-F needed in ws_release even WS is in default File System

[mflehmig]

Hello Holger,

as far as I can see, the option -F FS is mandatory for ws_release even for workspaces in the default file system. This is inconsistent behavior w.r.t. the other ws-commands, where the default FS is used if the option -F is not provided by the users.

Can this be fixed?

Best regards
Martin

[holgerBerger]

I think the idea behind that is to avoid releasing the wrong one in case the name exists in several FS.
I can have a look how complex it is to require -F only if the name is not unique.

[mflehmig]

I think the idea behind that is to avoid releasing the wrong one in case the name exists in several FS.
Good point. Then, it should be explicitly documented. Furterhmore, ws_allocate -x might print an appropriate error/warning(?) and not "Error: workspace does not exist, can not be extended!".

[holgerBerger]

I think the idea behind that is to avoid releasing the wrong one in case the name exists in several FS.
I can have a look how complex it is to require -F only if the name is not unique.

[holgerBerger]

I checked it, one does not need to give -F when the workspace is in default, only if it not in default.

I added debugging option (which was already in the code) to ws_release, and one can see the search process when giving --debug.

It searches if no -F is given (that happens in validation of command line) trough the workspaces, and if nothing is found it uses the default.

debug: no filesystem given, searching...
debug: searching ws1
debug: searching ws10
debug: searching ws2
debug: searching ws3
debug: fallback, using global default, ending search

[holgerBerger]

next major version will have a defined order for all search operations

[holgerBerger]

@mflehmig something we could check with v2 if it now behaves as expected.