The two bugs are explained in red in the work sheet.

The two bugs:
1. The local definition of the Σ(t) is viable though
not obvious for constructing formulas from classical
maths litterature.
2. Bézier-Berntein is parametric wrt 't' in the range
0..1____In Lobe, Smath 5346 did support 't:=0,0.01..1'
Smath 6179 does not. It needs to read some default value
for '0'. It raises the suggestion that 6179 should
have an implicit '0 guard'or there is a more sutile bug.
In Lobe, set t:=0,0.01..1, tabulate Lobe => "uncertainty"
caused by t[0]^0, true but pure theory not applicable in
range variable.

Spline Bezier_Bernstein [Lobe Project].sm (50 КиБ) скачан 48 раз(а).

Hello Jean,

Just to recall this once more Zero^Zero

For example, if I put the value 10^-17 or some smaller one for the first value in t range in Lobe variable, it seems that calculation will never stop (did not try to debug this).

Regards,
Radovan

EDIT: Actually, it seems it is due to "Uncertainty." error.

Wrote

Hello Jean,

Just to recall this once more Zero^Zero

For example, if I put the value 10^-17 or some smaller one for the first value in t range in Lobe variable, it seems that calculation will never stop (did not try to debug this).

Regards,
Radovan

EDIT: Actually, it seems it is due to "Uncertainty." error.

Same thing here: the smallest 't' is e-15.
More puzzling: Smath accepts (e-323)^0 = 1,,,(e-324)^0 = "Uncertainty"
The real question is why 5346 takes t=0 but not 6179 [only e-15 or >].
I thing 6179 is missing the default '0' which is quite real whereas no
numerical functions in Pentium reports more than 15 decimals.
Though they are all or most of them in 21 floating point, they are
intentionally truncated to 15 decimals. Thus meet 15 D tabulated references.

Thanks Radovan, Jean