there any?) in other ecosystems. Learning what worked for them is likely to save grand amount of design effort and improve chances to succeed.

IMHO matchmaking is an art, not a market. Contributing to a random open-source project in a meaningful way is hard, and finding a project per se is the smallest of hurdles.

4 Likes
#8

Yes, it’s more about advertising your ideas and making contributors excited… similar to how the project pitches at ZuriHac work.

A good mix of technical and social incentive.

I’ve been struggling to do this for my projects:

  • GHCup isn’t much of a fun codebase and design and reviews have to be very meticulous
  • core libraries are even worse in terms of contribution experience (for a reason)

Casual contributors burn more of the maintainers time than they get out of it and yet I’m willing to do it.

Maybe there could be some sort of matchmaking approved template, where projects add a section to their own README and then get linked on a curated page.

That section would explain things like:

  • motivation and vision
  • contribution workflow
  • issues classified by difficulty level (and ready to be picked up)
  • contribution guide (tech overview and navigation help)

A bit more elaborate than volunteering/projects.md at master · haskellfoundation/volunteering · GitHub

The matchmaking project would then pre-evaluate projects contribution experience (based on documentation, structure etc) and set some sort of minimal standard. And actual contributors could report back their contribution experience.

It’s my opinion that many crucial projects in Haskell do not explicitly design or think about their contribution experience other than a bit of CI shenanigans. That may partly be, because they almost never get explicit feedback.

3 Likes