Log2 and beta test failures

[Michael Goffioul]

On Tue, May 6, 2008 at 7:35 PM, John W. Eaton <jwe at bevo.che.wisc.edu> wrote:
>  A few fixes are probably OK, but I prefer to avoid littering Octave
>  code with workarounds for bugs in other software, especially when the
>  other software is freely available.

I personally can live with the fact that:

[a,b] = log2(Inf)

returns (NaN,-1) instead of (Inf,0). So whether this is fixed or not is
not really mandatory for me.

Michael.