Mise à jour terminée. Pour connaître les apports de la version 13.8.4 par rapport à notre ancienne version vous pouvez lire les "Release Notes" suivantes :
https://about.gitlab.com/releases/2021/02/11/security-release-gitlab-13-8-4-released/
https://about.gitlab.com/releases/2021/02/05/gitlab-13-8-3-released/

## Operators

- Bug #1257: QuasiIncompressibleSecondPiolaKirchhoffStressTensor was not good enough: policies were improperly implemented.
- Support #533: Assembling into only matrix or vector for hyperelasticity and non linear membrane operator is now fully supported.

### Needs further feedback to be closed

- Feature #1259: Implement new linear membrane operator.

## Tests

- Design #1264: Add embedded models in the CTest framework. And remove the former Python scripts that were very unwieldy to use.
- Support #1262: Set up CTest for the most recent tests.
- Support #1261: Fixture: one of the template parameter is no longer required following #1249

### Still in progress

- Support #1272: Add older tests to CTest. Done for all Core tests and most of Utilities ones.
- Feature #1245: Proper tests should be provided for all operators.

## Environment

- Feature #1268: Introduce a sort of internal environment variable which is very handy for tests of models.
- Design #1270: Remove the setenv calls from the code, which are not portable contrary to geteven which is within std namespace.

## Command line options

- Feature #1266: In MoReFEMData, add on command line options the possibility to give environment variable.
- Bug #1269: Fix the way to add additional command line options for a specific model.
- Support #1267: Update TClap library to v 1.2.2.
- Support #1265: Provide a new TClap specialisation to enable pair support.

## Miscellaneous

- Bug #1271: Modify the default path to third party libraries in XCodeConfig, which was not on par with the default used in ThirdPartyCompilationFactory.
- Support #1273: Fix warnings obtained with new version of LLVM clang (7.0).
- Support #1260: Add a functionality to check whether a type that satisfies a specific condition might be found in a tuple.
- Documentation #1224 Clean-up documentation directory. In fact it was done in a former tag but not properly reported.

### Needs further feedback to be closed

- Feature #1263: Allow boundary conditions for dimension N-2. This was formerly avoided by design, but is handy in some cases... provided you're sure what you're doing is mathematically sound. Needs to be confirmed by Reo before closing the ticket.
This tag has no release notes.