Useful Software Celebration Day - Program Abstracts
Peter Naur, Professor Emeritus, DIKU
In the design of Regnecentralen's Algol compiler for Gier usefulness was a dominating concern: (1) Explicitly and clearly defined programming language; (2) Fast, reliable compiler; (3) Extensive checking in the compiler of source programs that they conform to the programming language, and reporting of any number of errors in the source program in one process. These concerns led first to our participation in the description of Algol 60, and then to the use in our compiler of a multipass technique.
Parallel with the Moore's law explosion of computing power, a number of other things also had to change so that we can have so much useful software today: computers had to change, the user community had to change, and software had to be separated ("unbundled") from hardware.
Most of these transformative changes started in the 70's to blossom in the following decades. The talk will discuss the tipping points for some of these changes, with anecdotes about the first personal computer (Alto) and the first WYSIWYG editor (Bravo) and how the Personal Computer industry sprung up to exploit the tremendous opportunity therein.
A View on Information Technology
Mads Tofte, Rector, IT University in Denmark
DIKU was born just four years after Peter Naur had defined the term "datalogi". At the time, definitions of terms such as "datalogi", computing, computer science and informatics must, I imagine, have been particularly important, because one was at the beginning of a journey that needed direction and distinctness. Today, computing has evolved into a pervasive and global phenomenon that seems to defy brief definition. Parts of this phenomenon are carefully defined and deeply understood, but the phenomenon as a whole seems to develop even faster than the deeper understanding of it. The phenomenon is currently called Information Technology; presumably, other terms will emerge in the future as the phenomenon evolves further. In this presentation I describe not a definition but a view on Information Technology, which I have found useful for thinking about the developments that have taken place during the first forty years of DIKU's existence.
Reminiscences of Whetstone ALGOL
Brian Randell, Professor Emeritus, School of Computing Science, Newcastle University
The Whetstone ALGOL Compiler was one of two compilers produced in the early 1960s for the English Electric KDF9 Computer, the other being a very ambitious optimising compiler. In contrast the Whetstone Compiler was a fast one-pass compiler, with extensive facilities for program checking and testing. Its design drew heavily on that of the pioneering ALGOL Compiler developed by Dijkstra and Zonneveld at the Mathematical Centre, Amsterdam.
I will describe the background to the Whetstone compiler, whose design was detailed in the book "Algol 60 Implementation" that I co-authored with Lawford Russell, and copied by a number of groups elsewhere, as well as giving brief details of other early ALGOL compilers..
Lessons learned from Building Google Chrome
Brian Rakowski, Director of Product Management, Google
When we set out to build a browser, our only goal was to make something that would make using the web more useful and enjoyable. By now, Chrome is used by over 70 Million people and is growing rapidly, but is still under active development to help it reach more users.
Along the way, we stumbled upon a series of lessons that have helped inform our design and development philosophy. I'll review those lessons, giving examples of things that worked well for us, as well as things that failed.