Julia 1.0 has been released: https://julialang.org/blog/2018/08/one-point-zero. Video of the Julia founders talk about the path to Julia 1.0 and the live tagging of the release:
Binaries for 1.0 are available on https://julialang.org/downloads, and we have them for the usual suspects: Linux (i686, x86-64, ARMv7), FreeBSD (x86-64), macOS, and Windows (32-, 64-bit). Check it out!
Current efforts are largely focused on updating packages. Julia’s
master branch is now 1.1.0-DEV, which means for now we’ll be adding bugfixes (which will be backported to 1.0.x releases), and new features that won’t break existing code.
Thank you to all of the developers and the amazing Julia community who got us to this point!
Thanks for the hard work. I also noticed that docker registry is also updated.
Look forward to playing with it.
Awesome. I have a question. Do I gain any performance if I compile the Julia binaries from source in my system, instead of downloading the pre-compiled binaries from julialang.com?
Unlikely. (No for most cases.)
So what has changed? We still remember the 7 gotchas by @ChrisRackauckas, especially: Gotcha #7: Not Building the System Image for your Hardware.
One thing that is much better than a year ago is the support of the default openblas version for modern CPUs like Skylake. In the past the binary version couldn’t recognize modern CPUs and used a slow fallback mode.
A good number of Gotchas are now gone. 7 is gone. 1 is basically gone: globals now have to be made explicit so now it’s just “dealing with globals” instead of dealing with REPL loops.
Hmm, very nice, things seem to improve so fast. Julia really has a wonderful community, thanks every one for getting us there.
So is there any other reason to build system image?
The doc still contains a section about how to build system image and I just submitted a bug report against that for Mac
of course there is, if you want to precompile your own methods into the system image to circumvent compilation happening at runtime.
Maybe this was already discussed earlier somewhere, yet I am curious to know the future development plan of Julia from now on.
Currently, both v1.0 and v0.7 have been released and are being maintained. But what is the next step? When v1.0.1 is released, would the corresponding version (with deprecation warnings) of v0.7.1 be also released? If so, would this process of parallel upgrades continue for a while? Or is the current v0.7.0 is the last release of the v0.7.x series?
Great news, Thanks!
v0.7 is only for depwarns to help people upgrade. I would be very surprised if there was a v0.7.x
Julia 1.0 news is even at Phoronix!
This is unclear to me. The Julia repo README still says
You can also set MARCH=native for a maximum-performance build customized for the current machine CPU.
Is that advice out of date, or does it address a separate issue ?
A very big, big thanks to the whole team that make it happens!
It’s out of date