I know you have gone on record saying that Open Hardware licenses are not needed. I wonder if you have thought through the whole picture? Chips and boards are _made_ out of software these days. There are a long list of hypothetical and not-so-hypothetical questions on how to interpret the GPL in the context of embedded hardware, software, and firmware.
Real-world question: a company sells a net-centric appliance, like a camera, that has GPL'd code burned in ROM on it. What happens to the end-user's right to update buggy code then? I happen to think that right lies at the heart of the GPL.
I think future versions of the GPL need to clarify the concept of linking and control. What physical and legal barriers can be erected to keep end-users from exercising their rights? What are those rights, anyway? Lines between hardware and software are fuzzy and getting fuzzier.
Mr T already said: "Thanks again for your wonderful books. Any plans for AC 3rd edition? Maybe with AES covered?" I second the motion, and want to know if you might cover elliptic crypto? If not in AC3, can you recommend an on-line review?
I've measured it, and it goes so close to dark the difference isn't worth talking about. The phosphors are clearly very short lived (less than a millisecond). The trace from a photodiode looks like abs(sin(x)).
The bandwidth comment is certainly correct. My tests with boa (http://www.boa.org) show an "older Pentium" easily saturates a 10 MB/s Ethernet link. If you used Apache, you could probably do as well _if_ you load up with RAM. Boa (and thttpd and Zeus) give better performance with fewer resources, on static material. Webcams count as static, because they don't get recomputed per request. Rather, a "camera task" updates the images frequently.
Real-world question: a company sells a net-centric appliance, like a camera, that has GPL'd code burned in ROM on it. What happens to the end-user's right to update buggy code then? I happen to think that right lies at the heart of the GPL.
I think future versions of the GPL need to clarify the concept of linking and control. What physical and legal barriers can be erected to keep end-users from exercising their rights? What are those rights, anyway? Lines between hardware and software are fuzzy and getting fuzzier.
Mr T already said:
"Thanks again for your wonderful books.
Any plans for AC 3rd edition?
Maybe with AES covered?"
I second the motion, and want to know if
you might cover elliptic crypto? If not
in AC3, can you recommend an on-line review?
I've measured it, and it goes so close to dark the difference isn't worth talking about. The phosphors are clearly very short lived (less than a millisecond). The trace from a photodiode looks like abs(sin(x)).
The bandwidth comment is certainly correct.
My tests with boa (http://www.boa.org) show
an "older Pentium" easily saturates a 10 MB/s
Ethernet link. If you used Apache, you could
probably do as well _if_ you load up with
RAM. Boa (and thttpd and Zeus) give better
performance with fewer resources, on static
material. Webcams count as static, because they
don't get recomputed per request. Rather, a
"camera task" updates the images frequently.