Changes between Initial Version and Version 1 of Ticket #5332, comment 3


Ignore:
Timestamp:
2019-02-13T10:38:56Z (6 years ago)
Author:
Francesco Casella

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #5332, comment 3

    initial v1  
    11Regarding the inlining, my idea was that if you have a (pure!) function call whose arguments are all constants or structural parameters, you should evaluate it anyway in the frontend, regardless what the Inline annotation says, and replace it with its constant literal output. Of course if the function is inlined, the backend should eventually do this anyway, but why not doing this in the frontend outright? What do you think?
    22
    3 Inlining would be useful in those cases where the gravity acceleration is actually a function of the position, but that's not the case in most applications, that take place on the ground and are therefore subject to a constant downards acceleration of 9.81 m/s^2. Only space applications actually require inlining - I've played with them in the past, but it's rather the exception than the norm.
     3Inlining would be useful in those cases where the gravity acceleration is actually a function of the position, but that's not the case in most applications, that take place on the ground and are therefore subject to a constant downards acceleration of 9.81 m/s^2^. Only space applications actually require inlining - I've played with them in the past, but it's rather the exception than the norm.