[Freeciv-Dev] Re: (PR#2650) Overflow in military_amortize

["Gregory Berkolaiko via RT" <rt@xxxxxxxxxxxxxx>]

Quoting "Per I. Mathisen via RT" <rt@xxxxxxxxxxxxxx>:

> Well, there isn't a problem that can't be solved with a few extra
> parameters! We can pass pcity or find_city_by_id(punit->homecity) to
> military_amortize() and use that city's shield_surplus instead of an
> average value.
> 
> This avoids the pathological case of most-cities-have-zero-surplus, which
> means the one city with lots of surplus shields will think "gee, it takes
> 1243928 turns to build that tanks and drive it to the enemy, so lets
> not!", even though it could build it in 3 turns and drive it there in 1!

true true

but it also tells the city with two shield surplus to give up hope building
units although it couldbe useful too.

> See the attached patch which implements this.

Not attached :(

> I would guess the maintainer to commit such a patch would do some play
> testing and comparisons of it, though ;)

Testing?  No thanks :)

But,seriously, I can't really think of a way to test it other than visually. 
Maybe only how longer/shorter it takes the dominant civ to take over the world?

Also, maybe it'smy impression, but I feel the game gets slower and slower,
especially GUI...  Makes it very time-consuming to test-run things :(

> > > The big question is if we need this at all. Simpe amortize might just
> get
> > > the job done, too.
> >
> > With what delay?
> 
> We might just ignore build_cost altogether, as I think your comment in
> military_amortize() hinted at.

It does...
Mhhh...

I guess I was just scratching my head, asking questions and not hinting at
anything...

But I think your way of doing now it is the most scientific-like, so we
canremove the comment.

G.