[ih] "The Internet runs on Proposed Standards"

[Grant Taylor]

On 12/4/22 2:33 AM, John Gilmore via Internet-history wrote:
> The opportunity for 2-way, cached, distributed web linking is still 
> available today, if somebody wanted to write a little software! 
> (But this is a history list, so let's go back to looking backward. ;-)

I'm not as convinced that the opportunity you're describing is still 
available.

I distinctly remember effort in the last decade by multiple web browsers 
to alter the contents of the "Referer" (spelled as it is) header for 
privacy reasons.  From memory, there are four classes of -- what I'll 
call -- redaction:

1)  None - use full URL and query parameter(s)
2)  Remove query parameters - use full URL without query paramers
3)  Remove the file path+name - use the referrer's FQDN
4)  Remove everything - don't use a referrer header at all

I'm seeing different levels used in different situations.  Usually 
crossing origin removes more information than staying within origin.

I believe there's also a Content Security Policy parameter where a 
publishing site can hint at what amount of reference they would like 
back to them.

As such, I don't think that there is as much opportunity to leverage the 
Referer (...) header as you indicate.

Aside:  I -- /personally/ /as/ /a/ /consumer/ -- don't see much value in 
knowing what other pages are linking to a page that I'm reading.  /As/ 
/a/ /webmaster/ I see /some/ value in being able to glean some 
visibility into where traffic to my site is coming from.



-- 
Grant. . . .
unix || die