Re: The History of the ALGOL Effort

A.L. wrote:


> On 14 Aug 2006 15:17:24 -0400, HT de Beer <H.T.de.Beer@gmail.com>
> wrote:
>
>>Dear reader,
>>
>>this month I have completed my Master thesis in Computer Science and
>>Engineering about the history of the ALGOL effort. In this thesis I
>>argue that the ALGOL effort was the catalyst to the transformation of
>>the field of compiler writing from a craftsmanship into a science in
>>the early 1960s. The history of the ALGOL effort is told, starting
>>with the computational context of the 1950s when the ALGOL effort was
>>initiated. Second, the development from IAL (ALGOL 58) to ALGOL 60 and
>>the role the BNF played in this development are discussed. Third, the
>>translation of ALGOL 60 and the establishment of the scientific field
>>of translator writing are treated. Finally, the period of ALGOL 60
>>maintenance and the subsequent period of creating a successor to ALGOL
>>60 are described.
>
>>People who are interested can download the thesis as a pdf file at
>>http://www.heerdebeer.org/ALGOL/The_History_of_ALGOL.pdf (1.1
>>megabytes) or read the summary at http://www.heerdebeer.org/ALGOL/ .
>
>>I post this here because a) I think there may be people here
>>interested in the early years of the field of compiler writing, and
>>b) the more appropriate newsgroup, comp.lang.algol, appeared to be
>>dead,
>
> Maybe you should go to Wikipedia before writing your thesis and learn
> a bit about TRUE history of Algol 60.


First of all, as I have said earlier, my thesis is an historical
paper, that is, I read, select and interpret sources to write a story
in which I want to argue something. It is well known among historians
that there is no such thing as the TRUE history: every history is just
another interpretation of sources, meanings, facts, and so on. As a
result, writing history never ends and historians do have a job for
their whole life :-)


As a side note, I will not recommend wikipedia as a scientific
resource, especially not for subjective topics like history. Be aware
that every one writing history, if they do know it themselves or not,
give a personal and hence subjective interpretation of events. So, if
you want to use wikipedia, use it with care.


> Unfortunately, there are a lot of omissions in your thesis, but the
> most significant is not mentioning the KDF9 compiler known also as
> Whetstone compiler. This was first complete implementation of Algol
> 60 based on priority grammars and stack based virtual machine.


It was not my intention to be complete, to list all ALGOL 60
compilers, or to mention all people involved. I have tried to argue
that the problems of implementing the new features of ALGOL 60, like
recursion, procedures, blocks, and arrays with unknown bounds at
compile time, were solved soon after the publication of the ALGOL 60
report by several people (I have named Dijkstra, Irons, Naur, and
Feurzeig, but there were others too).


> Complete description of this
> compiler was published as a book by Academic Press (Rendall, Russel
> "Algol 60 Implementation"). This book contains complete description
> and pseudocode of both the translator and virtual machine. To my best
> knowledge, the role of this compiler was the same as the role of
> Amman's P4 Pascal compiler published years later: it was possible to
> purchase the book, code the algorithms presented here and get working
> Algol compiler. This book was the starting point of many other Algol
> compilers, and probably influenced the architecture of future Pascal
> implementations.
>
> It seems that Randall, Russell, KDF9 and Whetstone are missing in
> Bibliography of your thesis. This is absolutely significant omission.


Because I did not refer to this book in my thesis does not mean I have
not seen and read it: if you look at my bibliographic file at
http://www.heerdebeer.org/ALGOL/bibliography.html you will find this
book and many other books not listed in the bibliography of my thesis
which I have read. As said before, writing history is first and
foremost selecting and interpreting sources. Some fit in my history,
some do not.


On the other hand, it leaves the way open for discussion and
criticism, which is a good thing because it can be a step in the
creation of a historiography of history of computer science.


--
HT de Beer
H.T.de.Beer@gmail.com
.

Return to the comp.compilers page.
Search the comp.compilers archives again.