<URL:
http://bugs.freeciv.org/Ticket/Display.html?id=20606
>
I'd like to propose a solution to a longstanding problem
with
Freeciv, but I need some help to implement it. The problem
is the
punishing effect of corruption in the early stages of the
game.
Essentially Freeciv offers only one solution to the
effects of
corruption - change to another form of government. But what
if
you could reduce corruption before you were able to switch
to
another form of government?
Here's my idea: reduce the corruption losses in each
city by a
percent equal to the percent of happy citizens in that city.
Here's
an example with some made up numbers:
.You have a size 1 city that is located 6 tiles away from
the
capital.
.That city's center is on a spice tile.
.Its only citizen is working a wine tile.
.The government is still Despotism.
The base output of that city is 1-1-6 (1 food (3-2), 1
shield,
and 6 trade). But after the effects of corruption are
computed, the
output drops to 1-1-3.
Now change the tax rate to 40% Science / 60% Luxury. Now
that
city is producing 2 luxuries from its 3 trade points. 2
luxuries
change that city's only citizen from content to happy.
1 happy citizen in (a city of) 1 is a 100% happy
percentage. So
the corruption effects in that city are reduced by 100% (to
0). Now
the city produces 6 trade points instead of 3.
Now you can change the tax rate to 60% Science / 30%
Luxury /
10% Gold. That will be enough to keep that city happy and
to allow
you to reap a decent amount of science and gold from that
city (even
though its still governed under Despotism). No more need to
hurry
up to research a new form of government.
Size 2 cities could reduce their corruption penalties by
50%
if one citizen is happy or 100% if both citizens are happy.
Size
3 cities could get 33%, 66%, or 100% reduction. Larger
cities would
work similarly. As with celebration, all other citizens
have to be
content (or specialists).
The research flexibility this change offers is key. It
makes
numerous alternative research strategies viable. Without
it, there
is effectively only one strategy - research a new form of
government
as quickly as possible. Researching anything else only
extends the
duration of the effects of corruption (which overwhelm any
benefits
that other strategies might provide).
This simple mod that I've made implements the basic
function
that I've described; but it has some problems. I believe
those
problems result from where the mod is in the calculation
sequence.
I've tried to figure out where to place the mod so that
it works
reasonably and doesn't cause undesirable side effects. But
I'm not
getting anywhere with that. This is where I need the help
of the
big brains on this list.
I put my mod in city_waste. One of the symptoms of that
placement is that the first tax rate change (setting
luxuries to
60%) doesn't take effect until the following turn. By the
next
turn the happiness due to luxuries is apparently recalc'd.
Only
then does the corruption reduction take effect.
However, even on the following turn, you can't switch
that size
1 city's citizen to work a different tile. That's
because, as soon
as you remove the citizen from the tile that he's on, he
becomes an
entertainer.
The entertainer is *not* happy (because he's a
specialist) and
so the corruption effects instantly return. Placing the
citizen on
a new tile doesn't necessarily produce enough luxuries
(from the
corruption depleted trade output) to make the citizen happy
again.
That means that you have to bump the luxury rate back to
60% and
wait for the next turn again. Clearly this effect is
annoying.
I assume that if the mod were placed somewhere else
(perhaps
followed by a recalc of the effect of luxuries), then these
problems
could be fixed. But I don't know where the mod can safely
be
inserted into the existing code. Set_city_production seems
to be
delicately balanced to make the existing functions work the
way
that they are supposed to.
Can anyone suggest a more effective placement for this
mod (or
one like it)?
I also considered implementing this feature as an effect.
Unfortunately, I couldn't figure out how to do it. One of
the
pieces is already available: EFT_OUTPUT_WASTE_PCT.
Unfortunately, I couldn't figure out how to access the
city size
and the number of happy citizens to code them into
effects.ruleset
reqs. Is there a way to do that? (Also I would also need
to be
able to access the numbers of unhappy and angry citizens -
for
nreqs.)
Even if I could use effects to implement this, it would
probably
only be practical to add effects rules for city sizes up to
3. That
would require only 6 rules. Each additional city size would
require
N additional effects rules (for Nth city). Besides, by the
time
cities grow to size 4, many players may have already
completed the
research they need to change their form of government.
Regardless, here is the patch. Any advice or suggestions
on making
it work better will be appreciated.
-Eddie
Index: common/city.c
============================================================
=======
--- common/city.c (revision 12298)
+++ common/city.c (working copy)
 -2177,7
+2196,7
***********************************************************
***************/
int city_waste(const struct city *pcity, Output_type_id
otype, int total)
{
- int penalty = 0;
+ int city_size, happy_count, penalty = 0;
int waste_level = get_city_output_bonus(pcity,
get_output_type(otype),
EFT_OUTPUT_WASTE);
int waste_by_dist = get_city_output_bonus(pcity,
get_output_type(otype),
-2221,6
+2240,18
penalty -= penalty * waste_pct / 100;
+ /* reduce waste based on pct happy citizens in city (also
must
+ * have no unhappy citizens). This is particularly
helpful in
+ * the early stages of the game (before new government
types
+ * have been researched). */
+
+ if (pcity->ppl_happy[4] > 0 &&
pcity->ppl_unhappy[4] == 0
+ && pcity->ppl_angry[4] == 0) {
+ city_size = pcity->size;
+ happy_count = pcity->ppl_happy[4];
+ penalty -= (penalty * happy_count) / city_size;
+ }
+
return MIN(MAX(penalty, 0), total);
}
|