Verizon’s Vexatious VoIP Patents

Over at Ars, I’ve got an analysis of the patents in the Vonage case:

The “invention” disclosed by the patent is the concept of applying these “advanced” routing concepts to DNS servers. One of the additional services envisioned by the patent is the ability to translate among domain names, telephone numbers, and IP addresses—clearly essential in any VoIP system that interfaces with the traditional phone system. The server described by the patent would also be able to “condition name-to-address processing on certain parameters relating to a request for translation, such as the time of the request, the party or terminal making the request, the status of one or more potential destination terminals, etc.”

What follows is a detailed description of the functionality provided by this DNS server, listing a variety of situations and how it would behave in each. But hardly any of the details would be interesting from a programmer’s perspective. No algorithms or data structures are described in any detail.

The patent author seems to regard the lack of “advanced” routing features in the Internet’s routers and DNS servers as an oversight. But of course, the limited functionality of routers and DNS servers on the Internet is a reflection of a deliberate design choice. The Internet was built with a layered architecture, so that most of the intelligence in the network resides in the endpoints. That allows the routers at the core of the network to be simple, fast, and reliable. The DNS system, too, was designed to provide a relatively simple task efficiently and reliably. Loading DNS servers up with a lot of customized logic for VoIP applications would slow their performance and reduce their availability when performing lookups for other types of services.

In short, the “invention” described in these patents is combination of an excruciatingly obvious point—that it would be helpful for an Internet telephony application to have a mechanism for translating between phone numbers and IP addresses—and a series of “enhancements” to DNS servers that many network engineers would regard as a step backwards.

I close by making a point I’ve made here on TLF a few times: the language journalists use in patent cases, that asks whether the defendant “used” the plaintiff’s “patented technology” is deeply misleading. Even more misleading is talk of a “workaround” for the “technology” covered by a patent. In many cases, including this one, the patents don’t cover a “technology” in the ordinary sense of the term, but a broad category of functionality. Journalists’ standard terminology obscures this fundamental point, impeding public understanding of the issue and reducing public support for reform.

SHARE: