[Maxima] featurep(e, real)

[Stavros Macrakis]

------=_Part_33042_27193483.1147725630604
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On 5/15/06, Robert Dodier <robert.dodier at gmail.com> wrote:
>
> On 5/15/06, Barton Willis <willisb at unk.edu> wrote:
>
> OK by me. But can we suppress some of the asksign nuttiness
> that crops up in rectform ?


What "nuttiness" did you have in mind?  Do you mean you want to suppress it
in the featurep case, or in general?

What do you suggest the value of featurep(sqrt(x),real) be in the absence o=
f
asksign?  I can think of several reasonable answers:
     false  -- because of course for x<0 it is imaginary
     x >=3D 0  -- a boolean expression
     true  -- because we treat the domain of sqrt as real

The problem with the boolean expression is that most users of featurep woul=
d
want to dispatch on the value, and aren't prepared to themselves return a
conditional expression....

>
> > (4) Maxima should signal domain errors ASAP.
>
> Well, I'm not so sure about that.


Robert, I know that errors are inconvenient in non-interactive usage, but a=
s
Barton says, they catch the problem before it propagates (possibly giving
incorrect results).  If you need to catch the error, what is wrong with som=
e
error handling mechanism (preferably more sophisticated than errcatch...)?

          -s

------=_Part_33042_27193483.1147725630604
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On 5/15/06, <b class=3D"gmail_sendername">Robert Dodier</b> &lt;<a href=3D"=
mailto:robert.dodier at gmail.com">robert.dodier at gmail.com</a>&gt; wrote:<div>=
<span class=3D"gmail_quote"></span><blockquote class=3D"gmail_quote" style=
=3D"border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; p=
adding-left: 1ex;">
On 5/15/06, Barton Willis &lt;<a href=3D"mailto:willisb at unk.edu">willisb at un=
k.edu</a>&gt; wrote:<br><br>OK by me. But can we suppress some of the asksi=
gn nuttiness<br>that crops up in rectform ?</blockquote><div><br>What &quot=
;nuttiness&quot; did you have in mind?&nbsp; Do you mean you want to suppre=
ss it in the featurep case, or in general?
<br><br></div>What do you suggest the value of featurep(sqrt(x),real) be in=
 the absence of asksign?&nbsp; I can think of several reasonable answers:<b=
r>&nbsp;&nbsp;&nbsp;&nbsp; false&nbsp; -- because of course for x&lt;0 it i=
s imaginary<br>&nbsp;&nbsp;&nbsp;&nbsp; x &gt;=3D 0&nbsp; -- a boolean expr=
ession
<br>&nbsp;&nbsp;&nbsp;&nbsp; true&nbsp; -- because we treat the domain of s=
qrt as real<br><br>The problem with the boolean expression is that most use=
rs of featurep would want to dispatch on the value, and aren't prepared to =
themselves return a conditional expression....
<br><blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(2=
04, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>&gt; (4) =
Maxima should signal domain errors ASAP.<br><br>Well, I'm not so sure about=
 that.
</blockquote><div><br>Robert, I know that errors are inconvenient in non-in=
teractive usage, but as Barton says, they catch the problem before it propa=
gates (possibly giving incorrect results).&nbsp; If you need to catch the e=
rror, what is wrong with some error handling mechanism (preferably more sop=
histicated than errcatch...)?
<br></div><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; -s<br>=
<br></div><br>

------=_Part_33042_27193483.1147725630604--