Miro Kropáček wrote:
And who says modification time can't be updated while overwriting? Original scenario assumed *new* copy, overwriting existing file (i.e. its *content*) seems like perfect reason why to update modification time.
You are right, I messed the 2 scenarios.But with the MS Windows explorer, it works as you expect. The target modification time is updated with the modification time of the new content.
*But* the tricky case I was talking about is when the new modification time of the content of the .c file is still older than the modification time of the .o file. make does not see that something has been updated.
-- Vincent Rivière