On 20.09.2012 12:36, Michael T. Pope wrote: > On Thu, 20 Sep 2012 08:28:17 AM Michael Vehrs wrote: > >> This looks like a rather hairy problem, since we are basically doing >> everything right already. Except that in the case of starvation, we need >> to look into the future. >> > The higher level fix (starve.diff) is still an option if you think the > production code is doing the Right Thing, however ISTM that the problem is > that the decision to build is too simplistic. Rather than just `>= 200 food', > it should be `>= 200 + one-unit-consumption if net food production is -ve'. > Not elegant I admit. >
Not particularly. But not as bad as I initially thought, either. > > >> Also, I don't understand why >> ServerTestHelper.newTurn() doesn't update my test game properly. >> > No idea. Last time I looked (the teaching tests, ugh) it worked for me. > > Cheers, > Mike Pope > > I see that I was not using the ServerTestHelper correctly, now. I should have looked at the server model instead of the common model, of course. Regards Michael ------------------------------------------------------------------------------ Got visibility? Most devs has no idea what their production app looks like. Find out how fast your code is with AppDynamics Lite. http://ad.doubleclick.net/clk;262219671;13503038;y? http://info.appdynamics.com/FreeJavaPerformanceDownload.html _______________________________________________ Freecol-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freecol-developers
