Vassal 3.6.16 Released

VASSAL 3.6.16 Released

The VASSAL Team is happy to announce VASSAL 3.6.16.


  • NEW! 64-bit ARM packages: There are now MacOS and Windows packages for 64-bit ARM processors. If you have a Mac with an Apple Silicon CPU (which is a 64-bit ARM processor), we recommend using the 64-bit ARM build.
  • Code deprecated more than a year ago has been removed. Modules containing custom code using that code MUST be updated if they are to work with 3.6.
  • Things may be broken which worked previously. If you find a new bug, please report it.
  • Modules saved in 3.6 are not openable by earlier versions of VASSAL. We recommend keeping a backup copy of any pre-3.6 modules you plan to modify in 3.6 until you’ve verified that everything works to your satisfaction.


MacOS (64-bit ARM)
MacOS (64-bit x86)
Windows (64-bit ARM)
Windows (64-bit x86)
Windows (32-bit x86)

Changes since 3.6.15

Bug fixes

  • 12232: Notify user about mismatch between module and save version instead of throwing IllegalStateException
  • 12231: Check that all tiles exist for an image when tiling, not just 0,0@1:1
  • 12221: Fixed various missing searchability on various components
  • 12209: Report correct trait in expression Audit Trail
  • 12159: Correctly handle install paths containing spaces in
  • 12158: List Key Commands tool can be re-opened after closing with X button
  • 12152: Work around Java bug which prevents displaying Windows file dialogs

Changes since 3.6.14

Bug fixes

  • 12119: Restore vanished Load Comments
  • 12116: Pieces shouldn’t receive same GKC twice just because it sent them to a new map
  • 12105: Deck’s ‘Send-to-Deck’ subcomponent has access to properties (e.g., PlayerSide) and properly shuffles “always shuffle” decks
  • 12093: Fix disabling of toolbar items nested in multiple Toolbar Menus

Other improvements

  • 12140: Action Buttons on pieces that have just been drag-moved are suppressed for a half second for user-error avoidance
  • 12085: Screen capture default icon height adjusted to match the other 16-pixel-high icons
  • 12069: Tiling: Improved detection of and recovery from running out of memory; improved error dialogs
  • 9894: Improved error dialog for failure to start child process

See the Release Notes for descriptions of the changes in 3.6.

Minimum Requirements

VASSAL 3.6 requires Java 11 or later.

The Windows and MacOS packages have an appropriate version of Java bundled with them, so there is no need to install Java separately on those operating systems. On Linux, use your package manager to install Java 11 or later.

Help us test

Please report bugs in the Technical Support & Bugs category at the VASSAL forum.



1 Like

During the past several upgrades of 32-bit x86 Vassal (like, the last year or two), my AVG antivirus would typically give me a warning or two, during the installs, before deciding a file was safe, or letting me override the warning - so no problem.
This time, however, it came up with these:

Apparently, it actually identified a threat, or thinks it has.
I have not tried clicking on “Report as false positive”.

Was just typing the same Rob :frowning:

Has been warning for a few previous (and I’ve read all the previous threads re certificates etc, and I’m happy to trust Vassal), but this time, like you, it’s not letting me bypass it.

1 Like

Always a relief to see I’m not the only one suffering, hahaha. :laughing:

1 Like

I can see from VirusTotal that VASSAL-3.6.16-windows-x86_32.exe is flagged by a few of the AV vendors. There’s not much I can do about that, however. I recommend reporting it as a false positive to your AV vendor.

AVG’s web form for reporting false positives won’t accept anything larger than 50MB, so I can’t report it that way.


Yeah, I flagged it as safe, and also sent it back up for review.

Hopefully it’ll clear soonish (not sure what sort of response time they have?)

1 Like

Just noticed you’re seeing _32.exe being blocked - mine was the _64 one, for your info

AVG and Avast are reporting false positives for all three of the 3.6.16 Windows builds. It would be most helpful if they were all flagged as false positives by someone who is having the problem.


Done for 32-bit x86.