Informational Post: Releases 0.8.1 and 0.8.2 assign incorrect vdW parameters

[j-wags]

This morning, we become aware of a major bug (#807) introduced in OpenFF Toolkit 0.8.1 and that is also present in 0.8.2 which causes van der Waals sigma terms to be off by a factor of 2. This could affect environments created or updated after December 7. Unfortunately, we expect that this affected the main code path that most users follow to assign parameters.

Basically, all energies calculated using these two versions of the Open Force Field Toolkit should be considered incorrect.

We sincerely apologize for this.

The following actions are being taken:

• Versions 0.8.1 and 0.8.2 of the Open Force Field Toolkit have been removed from the main label on omnia
• A bugfix is being implemented and version 0.8.3 will be released later today
• A postmortem will be scheduled for early next week to establish how our development practices can be changed to prevent this from happening in the future

[j-wags]

We've released the Open Force Field Toolkit version 0.8.3. This fixes the major bug reported earlier today. Detailed release notes are available at https://open-forcefield-toolkit.readthedocs.io/en/0.8.3/releasehistory.html

Conda packages available now on the omnia channel.
https://anaconda.org/omnia/openforcefield/files

Single-file installers are available in the "assets" section of the GitHub release, at https://github.com/openforcefield/openforcefield/releases/tag/0.8.3

[mattwthompson]

This is two years old now - could we move it to Discussions or another venue?