GitLab

if a port number is given and we are not in "listen" mode, we append the client number in order to be able to manage several clients; it is however recommended to not specify a port number explicitly when not in "listen" mode, so that the port is automatically assigned by the OS

Listen custom socket cl

See merge request !391

try to parse socket only if it doesn't start with '-' (in which case we should go to the next option)

Mesh.{Line,Surface,Volume}Numbers -> Mesh.{Line,Surface,Volume}Labels, Mesh.Point* -> Mesh.Node* (old names are still supported, but deprecated)

{Point,Curve,Surface,Volume}Numbers -> {Point,Curve,Surface,Volume}Labels (keep old names, just deprecated)

in addition to the default camelCase function names, add snake case aliases in the Python and Julia API, as this is the recommended style in these languages

get{Edge,Face}Number -> get{Edge,Face}Tags for consistency + remove experimental warning, as this is here to stay

make edge identifiers size_t + fix arguments of mesh/getEdgeNumber + add mesh/getFaceNumber (these should be renamed...)

revert change to popup ONELAB context with right-click: we should only respond to double-clicks, otherwise we mess up panning with the right mouse button (cf. #1171)

Update meshRefine.cpp

See merge request !390

new (probably temporary) gmsh/mesh/createEdges and gmsh/mesh/createFaces (until we provide a facility to simply create and store mesh edges/faces as elements)

only return entities in output maps if the entities are bound: otherwise, when entities are unmodified by the boolean operation but are deleted, we would return non-existing entities (cf #1154)

put "My solver" widgets in the main Solver menu + show how to have a "3D" button for actions instead of the standard text

new "Aspect" attribute for string variables: if equal to "Button", draw a standard button instead of a simple clickable text

fix error in documentation: node and element tags in MSH4 are still ints (not size_t) in $NodeData, $ElementData and $ElementNodeData (cf. #1157). The switch to size_t is planned for a future revision of the MSH format, at the same time as the restructuring of the post-processing IO format to make it more efficient.