Federating Domains: Advancing a Tiny Google Agenda?

April 14, 2021

Years ago I documented some of Google’s aspirations in my monograph Google Version 2.0: The Calculating Predator. The research was funded by commercial enterprises, but I was given the green light to publish some of the data my team and I gathered. One of the findings was that Google’s founders had an image of Google as the Internet. The idea was that a user would come to Google, and Google would serve “answers” without having to send the user to the source Web site. No one paid much attention to my diagram of how this would work (included in my first Google monograph) the Google Legacy and then in word form in Google Version 2.

A harmonic response occurred when I read “W3C Technical Architecture Group Slaps Down Google’s Proposal to Treat Multiple Domains as Same Origin.” The write up explains the Googley approach. I noted this statement in the source article:

The proposal suggests that where multiple domains owned by the same entity – such as google.com, google.co.uk, and youtube.com – they could be grouped into sets which “allow related domain names to declare themselves as the same first-party.” The idea allows for sites to declare their own sets by means of a manifest in a known location. It also states that “the browser vendor could maintain a list of domains which meet its UA [User Agent] policy, and ship it in the browser.”

Why?

I have no great intuition for the 2021 Google. It seems to me that this notion of using a “”virtual domain” for multiple domains is a useful functionality. User tracking, advertising applications, and making the Google infrastructure the central authority for named virtual hosts, and other operations is a good one. By that I mean “good” for Google.

The notion of efficiency is central to Google business model. Due to the scale of the company, consolidating, federating, and controlling deliver both business and technical payoffs. Speeding up Web site response or reducing the time required for DNS operations translates to money savings for the Google. When coupled with tasks such as streamlined monitoring, there are probably additional benefits.

Will the rejection of the Google idea cause the idea to go away? In my view, this type of virtualization which is semi-possible within WordPress (a Google fave at the moment) is in operation in some Googley test set ups. Cross domain tracking is more efficient when the federated targets pipe data into the monitoring subsystem.

Rejection is irrelevant in my experience. Google does what Google does with or without permission or the blessings of committees. Azure and IBM are poking around the same functionality. Amazon AWS may be ahead by two years, an estimate offered by a senior AWS manager a couple of years ago.

Each of these outfits have one thing in mind: Control, revenue, and data collection.

Will anyone care? Sure, the Google competitors, but to users and advertisers my hunch is that the play is irrelevant until it is not.

Stephen E Arnold, April 14, 2021

virtual parking

domain mapping (Cloud Run)

named virtual hosts

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta