Re: feature request againts pubserver cheats

On Wed, 8 Oct 2003, Horn G=E1bor wrote:
> I'd like to ask if it's a big work to implement that after setting
> endyear a fixed time of countdown would start before the game actually
> ends, which would restart if an other player changes the endyear.

Maybe: Could it be possible to restrict endyear so that its minimum value
is current year + 1 (assuming it ends after endyear and not at endyear)?
This way you always get one turn to change endyear after it is set.

> he waited while we press turn done, the suddently luxed and set endy...

I do not like that it is possible to improve your score merely by
adjusting the tax rate. I suggest that either, when score is calculated,
everyone's tax is set to the same and specialists are transformed into
workers, or happiness/luxury is counted with the same weight as science
output, gold income and shield production.

> a, do the devs care about this? Or is it just me and a few other players
> who feels sad to see where pubserver's 'morale' is going to?

For our part, I do not think the problem is lack of care, but lack of
knowledge about how pubserver games work and what problems there might be.
I do not play much on pubserver myself (less than one game a month).

> b, is there a chance to implement that countdown (or any other solution)
> to stop endy cheats?

If we can agree on a well-designed and simple solution to the problem,
then I will be happy to implement it.

Here is something I have suggested before: We implement a voting mechanism
for all CTRL level commands done during the game. All players get INFO
level access instead. It works like this

=09/vote endyear 500

Then _next turn_ all players are notified that a vote is in progress, and
can vote yes or no like this

=09/vote yes

or

=09/vote no

If more 'yes' votes are counted than 'no' votes before turn ends, then the
command specified in the vote gets executed by the player who suggested
the vote as if that player had CTRL access.

The only thing I can think of that is not so nice about this is setting
timeout or a player to AI when there is no timeout, if not all players are
at their keyboard. Then the game cannot continue.

Note that it would be possible to

=09/vote cmdlevel ctrl player1

and make player1 a judge for the game, to avoid such situations.

What do people think about this idea?

> c, is there a chance for implement client authorization on pubservers
> sometime to stop the multi-nick cheats?

Client authorization is written for 1.15.0 and I believe it will be
implemented for pubserver when that version is released.

> d, is there a need for feedback/suggestion for rethink the ranking
> system to reflect the fact of alliances/gangs? (so if 8 player kills a
> good one in agangbang the 'winner one' wouldn't get even more points
> for this than if he'd do it in a duel...which would be MUCH harder and
> worth of reward than a stupid gang)

Teams members get their scores averaged. We could do the same for
alliances. This would mean that if you endyear when you are obviously
outgunned but the enemy alliance consists of many weaker players, you
could still win...

=2E..I guess this does not solve the problem, however. Suggestions on how t=
o
solve this are welcome.

  - Per