Re: [YANG] new pyang errors

Randy Presuhn wrote:
> The "with-defaults" view sounds like it would help, particulary
> if a default-ed value isn't known a priori.

Let's not forget that the handling of defaults effect much more then the result of the 
get-config. IMHO the basic question is simply "does a leaf set only due to it's default exist?" 
(in the conceptual database).

This effects the following:
1) Does it exist?

2) How does edit-config create/delete work on it? E.g. can you create a leaf that already 
exists(?) due to a default value?

3) If you later explicitly delete a leaf with a default value what happens. Will it cease to 
exist? Will it automatically pop-up with the default value again?

4) Is there any difference between a leaf set by the default and a leaf explicitly (edit-config 
replace) set to the same value?
I really hope not.

5) Will get-config return such a leaf?
It may or may not depending on the implementation or the with-defaults option

6) Will a non-presence container exist if it has leafs set by default? This will effect how 
edit-config none is handled.

7) Does a "must" statement consider such leafs to exist.
I assume yes.

Balazs

_______________________________________________
YANG mailing list
YANG@ietf.org
https://www1.ietf.org/mailman/listinfo/yang