Chi Zhang

I guess the reason is that if there is `(0.0, -0.0)` in a fact `abc`, and `abc` is a subgoal of a rule `abc(A, A)`, in this case 0.0 not...

I have a question that whether all comparisons under the compiler mode are float-point comparisons? But in compiler mode, in the first example, 0 equals to -0, and in second...

It is interesting! But I'm afraid I don't have time to fix this bug. We are primarily interested in designing and evaluating a new testing approach, I wonder whether it...

Hi, I meet another situation that have the same issue as this, but not the same reason. I guess the reason is there is difference in input handler between interpreter....

Hi, I find another case, I don't know whether they are the same case: ``` .decl a(A:float) .decl b(A:float) a(-0). b(A) :- a(A), 0.0 = A. .output b ``` I...

I found another case, not sure if they have the same reason, so just report here. In interpreter mode, the following program return `0`, but if I remove `inline` or...

I get a simplest test case: ``` .decl buyh(A:number, B:number, C:number, D:unsigned, E:unsigned) .decl xcgl(A:symbol, B:float, C:float) magic .decl eozn(A:unsigned, B:symbol) buyh(-2, -2, -2, 6, 6). buyh(0, 0, 0, 7,...

Maybe this simpler test case more useful: ``` .decl buyh(A:unsigned) magic .decl eozn(A:unsigned) buyh(6). buyh(7). buyh(3). eozn(E) :- buyh(E). eozn(E1)

I reduce it to a simpler version: ``` .decl bydd(A:number) .decl unoa(A:number, B:number) inline .decl ycuc(A:number) .decl lwug(B:number) bydd(-3). ycuc(0). unoa(B/B, B) :- bydd(B). lwug(K) :- ycuc(K), unoa(K, K). .output...

Hi, I reduce this test case: ``` .decl a(b:unsigned) .decl h(c:unsigned, d:unsigned) inline .decl i(c:unsigned) a(0). h(d, d) :- a(d). i(min(max(c,c),8)) :- h(c, 8). .output i ``` I guess it's...