Julia v1.5.0 has been released

Thank you to everyone who made this year’s JuliaCon great! As a parting gift, the Julia developers are pleased to announce the release of Julia v1.5.0, the fifth minor release in the 1.x series. Jeff and Stefan put together a blog post highlighting some of the most exciting new features in 1.5. Check it out!

As usual, binaries are available for all of your favorite platforms (Linux, macOS, Windows, and FreeBSD) at https://julialang.org/downloads.

As a minor release, v1.5.0 contains no breaking changes, only new features, performance improvements, and marginal, undisruptive changes in behavior. You can also see the NEWS file for the full set of changes.

Note that like 1.5, like its predecessor 1.4, does not have long term support. As of this release 1.4 has been effectively superseded by 1.5, which means that there will not likely be any further 1.4.x releases. Julia 1.0 is still currently the only long term support version.

We encourage everyone to give it a try. Packages can test with 1.5.0 on CI by specifying 1.5 on Travis, AppVeyor, Cirrus, and GitHub Actions. As always, let us know in the issue tracker if you run into any issues.

139 Likes

Looks like I got a major regression in performance with Makie when I have a plot loaded in v1.5.

While a plot is loaded, the contents of a vector is printed line by line slowly, and takes about 5 seconds to finish printing vector in the REPL. After closing the plot window, the printing of vectors returns to being fast.

Try loading a Makie like plot(1:10,1:10) and then displaying a vector rand(100,2) in the REPL before closing the plot. It causes the printing of vectors to be extremely slow. Never had this issue before v1.5 of Julia.

Trying to download the manual in PDF at https://docs.julialang.org/en/v1/ results in a 404 Not Found error.

1 Like

:point_up:

12 Likes

Is this release identical to the RC2?

Yes.

3 Likes

Argh, I finished running week-long experiments last week and now I want to re-run them with Julia 1.5 so I can say I used it in the paper, :laughing:.

I was hoping 1.5 would only appear after I finished the manuscript so I would not get these urges, but the development was faster than I expected.

12 Likes

This could be a recent update to Makie (I think the render loop has been changed if I interpreted the comments on slack correctly) since I get this behaviour on v1.4 as well as v1.5.

7 Likes

Holy Molly, all my random tests has failed. Does the random number generator generates a different set of numbers with the same seed?

Test Summary:                               | Pass  Fail  Total
All tests                                   | 1170    47   1217
  temporarily_setDefaultPrecision function  |    2            2
  random function                           |    1    47     48
  fromString function                       |   27           27
  derivative function                       |   28           28
  grad function                             |    2            2
  jacobian function                         |    2            2
  hessian function                          |    2            2
  newtonRaphson function                    |   36           36
  findmin function                          |    4            4
ERROR: LoadError: Some tests did not pass: 1170 passed, 47 failed, 0 errored, 0 broken.
in expression starting at /Users/ssiew/juliascript/pdfp/runtests.jl:132

1 Like

Yes, this is the contract for the default RNG’s. You want https://github.com/rfourquet/StableRNGs.jl.

11 Likes

Plots fails to compile on Windows, but Plots#master works well, if someone is wondering.

1 Like

It seems that the issue was solved on Windows and macOS (https://github.com/JuliaLang/julia/issues/36893). I don’t have any trouble in pre-compilation of Plots on Windows.

Thanks for such a big announcement. I have some questions: Do I understand that by installing this new version, all the packages have to be installed again? Also, once installed the new version, I have to update the path in the VS Code, right? In that sense, how can I do it without losing the package installed?

1 Like

All you need to do is copy the Manifest.toml and Project.toml files from ~/.julia/environments/v1.4 to ~/.julia/environments/v1.5. That will tell Julia 1.5 to use exactly the same versions of the same packages that you already had installed in Julia 1.4, and it will not need to download them again.

9 Likes

Perhaps, adding a command to Pkg that does this for you automatically makes sense.

5 Likes

The blog post was lovely. It makes the release much more exciting than the dry NEWS.md.

36 Likes

Julia docker images have been refreshed. :julia: :whale: :white_check_mark:
https://hub.docker.com/_/julia?tab=tags&page=1&name=1.5

  • docker pull julia:1.5
  • docker pull julia:1.5.0-buster
  • docker pull julia:1.5.0-alpine3.12

example:

:~$ docker run -it --rm  julia:1.5  
               _
   _       _ _(_)_     |  Documentation: https://docs.julialang.org
  (_)     | (_) (_)    |
   _ _   _| |_  __ _   |  Type "?" for help, "]?" for Pkg help.
  | | | | | | |/ _` |  |
  | | |_| | | | (_| |  |  Version 1.5.0 (2020-08-01)
 _/ |\__'_|_|_|\__'_|  |  Official https://julialang.org/ release
|__/                   |

7 Likes

Makie issue is now fixed.

9 Likes

When will version 1.5 make to debian and fedora?

I think Fedora is a bit better than Debian here, but since Julia needs some custom patches to LLVM, for example, it’s fairly difficult to package in a regular Linux distribution, which is why most distribution’s official repositories ship quite outdated versions of Julia. If you want the best Julia experience on Linux, I would heavily recommend using the official Julia binaries from https://julialang.org/downloads/, as linked above.

8 Likes