Skip to content

Should we consider an LWO implementation? #1169

@Silenus-the-Companion

Description

@Silenus-the-Companion

The question is what is the key issue that LWOs address for a driver like fluffos. I suspect one of the nice features is "name" hiding which means that the name isn't entered into the object table or name server. This prevents other parts of the system from monkeying with the code i.e. a destruct_object call for example or a call_other making the state more consistent and easier to control. Would it be better to have "fake" LWOs by having another form of the new("objpath") operator where the clone does not have a name? Or perhaps has a name that is only accessible if you have certain security privileges?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions