Release Candidate, 2020-06-19

Installers for R 3.6.1, 3.6.3 and 4.0.1 based on the 2020-06-19 version of the development branch.

Key questions: does it work? (you should just click the VisionEval.Rproj after you have R and RStudio installed in a suitable version).

Give me feedback on the new "docs" directory.

Each binary installer has an associated "source" installer (unzip into the same folder). Creates a 'src' directory that has all the built (but not package) VE framework and module sources, so you can load them up into RStudio to set breakpoints, debug, etc.

If you're really brave, the Github (either via the .zip for the source, or by cloning) has an integrated installer. No more VE-Installer or screwing with Git for Windows. You'll need RTools40 for Windows installed and on your Windows system PATH (edit user environment variables to set). See "build/Building.md". You do need a (fast) internet connection.

To try building via Git, double click VisionEval-dev.Rproj in the root of the github. It starts RStudio and loads the build helper functions.

Run "ve.build()" to create a runtime on your system. Run "ve.build('installer')" to create the runtime installer. To get exact replicas of the installers uploaded here, you'll also need to check out the VE-Wiki in a parallel folder to your VisionEval git installation (see build/config/VE-config-release.html) since I'm currently pulling over a couple of documents from there.

ve.build Should work on any system (Windows, Mac, Linux) if you have a development environment - Linux may require some additional package dependencies at the OS level (e.g. if some of the VE dependencies require additional "shared libraries"): pay close attention to any error messages as the dependencies build.

It would be great if someone could contribute Mac binary installers for this version.

• Dan says: I will make a Mac installer!
• If you need to upgrade to R 4. 0.1. on Windows, click here