I was thinking about the following situation: I become delivered a
Gemini page with links in it. These links are dynamically created by the
server and on this server coupled to my ip address. If I call the link,
the server knows, that I'm interested in this stuff in some way.

Restriction: This just works as long the links point to this (or another
associated) analyzing server, as it is with the web's technology.

A protection would just be, if no links are used at all or if data links
are used, so that all data is loaded at one with the page.

What do you think?

Martin

On 23.2.22 8.59, Martin wrote:

> I was thinking about the following situation: I become delivered a
> Gemini page with links in it. These links are dynamically created by the
> server and on this server coupled to my ip address. If I call the link,
> the server knows, that I'm interested in this stuff in some way.
>
> Restriction: This just works as long the links point to this (or another
> associated) analyzing server, as it is with the web's technology.

The Gemini protocol does not prevent this type of tracking. The server
knows what requests you make, and sees an IP address and perhaps some
identifiable characteristics of your SSL connection. The rest is up to
the user/client:

* Do not activate a client certificate unless required.
* Use a VPN service (that you can trust) to obfuscate your IP address
and approximate location. Never a bad idea if you ask me...
* Check the actual link URIs before clicking on them.
* Observe what redirects occur when navigating on pages.

Having a client option to make redirects require user acknowledgement
could be an effective way to uncover server-side tracking, although it's
at the cost of user experience.

--jaakko