Skip to content

GAM 7.00.03

Compare
Choose a tag to compare
@github-actions github-actions released this 18 Sep 21:59
· 143 commits to main since this release

GAM 7.0.3 includes:

  • Order of magnitude performance gains for Mac / Linux builds start time. GAM now uses a "one directory" PyInstaller build which offers better performance. Note that you'll see a lib/ folder extracted in the GAM install path also with many files under it. This is an expected part of the change and those files can safely be left as they are. On a Macbook Pro M1 Max / 64gb RAM / 1tb SSD GAM 7.00.02 takes ~7 seconds to run "gam version". GAM 7.00.03 runs the command in less than half a second. These optimizations will becoming to Windows shortly also (as soon as Jay can fix the MSI installer, ugh)

  • MacOS builds are now signed with an Apple Developer certificate and notarized by Jay. This should allow running GAM on MacOS without needing to deal with security warnings and disabling them. This did require me to register to be a Apple software developer and pay a $99/year fee. I have feelings about needing to pay registration fees to make open-source software run but that's just life in 2024 and we need to go along with it to keep GAM admins secure. I plan to follow the equivalent processes to sign GAM Windows builds in the near future. If you'd like to help offset these costs feel free to Venmo me a few bucks. See the wiki for details on verifying your GAM7 binary is legitimate and official (I'll add commands to validate MacOS signature soon).

  • MacOS 12 is no longer supported by Apple and Python 3.13+ won't run on it. Running GAM on unsupported platforms that aren't receiving security updates risks the security of your Workspace users and their data. Don't do it. This version removes support for MacOS 12.

  • Python 3.8 is reaching end-of-life Oct 1st. We've removed testing and listed support. This version probably still works with Python 3.8 but a future version is likely to break it. This only matters if you're running GAM from source and using a very old Python version (which might be the case if you're running an old Linux distro).