>Every so often, the NT file system folks dream of changing the deletion
>model to be more Unix-like, but then they wonder if that would end up
>breaking more things than it fixes.
Do you know the history of 'why' the current model was chosen? It seems to be more complicated to use, and I assume more complicated to implement. So therefore I suppose it must have been chosen deliberately, but why?
Aside: fans of arcana may care to read [MS-FSA].pdf for the gory details.