What this page doesn't mention is 301 or 302 redirects. SEO has made "old URLs staying live" more of a widespread concern than it was at the time. And WordPress etc ship with inbuilt redirects upon slug rename
So to a large extent this has been mitigated and not using the suggestion here, which is to create a permanent URL ontology upfront
That said eventually neglect, removal, reorgs (or simply websites going offline) still happens... but the way the suggested goal has been advanced is due to becoming a business priority and with redirects and CMSes as tech to help
That said one suggestion made here turned out to be very useful and indeed is the default in WP:
http://www.w3.org/1998/12/01/chairs
If you use date as part of the taxonomy then -- as Tim BL says here:
> A reason for using a topic area as part of the URI is that responsibility for sub-parts of a URI space is typically delegated, and then you need a name for the organizational body - the subdivision or group or whatever - which has responsibility for that sub-space. This is binding your URIs to the organizational structure. It is typically safe only when protected by a date further up the URI (to the left of it): 1998/pics can be taken to mean for your server "what we meant in 1998 by pics", rather than "what in 1998 we did with what we now refer to as pics."
The headers that push a bit into the margin are an interesting stylistic choice, but why on earth does it have to be random? Did something change that affected the layout somehow? (ironically)
Not that long ago I clicked on a link that Microsoft provided somewhere in Windows -- could have been the event log, I don't remember. I do remember it was for a specific support article, but it ended up at a generic landing page for something. I am not talking about a link from Windows 95, it must have been Windows 10. But regardless, it is all gone. Then again, it probably wasn't a cool URL to begin with.
It's amazingly many news sites that also seem to scrub their URLs every time they do a redesign. Also uncool.
Very uncool; Microsoft even uses an URL shortener for most of their links (from windows etc) so that they can theoretically change where the user ends up at if the original page is replaced for whatever reason.
But they shouldn't. Microsoft especially should make all of their documentation pages plain static HTML pages and keep them available forever.
But also, any Windows documentation can live offline in the OS itself. At least nowadays, I get how in the windows 98 / XP era they wanted to reduce disk usage.
If you run a statically generated website, I recommend append-only generation, where you keep your `dist/` (or whatever web root) stateful between builds. This guarantees you don't break URIs even if the static generator changes or the source content is removed.
You can even make an orphan branch and check out that branch into `dist/` as a worktree to keep it under version control.
This is something I work really hard to maintain on my websites. I often refer to government websites as sources for my guides, and these URLs break at an alarming pace. It's as if the German government is moving reference pages and services around just for giggles. It creates a significant maintenance burden on my end.
When I worked for an ecommerce website, we spent weeks making sure every URL worked after a migration.
Nowadays it feels a little quaint because most of my 404s are from LLMs hallucinating URLs that never existed, sometimes for topics I never covered. I wonder what nonsense it attributes to me.
I've been trying to keep all URLs backwards compatible forever on swizec.com and it's surprisingly hard. I _think_ anything from 2010 onward should be good. The domain started in 2005.
I find that every time major companies as well as most authors switch blogging systems even if they maintain their URLs broadly, their RSS feed url breaks.
The concept of hooking onto a URL, for years, is a bad thing. It prohibits the URL owners from changing it. It's not 90's. You have search engines now to get to content. It's almost same as bookmarking an IP address, and complaining that the change has frustrated you, or someone scribbled the IP address on the margin of a letter.
I usually bookmark a link only to come back to a few days/weeks later. I don't expect a bookmark to work after years. I let the URL owners to have freedom to change it.
A thing that changes or goes away, feels natural. An URL that didn't change in 20 years, actually freaks me out, like a non-degradable plastic.
That's mostly about the social contract: if you publish something on the Internet, should it be made easier for the readers/users to access, or for you to move the thing around for whatever silly reason?
> I let the URL owners to have freedom to change it.
You know, you used the term "URL owners" twice, and I think I have a fundamental problem with this concept. The publisher ought not to own the URL, it really ought to belong to the consumers.
> A thing that changes or goes away, feels natural.
Yeah, but the URL is not a thing of nature, it's manufactured, and its original intent was that it was manufactured to last.
> An URL that didn't change in 20 years, actually freaks me out, like a non-degradable plastic.
A URL that doesn't change shares zero problems that non-biodegradable plastic has. In fact, the constantly changing URLs do have a couple of problems similar.
Also, you can try a different metaphor, like e.g. a rock or a mountain. It's a freaky thing, innit? Always there, unchanging... until people with excavators show up and turn a mountain into a quarry (which happened to several mountains, actually). Take it, nature!
URLs are URIs and URLs include "how to access this" which fundamentally makes it hard to prevent them from changing. If it were an abstract ID, like a UUID, or a hash or something I'd get it. So are all URLs bad URIs?
One of my pet peeves are pages that detect that I am from another region and then redirect: not to the corresponding page in my language: but to the site's front page with a different region/language setting. Sometimes it is another domain altogether.
This makes it difficult to share links to info on some products in forums, chat or a Wiki -- because the recipient may not be from the same country as I am.
30 comments
[ 3.3 ms ] story [ 22.9 ms ] threadSo to a large extent this has been mitigated and not using the suggestion here, which is to create a permanent URL ontology upfront
That said eventually neglect, removal, reorgs (or simply websites going offline) still happens... but the way the suggested goal has been advanced is due to becoming a business priority and with redirects and CMSes as tech to help
That said one suggestion made here turned out to be very useful and indeed is the default in WP:
If you use date as part of the taxonomy then -- as Tim BL says here:> A reason for using a topic area as part of the URI is that responsibility for sub-parts of a URI space is typically delegated, and then you need a name for the organizational body - the subdivision or group or whatever - which has responsibility for that sub-space. This is binding your URIs to the organizational structure. It is typically safe only when protected by a date further up the URI (to the left of it): 1998/pics can be taken to mean for your server "what we meant in 1998 by pics", rather than "what in 1998 we did with what we now refer to as pics."
It's amazingly many news sites that also seem to scrub their URLs every time they do a redesign. Also uncool.
But they shouldn't. Microsoft especially should make all of their documentation pages plain static HTML pages and keep them available forever.
But also, any Windows documentation can live offline in the OS itself. At least nowadays, I get how in the windows 98 / XP era they wanted to reduce disk usage.
You can even make an orphan branch and check out that branch into `dist/` as a worktree to keep it under version control.
When I worked for an ecommerce website, we spent weeks making sure every URL worked after a migration.
Nowadays it feels a little quaint because most of my 404s are from LLMs hallucinating URLs that never existed, sometimes for topics I never covered. I wonder what nonsense it attributes to me.
archive.org perhaps have/could have some interesting insights on it.
Unfortunately I'll never know for sure.
https://doi.org/
https://perma.cc/
https://en.wikipedia.org/wiki/OpenURL
But today is not that, it is the opposite of that with stuff being put online that really doesn't need to be. This is permaweb vs slopweb.
I usually bookmark a link only to come back to a few days/weeks later. I don't expect a bookmark to work after years. I let the URL owners to have freedom to change it.
A thing that changes or goes away, feels natural. An URL that didn't change in 20 years, actually freaks me out, like a non-degradable plastic.
That's mostly about the social contract: if you publish something on the Internet, should it be made easier for the readers/users to access, or for you to move the thing around for whatever silly reason?
> I let the URL owners to have freedom to change it.
You know, you used the term "URL owners" twice, and I think I have a fundamental problem with this concept. The publisher ought not to own the URL, it really ought to belong to the consumers.
> A thing that changes or goes away, feels natural.
Yeah, but the URL is not a thing of nature, it's manufactured, and its original intent was that it was manufactured to last.
> An URL that didn't change in 20 years, actually freaks me out, like a non-degradable plastic.
A URL that doesn't change shares zero problems that non-biodegradable plastic has. In fact, the constantly changing URLs do have a couple of problems similar.
Also, you can try a different metaphor, like e.g. a rock or a mountain. It's a freaky thing, innit? Always there, unchanging... until people with excavators show up and turn a mountain into a quarry (which happened to several mountains, actually). Take it, nature!
There are people in MS who knows this is an issue. But knowledge is not equal to behaviour.
This makes it difficult to share links to info on some products in forums, chat or a Wiki -- because the recipient may not be from the same country as I am.
See also: https://xkcd.com/869/ for a variation of the same problem.