It's HEDLEY!

The stories of Hedy Lamarr and the computer women at NASA get repeated endlessly as if they're the only contributions women have ever made to CS, which must be absolutely discouraging to the young girls who are interested enough to dig deeper and find out that Hedy's contribution was essentially negligible.

It'd be much better to talk about Liskov, Goldberg, the women at Bluetooth SIG, or the countless other examples available.

Bizarre “research.” Why go out of your way to cast doubts on Lamarr’s contributions to the patent. He offers no evidence, or even a compelling theory as to why she would wrongly have her name on it.

> Lamarr … may have contributed little to the insignificant invention actually taught by the Lamarr-Antheil patent.

Sounds like he’s got a chip on his shoulder over… something. Despicable stuff.

George Antheil’s autobiography, Bad Boy of Music, is quite entertaining. In it he recounts his adventures related to this patent. If I recall correctly (I read it decades ago but I think I’m right here), he describes a close and co-equal collaboration with his friend Hedy Lamarr on this invention. Therefore I think these remarks by the author:

‘Since the actual invention is a player-piano-like mechanism, and since experimental musician George Antheil had expertise in the inner workings of player pianos, and further since Hedy Lamarr evidently had no such expertise, it may be more appropriate to call the Lamarr-Antheil patent “Antheil’s patent.”’

are inappropriate and unjustified.

Look at the prior art. The key prior patent cited is BAESECKE, (2,134,850). This has a "wobbler" to change the transmitting frequency, which is fine, but that patent is vague as to how the receiving end stays in sync. Markey gets that syncing up transmitter and receiver is the real problem, and talks of it as a solved problem from television. (This is 1941, experimental TV exists.) Some mechanisms are discussed, borrowed from multiple player piano synchronization, which is where this technology came from.

Voice cryptosystems of the pre-WWII period included the Western Electric A-3 scrambler [1] There's one on eBay! [2] That split audio into a number of frequency bands, which were shifted and reassembled. At the receiving end, the process was reversed. The shift pattern changed periodically, on the order of tens of seconds. That was slow enough that keeping the thing in sync was possible with clockwork of the period. Note that this is working on the audio, not the RF; it's a scrambler, not a frequency hopper. This is what AT&T used for transatlantic commercial voice. Worked OK, mediocre security. Only 6 different frequency shift patterns were in use at a time. The Germans cracked it.

This is not the better known SIGSALY. That's a similar concept, but with a lot more audio channels, much more frequent changes, a one-time key for the changes, and more hardware than some mainframe computers. The A-3 was a desk-side wooden box. Neither system does frequency-hopping of the RF signal.

Hedy Lamarr's first husband was an arms manufacturer, and she apparently paid attention when visiting the factories. Hence the radio-controlled torpedo, which is close to Tesla's radio-controlled boat.

MARKEY (2,292,387) describes a cross between a radio-controlled boat and a set of synchronized player pianos. There's some handwaving around the sync problem. The trouble with syncing a frequency hopper is that you have trouble even finding the signal to get started. But if you're launching a torpedo or a bomb from a larger craft, both ends of the connection can be started in sync and will probably stay in sync long enough for the bombing run. Doing this with a player piano roll reader, vacuum pump and all, is probably not the right approach.

Trying to get things to sync up reliably has a long history. Edison's first useful invention was a way to get stock tickers to sync. There's a long history of clunky mechanisms, early ones involving flywheels or big tuning forks, and later electronic ones with too many screwdriver adjustments. Not until the invention of phase locked loops did it really Just Work. (I'm into restoring early Teletype machines, and I'm way too aware of the early days of sync problems.)

Markey was just too early. Reasonable idea, but not practical at the time due to lack of supporting technology.

[1] https://chris-intel-corner.blogspot.com/2012/02/intercepted-...

[2] https://www.ebay.com/itm/205373065319

Ouch.

From the article: "A letter on 3 October 1941 from the Lyon and Lyon attorney to Lamarr and Antheil says '...we rather doubted at the time that method claim 7 would be considered patentable, since the invention appears to reside more in a new apparatus than in a new method.' Thus, the attorney representing the applicants agreed with the patent examiner that the evidence was against Lamarr-Antheil’s definitive method claim to FHSS, which was claim 7."

This analysis makes it pretty clear that EFF's 1997 assertion that she and Antheil "developed and [...] patented the concept of 'frequency-hopping' that is now the basis for the spread spectrum radio systems" is flatly untrue.

This isn't to say that she wasn't an inventor or innovator, or didn't put together existing known techniques in a new way to address a relevant and interesting problem.

I find the article unconvincing, although I'm open to being convinced. With historical hindsight, it should be easy to see if the Lamarr et al patent seems novel. Just because an examiner doesn't allow a claim, I don't see that as strong evidence it wasn't novel at the time. They always are rejecting claims, sometime for good reason, sometimes not.

A more convincing article would focus on purported prior art patents, and let the reader judge if really anticipated frequency hopping.