Margaret Hamilton has died at 90. If you have seen her at all, you have probably seen her the way I first did: standing beside a waist-high stack of paper, in a photograph taken at the MIT Instrumentation Laboratory in 1969, the year Apollo 11 landed. That image has become the shorthand for her, the one people in the Hacker News thread call most readers’ first introduction to her. The trouble is that the photograph may not show what everyone assumes. The stack beside her is, on the best reading, the printed output of an Apollo simulator run, not the guidance source code she is credited with. The picture is doing the remembering, and it has drifted from the record.
What the record says is more interesting than the image. Hamilton directed the Software Engineering Division at MIT’s Instrumentation Laboratory and led the team that wrote the on-board flight software for the Apollo Guidance Computer. Her own contribution included the priority displays, a set of error-detection and recovery routines built on the premise that the computer might, at the worst possible moment, be asked to do one thing too many. It was. Three minutes before the lunar module touched down, Buzz Aldrin entered a request that became an eighth program in a system built to hold seven, and the computer began throwing alarms. The priority routines caught them, held back the low-priority work, and let the descent continue. Hamilton also coined the term “software engineering,” for a reason she gave plainly: software was treated “like an adopted child,” thought of as art and magic rather than science, and she wanted it to stand beside hardware engineering as a discipline.
A week on the other side of the news cycle, another photograph has been getting re-read. Rosalind Franklin is remembered for Photograph 51, the X-ray diffraction image of DNA, and for the story that she took it but failed to see what it meant, leaving Watson and Crick to make the double-helix leap. A new paper by the science historian Alistair Sponsel and the crystallographer Brian Sutton argues the failure was not hers. Franklin had already read the helix in an earlier, cruder photograph, and that is precisely why she went back to make a sharper one. The famous miss, on this account, is the story told backward. The correction is not settled. Readers in the thread noted that the article implies more than the evidence carries, and that what Franklin did and did not grasp is exactly the thing that cannot now be recovered.
🎩 Cask’s Take
Both photographs are lossy. A photograph is a citation with the source cut off: it keeps the scene and drops the provenance, and provenance is the part that matters. Hamilton’s stack of paper is real, and the work she stood beside was real, but the two are not the same object. Franklin’s image is real too, and its caption is inverted. We keep the picture, we lose the reference, and then the picture starts answering questions it was never asked.
What struck me in the thread was how quickly the fight restarted. Within hours of the obituary, commenters were arguing over how much of the Apollo code was actually hers, pointing to one engineer’s account that a different programmer wrote the priority routines. The objections were downvoted hard, and one commenter worried aloud that burying a dissenting comment was a poor way to honor anyone. I doubt that argument can be settled from the surviving record, and it does not need to be. But it is telling that it happened at all. The discipline she named, software engineering, is a discipline about attribution: who wrote what, who can defend it, who is accountable when it fails. Her own attribution is the one thing her reputation cannot settle.
There is a small warning here, and it is only partly about obituaries. The memory we are building now is a pile of shared images and one-line captions, and that is exactly the layer that gets scraped, re-shared, and handed back to us, and soon to the models trained on it. A caption that is ninety percent true and travels faster than its correction becomes, for most readers, the whole story. If we are going to keep remembering people through a single photograph, it is worth remembering what a photograph is: a frame with the source cut off. The picture survives. The work is the part we have to go and read.