RECYCLING PREVIOUS DOCUMENTS
FOR DISTANCE EDUCATION
Jean-Michel Hufflen
LIFC (EA CNRS 4269), University of Franche-Comt´e, 16, route de Gray, 25030 Besanc¸on Cedex, France
Keywords:
Presentational education, Distance education, Course text, On-line course, Case study, L
A
T
E
X, PDF.
Abstract:
Given a course document concerning a teaching unit in Computer Science and written for presentational
education, we explain how we took as much advantage as possible of it for the same teaching unit, but adapted
to distance education. In particular, whenever we are building a new version, we are able to update this
document once, so that changes are automatically applied to both versions, for presentational and distance
education. If the original document is clearly written and well structured, the adaptations we propose should
be easy to be put into action.
1 INTRODUCTION
As result of greater and greater interest in distance
education, most universities have increased such of-
ferings. Since this field aims to deliver education to
students who are not physically on site, it is of inter-
est for students who have a full-time job or are very
distant, possibly living in another country. As an ex-
ample of an institution delivering distance education,
the CTU
1
, part of the University of Franche-Comt, al-
lows students to get the whole of the units of a master
in Computer Science. Of course, this university still
provides curricula in presentational education, which
remains the ‘traditional’ way of teaching.
We are in charge of a teaching unit for four-year
university students in Computer Science. First this
unit has been launched in presentational education,
then it has also been offered as part of the curricu-
lum in distance education. This article aims to explain
how we take as much advantage as possible of doc-
uments written for ‘traditional’ students, in order to
recycle them for ‘distant’ students. These documents
were written using L
A
T
E
X (Mittelbach et al., 2004), but
reading this article only requires basic knowledge of
this typesetting system. First we describe the situa-
tion when our distance education unit began, then we
explain our choices and give an overview of our tools.
After a short mention of alternativesolutions, we con-
clude with sumarising the experience we have got.
1
Centre de Tl-Enseignement, that is, Centre for Tele-
teaching.
2 OUR TEACHING UNIT AND ITS
DOCUMENTS
Our teaching unit is entitled Advanced Functional
Programming, PFA for short
2
. Let us recall briefly
that functional programming emphasises functions’
application, whereas imperative programming—the
paradigm implemented within more ‘traditional’
languages—emphasises changes in state. Many uni-
versities include courses about functional program-
ming, examples being reported in (Thompson and
Hill, 1995). Going back to the title, ‘advanced’ means
that this unit is not for beginners in programming,stu-
dents are supposed to be experienced. Practically,
most students attending this unit have already pro-
grammed in Java (java, 2008), Scheme (Springer and
Friedman, 1989), and C++ (Stroustrup, 1991).
Functional programming languages have a com-
mon root as the λ-calculus, a formal system devel-
oped in the 1930’s (Church, 1941). However, these
programming languages are diverse, some—e.g., the
Lisp dialects
3
—are dynamically typed, some—e.g.,
2
Programmation Fonctionnelle Avance, in French.
3
‘Lisp’ stands for ‘LISt Processing, because major
structures are linked lists. The first version came out in 1958
(McCarthy, 1960) and has many descendants, the most used
nowadays being COMMON LISP (Steele et al., 1990) and
Scheme.
469
Hufflen J. (2010).
RECYCLING PREVIOUS DOCUMENTS FOR DISTANCE EDUCATION.
In Proceedings of the 2nd International Conference on Computer Supported Education, pages 469-472
DOI: 10.5220/0002801604690472
Copyright
c
 SciTePress

Standard ML
4
(Paulson, 1996), CAML
5
(Leroy et al.,
2004), Haskell
6
(Peyton Jones, 2003)—are strongly
typed and include a type inference mechanism.
Our unit’s first part is devoted to the λ-calculus’
bases (Hufflen, 1998). Then all the practical exer-
cises are performed with only one language, Scheme.
The starting point of the most important part: when
we begin to program, the language we are learning is
always shown as finite product. It has precise rules,
precise semantics, and is consistent. According to
the language used, some applications may be easy
or difficult to implement. When you put down a
statement, running it often results in something pre-
dictible. That hides an important point: a language
results from some important choices our unit aims to
emphasise. For example, if the language is lexical
(resp. dynamic), what kinds of applications are eas-
ier to implement? Of course, answers of such ques-
tions depend on the programming languages consid-
ered. Our strategy consists of explaining the choices
of Scheme, and we demonstrate alternate solutions
using other functional programming languages such
as COMMON LISP or Standard ML. After this main
part, our unit ends with some advanced features of
Scheme: delayed evaluation, continuations, macros.
There is a big document grouping what is taught
within this unit, the first version was (Hufflen, 1997).
It consists of six chapters. Each chapter includes ex-
ercises, given with model solutions. These chapters
are followed by several appendices, making precise
some extra information or devoted to lab classes done
by students. The whole document is approximately
400-page long. It can be viewed as a textbook, even if
its diffusion is limited to this unit’s students. As men-
tioned in the introduction, we wrote it using L
A
T
E
X,
which seems to us to be the best typesetting system
for large documents: cross references are widely used
throughout this textbook, and there is a rich ‘Bibli-
ography’ section. Students progressively are given
the successive parts of this document, but it is or-
ganised as a whole, with precise architecture. Of
course, it contains not only texts—in the sense of suc-
cessive paragraphs—but also many examples of pro-
grams and some mathematical formulas, even if it is
not really a textbook in Mathematics.
4
‘ML’ stands for ‘MetaLanguage’ and has been initially
developed within the formal proof system LCF (Logic for
Computable Functions) (Gordon et al., 1979). Later on, it
appears as an actual programming language and its stan-
dardisation resulted in the Standard ML language.
5
Categorical Abstract Machine Language.
6
Named after Haskell Brooks Curry (1900–1982).
3 ADAPTATION TO DISTANCE
EDUCATION
3.1 Situation
When the distance master was launched, its curricu-
lum obviously resembled master’s in presentational
education. But a unit common to these two curricula
was not necessarily in the charge of the same teacher.
In other words, we have been in charge of the PFA unit
within both presentational and distance education, but
this arrangement did not hold true for all the units.
Besides, we were still in charge of the ‘presentational’
unit. So we were interested in a method that would
allow us to derive the two versions—printed and on-
line—from the same source. Some slight mistakes,
especially typing ones, should be fixed, we wished to
add more examples. In addition, the version of stan-
dard Scheme changed (Kelsey et al., 1998), so we
ought to adapt some existing examples. If we con-
sidered our text, there were only two differences to be
managed. The first difference was located at the in-
troduction to Scheme: since most of presentational
students attended a unit for beginners in functional
programming for the 2nd-year university degree, this
introduction was just some revision. On the contrary,
most of distant students do not know Scheme, and a
suitable introduction should be more progressive. But
this point was not really difficult since L
A
T
E
X allows
the definition of conditional texts; for example
7
:
\ifpfaforde ... (For distance education students.)
\else ... (For presentational education ones.)
\fi
The second difference is related to exercises. Presen-
tational students get the successive texts at the end of
each chapter, so model solutions may be given after
each exercise, especially if this exercise has already
been proposed at classes. That cannot be the same
for a document devoted to distance education: model
solutions should be grouped at the end of each chap-
ter, or provided in separate files. Here also, if these
model solutions have been put into separate source
files, an ‘ \if... ’ command of L
A
T
E
X may allow us
to put model solutions at distinct places, according to
the document we are building, for presentational or
distance education.
7
To avoid clashes among L
A
T
E
X names, the new com-
mands related to our adaptation are prefixed by ‘ \pfa... ’
or ‘ \ifpfa... ’.
CSEDU 2010 - 2nd International Conference on Computer Supported Education
470

3.2 Difficulty
When distance education was launched, teachers were
obviously asked to put on-line documents on the Web.
Some teachers put documents using HTML
8
. How-
ever, such a choice seemed to us not suitable for sci-
entific documents: the look of resulting Web pages
depends on the browser used; in addition, format-
ting mathematical formulas and program fragments
often results in poor-quality output. In our case, this
last point was essential about the fragments given in
languages other than Scheme. We could perform
some demonstrations during the lab classes of pre-
sentational students, so they could observe these other
programs’ behaviour. The same modus operandi was
impossible for distant students, and it was difficult to
ask them to install many compilers or interpreters.
So the solution was to ask them for exercises only
in Scheme—as done for presentational students—but
the examples given throughout our text must be ex-
plicit, in order for these students to understand with-
out running them.
As abovementioned, our document was typeset by
L
A
T
E
X, so an acceptable solution was to use pdfL
A
T
E
X,
able to produce PDF
9
files. In addition, if the hyperref
package is used, PDF files produced by pdfL
A
T
E
X can
support hyperlinks, as in HTML. But obviously, we
could not provide a single document, as a huge PDF
file. It is preferable for distant students to get separate
medium-sized PDF files, according to the steps of their
planning. Besides, let us not forget that these files are
downloaded: students cannot be asked to download a
huge file again if only some typing mistakes have just
been fixed. Splitting this big document into separate
files induces a precise organisation of cross-reference
links throughout the original version.
3.3 Our Adaptations
Let us assume that the chapters, sections, etc. of
the two versions—printed and on-line—are numbered
identically. Besides, L
A
T
E
X allows each chapter of
a document to be associated with its own auxil-
iary (.aux) file, containing information solving cross-
references
10
. So we can compile a chapter for the on-
line version by using the auxiliary files of the doc-
ument’s other chapters. A cross-reference put by
L
A
T
E
X’s \ref command is implemented in pdfL
A
T
E
X as
an internal hyperlink, what is fine for cross references
8
HyperText Markup Language. (Musciano and
Kennedy, 2002) is a good introduction to it.
9
Portable Document Format, Adobe’s format.
10
That can be done by the commands \includeonly and
\include .
within the same chapter. For cross-references to an-
other chapter’s part, we define a new command:
\pfaexternalref[ chapter-file ] {label
0
}
If the big document for presentational education is
generated, this works like \ref {label
0
}. If the
chapter is generated as part of the on-line text, an link
to the PDF file chapter-file is put. In both cases,
the same text is displayed. That means that label
0
is
a label identifying a resource belonging to a file used
to build the file chapter-file. This file has been
declared by the \pfaexternaldocument command,
so label
0
is known as a label. Of course, when we
started this task, such a choice led us to look for all the
occurrences of the \ref command and change some
into \pfaexternalref ones. In practice, that was not
difficult, because a good technique is to prefix labels’
name by an identifier for the corresponding chapter.
So the file name to be put was not difficult to supply.
We use a similar technique for cross-references to the
bibliography, and to footnotes belonging to another
chapter. All these new commands have been grouped
into a package.
4 DISCUSS
4.1 Students’ Opinion
As far as we know, students’ feedback is globally pos-
itive. In fact, they quickly perceive that PDF files al-
low them to watch exactly what teachers want to ex-
press, like in a book or blackboard. Our document
giving many ‘cultural’ complements, we had to de-
fine typographical signs to mark up what is important
and what may be skipped in a first reading, but this
task can be performed progressively. It also seems
that the hyperlinks pointing to a part of the current
chapter are most useful, so pointing to the beginning
of another chapter does not cause much trouble.
4.2 Comparison with other Methods
We were obviously interested in reusing our first ver-
sion written in L
A
T
E
X. There are some converters from
L
A
T
E
X to HTML (Goossens et al., 1999), and theyallow
a base document to be split into several Web pages.
However, these converters are not suitable when we
update an existing text since the names of generated
Web pages are generated, too; it may be difficult to
point just the HTML files that have been changed.
If we proceed from scratch, an interesting method
RECYCLING PREVIOUS DOCUMENTS FOR DISTANCE EDUCATION
471

could be to specify our input files using XML
11
, which
has become a standard for information exchange and
provides a rich toolbox. XSLT
12
(W3C, 2007), the
language commonly used for transformations of XML
texts could be used to derive texts for L
A
T
E
X, or in
XSL-FO
13
(W3C, 2006), an XML language that aims
to describe high-quality print outputs. However, the
current XSL-FO processors—generating PDF files—
are not complete yet, even if they implement most of
this recommendation, so using XSL-FO is interesting
for experiment, but not for intensive use by students.
5 CONCLUSIONS
As abovementioned, the first complete version of our
course text came out in 1997. Then it has evolved
deeply—chapters and appendices have been wholly
revised—and continuously, since we have applied
some changes each year. We did it successfully, so we
can think that our system is reliable. Of course, even
if our new commands could be applied throughout
any document, it may be noticed that this document
must be a L
A
T
E
X source text. This not too restrictive
for documents in Computer Science or Mathematics,
since L
A
T
E
X is widely used within these communities.
Further expriment should be made about documents
concerning other topics.
ACKNOWLEDGEMENTS
I thank the distance education students who addressed
me very constructive criticisms. Year after year, they
indirectly helped me improve my tools.
REFERENCES
Church, A. (1941). The Calculi of Lambda-Conversion.
Princeton University Press.
Goossens, M., Rahtz, S., Gurari, E. M., Moore, R., and Su-
tor, R. S. (1999). The L
A
T
E
X Web Companion. Addison-
Wesley Longman, Inc., Reading, Massachusetts.
Gordon, M. J., Milner, A. J., and Wadsworth, C. P. (1979).
Edinburgh LCF. Number 78 in LNCS. Springer-Verlag.
Hufflen, J.-M. (1997). Programmation fonctionnelle
avanc´ee. notes de cours et exercices. Polycopi´e.
Besanc¸on.
11
eXtensible Markup Language. (Ray, 2001) is a good
introduction to this meta-language.
12
eXtensible Stylesheet Language Transformations.
13
eXtensible Stylesheet Language—Formatting Objects.
Hufflen, J.-M. (1998). Introduction au λ-calcul (version
r´evis´ee et ´etendue). Polycopi´e. Besanc¸on.
java (2008). Java Technology. http://java.sun.com.
Kelsey, R., Clinger, W. D., Rees, J. A., Abelson, H.,
Adams iv, N. I., Bartley, D. H., Brooks, G., Dyb-
vig, R. K., Friedman, D. P., Halstead, R., Hanson, C.,
Haynes, C. T., Kohlbecker, Jr, E. E., Oxley, D., Pit-
man, K. M., Rozas, G. J., Steele, Jr, G. L., Sussman,
G. J., and Wand, M. (1998). Revised
5
report on the
algorithmic language Scheme. HOSC, 11(1):7–105.
Leroy, X., Doligez, D., Garrigue, J., R´emy, D., and
Vouillon, J. (2004). The Objective Caml Sys-
tem. Release 0.9. Documentation and User’s
Manual. http://caml.inria.fr/pub/docs/manual-
ocaml/index.html.
McCarthy, J. (1960). Recursive functions of symbolic ex-
pressions and their computation by machine, part I.
Communications of the ACM, 3(4):184–195.
Mittelbach, F., Goossens, M., Braams, J., Carlisle, D., Row-
ley, C. A., Detig, C., and Schrod, J. (2004). The L
A
T
E
X
Companion. Addison-Wesley Publishing Company,
Reading, Massachusetts, 2 edition.
Musciano, C. and Kennedy, B. (2002). HTML & XHTML:
The Definitive Guide. O’Reilly & Associates, Inc., 5
edition.
Paulson, L. C. (1996). ML for the Working Programmer.
Cambridge University Press, 2 edition.
Peyton Jones, S., editor (2003). Haskell 98 Language and
Libraries. The Revised Report. Cambridge University
Press.
Ray, E. T. (2001). Learning XML. O’Reilly & Associates,
Inc.
Springer, G. and Friedman, D. P. (1989). Scheme and the
Art of Programming. The MIT Press, McGraw-Hill
Book Company.
Steele, Jr., G. L., Fahlman, S. E., Gabriel, R. P., Moon,
D. A., Weinreb, D. L., Bobrow, D. G., DeMichiel,
L. G., Keene, S. E., Kiczales, G., Perdue, C., Pitman,
K. M., Waters, R., and White, J. L. (1990). COMMON
LISP. The Language. Second Edition. Digital Press.
Stroustrup, B. (1991). The C++ Programming Language.
Addison-Wesley Publishing Company, Inc., Reading,
Massachusetts, 2 edition.
Thompson, S. and Hill, S. (1995). Functional programming
through the curriculum. In FPLE ’95, pages 85–102,
Nijmegen, The Netherlands.
W3C (2006). Extensible Stylesheet Language (XSL). Ver-
sion 1.1. http://www.w3.org/TR/2006/REC-xsl11-
20061205/. W3C Recommendation. Edited by Anders
Berglund.
W3C (2007). XSL Transformations (XSLT). Ver-
sion 2.0. http://www.w3.org/TR/2007/WD-xslt20-
20070123. W3C Recommendation. Edited by
Michael H. Kay.
CSEDU 2010 - 2nd International Conference on Computer Supported Education
472