Maple bugs: idea by Peter Luschny
Vladimir Bondarenko
This beats all. Even me, a connoisseur of Maple bugs, did not
expect this. I had no idea to look for any plot bugs, I just
was sitting and placidly analyzing the Maple's failure with
maximize(z*ln(z^2+1)/(z^2+1), z= 0..infinity);
Then a shock came, bang!
Only Maple V Release 3 of 1994, and Maple V Release 4 of 1995
can draw smooth graphs. All next Maple versions produce a ladder
with more than a dozen conspicuous rungs.
plot(z*ln(z^2+1)/(z^2+1),z= 2.93..2.94);
All this does not help...
plot(z*ln(z^2+1)/(z^2+1),z=2.93..2.94, numpoints=2000);
plot(z*ln(z^2+1)/(z^2+1),z=2.93..2.94, numpoints=20000);
plot(z*ln(z^2+1)/(z^2+1),z=2.93..2.94, numpoints=20);
plot(z*ln(z^2+1)/(z^2+1),z=2.93..2.94, numpoints=10);
Consider MuPAD 3.1 - a nice smooth picture.
plotfunc2d(z*ln(z^2+1)/(z^2+1),z= 2.93..2.94);
Consider Mathematica 5.1 - a nice smooth picture.
Plot[z Log[z^2+1]/(z^2+1), {z, 2.93, 2.94}]
Consider Derive 6.1 - a nice smooth picture, too.
Peter Luschny wrote on Dec 23 2003 in the message Re: euler(0,1) shows
Maple bugs are really ubiquitous (comp.soft-sys.math.maple)
PL> Yes, Vladimir's marketing is not perfect, yet. But what
PL> about the slogan: "Maple - unsafe at any evaluation"?
For a year, I had been resisting this slogan sturdily...
But after this eerie plot, I think I should accept it.
Peter, how much does it cost to use your idea? ;)
I was wondering do you agree for 51% of net profit? ;-)
Best wishes,
Vladimir Bondarenko
http://www.cybertester.com/
http://maple.bug-list.org/
http://www.CAS-testing.org/
............................................................
Joe Riel
> Hello symbolic computation community,
>
> This beats all. Even me, a connoisseur of Maple bugs, did not
> expect this. I had no idea to look for any plot bugs, I just
> was sitting and placidly analyzing the Maple's failure with
>
> maximize(z*ln(z^2+1)/(z^2+1), z= 0..infinity);
>
> Then a shock came, bang!
>
> Only Maple V Release 3 of 1994, and Maple V Release 4 of 1995
> can draw smooth graphs. All next Maple versions produce a ladder
> with more than a dozen conspicuous rungs.
>
> plot(z*ln(z^2+1)/(z^2+1),z= 2.93..2.94);
Actually, all versions of Maple can produce the smooth plot,
they just diplay it improperly in the GUI (which is a bug).
The problem lies in the plot driver used by the Maple GUI.
If you use, say, the postscript or x11 plot driver then
the rendered plot is very smooth.
Joe
Peter Luschny
original post of Vladimir. This reply is meant for him.
------------------------------------------------------
> "Vladimir Bondarenko" <v...@cybertester.com> wrote:
> Hello symbolic computation community,
> Peter Luschny wrote on Dec 23 2003 in the message Re: euler(0,1) shows
> Maple bugs are really ubiquitous (comp.soft-sys.math.maple)
PL> Yes, Vladimir's marketing is not perfect, yet. But what
PL> about the slogan: "Maple - unsafe at any evaluation"?
> For a year, I had been resisting this slogan sturdily...
> But after this eerie plot, I think I should accept it.
> Peter, how much does it cost to use your idea? ;)
> I was wondering do you agree for 51% of net profit? ;-)
Hi Vladimir,
I allow a 90 per cent discount :)
Well this slogan is not completely new, in fact it is the
paraphrase of a famous one, coined by Ralph Nader in 1965.
Nader first gained national attention in 1965 as the
thirty-one-year-old author of "Unsafe at Any Speed".
The book indicted unsafe automobile design in general
and General Motors' Corvair in particular. When it
became publicly known that General Motors had hired
private detectives in an attempt to dig up information
that might discredit Nader, a Senate subcommittee
looking into auto safety summoned the president of
General Motors to explain his company's harassment
and personally apologize to Nader.
from: http://www.nader.org/ecm.html
Somehow the case Bondarenko versus Maple reminds me
to Nader versus GM et alia. And if you succeed and
eventually a better Maple is build, you will become
famous like Nader and might some day be candidate for
the presidential election of the Ukraine :)
Ok, but let me make clear once again, what my intention
was, to propose such a provocation: Besides being true,
of course, it should emphasize, that I would like to see
a public bug-tracer for Maple, like Sun (Java) has, like
in the meanwhile even Microsoft maintains, or like the
competitor Mupad: http://research.mupad.de/bugs.shtml
Such a public bug-tracing system has a great many benefits.
The efforts of the users in discovering and describing
bugs will reflected, the efforts of the Maple development
team will be reflected, and - last not least, other users
will be warned and enabled to avoid known bugs.
I do believe, that such a bug-tracer would *not*
diminish the public respect for Maple, on the contrary,
the respect would increase. It is a kind of modern
public relation between the manufactures of software
systems and their users. Only old-fashioned stalinistic-
capitalistic boneheads are not able to appreciate this.
Yet another story, just because it happened two days ago
and is related to the thread you refer to. There I wrote:
In the news "Constant polynomial has multiple values"
I observed the fact that euler(0,1); -> -1, with Maple(V,5).
Robert then said: "It's still in Maple 9."
Now I found the solution to ugly bugs like these:
Use Maple and Mathematica simultaneously!
Maple(V,5) bernoulli(1,1); -> -1/2
Mathematica (5.0) BernoulliB[1, 1] -> -1/2
Note, this was a thread in comp.soft-sys.math.maple
which, by the politics of comp.soft-sys.math.mathematica,
never could have been published there!
The bigger was my surprise yesterday, when I found in
my email box a letter of Roger Germundsson, Director,
Research & Development, Wolfram Research, Inc.
Roger refers to and quotes the above statements! Then
he says:
One of our continuing objectives is to resolve
even the smallest problems in Mathematica. In
addition to our own large-scale testing effort,
we rely on users like you to bring problems to
our attention. The new developments in Mathematica
5.1 made it possible for us to resolve a considerable
number of open issues, and I am pleased to be able
to tell you the problem you reported was among them.
Please let us know if you have other feedback or
suggestions.
I really liked this. I never-ever saw a comparable response
to the many bug reports here on this list from official
Maple team members. In fact, last year, after this
disastrous Maple release 9, I switched to Mathematica 5.
Clearly, I don't say Mathematica is pure gold. But such
responsiveness to bug reports is a reason which might
keep me stay with their product, or with Mupad.
Regards Peter
P.S. Vladimir, note that it is good practice in newsgroups
not to put names of persons into the subject - at least,
if they are not persons of the 'Zeitgeschehen', persons
of public interest, like, for example Ralph Nader,
Witali Klitschko, Wiktor Juschtschenko and certainly
in the near future, Vladimir Bondarenko :)
--
Look what THEY have done to Juschtschenko:
http://www.spiegel.de/politik/ausland/0,1518,grossbild-415969-332482,00.html
The Bush-administration uses the GM technique against Nader today
against Al-Baradei. See the headline [December 12, 2004, 11:45 a.m.]
http://www.washingtonpost.com/ or
http://www.spiegel.de/politik/ausland/0,1518,332492,00.html
So Vladimir, I whish you good luck with your campaign, but
be careful, please.