You must be referring to the foam absorbers used in an anechoic chamber, these are used for absorption. I was referring to finger stock that's used to minimize leakage out (or into) an enclosure with a door.

802.11a devices (operating at 5.45 GHz) are already supposed to detect radar signals and switch channels if one is found. This is particularly a problem in Europe, where most weather monitoring radars are C-band, and share the same frequency band as 802.11a.

Perfect anything is impossible to achieve when dealing with microwave devices, in my opinion. Particularly shielding - you can reduce it down to a point, and it gets expensive the lower you go. For instance, "quiet chambers" used when testing for EMI compliance typically have doors with beryllium-copper fingers that try to create a faraday shield to keep out external interference. I've never seen any commercial microwave oven that uses these - for good reason too, since they're expensive and won't last very long in a kitchen environment.

Most microwaves would try to ensure that the gap between the metal door and the body is small enough that the waveguide thus formed would have significant attenuation at 2.45 GHz. Any waveguide has a "cutoff" frequency, below which propagation can only occur through evanescent waves, which decay very rapidly. However, the relative power levels involved (microwaves generate 1 kW, or about 60 dBm, while WiFi receivers are sensitive down to about -90 dBm, or 1 picowatt) means that evanescent waves that escape can cause interference.

A winmodem is basically a sound card with an analog telephone line interface. This means all the modulation and channel coding that needs to be done to send/receive data is done in software. Real modems use a DSP to implement the modulation, and have the firmware for the DSP in a Flash chip. The only cost-cutting involved would be to do away with the Flash chip and have the device driver download the firmware each time the modem is power-cycled. Nobody did this with any modems I'm aware of, partly because being external from the PC case, it's hard to say if the modem DSP has already booted and is running firmware or not. With an expansion card, this is not an issue. Back in '98, I was able to update the firmware on a Zyxel 28.8k modem and have it support 56k, since it used a programmable DSP and allowed Flash reprogramming. Greedy marketers would usually prevent this from happening so I'd spend more $$$ to buy the 56k modem - somehow Zyxel were different at the time.

Some of the better APs will let you set the signal strength. I set mine down to -10 dBm (lowest available), but I can still get a signal from several hundred feet away.

What I'd like to see is what Red Hat will do when it has to release RHEL 7 - will it go with Gnome3 or stick with something sensible? I can't imagine running a workstation on which one is expected to do actual work with Gnome3.

He would also write about defensive driving on occasion in Pease Porridge, his column.

What? Jim Williams too??? Who will write our humorous app notes with the doodles on the last page? :( A dark day indeed. Williams' application notes (and his tales of learning from fixing broken test equipment) are one of the reasons I'm a EE today.

Sure, because a nuclear engine is so much more easier to fix, right?