If you use Clojure for your business, please consider funding this effort and also directly developers who work on software that you use. It makes for a sustainable ecosystem.
The Clojure community is very mature and incredibly nice, so things are not bad as they are, but they could definitely be better.
I try to set aside a portion of my business revenue (I call it a "sustainability fee" in my P&L) and spread it among the authors of libraries that I use. It's not much for each author, but if everybody did this, many authors could work on open source libraries full time.
what needs funding is clojure core, Clojurists Together is a nice effort to try to fill a gap but what my business needs, and other business owners tell me they need, is for the core language (clojure and clojurescript) to not feel like it is falling apart. I want to invest my money in targeted, specific high value problems and get leverage on my money by sharing the cost with other business owners who need the same issues fixed. Until such a vector exists, the next best thing (for my business) is to fund a few key maintainers directly via github sponsorship. But because it is indirect and not outcomes-driven, the money amounts will be smaller. My business can afford to invest more than the $2400 per year that we have been donating since the release of Electric. My business has employed 4 devs for 5+ years, I can find more budget for investments in Clojure. Businesses pay for things (unlike individuals). We want to pay. We want our key dependencies to thrive. But there is no vector to invest in the specific core issues that would benefit my business to improve. And this makes me sad because over the last few years I find myself leaning more and more away from Clojure, even repelled (as if by some invisible force), instead of leaning in.
Well, you can fund individual developers on GitHub (like David Nolen), which I do. I would also gladly fund the core, although I do not feel it is "falling apart", quite the opposite in fact — I am very happy with what it provides and how it is maintained. My business is based on it.
Not sure this type of response helps and referring to someone who is active in the Clojure community and tries to build some cutting edge stuff as 'this guy'. (Did some digging and hyperfiddle looks pretty cool).
I think it would help to understand why he is talking about this 'every two months' and then try to address his concerns.
Would be a shame to lose people who are active in the community but not heard.
Of course. But for people not in the know and looking outside in, they don't know what's going, i.e. why is someone who's been using Clojure for years and building a business with it making these points every few months as you say?
To counteract FUD it would be useful to give a few more details.
Might be some personal beef, but I don't know really and don't care enough. He makes these authorative claims mixed with personal sentiment and then doesn't back them up.
Thanks for that. These details are important for newcomers, because it seems the original commenter isn't happy with the core team's process and 'feels like it's falling apart' is hyperbole and subjective.
For future readers: Clojure is known to be extremely stable and backwards compatible. It doesn't suffer from the same churn, fatigue and breaking changes that the JavaScript ecosystems experiences for example. To get to that backwards compatibility it requires a different process.
Given that software is composed of a hierarchy of dependencies, I would like to see a funding approach that works at the dependency tree level to support an entire tree or sub tree.
There is a huge freeloader problem where business don't contribute any support for their core dependencies.
I wonder if there is a role for an organization that could act as the interface for corporate support at the dependency tree level. It could offload maintainers (or fund them) to handle certain compliance requirements and provide an official sanctioned entity for purposes of corporate policies.
There should be a way to garner support broadly for risk management and specifically for security in the corporate context.
I wasn't aware of the Gloat project before this. It's a compiler that turns Clojure into native binaries by first transpiling to Glojure (which I'd also never heard of before this), which in turn targets Go. This is rather than using a GraalVM native image, which as I understand it is at this point the better-explored mechanism of doing that for JVM-based stuff (but has its own trade-offs).
Clojurist makes it sound like adherents to a belief, which does imply identity though Clojurian more directly so. I’ve also found Clojurian most often. If only we had software that indexed the popularity of words…
Neither of them is AI genetated software though. One is a fast Clojure LLM inference library, and the other a MCP server, both coded by humans, without AI.
If you use Clojure for your business, please consider funding this effort and also directly developers who work on software that you use. It makes for a sustainable ecosystem.
The Clojure community is very mature and incredibly nice, so things are not bad as they are, but they could definitely be better.
I try to set aside a portion of my business revenue (I call it a "sustainability fee" in my P&L) and spread it among the authors of libraries that I use. It's not much for each author, but if everybody did this, many authors could work on open source libraries full time.
what needs funding is clojure core, Clojurists Together is a nice effort to try to fill a gap but what my business needs, and other business owners tell me they need, is for the core language (clojure and clojurescript) to not feel like it is falling apart. I want to invest my money in targeted, specific high value problems and get leverage on my money by sharing the cost with other business owners who need the same issues fixed. Until such a vector exists, the next best thing (for my business) is to fund a few key maintainers directly via github sponsorship. But because it is indirect and not outcomes-driven, the money amounts will be smaller. My business can afford to invest more than the $2400 per year that we have been donating since the release of Electric. My business has employed 4 devs for 5+ years, I can find more budget for investments in Clojure. Businesses pay for things (unlike individuals). We want to pay. We want our key dependencies to thrive. But there is no vector to invest in the specific core issues that would benefit my business to improve. And this makes me sad because over the last few years I find myself leaning more and more away from Clojure, even repelled (as if by some invisible force), instead of leaning in.
Clojure core has always been highly protected, stable and curated
The downside of that is it's slow and difficult to get changes in
Instead of talking about throwing money to get the contributors to fix what you want it would be better to publicly talk about the problems
Are they tracked or known, what do they stop you from doing, how long have they been in the queue have you had any feedback etc?
Well, you can fund individual developers on GitHub (like David Nolen), which I do. I would also gladly fund the core, although I do not feel it is "falling apart", quite the opposite in fact — I am very happy with what it provides and how it is maintained. My business is based on it.
What do you mean by "falling apart"?
Clojure is the most stable and robust language I've ever used, and I've been using it in production for 15 years now.
Cognitect is a holding of NuBank (since 2020 I believe).
Doesn't that mean the core developers are now funded by NuBank?
Isn't it in their interest that 'the core language (Clojure and Clojurescript) to not feel like it is falling apart'?
And what do you mean with falling apart?
I'm new to Clojure, so the above is a genuine question.
There is no falling apart. This guy just tries to spread this narrative every two months or so, idky. Check out latest dev call https://www.youtube.com/watch?v=ngyvDkZA3o0
Not sure this type of response helps and referring to someone who is active in the Clojure community and tries to build some cutting edge stuff as 'this guy'. (Did some digging and hyperfiddle looks pretty cool).
I think it would help to understand why he is talking about this 'every two months' and then try to address his concerns.
Would be a shame to lose people who are active in the community but not heard.
Of course calling out baseless FUD is helpful. Otherwise it just discourages people from engaging with Clojure, which is not helpful.
Of course. But for people not in the know and looking outside in, they don't know what's going, i.e. why is someone who's been using Clojure for years and building a business with it making these points every few months as you say?
To counteract FUD it would be useful to give a few more details.
Might be some personal beef, but I don't know really and don't care enough. He makes these authorative claims mixed with personal sentiment and then doesn't back them up.
Here for instance the core team supposedly has "lost control of the compiler" https://x.com/dustingetz/status/1947643515480965346
Work on the compiler had been pushed a month earlier https://github.com/clojure/clojure/commit/6a3e0f02c6a96fcef2...
Thanks for that. These details are important for newcomers, because it seems the original commenter isn't happy with the core team's process and 'feels like it's falling apart' is hyperbole and subjective.
For future readers: Clojure is known to be extremely stable and backwards compatible. It doesn't suffer from the same churn, fatigue and breaking changes that the JavaScript ecosystems experiences for example. To get to that backwards compatibility it requires a different process.
Given that software is composed of a hierarchy of dependencies, I would like to see a funding approach that works at the dependency tree level to support an entire tree or sub tree. There is a huge freeloader problem where business don't contribute any support for their core dependencies. I wonder if there is a role for an organization that could act as the interface for corporate support at the dependency tree level. It could offload maintainers (or fund them) to handle certain compliance requirements and provide an official sanctioned entity for purposes of corporate policies. There should be a way to garner support broadly for risk management and specifically for security in the corporate context.
What's core issues are you thinking of?
I wasn't aware of the Gloat project before this. It's a compiler that turns Clojure into native binaries by first transpiling to Glojure (which I'd also never heard of before this), which in turn targets Go. This is rather than using a GraalVM native image, which as I understand it is at this point the better-explored mechanism of doing that for JVM-based stuff (but has its own trade-offs).
Very cool!
I can see why people try this way: graalvm is limited outside the entreprise version, and it’s quite slow to compile with.
I think the aot compilation story on the JVM lacks fast tooling with good UX compared to go.
How is the demonym not Clojuristas?
The demonym I've most commonly seen is Clojurian.
Clojurist makes it sound like adherents to a belief, which does imply identity though Clojurian more directly so. I’ve also found Clojurian most often. If only we had software that indexed the popularity of words…
In that case, Clojurians can be used for new initiates and Clojurists can be used for the most dogmatic Clojure supremacists, like myself.
Maybe clojure wouldn't have had it's nubank "exit" if we'd called ourselves clojuristas but this old clojure lefty loves it.
Reminds me of S.P.A.T from about a boy.
Two of these projects are just AI. This is not very promising.
Neither of them is AI genetated software though. One is a fast Clojure LLM inference library, and the other a MCP server, both coded by humans, without AI.
I know at least one of them and it has already existed for a decade or so, and is a legit framework, so this isn’t some hype based thing.
And Clojure is legitimately good at data processing.
The MCP server is a relatively simple project, don’t think that’s a bad idea.
> are just AI.
You realize that the term "AI" didn't come to life in 2009, it's been around for over 70 years?