Wow, that kinda stinks when linking to articles. I usually link to AP content on Yahoo! news. Do their articles expire in 30 days too?

The link in the post to Blogoscope's September 2003 archive led me to this one:

"Not All Google Answers Questions Make Sense" – Sep 14 2003
http://blogoscoped.com/archive/2003_09_14_index.html#106356362942374834

I was surprised to find that the link from that story to pinkfreud's Google Answers question is now broken. It turns out that Google Answers no longer works for URLs beginning with "www", e.g.:

www.answers.google.com/answers/threadview?id=255205

but it still works for the same URLs without the "www", e.g.:

answers.google.com/answers/threadview?id=255205

I'm not sure when this changed, but I think it's in the last month or two because I have used URLs with the "www" prefix not too long ago.

[links unlinked to show www vs non-www... -Philipp]

Yeah, news articles do tend to disappear quickly. That's why I use Usenet newsgroups (archived by Google Groups) to post quotes from news articles in addition to mentioning links to the articles. Then, years after, I still can refer to my own posts even when the news article has already been removed.

Interesting. I'd have thought that with Google's desire to ensure all knowledge is up and forever that they'd hold onto this stuff until the dawn of a new eon. It can't be an issue of storage, I'd think; must be something about the agreement to republish the AP material. Ideally, all the content would exist forever.

Tomi, mixing news by news agents (professional journalists) and news (messages) from newsgroups is off-topic here.

Andy, the blog post is about how quickly we lose news information. My comment was about archiving news, so I don't see it off-topic here.

Thanks for sharing information about the Google's new policy here, now we are aware that we will meet non-working links at Blogoscoped Forum too :).
Personally I have posted these kind of links recently;)

Whoa that's some useful information indeed! I wonder how many web publishers and authors realize that their precious links now lead to nowhere? I for one consider that a valuable lesson learned!

Aren't URLs supposed to be persistent, with URI-Queries the non-guaranteed value? Or am I being super-old-school?

There are many possible reasons for this – from storage concerns on the Google side, to licensing issues with all or some of the original sources (some who may make money from access to archives), to a simple cost/benefit analysis based on traffic falloff and general usefulness of archived news.

Important to remember that Google!= Archive.org

Not that the Wayback Machine stores that much news content...

Ancient Greeks used to say "Pantha rei," [everything floats/ nothing is ever immobile], and this seems to apply also to webby matters: the AP news link rot (also spelled together, linkrot), must have its causes in mutual planned news-obsolescence agreements of Google and Yahoo, and not in any technical or data-storage space limitations.

Unfortunately, today's web-accessible and largely textual docuverse has not been designed with persistence in mind. So individual linking to a more responsible, time-persistent website, rather than a fluxy AP-newsrepository, isn't really the solution – the former may for the moment look steady enough, but it only requires one misguided corporate asshole with a calculator, for whom miniscule short-term gains due to abandoning current "eternal storage" website policy outweight everything else, to make these archives equally patchy/ unusable in the end. The web isn't 20 years old yet, and hardly anybody cares about persistent data availability aspects of it.

In the end I think the ever-growing web (etc.) will have to evolve some dimension of Xanadu-like perpetual storage ability, or we won't be able to retrace the steps other people took earlier to arrive at conclusions that may be of importance to future-here and -then selves....

While web archeologists will have a tough time in the future, maybe the conclusions of all these online discussions are still preserved: site A covers something, site B and C quote from it, site D comes to a conclusion based on the quotes from B and C, site E corrects D's conclusions, sites E, F, and G all copy the core of the corrected conclusion by D through paraphrasing, and so on, a thought continuing to evolve over time. In the end perhaps A has died but A's genes (well, memes) haven't gone extinct – they've multiplied online.

Nevertheless, it would *still* be of interest in the end to analyze what A originally said, because that could offer ways to reflect on the currently breaking meme X, and with the help of knowing what A led to, we could analyze X in a new light and extrapolate into the future. We could also analyze which strategies lead to worthwhile ideas, which could help us focus on the right Y's and Z's of our time, all which could lead to more progress, perhaps...