Follow-up Comment #6, patch #3345 (project freeciv):
Good catch. I looked at the compilation output again and found the mentioned
warning now. I overlooked it among the information from make, translations,
etc. (I don't have much experience concerning developing in C.) Seems like I
(re)read the testing tips a bit to fast. (The AI vs AI games I ran tested no
effect, the effect activated by nationality and activated by a wonder I gave
using the editor) I'll use --enable-debug from now on. (CFLAGS=-Werror
./autogen.sh && make stops before it reaches aicity.c)
As far as I understand the function if aicity.c is an utility function for
effects. Would something like the number of known foreign cities not hidden by
a fog of war (perhaps divided by the number of player cities) weighted to be
reasonable be an appropriate measure here? Is this function called often
enough that its really important to be fast vs accurate? Are the built in AI's
still so omniscient (as the other comments indicate) that the function still
is about pretending its valuable?
I misunderstood what qualifies as a change to the network protocol. It will be
removed. My bias had me consider the problem from an angle closer to "will it
make it quick and easy for software to know things are different assuming the
other party isn't lying" than to "will it let human players see if they can
play without installing a new client".
Reply to this item at:
Message sent via/by Gna!
Freeciv-dev mailing list