Poor interaction between resolution caching and FastNonminimal

[jkyang92]

Somewhat annoyingly, the following code yields an error

needsPackage "FastNonminimal"
R = ZZ/101[x,y,z];
I = ideal(x,y,z);
res I;
r = res(I,FastNonminimal=>true);
betti(r,Minimize=>true);

Running this gives

error: cannot use Minimize=>true unless res(...,FastNonminimal=>true) was used

This seems to be because of caching.

[mahrud]

Interesting. It makes sense to me that res should cache and return the better result, but indeed annoying that betti would not realize that it can simply ignore the Minimize=>true option.

ps: I suspect you know, but res(ideal I_*, ...) is a good way to clear the cache.