Five Years of CSC Mathematical Topics
This article was originally published in the CSC News magazine (Vol. 9, No 3, October 1997).
"If you build it, they will come" (quote from the movie Field of Dreams)
CSC Mathematical Topics is a popular World Wide Web service on mathematics. This service was originally established in 1992 as a Gopher service called simply "Mathematical Topics". Since then both the usage and contents of the service have grown steadily.
I started in 1992 by offering a list of interesting and useful Gopher links. This service was used about 200 times each day in the end of 1993. In fact, the usage was more frequent than I anticipated. There seemed to be a real need for a mathematics information service, and moving to the Web system made sense already in 1993.
Now, in the middle of 1997, CSC Mathematical Topics is listed in about 700-1000 documents outside the Web servers at CSC, according to results of AltaVista searches. CSC Mathematical Topics gets about 2600 visits each day, or about 80,000 in each month. Of course, many of the visits are due to search robots such as AltaVista. On the other hand, many visitors are not listed due to local Web caches and proxy servers.
Contents of the service
What does CSC Mathematical Topics offer to the Web user?
Starting in 1994, I have archived the NANET Digest mailing list for fast local access. Combined with a search service, this archive makes a very useful information resource for the numerical analysis community.
Furthermore, CSC Mathematical Topics contains detailed descriptions of CSC guidebooks published since 1992. The topics of the guidebooks include mathematical software, numerical methods, programming, usage of computers etc. Some of the guidebooks have also on-line versions. In addition, example codes from the guidebooks are available on-line. Most of the guidebooks are written in Finnish.
An English-Finnish-Swedish mathematics dictionary has become quite popular, since you can search interactively for translations of mathematical terms. Figure 1 shows an example of using the mathematics dictionary. The database contains about 2000 mathematical terms in English and Finnish, and some have also Swedish translations.
![[vectorfig.gif]](http://radio.weblogs.com/0112083/images/2002/08/22/vectorfig.gif)
Figure 1. Mathematics dictionary.
Another popular resource are examples of mathematical visualizations. This part of the Web server has also been included on CD-ROMs published in such places as USA and South Korea.
The original contents of CSC Mathematical Topics were lists of mathematical Web servers organized by topic. This part of the server has become less important, since there are now several exhaustive search services and well-maintained topic indexes. Also, maintaining a list of Web servers tends to be a time-consuming project, so I have tried to minimize the work by only including well-maintained and important sites. Instead, I have put more emphasis on the local documents and resources.
How has Internet evolved?
I wrote an article titled "Using the Electronic Highway in Mathematics" for the October 1994 issue of CSC News (Vol. 5, Nr 3). At that time the Internet was still something new, and the main emphasis of the article was on the possibilities offered by the World Wide Web system.
Dissappointingly, many of the predictions I made in that article failed to happen, or were only partially successful. One of the biggest reasons for this is the commercialization of the Web. As an example, the current browser vendors do not emphasize the needs of the scientific community.
In 1994, the main Web browser was NCSA Mosaic, and the academic community was very visible on the Web. At that time, all browsers could show (almost) all HTML pages on the Web. Nowadays there are basically only two players in the field: Netscape and Microsoft. However, both companies have added their own features to their respective browsers, and the Web has split in two. There is a multitude of HTML variants, and you may need specific system and software versions to be able to view a particular Web page.
Some pessimistic predictions I made in the 1994 article are all too true today. For example, unwanted e-mail has become a much bigger nuisance than it was in 1994. Also, Usenet News has become, for all practical purposes, a gigantic chat room where there is very little useful information. In 1994 I actively followed about a dozen newsgroups - today, I follow half a dozen, mainly when I have a current interest in a particular subject.
Java: a solution or a problem?
Recently I added a couple of Java applets to the visualization gallery at CSC Mathematical Topics. Previously there was a collection of MPEG movies for illustrating similar phenomena. In contrast, Java makes it possible to study a topic by yourself using a Java applet as a tool.
Figure 2 shows an example of visualizing one-dimensional cellular automata. The user can select the properties of the cellular automaton using the standard notation developed by Wolfram, and the applet shows how the system evolves. The Java source code is also included for others to use.
![[ca1d.gif]](http://radio.weblogs.com/0112083/images/2002/08/22/ca1d.gif)
Figure 2. Visualizing one-dimensional cellular automata.
A potential problem with Java is the same as with HTML: already there are several different flavours. For example, Microsoft tries hard to keep developers writing code solely for Windows, and apparently sees Java as a threat. For this purpose, other vendors have established a "100% Pure Java" initiative to keep Java implementations and programs compatible.
If you are interested in providing tools for studying and learning scientific topics, I recommend having a look at Java. A possible inconvenience is the sparsity of support for numerical and symbolic computing in Java. Mathematica, MATLAB, or Fortran 90 have better facilities for that kind of work. Hopefully the situation improves, as Java tools get better. Of course, Java will never be as fast as Fortran, but for demonstrations and small-scale experiments the speed may be enough.
A standard for mathematical documents?
Currently there isn't any universally accepted standard for publishing mathematical documents on the Web. This is one of the biggest problems of using Internet for scientific communication. Some possible solutions are PostScript, Adobe Acrobat (PDF files), HTML with GIF images for equations, or TeX/LaTeX.
All of these approaches have their drawbacks. For example, PostScript and Acrobat files are often best viewed on paper, and HTML with GIF images gives often quite unesthetic results. Also, there are several different versions of HTML in use, such as HTML 1.0, 2.0, 3.2, and 4.0. Furthermore, LaTeX works only if you use standard document classes and fonts, and have access to a high-quality LaTeX implementation.
A recently proposed standard called MathML (Mathematical Markup Language) may be a good enough solution for publishing mathematical texts on the Web. MathML makes it possible to show high-quality mathematical typesetting within a Web browser. MathML could also be used as a medium for inputting mathematical expressions to Mathematica or some other software for symbolic computing.
MathML contains both visual and semantic markup. Because of this, documents are relatively easy to translate to and from MathML. For example, a LaTeX translation is not too difficult to define, and is in fact already in planning stages.
MathML is based on the XML markup language (Extensible Markup Language), which is also used by chemists to define the Chemical Markup Language. XML is a subset of SGML (Standard Generalized Markup Language), an international standard for defining document types.
If you have an interest in mathematical typesetting and publishing, please consider supporting the MathML project. See the Web address http://www.w3.org/XML/ for more information about XML. The MathML standard is reviewed in the Web address http://www.w3.org/Math/.
