My experience with working on Julia code in VS Code (the supposed best coding environment for Julia) has been very hit-or-miss and there are very few signs of progress in the last year.
A while ago, I encountered an issue where the linter doesn’t recognize the modules added through using even though they’re available in the global environment. This is an issue that plenty of people have come across. I’ve reported it two months ago and no reply has been given yet.
Since then, I’ve resorted to setting up custom project environments where I add the necessary dependencies. Lately, I’ve come to the sad realization that even that also occasionally doesn’t work. These are issues that have been open since 2019 with no sign of when they’ll be resolved. To be clear, this is not only an issue of false positive error warnings, this also causes the linter to be completely blindsided by all the functions used from the packages as well.
In fact, the last resolved issue for LanguageServer was over 6 months ago. I’ve also encountered a significant amount of bugs in the VSC extension and many, many inconveniences that one would guess would’ve been ironed out at this point. I’m guessing having a dodgy work environment like this can’t really be helping with language adoption.
We all know that the Julia VSCode extension is a weak point of Julia development…
From my point of view only a major investment that allows to hire a few new developers can fix that… If you have an idea how to:
get major funding from companies
get major funding from research institutions
get major funding from private people
please share your ideas…
We could start a Julia Foundation… There is a Mozilla Foundation, other open source projects are doing that… But this would make only sense if we are pretty sure that we are able to collect at least 10 000 EUR per year, otherwise the overhead is not worth it…
I obviously don’t know much about fundraising. From my understanding, JuliaHub has been the major inlet of funding for language maintenance, so they might have better answers.
I’m not trying to bash the language or its maintainers. I think the language itself is great and has a lot of potential. It’s just when I see discussions about what’s holding back mass adoption, people usually refer to it being unpopular and not having enough libraries when, in fact, I’d find it hard to recommend the language with such significant annoyances in the coding environment.
Well, as far as I understand JuliaHub is a company that earns money with closed-source development, not with open-source development.
From this point of view I think launching a foundation that allows to contribute and not pay taxes might be useful for open source tooling…
I think contributing to the Julia VSCode plugin is currently very hard for the normal spare-time open source developer due to the lack of documentation of the internals and the protocol…
I really don’t think it’s a funding issue… Julia development is quite active: Commits · JuliaLang/julia · GitHub (not to mention the ecosystem of packages).
It’s moreso that most core Julia developers do not have this as a priority. Which I also would agree is something that could use some realignment. The VSCode extension & LanguageServer, while awesome, are indeed not as ergonomic as in other languages, and this hurts adoption.
A lot of things are ultimately funding issues. Much of the work being done there is ultimately justified either by grants or supporting some companies products one way or another.
People aren’t working on things just because they feel like it or would be nice to have.
One public example was the PkgImages, which really improved compile times for Julia 1.9+, was funded by a CZI grant.