With JuliaCon in full swing, Julia IDE Juno developer Sebastian Pfitzner took to the virtual stage to brief users about the state of things and present them with… Julia for VS Code 1.0.
After Microsoft bought GitHub, development on the Atom editor unsurprisingly slowed – possibly because Microsoft has already put resources into Visual Studio Code and can live without homemade competition. According to Pfitzner, the future therefore lies with VS Code, which is why he and his colleague decided to focus on LanguageServer.jl and join the efforts of the team behind VS Code for Julia.
As the IDE extension is almost at feature parity with Juno, switching should be comparatively easy, while also adding useful capabilities such as “proper” linting, as Pfitzner put it. Other benefits include better stability, while the opportunity to put more work into the language server is hoped to help the whole editor ecosystem. Juno is still said to be “perfectly usable” but will switch to maintenance-only mode.
Pfitzner also took the opportunity to start a naming discussion for the new project, debating if Juno would be a good moniker for its reincarnation and asking the community for feedback. A new name could help those well versed in the old IDE develop a better sense of ownership for the new project, so it probably wasn’t a bad idea to step away from the somewhat longish current designator. The only criteria are that the new name should start with a J, be short and recognisable, and have some sort of relation to either Juno or Jupiter.
During his talk, the team also launched version 1.0 of the project, which came packed with useful new features. Those included, maybe most importantly, a reworked debugger and some basic profiler implementation, which might still take a while before it is ready for prime-time. It also got a remote extension and an option for inline results that developers more used to Juno will surely appreciate.
In addition, the continuous delivery system for new versions has been revamped so that upcoming releases are put into an insider channel first. If all is well, and there aren’t any issues with crash reporting, the new version will then be pushed into the release channel.
To emphasise the team’s commitment to move the project along, it also gave a short walkthrough upcoming features. Amongst other things the roadmap plans for more Juno-like dynamic features, and a quite advanced notebook integration with debugger and autocompletion, however, these might take a while. To tackle the apparent lack of documentation, the team said it is currently waiting for the Google Season of Docs for which it applied. Once that has been taken care of, a Juno-style documentation browser is on the todo list
Julia is a general purpose language first introduced in 2012. Its creators intended to develop a language that combined the speed of C with the dynamism of Ruby, usability of Python, mathematical notation of Matlab, Lisp-like macros, and good usability for statistics, which is something they associate with R. Today it is indeed mostly used for numerical analysis scenarios, it seems, with users ranging from NASA to investment management company BlackRock.