Haskell Foundation 2026 Update

Haskell Foundation
1

The Haskell Foundation has been at the center of a flurry of changes, all of which I’d like to share with the community today.

First, José, our executive director for the past few years, has decided to step away from his role in June 2026. José has been the longest-serving executive director at the Haskell Foundation, spending his time interacting with and supporting the community in many ways, often away from public view. Through personal sacrifices, José has allowed the Haskell Foundation to survive some tough times; today, the Foundation is in good shape. The Board of Directors, and I in my personal capacity, are deeply grateful for José’s work and dedication over the years. We wish him the best of luck for the future.

Second, the Board of Directors has decided to take this opportunity to restructure the Haskell Foundation. The principal goal is to dedicate most of the Foundation’s financial resources to technical work. The Foundation wants its members to be able to see a clear causal connection between their contributions and a series of improvements in the Haskell ecosystem. We want our members to have a voice in which projects we undertake, and thereby to take on a much greater sense of ownership of the health of our shared ecosystem.

Our use of the word member, rather than donor or sponsor, reflects a deliberate shift in tone. We want to build a greater sense of partnership and agency among our members. Rather than merely throwing money over the wall, we want membership to feel engaged in an inspiring shared mission, and that by sharing our resources we can accomplish much more than we can accomplish individually.

A new Haskell Foundation committee will direct resources toward a unified technical vision, and wield most of the Foundation’s resources to make this vision a reality. This plays to the strengths of the Haskell Foundation’s Board and collaborators, and will also include strong representation from the members of the Foundation.

To make this possible, the Haskell Foundation will remain without an executive director for the foreseeable future. Instead of having a full-time employee in charge of fundraising, events, coordination, mediation and much more, we will split these responsibilities between the Board and a new, part-time role dedicated to the Haskell Foundation’s financial sustainability.

We will share more concrete updates in the coming weeks and months.

Finally, the Haskell Foundation’s Board of Directors changed its composition last month. Departing directors are Andres Löh, former Chair, as well as Hazel Weakly and Josh Meredith. The Board warmly thanks Andres, Hazel, and Josh for their service and support over the past three years. The Board has elected Dominik Schrempf and Simon Marlow to join as new directors.

Expect more regular communications over the near term!

Laurent P. René de Cotret, Chair
on behalf of the Haskell Foundation Board of Directors

52 Likes
After 7 years in production, Scarf has reluctantly moved away from Haskell
2

So to see Jose step away, but I’m glad that we worked together.

Thanks a lot, I look forward for the next steps of HF!

3 Likes
3

I also want to thank @jmct for all his excellent work and juggling the many challenges that the HF brings with it, allowing us to survive for this long. :heart:

At the same time, the proposed organziational structure makes me somewhat skeptical. From what I understand:

  • there will be a half-time employee looking for funding
  • infrastructure will be done by volunteers
  • outward communication will be done by volunteers
  • events will be done by volunteers
  • maintaining a vision will be done by volunteers (a new committee it seems)

How many of those volunteers will have to be recruited from the board itself?

It’s also not clear to me how that new commitee will relate to the TWG.

12 Likes
4

Thanks @hasufell :slight_smile:

Part of the restructuring is indeed that the board will be taking on more work. This is something we’ve been working through internally, figuring out who is going to do which aspects of the ED role not covered by the new part time hire.

As for TWG, this will replace it. We made this plan after consulting with the TWG and trying to figure out what works and what didn’t. There was no objection to the new plan from the TWG and many of them are going to take part in this new framework.

17 Likes
5

Thanks @jmct, @kosmikus and Josh , it’s been a real pleasure meeting and working with you.

Congratulations to the new directors and I wish you all the best of luck :slight_smile:

5 Likes
6

@jmct It’s been a delight to serve the Foundation and the community under your stewardship, and I hope you get the rest that you deserve. :slight_smile:

3 Likes
7

Thanks for everything @jmct!

1 Like
8

Hello,

I feel I have to say something here. But, I will first acknowledge my bias, given that I applied for the ED role at the same time as Jose (and was unsuccessful; cleary :sweat_smile:.)

I think this is a serious mistake. The Haskell Foundation should absolutely not focus on technical work; it should focus on community work. The Haskell Ecosystem should certainly continue doing deep technical work (and other work as they see fit/supported), but merely electing to have no leader is doing a huge disservice to the community.

My perspective is over a few areas:

  • “Dedoption”: Organisations are moving away from Haskell. Everyone who is applying for work in Haskell out there knows it. Yes, we might still have good attendance at certain conferences; but many other languages are taking over our space, and we are doing nothing about it.

  • Funding: I haven’t seen the Haskell Foundation get any new funds, or anything innovative (NLnet or others). I have seen rich private funders funding other ecosystems; I haven’t seen anyone new fund the foundation. If I look at the list of funders now, it’s the same that it has always been. I think at least part of the reason for this is that we as a community are not offering anything new; and certainly “further technical work” is going to be of interest to exactly no (new) funders.

  • Missing the AI boat: Individuals should make their own AI choices; but to me it’s a huge missed opportunity not to talk about the extremely useful benefits of Haskelling with AI; the rich types and strict boundaries are extremely useful; best-in-class testing over basically any language is wildly undersold and undervalued.

  • Inability to engage non-Haskell community: For those of us who are extremely passionate about Haskell, I think we can all agree that it’s baffling the inability of our community to connect kindly with others. Ask almost any non-Haskell user what they think of Haskell and they say the same thing: confusing, too hard, difficult community. I find this at odds with almost all of the Haskellers I know: kind, lovely, thoughtful people who care about connecting with others. What’s going on here? How can this be changed? What initiatives have been tried? What’s next?

  • Lack of non-technical vision for the community. Yes, it’s a given that our technical achievements are a huge part of our work. Yes, technical improvement to Haskell is essential and highly valued. But, it’s not the only thing we should be doing. We need to focus on the other aspects of life that people care about: people, communities, newcomers. What are people using Haskell for? Can we influence that? In which direction should we push?

I hope that the Haskell Foundation can find a way to address these concerns; otherwise I personally am certain that Haskell will fade away; and to me that is a great shame; precisely because of all the extremely useful technical benefits that come from using it!

The only way to ensure Haskell’s continued adoption is to push past the Foundation focusing on technical work; you need to connect with someone who has a vision for Haskell (maybe different from mine) and pushes that forward.

It’s time to stop talking only to each other and start to expand out; improve our reach, improve our listening, improve our connection. This will improve our language. This will improve our community. And this then, maybe, will have a positive impact on other communities; strengthen bonds, and grow together.

It’s time to focus on more than the technical.

24 Likes
9

Hi Noon, thanks for posting your thoughts!

(The following is my own personal opinion and not necessarily representative of the board.)

I’d love for you to suggest specific kinds of community work could be funded by the HF. Most of HF funds in the past few years were spent precisely on non-technical work, either directly, or indirectly via the director’s compensation: communications, event planning, coordination between separate groups that are governed independently, learning tools such as the Haskell guide to performance, etc.

During that time, and even today, some people suggested focusing more on outreach and evangelism. I think this remains important, and the way I’ve seen it work in practice is events. For example, at our first Amerihac event this past February, a sizeable portion of attendees were neophytes.
If you have specific suggestions on this front, I’d love to hear them. I’ve proposed the creation of something like the Rust Book but for Haskell, for example, but reception was lukewarm. What else can the HF do on the community front?

Now, the lifeblood of the HF is funding, and the goal of this funding is to be spent to make the community stronger – whether that’s for technical, or non-technical work. You touched on this: the sponsors of the Foundation have been a pretty stable set of companies over the past few years, and even if this set, contributions are heavily skewed towards a few large sponsors.

We’d like to expand the set of sponsors! Some of our sponsors support us because they feel it’s the right thing to do. However, we keep ending up in the same conversation with prospective sponsors: “what do I get in exchange for the money”? That’s where the focus on technical work comes in. The HF is uniquely positioned to address community pain points that don’t necessarily make sense for a single sponsor to tackle (e.g. the work on Botan, or the Cabal exact-print parser).

One more thing I’d like to add: no executive director does not mean that the HF is leaderless, and it certainly does not mean that the executive director’s duties are no longer performed. The HF board has become more active to cover executive director duties, in addition to new efforts that we’re trying out. The board is also quicker to delegate authority to make decisions faster and more coherent (since far fewer people are involved).

Finally, this is an experiment. The HF is trying something out amidst a brutal fundraising environment – for all software foundations. If you, Noon, or others, have ideas, please do not hesitate to share them either publicly or in private with me.

8 Likes
10

Thanks for the response @LaurentRDC

Let me just address one point in particular:

While I think that some of the sponsors maybe get value from the HF directing some money to specific tasks, I think by and large if any of those organisations cared deeply about some technical thing, they would fund it themselves (and indeed, many of them do, hiring many lucky Haskellers!)

But I think most funding is mission/vision aligned; i.e. see this post from mitchellh regarding his family’s donation to Zig: https://mitchellh.com/writing/zig-donation-2026 .

My point is that (in my opinion) Haskell is sorely missing an inspiring vision and a leadership to carry it out; and this makes getting funding hard.

1 Like
11

My point is that (in my opinion) Haskell is sorely missing an inspiring vision and a leadership to carry it out; and this makes getting funding hard.

I have spent quite a bit of time over that last 18 months talking directly to companies who use Haskell, most of whom are not directly contributing (in a financial way) to the Foundation. My conclusion is different to yours. Mostly what I heard is

  • I personally am totally signed up to contributing to the health of the Haskell ecosystem
  • Moreover my company depends, in a mission critical way, on that ecosystem.
  • And yet I am unable to justify any direct contribution to the Foundation, to my Chief Financial Officer.

I have heard this over and over again; it’s not the lack of vision. Rather, the only way for a compan to make a sustainable contribution to the Haskell ecosystem is if the company perceives such support as buying them something of value. (By “sustainable” I mean a commitment that is not reliant on a single individual who may move on.)

I have considerable sympathy with this position. CFOs want their companies to succeed, are deeply relunctant to spend any dollars that they don’t have to spend. Individuals, such as Mitchell Hashimoto, are motivated by vision; companies not so much!

I think by and large if any of those organisations cared deeply about some technical thing, they would fund it themselves

For some technical things that is absolutely true; but for many it is not. It may well be that a company would gladly pay 10% of the cost of some improvement in the ecosystem, but would absolutely not pay 100%. The HF’s USP here is that we can get more done together than we can individually.

As Laurent says, it’s an experiment. If we commit to spending a large fraction of HF-member contributions on making their lives better every day, perhaps that will make it easier (maybe even a lot easier) to persuade that CFO. We’ll have to see!

Simon

9 Likes
12

My view is that this is just a classic (marketing) falacy; (and an in general tech-company founder problem: focus on the technical and the rest will follow; related in general to the mistake of selling the features rather than the feeling.)

Yes perhaps it’s true that rich individuals give money for different reasons; but I think that in reality sponsorship monies rarely comes from a desire to pay for part of a mild improvement in compilation times; that’s not going to convince XXX to give 100k. They need to be convinced because they see that the Haskell community is so technically strong; on such a great trajectory; consistently delivering on (self-determined) technical goals, and with such a strong vision that they want to see it succeed (and continuing using it).

Don’t get me wrong - I don’t think that the HF should have no technical work coordination component; but it absolutely cannot be the only thing it does; if it wants to continue to exist.

(Note: I’m not saying it’s easy to get sponsorship; nor that it should be the only source of funding; but I am saying that a lack of vision and leadership makes it much much harder.)

4 Likes
13

In this update, the HF is committing to spending a majority part of its financial resources to technical work, but, as you suggest, this is not the only thing it can (and should) do with its financial resources.

I personally think that events are another great way for the Foundation to spend its money, but these events end up either being sponsored (e.g. Amerihac being possible thanks for the generous hosting from Jane Street) or self-funding (e.g. the Haskell Ecosystem Workshop is funded by ticket sales).

In short: happy to hear about other ways the HF can support the community in non-technical ways, because there is space in the budget

4 Likes
14

The cases of dedoption I’ve heard e.g @avi had a lot to do with technical gaps (no libraries for common tasks, long compile times that affect CI and deployment velocity, spotty telemetry). If dedoption is a primary concern then focusing on technical gaps helps. Getting new users (outside of the Haskell community) is hard for various reasons but I think having a strong (primarily technical) vision is an important prerequisite. I’ve been doing a bunch of dev rel type work for DataHaskell. It’s easy to pique people’s interest but retaining users is almost always an ecosystem thing. I’ll do a whole vision pitch and show some examples at local python meetups and eventually people will ask about big query (or some other data lake/file format) support, AI tooling integration. After enough caveats they don’t come back. Plus a broadly technical direction has an easier story around milestones. Vision is hard to operationalize.

Edit: the Julia ecosystem concerned itself with time to first plot for the longest time. I think that was a good choice.

8 Likes
15

People these days don’t really want to learn anymore. But that is just my perception. The days of intense IRC conversations in #haskell are over. The Matrix channel is semi-dead and this discourse forum only has occasional sparks of interesting conversations about actual Haskell.

Open source is in decline, collaboration is in decline. There’s many reasons, but there’s also an elephant in the room.

I’ll be honest. When I learned Haskell, I was in university… it was close before summer break and one of my online friends who I met in IRC through the gentoo community was learning this weird new language that would revolutionize all of software engineering.

I was upset and said “no way you gonna do this without me”, so I spent 6 weeks during summer break teaching myself Haskell via the famous CIS 194 by Brent Yorgey course and getting real-time support in the IRC #haskell channel. This was exciting. There was learning material, my friend was motivating me and people were mentoring me. I was doing something that seemed meaningful: learning things and building knowledge about something that might be the next stepping stone in software engineering.

If I was in university today… I cannot imagine that anything of this would have happened. Human collaboration is at an all-time low, no one is going off with esoteric languages in the age of LLMs to prove something. I don’t hear “I tried this amazing new library yesterday” anymore. It’s more like “I asked Claude yesterday” or “I built a website in 5 minutes with Claude”. It’s boring as hell. The magic is gone. No one is mentoring anymore… I can just have Claude as my pair programmer, teacher, reviewer, colleague…

I’m sure I would not have learned Haskell if I grew up with agentic coding, no matter the typesafety.

And this is exactly where the difference is in vision: Zig has this “non-technical” vision… they believe in mentoring, learning and making each other better programmers. And interestingly, they decided that banning LLMs is the way to maintain this vision.

Does Haskell have to do the same? Not really. I can’t link to the issue on the GHC tracker, because gitlab is broken again… but there was a discussion where someone said:

I think we should aim to build an ecosystem where nobody feels tempted to use an LLM in the first place.

I think this is quite an interesting proposition. It doesn’t take sides in the yes/no discussion, but it formalizes the goals very clearly and is very actionable.

E.g. building an equivalent of the Rust Book would be a stepping stone towards this:

  • yes, the material would end up in the training data and improve LLM output
  • yes, it also means I don’t actually have to use an LLM to get basic information, because it’s well organized now

This is also my experience. Whenever I would reach for Claude in frustration, it was because the thing I’m dealing with is a mess, chaotic, underdocumented, full of pitfalls and simply inaccessible. Software engineering is exposed: it’s an utter mess and barely an engineering science. We failed. But there must be an alternative reality where we suceeded.

To me, Haskell is about maintaining that spark, that excellence, that adventure.

Maybe the only way to survive is to keep both sides engaged. And this seems like one approach.

32 Likes
16

I think it is correct to note that many students do cheat, we must refrain from being overly pessimistic. I hang around in the Discord “server” of my alma mater’s computer science students’ society, and there’s still an active channel of side projects, people learning modern languages, etc. While Haskell specifically isn’t the attractor it used to be, the wave of newer languages gaining popularity has been reflected in an uptick of “recreational PLT” in that student community.

While there are a lot of students hitting the LLMs to cheat, there are a also a fair number of students trying to do the right thing, still trying to learn how to learn, and then learn the material properly. A couple of days ago I stumbled upon this thread by a professor at Dartmouth, which showcases some of those student responses:

9 Likes
17

Not to derail this thread, but since “something like the Rust book” came up a few times now: I’d actually like that. One of the annoyances of learning Haskell was that I had to continuously find the right blog posts and comments that explained things at a level I could understand at that time. It’s hard to find these in the correct order, plus you’re reading 12 year old reddit threads and have no idea why things don’t compile anymore, some of the suggested readings will 404, etc. An official guide/book could mitigate that.

If it is of any help, from the top of my head these were resources that particularly stuck with me that I remember were helpful:

  • Learn You A Haskell was a good starting point (I still had no idea how to write an actual application afterwards though and some of it seems quite contrived)
  • tsoding’s “JSON Parser 100% From Scratch in Haskell” to grasp how Parsers in Haskell work
  • @lsmor 's snake-fury to get a feeling for monad transformers vs. mtl vs. mtl with type classes
  • Rebecca Skinner’s “Effective Haskell” didn’t exist at that time, I read it later and loved it, but I don’t know how a beginner would see it. It seems entry-level enough though.

I’m still very much a Haskell beginner in a sense so take this with a grain of salt.

19 Likes
18

Precisely someone we should listen to on this matter!

3 Likes
19

Good to know someone benefited from that resource. I’ve been thinking about updating it and add docker/notebook support so learners can jump into it more directly

1 Like
20

I have no idea how to popularize Haskell, the diagnosis/prognosis behind its unpopularity, or its future.

Academically, Haskell seems to be thriving using Google Scholar as a metric. There are lots of papers where Haskell or a Haskell-like language is used in technical expositions.

One one hand, I guess where do you practicably start in promoting a functional programming language that’s generally opaque at first glance to imperative coders. And, thereafter all the idioms and cognates that an imperative programmer fought tooth and nail to understand have evaporated.

On the other, due a lack of network effects, there will always be friction when it comes to using Haskell in production. Good production idioms for Haskell to either avoid or mitigate some of its pitfalls are not common. And so, you’ll be forced to learn them. Some teams are adverse to learning new things when they know and have done all those things in say Python, C/C++, Java, Javascript, etc.–and that’s perfectly understandable.

And yet, Haskell, out of all the FP languages locked away in academia has escaped. It’s surviving and can be (and is) used in production provided you have the right mindset and perseverance to do so, When I learned Haskell, cabal hell was a thing, there was no HF, and not many useful Hackage packages for production, It’s far better today than ever before. I believe that as long as HF and organizations (human beings) care and contribute to Haskell’s ecosystem (DataHaskell is certainly one), Haskell will survive. It has always been about people. And, it will always be despite what some CEOs say.

I embrace “avoid success at all costs.” Perhaps, we’re approaching that “success” given some of the ad-hoc GHC proposals that are incremental or cosmetic at best.

I believe that should Haskell die, FP communities will never die. Perhaps, it’ll be Lean or Agda/Roqc to C or Javascript or native–or some spanking new FP language.

10 Likes