Grace Hopper

Grace Hopper looked down at the instructions.

She had seen some of them before.

Written them before.

Checked them before.

Now here they were again.

It was the early 1950s. Grace worked with a computer called UNIVAC. Its tall cabinets and equipment took up a whole room. It could work through numbers much faster than a person.

But first, people had to tell it exactly what to do.

Every step. In the right order.

Those instructions made a program. Preparing them took patience, and copying an old set of instructions into a new program could introduce fresh mistakes.

One wrong number could send the computer to the wrong place.

Grace and other programmers had already found a useful trick. They saved little sets of instructions in notebooks and borrowed them from one another. Each set handled a particular job.

These small, reusable pieces had a big name: subroutines.

Imagine having to write out an entire recipe every time you wanted to bake a cake. Even the part explaining how to make the frosting.

Grace’s notebooks were like a collection of recipes.

But the computer could not reach over and open a notebook.

People still had to copy.

Grace wanted to get on with solving new problems. Instead, familiar instructions kept getting in the way.

She looked at what UNIVAC could already do. It could store information on magnetic tape. It could find information and bring it into its memory.

What if the instructions went on the tape, too?

Then the computer could help gather the pieces for a new program.

It was an exciting idea. It was also another programming problem.

The machine would need careful instructions for finding instructions!

Grace broke the job into parts.

First, there had to be a collection of useful routines on tape.

Then each routine needed a number so the computer could find it, rather like a book in a library.

And the pieces had to fit together. A routine needed to know where to find the numbers it would use and where to put its answer.

Giving every piece a number was only the beginning.

Grace worked on the system that would gather those pieces and arrange them for use. There were details to check, steps to work through, and results to examine.

The giant machine could hurry through calculations.

The people building its programs still had to think carefully.

By 1952, Grace had a working system called A-0.

Now a programmer could request routines by their numbers. The computer could retrieve them from tape and help assemble the program.

A useful piece of work no longer had to begin as another page of copying.

But a working system was not the end.

Some people doubted that computers could help prepare programs. Grace explained her work and encouraged people to try it. She and her colleagues kept improving the tools, producing versions called A-1 and A-2.

There were more problems ahead.

There were also more pieces they could build with.

The notebooks had held something valuable: work that people had already figured out.

Grace had found a way for the computer to help them use it again.

And that left people more time for the next thing they had not figured out.