#fran allen

[taonf]

Hired by IBM in 1959, Fran Allen became an expert in optimizing compilers. She was assigned to work on Harvest, a code-breaking project for the National Security Administration (NSA) on the Stretch computer. Fran worked with a program called Alpha, and her assignment was to write the acceptance test for it. She says:

I knew that I couldn’t leave Fort Meade until that was done and I dreaded it for the whole year I was there.

Fran Allen was at Fort Meade during the Cuban Missile Crisis in 1962, working on code breaking. She remembers that all the computer printers were draped with black cloth so people couldn’t see what was going on and there were Marines everywhere.

Far from being “hidden figures,” women have been working in computing since the earliest days of the new field. They’ve often faced gender b

[elvisomar]

Fran Allen

(born 1932) Computer scientist

Frances Elizabeth Allen is a pioneering computer programmer specializing in code compilers and code optimization. Allen was the first woman to be named an IBM Fellow, and was the first woman to be awarded the prestigious Turing Award, an honor conferred by the Association for Computing Machinery sometimes referred to as the Nobel prize of computing.

Number 126 in an ongoing series celebrating remarkable women in science, technology, engineering, and mathematics.

[jesseriggins]

fran, tokyo, 2014

[jesseriggins]

fran, toronto, 2014

[lunasspecto]

Ken Thompson and Fran Allen on abstraction

I've been reading Peter Seibel's collection of interviews Coders at Work: Reflections on the Art of Programming and I decided to read Fran Allen's interview directly after Ken Thompson's. This turned out to be an interesting way of reading the text because the two of them directly contradict each other on some big-picture stuff.

First you have Ken Thompson, who left university in the mid-1960s to work at Bell Labs, and by 1970 had co-created Unix and created a direct precursor to the C programming language, which is still the most popular programming language today. Thompson considers computer science to be at something of a dead-end. As he sees it, hardware will continue to become more efficient while programming will continue to become more abstracted and less efficient, allowing hardware advances to counteract the flaws in new software. Thompson did much of his early work in assembly language and finds the lack of familiarity new programmers have with low-level hardware operations "scary", even considering the complex interfaces and end-user functionality enabled by abstraction.

I found Fran Allen a lot more forward-looking. She joined IBM as a programmer in the late 1950s and oversaw the creation of some important early compilers designed to allow programmers to write portable code which would be optimized for computational efficiency at compile time. In her view, the industry adoption of the C programming language as an application programming language (and not just a systems programming language) was a "big blow" because it placed the responsibility of certain low-level tasks like memory management squarely on the shoulders of the application programmer. Allen argues that optimization and memory management at the compiler and hardware levels, combined with very high-level languages, will allow computer science to tackle the challenges of managing data in global networks in unified, efficient, and innovative ways that improve end-user experience.

I generally find Allen's arguments a lot more compelling, though I'm with Thompson when he says C++ overcomplicates things to the detriment of code readability.

[lunasspecto]

We have seriously regressed, since C developed. C has destroyed our ability to advance the state of the art in automatic optimization, automatic parallelization, automatic mapping of a high-level language to the machine.

Fran Allen on C as an application programming language, interviewed by Peter Seibel in Coders at Work: Reflactions on the Craft of Programming.

[holyghostzine]