Follow-up Comment #3, bug #22381 (project freeciv):

That the source tile isn't checking can_exist_at_tile() now is an oversight
(and a bug).  I'm unsure of the right solution for the destination tile
though.  Should the source bug just be fixed, and we can discuss the
destination more?  Should can_exist_at_tile() be refactored so that there is
another function that includes everything by the safety test that can be
called from can_exist_at_tile(), and that used, so everything is a single
commit?  Is there a narrative reason UTYF_TRIREME units should not attack
unsafe terrains?

( I almost added a patch for this as soon as it was raised, but suddenly
became uncertain regarding these questions, so commented instead).

As for consistency with spy actions: I strongly believe we need to determine
the right answer, and then have that apply in both situations (ignoring
facilities for ruleset flexibility that may permit ruleset authors to adjust
this), rather than that we ought make this change for the value of consistency
(that the current actions implementation happens to be right means that I
actually want consistency, just not for the sake of consistency). 
Unfortuntately, I don't have capacity to think about this deeply currently, so
can't help that much (and would be happy with the conclusion of anyone else
who is able to think about it sooner).


Reply to this item at:


  Message sent via/by Gna!

Freeciv-dev mailing list

Reply via email to