h a l f b a k e r y
The halfway house for at-risk ideas
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
A website, consisting of two
user interface pages, a crawler, and a database.
Main service (Page 1): Users can enter URLs that no longer work. For a small number of them, the website presents one or more suggestions for replacement URLs, maybe with comments.
(This functionality could be exported
as a machine-accessible service as well.)
These replacement URLs are found by two means: automatically and manually.
Automatically (Crawler + Database): A search engine that notices when different places (e.g., blogs) refer to the same URL at one point in time, and when, at a later time, some of those references have changed. (The surrounding text is still the same, but the URL has changed). If the search engine detects such a case, the second URL is a potential replacement for the first.
Manually (Page 2): A place where the human owner of the resource or a knowledgable, identified person can suggest a replacement for an URL. (E.g., "I have lost access to this subtree. The new, updated version is _here_."
Source of the name.
[contracts] used it as a sample nonexistant website, prompting me to search for an excuse to create it. [jutta, Nov 13 2004]
Always a good fall-back, but I don't want a copy of the original document; I want the a new address of the live, maintained, document. [jutta, Nov 13 2004]
Proof of concept of a fully open-maintance collaborative database [contracts, Nov 13 2004]
Krelnik's content-based angle.
[jutta, Nov 15 2004]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
||What about a modified form of the second means, where helpful (but less authoritative) people can point out other possibilities?
||[jutta] said //a knowledgable, identified person can suggest a replacement// and [detly] said that helpful //people can point out other possibilities//
Sounds like you're both describing a sort of 'Linkipedia'[+]
||//create a monstrous library db//
It may not be indexed precisely for the purpose of this idea, but I think said database already exists. Google.
||My idea "Link Rover" in this category was a different approach to this same problem: capture the target page text at the time the webmaster creates the link initially, so that later when the link breaks a replacement can be found in an automated fashion.
||Perhaps this idea, implemented as a web service instead of a web site, would be the back-end of a client-server version of Link Rover.
||I think your Rover and my Emporium
would augment each other well - one
looks at context, the other at content.
(For the prototype, we can use
archive.org as your monstrous