On the production of the ACM book
Providing Sound Foundations for Cryptography:
On the work of Shafi Goldwasser and Silvio Micali
This book, edited by me (Oded Goldreich), will
appear in ACM's book series.
As stated in the book's webpage,
ACM production of this book included re-typing the original papers
(for Part II), rather than using facsimiles of these papers,
and changing various aspects of the texts of Part III
(e.g., the bibliographic conventions
and the numbering of theorem-like environments).
These production decisions were forced upon the editor,
who strongly objected them both for reasons detailed below,
and due to the likelihood of errors caused by implementing them.
- Re-typing the original papers.
Original scientific works of the past are typically
valuable only for historical reasons.
Their actual contents is typically improved if not superseded
with time, and the exposition is also improved once the community
internalizes the novelties and finds better ways
of presenting its core ideas and details.
Hence, both experts and novices to the area are advised
to study the contents of classical works based on later
expositions. The value of the original texts is historical;
one can see in them the state of the area at the past
as well as the authors' ways of revolutionizing it.
These are best reflected by facsimiles of the original works,
rather than by re-typing that alter the layout and the style,
let alone the error that are bound to be introduced.
- Changing the bibliographic conventions.
Most authors in our community prefer either numerical
or alphabetic labels in referencing to bibliographic items,
and dislike the author-name+year format forced by the publisher.
Talking for myself, I find the latter style extremely imposing;
it is annoying to the reader when providing a citation list that
contains several papers, and it does not allow to provide
proper references without distracting the readers attention
(as I often like to do). In any case, the publisher should
respect the authors' autonomy and choices rather than
intervene in their actual personal (or communal) style.
- Changing the numbering of theorem-like environments.
I believe that in long texts, it is better to use a single
numbering systems for all theorem-like environments,
because it facilitates searching for a given item that
is being referenced. (Indeed, it is `logical' to use
a different numbering for each type of element,
but this convention makes finding a specific element quite hard.)
Again, the publisher should respect the authors' autonomy and choices
rather than intervene in their actual personal (or communal) style.
Neither the editor nor the authors assume responsibility
for the errors introduced by the publisher in the process
of implementing the foregoing (wrong) decisions.
The reader is advised to fetch the original versions
of the surveys appearing in Part III from the authors
(mine are available here).
Back to Oded Goldreich's homepage.