Complete.Org: Mailing Lists: Archives: freeciv-dev: December 2002:
[Freeciv-Dev] Re: (PR#2631) [Fix] ACTIVITY_EXPLORE infinite cycle
Home

[Freeciv-Dev] Re: (PR#2631) [Fix] ACTIVITY_EXPLORE infinite cycle

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
Cc: freeciv-dev@xxxxxxxxxxx
Subject: [Freeciv-Dev] Re: (PR#2631) [Fix] ACTIVITY_EXPLORE infinite cycle
From: "Gregory Berkolaiko via RT" <rt@xxxxxxxxxxxxxx>
Date: Mon, 23 Dec 2002 09:55:06 -0800
Reply-to: rt@xxxxxxxxxxxxxx

On Mon, 23 Dec 2002, Per I. Mathisen via RT wrote:

> On Sun, 22 Dec 2002, Gregory Berkolaiko via RT wrote:
> > Attached are two possible fixes for PR#2631.Both work, both are kludges.
> >
> > Actually, this bug is _very_ rare, because you'd need to manually set a
> > AI-controlled unit to be explorer and only then you'd have it.But still
> > it is a bug.
> 
> I'd prefer a real fix, then. Separate out the
> is-there-more-to-explore-or-huts-to-find part of ai_manage_explorer into a
> function of its own, which can be called from within
> handle_unit_activity_dependencies without risks of recursion. This would
> mean a little redesign of ai_manage_explorer, but then, someone was
> working on that already...

This is neither "real" nor a "fix".

The problem is that we _need_ to call ai_manage_explorer from
handle_.._request.  And this call must really make explorer walk around,
not just speculate how many huts are still standing, how many native men
are still heathen and how many native women are still not raped.

This is because when we hit "X", we expect the explorer to go do it.
Compare with what happens when you hit "A" on a settler (nothing happens, 
until end of turn IMO).

Fix#1 is more proper of the two, of course, as it distinguishes between 
requests generated by hitting "X" (which ask for immediate action) and 
requests routinely generated by update code.

G.




[Prev in Thread] Current Thread [Next in Thread]