Tue, Apr 21, 2020 at 04:17:03PM CEST, pgagne(a)redhat.com wrote:
I think I asked Ondrej at some point what his thoughts where on type
hinting in LNST, and my recollection was it was something he wanted to move
gradually to so I decided to put those there, since in the case of those
functions the type is very important.
From my understanding, as we move from XML recipes to python ones, we
probably want to be a little more strict about the typing of parameters.
I think it makes it cleaner just from an inline code documentation
perspective, especially in this case where the parameters were previously
of an ambiguous type, and now they are more concrete.
Related to the static analysis it sure makes IDE code completion ( and
similar things) a lot more concise.
Makes sense. I asked Ondrej about his opinion and he agreed that he'd
like to have this (eventually) in all of the LNST code.
Thanks for explaining this, I have no objections. The patch looks ok.
-Jan