Gemini IRAF VM tutorial¶
Instructions for running Astroconda (32-bit) IRAF on MacOS 10.15 or later, under an x86 Linux virtual machine.
A link to the previous instructions for VirtualBox can be found below, but they do not work on newer M1/M2 (ARM) Apple machines, nor can VMWare or Parallels run Gemini IRAF on those machines.
This method uses QEMU to emulate x86_64 hardware on ARM64, which works well, producing the same results as on a Linux machine, but with a large performance penalty. A test case using GMOS multi-slit data took 14 times as long to run on an M1 machine as when running natively on a recent i7 processor, with the overhead varying considerably between different processing steps. This is, however, the only way of running the necessary IRAF packages on new Apple CPUs (at least prior to MacOS 13).
Compared with the original 2020 release for VirtualBox, the new Gemini IRAF VM image has the following features:
- A new control interface, GemVM.
- A simpler installation method, using Anaconda / Miniconda.
- The latest Gemini IRAF 1.15.
- The latest DRAGONS 3.0.3. 
- A new version of PyRAF (2.2.1) that is fully compatible with Python 3.
- More recent Anaconda (2021.11) package versions.
- The (unofficial) PyFU scripts for mosaicking IFU datacubes.
- Support for compiling IRAF packages.
- Miscellaneous OS updates etc.
- Installation & setup
- Starting the VM & logging in
- Sharing data with the host
- Using the VM
- Shutting down the VM
- Other things to know
- Old VirtualBox instructions
|||You should run DRAGONS directly on your host machine instead, unless you really need to use it in conjunction with IRAF.|