Follow-up Comment #1, patch #4401 (project freeciv):
> Anyway, question: what is the current situation with nreqs vs
> present=FALSE (previously negated=TRUE) after all the recent
> work by Marko and Emmet Hikory?
Emmet stated that he had plans to work on that field, so I mostly left it for
him. He has been gone for quite a long time already, though, so maybe someone
else should pick this up.
Ignoring Emmet's more advanced dreams, and considering the absolutely
necessary basics only:
- It's basic boolean logic that use of right combination of ANDs & NOTs
("present = FALSE" in main reqs list) can achieve exactly same things as any
combination of ORs & NOTs ("present = FALSE" in nreqs list). Having them both
available would only serve to make it easier to non-logicians to formulate
- Goal was to get rid of 'nreqs', but as it has been always the standard way
of writing negated effect requirements, and "present = FALSE" has never been
used there, it's practically guaranteed not to be handled correctly in all
parts of the code, i.e., there's bugs.
For AI I had one simple fix in mind: Currently AI "handles" nreqs by
completely ignoring them. Having it to do same for "present = FALSE" reqs
would retain current AI functionality even if ruleset is written using
"present" rather than "nreqs".
Reply to this item at:
Message sent via/by Gna!
Freeciv-dev mailing list