[Freeciv-Dev] (PR#9409) Rollback: improvements to set_ratio

["Marcelo Burda" <mburda@xxxxxxxxx>]

<URL: http://rt.freeciv.org/Ticket/Display.html?id=9409 >

Note to Jason after rollback the buged patch.
New version has bugs

1)some times the xsize/ysize <> Xratio/Yratio this is a bug, this is
there because the round to int are made in different way for x/ysize the
right way is the original one!
2)map_temperature assume size never are odd, change it force us to
review temperature math, (see the foldering math)
3)hex code assume (i think in us last exchange) assume y size never are odd
4) i make this assumsion in all of my patches, mmm maybe the overview code

conceptually wrong (but not exact a bug)
A)when making a init map code we has the choice to get priority on ratio
or on size, I choice ratio then some time size is not exact  but near
the expected one. the size 1 is not exact 1000 tiles but near this.
and this not need to be different, size 1 is a very small map only
suitable for testing. You useless  spend time on it. get exacts size for
so extreme small maps are not possible when the even numbers are needed.

b)allow odd numbers for size are unnecessary and allow to make maps and
scenario useless in some topologies needing even sizes!  there is best
if all maps has even from today!

c)You delete the coding testing size arguing these code are never used.
this is wrong, if you change the default ratio of one topologies or
include a new one with a bigger ratio you will get a assert! this force
you to get a smaller maximal size for all topologies, this is the exact
opposed of a improvement. if code are ugly, you can improve it not
delete it.  

Then the new path has bugs and this improve nothing. 
I am surprised you not asked me about my code, probably i has comment to
you before make changes. Or you not want these comments? 
Best regards
 Marcelo