Julia, a zippy programming language for data scientists and machine-learning experts, has been updated with improved multi-threading, new library features, and tweaks to the build system.
The language has been embraced by some programmers for its C-like speed. Its makers aimed for it also to be as easy to use as Python, with the best qualities of R for statistics and Matlab for algebra.
The new features arrive in Julia version 1.4, the fourth minor release since it reached version 1.0 in August 2018 – a milestone it hit six years after it debuted with the help of MIT’s Computer Science and Artificial Intelligence Lab (CSAIL).
While Julia has far fewer packages than Python, Julia developers say the language has been downloaded over 13 million times, while there are over 2,800 Julia packages, up from 1,900 registered packages two years ago.
According to a Julia contributor, this release doesn’t contain any breaking changes or changes in behavior that cause major disruptions for the user.
Nonetheless, there are a long list of changes to language features, to the language itself, multi-threading and changes to the build system, as detailed in the release notes for version 1.4.
Julia developers have a note for users who code on Windows machines. “Windows build installer has switched to Inno Setup. Installer command line parameters have thus changed. For example, to extract the installer to a specific directory, the command-line parameter is now /DIR=x:\dirname. Use julia-installer.exe /? to list all new command-line parameters,” the release notes state.
There are several new library functions, such as ‘evalpoly’, which serves the same role as the @evalpoly macro and will likely replace it if Julia users like it. “The function is just as efficient as the macro while giving added flexibility,” Julia developers note.
The update also includes new library features as well as changes to the standard library.
Only Julia 1.0 has long-term support so that means version 1.4 supersedes version 1.3 and there are unlikely to be any further improvements made to version 1.3.
As noted on Devclass, Julia version 1.4 drops support for the (;) syntax for empty block expressions, apparently because users found it more confusing than helpful.