From: George N. White III on 25 Feb 2007 07:30
On Wed, 21 Feb 2007, mecej4 wrote:
> analyst41(a)hotmail.com wrote:
>> On Feb 20, 1:42 pm, Dan Nagle <danna...(a)verizon.net> wrote:
>>> analys...(a)hotmail.com wrote:
>>>> I suppose this applies to any programming language.
>>>> What does the standard say about pictures as comments ?
>>>> I think it would be a great idea.
>>> You may place any characters you want following a !,
>>> so a url to a graphic is as ok as anything else.
>>> Dan Nagle
>>> Purple Sage Computing Solutions, Inc.
>> Well,pushing the idea further, we should be able to put both the
>> source code and text and graphic comments all together at an editable
>> web page and some kind of IDE interface should be able extract only
>> the source code for compilation.
> The idea is neither far-fetched nor new. Knuth the Great designed his WEB
> system nearly four decades ago with similar objectives. He published some
> papers on "literate programming". The WEB system did not become as hugely
> successful as TeX, however.
What do you mean by "success"? TeX is still maintained as WEB
sources, so if you think WEB was a key element in the success of TeX, then
WEB has be considered a success. If you think that WEB has hampered other
implementations of TeX-quality typesetting then WEB could be considered a
A number of really important systems (e.g., R, the S-Plus "clone") use
LaTeX for documentation. Many journals are published using TeX. Since
WEB is vital to maintaining TeX, the fact that only a few dozen people are
currently programming in WEB is not a reasonable measure of success.
I used fweb to great advantage when I was given the task of organizing a
legacy collection of Fortran codes. I was at least the 3rd person given
the task, but the first to produce a useable library. The ability to
include mathematics in the text and cross-referencing were a huge
advantage, but in the long term maintaining the system is a headache
because you have to get fweb running on the platform du jour.
There are a number of literate programming systems still being actively
used. I like ProTeX:
Eitan Gurari, "TeX and LaTeX: Drawing and Literate Programming".
McGraw-Hill, 1994. ISBN 0-07-025208-4 / 0-07-911616-7 (out of print)
ProTeX doesn't need tangle and weave -- all the heavy lifting is done by
TeX (remember that Knuth developed TeX back when many of us were still
using punched cards and had to write code that could run in 32MB). To use
ProTeX on a new platform you only need to get TeX working. ProTeX can be
used with ConTeXt as well as with LaTeX. Indexing does, however, require
external code to sort the list of variables. The big benefit for me is
that ProTeX lets me use more than one language. For some teaching
documents it is useful to present the same algorithm coded in Fortran,
Matlab, IDL, and S-plus because different audiences will generally know
only one or 2 of the languages. Real-world systems often use more than
George N. White III <aa056(a)chebucto.ns.ca>