If we chalk up the "browser vendors should not help advertisers learn more about the user for no end-user benefit" as a difference in opinion, I specifically object to adding parameters to a markup language standard to support a domain-specific problem that a (possibly transient) industry faces. The markup language should not be facilitating this tracking, that's an application workflow - it should be way beyond the scope of concerns of the HTML.
If they really wanted to standardize something, why not make a Privacy Preserving Tracking as a Service endpoint and hit it with standard snippets of JS, instead of baking this controversial junk into the browser's core.
So HTML should not be concerned with supporting web application workflows? What should people be building web applications with if not HTML/CSS/etc.? If HTML doesn't support common use cases (by supporting additional markup) then everyone will have to roll their own. And if the only tools they have to do that are ugly privacy invasive monstrosities like fingerprinting, what will they use?
What should replace internet advertising as a way of funding content? This is not an advertising industry problem -- it also affects publishers. Will they have to start charging everyone micropayments to view their content? How will this affect the ability of third parties to index and archive content?
> And if the only tools they have to do that are ugly privacy invasive monstrosities like fingerprinting, what will they use?
They use that because the privacy invasive monstrosities they previously relied on are becoming less reliable as public awareness of being spied on and manipulated grows and appropriate countermeasures are being taken. So now, every aware person is using an ad blocker and user-friendly privacy laws are being implemented. Sucks for the web advertising industry, but that's what they get for pissing upwards.
> What should replace internet advertising as a way of funding content? This is not an advertising industry problem -- it also affects publishers.
I will not be sad if platforms that so fundamentally rely on advertising that they can't find other means of funding their businesses die off completely. If people aren't willing to pay (users through donations and subscriptions or publishers out of their own pockets), maybe the content just isn't very valuable.
The point of this feature is to provide limited access to data that webpage JS is not normally permitted to have – namely, information about the user’s interaction with other sites. (Although data sharing between different origins is not always forbidden, Safari’s “Intelligent Tracking Prevention” tries to block it for origins that are used for tracking.) Therefore, making it an “application workflow” implemented by the page JS would not work, whether or not a remote server is involved.
If they really wanted to standardize something, why not make a Privacy Preserving Tracking as a Service endpoint and hit it with standard snippets of JS, instead of baking this controversial junk into the browser's core.