Hacker News Reader

Comment by TheDong | Hacker News Reader

These patents sound like they loosely describe a worse version of ssl/tls.

They talk about the client having a whitelist of IPs to form secure channels to, and using a combination of a dns proxy and cooperating client to form secure communication layers.

The actual described thing sounds a little different (i.e. 'requests for 192.168/24 have a transparently encrypted link' rather than tls handshakes), but not different enough that it's not obvious.

From the existence of SSL, dns, and a VPN, this idea seems quite obvious to me. In 1998, all of those things existed. The existence of SSL (in 1995), should have by itself invalidated this patent entirely IMO.

From the case filings, it sounds like the supposedly infringing part of Apple's tech is "VPN On Demand" and "FaceTime".

I am not a patent lawyer, I likely don't know what I'm talking about.

View on HN

> I am not a patent lawyer, I likely don't know what I'm talking about.

The point of a patent nowadays is to be as vague as possible while still unmistakably protecting the core invention, so that a future patent infringement lawsuit receives the benefits of claims with broad scope. My cofounder and I were once told by our patent attorney (himself a startup founder) that he views a patent application as a failure if the USPTO approves it on the first attempt, because that means he was too specific with his claims.

Frankly, as ugly as it sounds, I don't blame the attorneys for working that angle. The patent system itself is structured in such a way as to reward gamesmanship.

Technologists are used to viewing clarity and conciseness as positive attributes for technical writing, but straightforwardness is kind of naive and undesirable in the patent world.

> I am not a patent lawyer, I likely don't know what I'm talking about.

I'm not so sure patent lawyers, and more importantly judges and jurors, know what they are talking about when it comes to technical details.

the solution for routing voip on a mesh network wasn't obvious.

data? yeah, albeit p2p sharing was popularized a year late, aggregating multiple channels for data transfer was something common.

but voip has some unique challenges, as you cannot reconstruct data as easily, you have strict time constraint as your call would exhibit disruptive latency otherwise, so your mesh cannot pump packet downstream at their own leisure, the goal is not to saturate the channel but to provide a ordered stream whee packets arrive in a timely matter.

as a matter of fact the first consumer available p2p voip app was skype, 5 years or so after this patent, and they couldn't stabilize it with a true mesh so their software elected supernodes across their own network to act as relay.

Reply on news.ycombinator.com