Re: Are websites closing down en masse? (distributed free standards and tools)

[Paul D. Fernhout]

A search on "average lifespan of a web page" (and similar) produces 
estimates ranging from forty days to just under three years (including 
an estimate supposedly derived from archive.org at some point). So, in 
general you are right that most web pages don't last very long, but this 
is not especially a new thing.

Although I can wonder if there are recent trends that might make this worse?

Part of the issue may be that Google tuned its results several years ago 
to prioritize recent content over older content (i.e. "Freshness 
Signals"). Also, social media companies tend to promote new posts. 
Partially as a result, most web pages of various sorts get their 
greatest number of views in hours or days after they are posted. Thus 
there are diminishing financial returns to advertising-funded content to 
keep it up past a few days. Website design companies also make money 
promoting "refreshes" to their clients every two years or so. So, there 
are a lot of obvious financial incentives for various actors to put up 
new web pages and fewer incentives to keep up old ones.

That said, I personally like the idea of permanent URLs. I try to keep 
webpages up that I create at the same URL. Some have been up for over 
twenty-five years. But at some point, after I pass away (or just enter 
old-age and perhaps poverty), will someone else want to keep those 
websites up? Priorities by individuals and communities can change over time.

One suggestion I've seen is for "permanent" URLs is to typically 
including a date of creation in URLs for things like blog posts. Then 
even if you change your content hosting platform you can more easily 
keep up the old URLs. But all too often I see websites redone with all 
the old content discarded or moved so old links are broken.

Methods using the hash of content as a link may help preserve public 
content if we have more distributed systems. This is because the content 
is not tied to a specific domain that may expire or a specific server 
that may be retired.

Archive.org and similar are amazing resources for keeping old content 
available. Wikipedia has some interaction with them to provide 
archive.org links for items linked from Wikipedia. I try to ensure 
content I put on the web on personal sites is archived there. But, while 
I hope archive.org ans similar will prosper for decades to come, there 
is no guarantee that archive.org will be around a long time either due 
to risks related to funding issues, technology issues, management 
issues, and/or legal/political issues.

I like email as a way to personally archive some forms of content. It's 
been sad over the years to see Mozilla short-change Thunderbird 
(implicitly a distributed content system) in favor of Firefox (generally 
used to access centralized content) with how they have spent about a 
billion dollar a year they have gotten from Google and similar funders.

Glad to read about "hyperdrive.el" by another poster as one more 
alternative for distributed content. I personally enjoy working on 
software in that area myself sometimes in my spare time.

A deeper issue, however, is the need for wide adoption of free standards 
for persistent distributed content (like HTTP became a widely adopted 
standard). While coding is fun and potentially useful, such standards 
are more important for social software than good free implementations 
(even if they ultimately go together and benefit each other, and a 
really good implementation widely adopted can define a de-facto social 
standard).

I made a Lightning Talk for LibrePlanet 2022 related to that issue:
https://media.libreplanet.org/u/libreplanet/m/lightning-talk-free-libre-standards-for-social-media-and-other-communications/

And I say there, I think helping define and promote such free standards 
for free distributed content and related tools is a valuable role the 
FSF could play in fostering a more libre planet.

Some related content by archive.org:
https://blog.archive.org/tag/distributed-web/

--Paul Fernhout (pdfernhout.net)
"The biggest challenge of the 21st century is the irony of technologies 
of abundance in the hands of those still thinking in terms of scarcity."

On 12/3/23 22:53, Akira Urushibata wrote:
> Recently I feel I frequently encounter defunct links.  Links to
> external material toward the bottom of Wikipedia articles often turn
> out to be unavailable.
>
> I don't know if there is any empirical data on this.