[Freeciv-Dev] Re: (PR#7252) further has_poles fixes

["rwetmore@xxxxxxxxxxxx" <rwetmore@xxxxxxxxxxxx>]

<URL: http://rt.freeciv.org/Ticket/Display.html?id=7252 >


These changes are just a further step down the wrong road which will make
future proper map generation and handling more difficult and full of
maintenance headaches. Syela coding practices are deprecated, I hope.

They should be rejected along with previous steps in this direction and
the few minor issues corrected as in the corecleanups or built in
conformance to a proper extensible form.

Jason Short wrote:
> <URL: http://rt.freeciv.org/Ticket/Display.html?id=7252 >
> 
> This patch includes three fixes for maps with no poles.
> 
> - In make_forests, we only want a bias toward equatorial forests if the 
> map has poles.

Linking presence of poles to latitudinal probabilities for forests is not
the way to go.

> - In make_rivers, we don't need a check for polar positions since there 
> is already a check for arctic terrain.  Unless the map has no poles, in 
> which case we don't want this check anyway.

Building hardcoded oddball special case behaviour into terrain types that
are user configurable by rulesets is not the way to go.


> - In make_fair, we again don't need a check for polar positions since 
> these will fail the terrain_is_clean check anyway.

Assuming conditions which are not obviously or required for given behaviour
is one of the incedibly poor programming practices that means a later
change is almost certain to break this.

> The latter two changes also help pave the way for iso-maps.

Nonsense. The concepts are again unrelated and should not be linked.

> jason

Please go back and reread the discussions about special casing separatepoles
behaviour form generic map features which may exist currently or in a prper
future RFC for map generation.

Please stop this stupid practice of adding hack on hack based on ignorance
of the underlying code base and rationale for it.

Cheers,
RossW
=====