Re: Mathematica: Long divison for polynomials

Nasser Abbasi wrote:
et an answer from the mathematica newsgroup
was the right thing to do.
That newsgroup seems to be unsatisfactory for some reasons;
MathGroup is simply moderated.
Unfortunately, that moderation slows discussions to a crawl.



<snip good points>

I am 100% sure that the reason it is moderated has nothing to do with spam. That is just an excuse.

You are probably right about the excuse.

There seem to be a trend for technical companies to want to control discussion about their product, and this is just simply that.

A Wolfram Research employee have been heavily critisised on Wikipedia for the way the Mathematica article was constantly edited to remove anything that was not felt to be in Wolfram's interest. I think the Wolfram employee who was doing the editing has stopped it now though.

An exception I see is Mathworks. Matlab news group comp.soft-sys.matlab is alive and well and serve the Matlab community very well and is not moderated, and there is no attempt by Mathwork to control it or moderate it or create a private forum inside their own control to talk about Matlab.

I would agree with you here. Whilst I don't use Matlab myself, it is clear comp.soft-sys.matlab works well, with numerous posts (*far* more than on comp.soft-sys.math.mathematica). The level of junk is close to zero.

Another very noteworthy exception is Sun Microsystems, who are very open, with very good results - some of this openness has directly benifited Mathematica (see below).

There are several very active Sun related newsgroups:

comp.unix.solaris
comp.sys.sun.hardware
alt.solaris.x86

Sun don't control any of them. Sun's employees contribute regularly to them, often with @sun.com email addressees, but also gmail and similar.

I am aware of at least two Mathematica issues which were solved on the Sun newsgroups.

1) Some years back, when Mathematica ate up loads of CPU time in Solaris 10, the reason was found by a Sun employee (Casper ***), after numerous rapid exchanges between Casper and myself, with me and others posting various outputs from 'truss' which showed the system calls Mathematica was making. (Wolfram Research's support did not know the answer).

IMHO, the delays introduced by a moderated group would have prevented that problem being solved.

2) With a hack I posted a week or so ago, Mathematica will run on Solaris x86 on Intel chips, despite the fact it will only run on AMD CPUs on Solaris x86 as shipped by Wolfram Research.

Again, it was the rapid exchange of views from many people that pointed me to the two libraries that needed changing which allowed Mathematica to run on Solaris x86 using an Intel CPU.

Whilst Sun used to sell SunOs and then Solaris, they now give it away free and have made most of the source code open too. Sure they have set up tons of forums on opensolaris.org

http://opensolaris.org/os/

but posts appear immediately - no moderation takes place.


I really think that trying to control technical discussions about a company product is not healthy in the long run. It will turn off many people about your products, since many will see anything good said about it there as biased, and it will hurt those who you are trying to sell or use your products.

Exactly. As happens on Wikipedia. There were various 'tags' attached to the Mathematica article related to its bias.

I think Mathematica is a great product, one of the best (even if some here do not think do), but unfortunately there is no open and free forum for the Mathemetica community to discuss this excellent software. (having to wait 48 hrs to see an answer to a techical question is simply not acceptable).

Agreed.

.