Subject: | Bug in Math::BigFloat - gives wrong results for bpow and bexp |

Hello,
I'm working with Perl 5.10 and tried it on Ubuntu and Feodora.
bexp gives wrong result for high numbers (negative) and takes ages to
calculate, if it is a real negative number such as
-394.84010945716266885 (or even -394.8). bpow gave also wrong result for
power -394.8 of e.
For numbers like -394 it would give the right result.
The code:
my $i=Math::BigFloat->new(-394.84010945715266885, 20)->bexp();
print ("exp: ".exp(-394.84010945715266885).", BigFloat exp: $i\n");
I'd expect it to give the same results, no?
The output:
exp: 3.33517962278651e-172, BigFloat exp:
19871330168979164802000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
0000
Would be happy to hear from you.
Best regards,
Nir.