Follow-up Comment #9, patch #1341 (project freeciv):
once this is in we can clean up rulesets for not protecting
against the rare case of city having two Barracks at once
(guards against getting double effects)
Now patch #5650
___
Update of patch #1341 (project freeciv):
Status: Ready For Test = Done
Open/Closed:Open = Closed
___
Reply to this item at:
Follow-up Comment #6, patch #1341 (project freeciv):
So (2) has behavior changed. That's probably only good, and means that once
this is in we can clean up rulesets for not protecting against the rare case
of city having two Barracks at once (guards against getting double effects)
Follow-up Comment #7, patch #1341 (project freeciv):
Ah, heh, indeed. Behaviour (2) seems so obviously correct I hadn't realised
the bug was present before. Do we have a bug number for that? Does it
exist/should it be fixed for 2.4.3/2.5.0?
Update of patch #1341 (project freeciv):
Status:None = Ready For Test
Assigned to:None = persia
Planned Release: = 2.6.0
Summary:
Follow-up Comment #3, patch #1341 (project freeciv):
Does this cause behavior changes?
What happens in these situations:
1) You learn Gunpowder, but haven't built any Barracks II yet (all existing
Barracks should be sold)
2) You conquer city with Barracks II, but don't know Gunpowder yourself,
Follow-up Comment #4, patch #1341 (project freeciv):
I'll test these conditions to verify, but from code interpretation:
1) Because, unlike other requirements vectors, obsolete_by is disjunctive,
Barracks is interpreted as obsolete by improvement_obsolete(), so handled
appropriately. Thinking
Follow-up Comment #5, patch #1341 (project freeciv):
1) is now tested, and indeed works as expected: learning Gunpowder causes one
to sell all Barracks, even if one has not built any Barracks II
2) Is also tested: attempts to force worklist inclusion of the obsolete
Barracks I in a city
