>From tech-net-owner-track=Plectere.com@NetBSD.org Mon Nov  8 22:39:21 2004
>Delivered-To: tech-net@netbsd.org
>Date: Tue, 9 Nov 2004 00:38:45 -0600
>From: David Young <dyoung@pobox.com>
>To: tech-net@NetBSD.org
>Subject: Re: slashdot on 'OpenBSD Activism Shows Drivers Can Be Freed'
>Mail-Followup-To: tech-net@NetBSD.org
>References: <200411031745.iA3HjYqE014875@Plectere.com>
>Sender: tech-net-owner@NetBSD.org
>Precedence: list
>
>On Wed, Nov 03, 2004 at 09:45:34AM -0800, List Mail User wrote:
>> 	A very fine point:  The manufacturer is barred from providing
>> "software or drivers" with a particular capability, but can allow a
>> third party to develop exactly that!?  I am not a lawyer, but I wouldn't
>> want to test that interpretation.
>
>I don't think the FCC could regulate device datasheets if it wanted to.

	I'm not so sure;  What about the prohibition on selling both
non-conforming "instructions" and parts from the same vendor?  For
examples just look in the back of "hobbyist" magazine (e.g. Nuts & Volts)
and notice the variety of ads which sell ONLY "plans" or even "kits" without
any instruction (of course from different vendors).  I you buy the "plans"
from one vendor and the parts from another, and assemble the unit in question
yourself, you may or may not be covered under the "hobbyist" exceptions.
Clearly, construction of at least some of the devices whose "plans" are
advertised would NOT be legal (at least in the US).

	Really, what is a "datasheet" and what is an "application note";
The entire electronics industry confuses and mixes these two somewhat
different types of documents (just try to use or write a driver for
an Intel chip from just a "datasheet" - and they are about the best
I've found).  What can be regulated are the instructions on how to use
the device, I believe they can't regulate things like electrical
specifications or pin-outs (but I've seen examples of modified pin-outs
of MicroChip devices being proposed for Part 15 compliance - e.g. the
standard ICE and programming tools wouldn't then work on the device
without modification).

>
>> P.S.  Bugs in all existing Window's drivers for Atheros chips seem to allow
>> "illegal" active scanning on some channels and some cards (e.g. anything
>> Netgear and many others), and "illegal" associations to access-points or
>> routers on those same channels (12-14, & 100-140) but I don't think that
>> it is on purpose (except, possibly the D-Link channel 52 mess - They already
>> had WHQL so seemed to ignore the FCC and their own filings, about removing
>> it, and operation in turbo mode on channel 50 seem to contradict their own
>> documents - though it works with their supplied drivers).
>
>These slip-ups are common enough.  It begs the question, "why bother?"
>
	I did state that I agree with you on many points.  This is one in
particular - If Atheros's own Window's reference drivers can access ANY
channel on many shipping cards (e.g. most Netgear US equipment), then I
also don't see what the point is either.

>> P.P.S.  As to "tamper-proofing" PART 15 devices, my drawer full of "security"
>> bits and screwdrivers attests that many manufacturer do attempt to prevent
>> physical access to the insides of the devices they sell (often at the FCC's
>> request - I can dig up other references if you like, but a search on PART 15
>> transmitters will give you all you should need).
>
>This jives with my argument that the FCC's aim is to prevent illegal
>operation that is casual or inadvertent.  Clearly these devices have not
>been "tamperproofed" if you have a drawerful of bits & screwdrivers to
>open them!  These are rudimentary barriers to disassembly.  A determined
>person can and will tamper with the device.

	Actually, these bits and screwdrivers are somewhat difficult to find;
Though until recently (they may be just "out of stock"), you could get many
of them at Fry's (though they weren't labeled as such) - I collected mine
over the years, mostly from specialty tool suppliers.  Basically, the concept
is simply that you shouldn't be able to walk into a Sears, Home Depot or
OSH, etc. and pick these up off the shelf (NOTE: Most OSHs, which is owned
by Sears, do have "security bit sets" hidden in the back corner of the tool
section - or you have to ask).

	The "real" use for these `strange' bits is to keep the casually
interested person out of pinball machines, card readers, gambling equipment,
telephone and cable company field boxes, etc..  But just like RP & RT
connectors the FCC accepts them as "good faith" evidence of attempting
to comply with Part 15 rules.

	Clearly, the rules are somewhat silly when the most popular RF
connector after type N, is the RP-SMA, which is "supposed" to make it
"impossible" for a typical user to change the antenna (and maybe, by now,
there are even more models or units sold with the RP-SMA than with type Ns
or PL-25Xs).  Every retail store sells them, nearly every vendor publicly
advertises them (including "compatibility" lists) and Hawking even includes a
RP-SMA <-> RP-TMA adapter with each add-on antenna.  The easiest way I know
to violate the FCC rules, is to walk into Fry's, buy a 23dBm AP and a Hawking
15dBi antenna, then connect them and turn it on - you'll be 3dB too high in
EIRP to be legal. (actually, I think Hawking ignores cable loss in their
specs, but it shouldn't be too much at 2.4GHz - it looks like the cable is
similar to LMR-100 and it is relatively short - The antenna "looks" like it
is actually made by Nearson, and my experience is that their specs. are
accurate).

	Once there are both removable antennas and an "after-market" for
replacement or even "enhanced" antenna, I believe even the "casual" user
will likely be tempted to violate the rules (in Fry's, most of the antennas
are displayed in the same aisle, next to the wireless routers and APs).
Still, the rules (which HAVE just changed) do NOT allow the end user to
change the antenna unless that particular configuration is "certified"
(this is the part that just changed, some new equipment can be "pre-certified"
for a variety of devices, not all of which must be specifically listed, but
their characteristics must be specified - and of course more rules apply
also and I'm greatly simplifying the single point I've chosen to expound).

>
>Dave
>
>-- 
>David Young             OJC Technologies
>dyoung@ojctech.com      Urbana, IL * (217) 278-3933
>

	Put simply, I FULLY agree that many of the "rules" are either silly,
ineffectual or both.  Also, the "software" radio requirements are indeed a
choice made by the manufacturer (NOTE: of the end product, not necessarily
Atheros who makes the chip -- The application could come from an Atheros
customer and then they would have to choose between making the sale or
requiring the more extensive certification process - which they might not
be able to pass!).  Just that the FCC rules do exist causes the problems
with some of your arguments -- You are trying to apply logic to government
regulations and laws and that just doesn't always work (though I believe it
should and thus all your arguments SHOULD be valid).


	Paul Shupak
	track@plectere.com