On 10/21/2015 10:04 AM, Bernd Schmidt wrote:
On 10/21/2015 05:56 PM, Jeff Law wrote:
The problem is the existing ENABLE_CHECKING conditions.
Anything which is #ifdef ENABLE_CHECKING will have its behavior changed
if we change things so that ENABLE_CHECKING is always defined with a
value.
So if we wanted to continue to use ENABLE_CHECKING, but with a value,
then every #ifdef has to be fixed all-at-once.
But why change it to have a value? Just keep it as-is, and use
flag_checking for runtime tests.
To avoid conditionally compiled code. I'm of the opinion we should be
stomping out as much as we reasonably can.
Essentially each blob of code that is conditionally compiled represents
two paths that can't be tested for basic syntax correctness together.
As a result we regularly see one of the two paths being broken from a
syntax point of view.
By stomping out the conditionally compiled code a standard build will
test both paths for basic syntax correctness. It'll take time to get
there, but that's where things, IMHO, need to go to help reduce the long
term cost of maintenance.
jeff
