Hacker News Reader

Comment by minipci1321 | Hacker News Reader

> It's such a blatant lie.

yes and no. Most today's CE devices lose 3/4 of their utility when disconnected from their shared medium. People running such mediums usually require some sort of certification from the participant devices, and in case the device has been modified, a re-certification after modifications is usually required.

Even replacing a couple of diodes by a looking-close-on-paper analogue might harm performance for neighboring participants in the communication. I let you imagine what one could do with arbitrary firmware images on a fully unlocked device.

I'd be curious to see what solutions to that the infrastructure people work out when these laws are eventually adopted.

View on HN

Firstly, one wouldn't have to use the "close on paper" if spare parts, or real specs, were actually available.

Secondly, as analog electronics, where the non-compliance argument might matter, are generally built of devices with 5-20% tolerances. Close enough is usually close enough, if you can get the right specs to start with.

Thirdly, the parts of modern communication networks that can actually be affected by component choice is a miniscule part, of the complete systems. Sincr almost no modern communication device is built with more than a few discreet devices around a special RF chip, you really can only repair it with the same chip, and trivially replaceable components.

I can go on. Yes, there are certifications, sometimes they are important and necessary, but usually they are there for business reasons. To create lock-in, or otherwise increase revenue, or to avoid blame by making it look like you care.

I have yet to see an industry certification which has not clearly put in place for a business reason.

For non-industry certifications, there would be no difference in who does the repair, re-certification would be necessary, and society aldready handles this in areas where it's actually needed for safety reasons.

> I can go on.

I'll skip that too. I have a counter argument literally for every your affirmation. Big part of modern CE devices do not fit into your description (or at least, we clearly have in mind very different ones). The more expensive a device, the more complex it generally is, and the higher is the motivation to repair it. I don't think we need the law only to repair car fobs and wireless door keys.

> I have yet to see an industry certification which has not clearly put in place for a business reason.

Since when business reasons became intrinsically bad? Aren't we on HN?

>> I have yet to see an industry certification which has not clearly put in place for a business reason.

> Since when business reasons became intrinsically bad?

The way I read it, it's a shorthand for "has nothing to do with a fair trade of value in exchange for money".

> Aren't we on HN?

HN is not a place frequented by wolves of Wall Street.

> HN is not a place frequented by wolves of Wall Street.

Errr... I hope they manifest themselves right away ;-)

If you design a system where remote users are impacted by other remote users' malicious actions, you've just created a bad system. "Criminals" can still do anything they want to the firmware of a JD, it's just illegal. Nothing actually stops them. But if JD designed a system that prevented local modification from affecting others, then they wouldn't have this problem.

Absolutely agree! Also in that world, all engineers would be smart, given all the necessary information, all equipped with all expensive measurement tools, and be given enough time to do good work (and motivated to do that). After coming back home at 5pm, they would grab the same expensive measurement equipment and start hacking on their devices, putting in the same high quality and knowledge and high conscience.

Illegal? Criminal? Mind you, a noob replacing a diode without knowing better, like I said.

> let you imagine what one could do with arbitrary firmware images on a fully unlocked device.

You could .. avoid paying the manufacturer's fees for re-enabling features that the firmware has disabled? You could refill your ink cartridges?

Bad actors at this layer seem to be quite rare. And the whole reason everyone uses the unlicensed bands is the freedom to transmit (and expectation of recieving) arbitrary signals.

> Bad actors at this layer seem to be quite rare.

I think our experience is rather limited for the moment to tell (ok, barring refilled cartridges ;-), since bad acting is inhibited by existing mechanisms. (Even with cartridges -- what if your new yellow ink breaks that hidden watermarking that certain people have come to rely on?)

Compliant device receives an accepted designator (certificate), non-compliant device is not expected to have it. This model can easily be broken, unless it is being taken care of somehow in the new model.

Unlicensed bands are a tiny part of the entire aired RF spectrum -- and we didn't even start speaking of devices sharing a wired RF medium.

Reply on news.ycombinator.com