Julia 1.0.1 has been released

The first patch release for the Julia 1.0 series of releases, version 1.0.1, has been released. Binaries are available on Download Julia for Linux (i686, x86-64), FreeBSD (x86-64), macOS, and Windows (32-, 64-bit). Linux binaries for ARMv7 will be available soon™, but we’re currently having issues with the infrastructure that builds on ARMv7.

As a patch release, version 1.0.1 introduces no new features or breaking changes; it contains bug fixes, performance improvements, and documentation updates only. Thus we recommend anyone currently using 1.0.0 to upgrade to 1.0.1. A summary of the commits included in this release is available in the release notes on GitHub. Note that 1.0 on CI now points to 1.0.1.

Enjoy!

49 Likes

A huge thank you to @kristoffer.carlsson for his work on this release. He set up automatic backporting for commits that seriously streamlined what is an otherwise painfully manual process, and he set up the automatic generation of release notes. (Kristoffer, I apologize if there’s more I’m forgetting.) Hooray for Kristoffer! :clap: :clap: :clap:

61 Likes

huge thanks to the effort!

by the way, would there be a “0.7.1”? I mean, I’m still occasionally receiving depreciation warnings using 0.7.0 for development, so I guess a version of 0.7.1 would be nice.

AFAIU there won’t be a 0.7.1 version because 1.0.1 had bug fixes and minor improvements compared to 1.0.0 but nothing got deprecated and no breaking changes were introduced. So nothing should change in your code and if you fix your code in 0.7 to work with 1.0, it should also work fine in 1.0.1. Any problem related to upgrading to 1.0.1 would not be noticeable by working in 0.7. In short, the extra effort to release a 0.7.1 isn’t worth it. That’s my interpretation.

4 Likes

This is correct.

3 Likes

3 posts were split to a new topic: Random not defined

reasonable.

just for curiosity: so when there’s a 1.1.0, would 0.8.0 be released?

No; the purpose of 0.7 is only to act as a bridge between 0.6 and 1.0 that gives information about deprecations and how to solve them in 1.0. Once you have solved those deprecations, you should upgrade to 1.0.1.

3 Likes

I downloaded the JuliaPro v10.1.1 release and it now has an authentication feature which blocks me from using it my offline machine. I understand the power of hooking up software updates to online repositories but there needs to be a mechanism for using this offline.

Also it was my understanding that JUliaPro ships with a collections of packages including Plots which I was trying to use. However, no matter what package I choose (I tried Plots, DataFrames…) an authentication is required.

Continuing some experimentation online, I found that these packages once downloaded take an inordinate amount of time to compile ie Plots did before they can be used. Couldn’t these be pre-compiled into the image to make JuliaPro run fast right out of the box with an option to update them later?

With respect to plotting, I noticed that the plot window is very small and I can’t seem to resize it for a larger view. I ran the Lorentz attractor example and discovered that I was missing ffmpeg. I used conda to get it installed but felt there should have been more direct scheme not dependent on conda. (user suggested using conda).

Once the Lorentz attractor started working I noticed windows popping up and disappearing repeatedly for at least 20 to 30 times and then nothing displayed but an image address. Using the browser I was finally able to see the animation that was created.

I bring these issues up as I think its important for the overall user experience. My comparison is based on Netbeans and Matlab which are completely usable once installed. I also noticed that JuliaPro brings its own copy of Julia with it and doesn’t seem to recognize that you have it installed already. Netbeans supports multiple version of Java as an example, something you may need to add as Julia matures.

Bottomline, I can’t introduce JuliaPro to my group until it can be run offline and I’ve already proposed a few projects where it would be useful. We are primarily a Matlab shop and so there’s going to be resistance to even consider it. However, not being able to run offline is a major show stopper, with the precompile issue number #2 and a local repository for Julia packages #3.

Please take these as positive criticisms as I really like the Julia paradigm and I really want JuliaPro to succeed.

It is not clear why you are posting this in the Julia announcement topic. Note that you don’t need to use JuliaPro to use Julia. Just get the Julia binaries from the link above, install your favourite IDE, some packages, and you are good to go. You should be able to use everything offline (if you also want to install to a machine that is not on the net, even intermittently, search this forums for a solution — I would recommend a Docker image).

Compilation time issues with Plots.jl are known, but there are other plotting packages which load very fast. You may want to explore them. Again, search this forum for options.

My apologies, I thought this forum was for JuliaPro. I will direct my comments to them.

Yes, I’ve used Docker too. The problem we have is the use of official images only as user developed can possibly have malware. There was a recet case of 17 tainted docker images containing cryptocurrency malware.

The compilation issue was primarily for JuliaPro as their website implied that they ship with these packages ready to go. I realize that Julia precompiles things the first time around.

Once again, my apologies.

THis was the link the link that misdirected me to your site:

https://juliacomputing.com/blog/2018/10/12/october-newsletter.html

The First line says JuliaPro but the first clickable link is for Julia both are using the same version number so you might get more misposts like this.

I think you are at the right site (if you want to solve a particular problem with Julia), just possibly the wrong topic (but that is harmless, except it may delay getting an answer). In particular, see

But again, I would just first try plain vanilla Julia.

Regarding Docker images: I am not sure that malware is a concern if you make them yourselves (if you trust the binaries from the official site; you can also compile it yourself). But that is no different for Julia than any other piece of software.