Parallel use of Julia 0.7 and 1.0


Is there any way to use versions 0.7.0 and 1.0.0/1.0.1 on the same computer (Windows 8.1 Pro with Atom/Juno) in parallel?

I have some old code of 0.6, which I need to transfer, so I need version 0.7 for the deprecation warnings.
But my new projects should be v1.0 from the start.

Currently, I have to switch the environment variable JULIA_BINDIR every time I want to switch between 0.7 and 1.0.
Version 0.6 works fine together with the others, because it relies on JULIA_HOME, which is no longer used.

Thanks in advance.

1 Like

As a Julia user, I use win 10 and have both versions 0.7 and 1.0 installed. And I am able to use without resetting the binder variable. I can open two separate instances of Julia simultaneously so don’t understand the reason behind resetting the binder environment variable.

Perhaps the issue is that if you have set a hard JULIA_BINDIR environment variable, you will have to manually change it to switch. If you allow it to be defined implicitly, then you can have different Julia executables, each of which will do its own implicit definition.

@Ajaychat3 Did you install into the default location (User-folder)? I did not do that, because I am in a corporate environment where populating the user folder with 10s of thousands of files severly slows down login/logout procedures.

@js135005 How would I perform an implicit definition?

@Moses: Yes I installed it to the default location. I am also on corporate environment but I don’t see any lag in login/logout timing.

@Ajaychat3 Lucky you. :grinning:
Our network and servers are pretty old and as the user folder gets synchronised on every login/logout, installation of julia into the user folder severly slows down the process (I had it installed there initially as well, but then one computer restart took 20 minutes).

There is no synchronisation of user folders in our case. Synchronisation is basically for backup purposes as I understand. You can explore whether this can be disabled at startup subject to organisation policies.