Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Backlinks and Annotations

By dazzle in Internet
Fri May 17, 2002 at 11:35:23 AM EST
Tags: Technology (all tags)
Technology

The basic idea of the web is one of movement. Most websites attempt to keep you on their site rather than help you to move around the web. They see themselves as the be all and end all of a surfers web experience. By using some simple tools - detailed near the end of this article - the web user can begin to control his, or her, web environment, and help to influence other web users in the process.


The idea for this article came about the other day while I was viewing some of the recent posts on drop.org. I followed a link to an article, 'Blogspace Under the Microscope', on the O'Reilly Network. The author was talking about a 'mutation' and the next 'phenomenon' in weblogging. He was referring to 'backlinks' - links which point back to the referring web page - which the web site Disenchanted has been using for about a year. The article consisted of very few links and no historical or background information to the idea of 'backlinks' and implies that Disenchanted invented this idea; which is dismissed by an accompanying article on Disenchanted.

After a quick search on Google I came across an early mention of 'backlinks' - going back to 1997. This led me onto finding CritSuite and the whole idea of annotations - comments which can be 'attached' to a web page but are stored independantly. CritSuite also implements a 'backlink' idea in allowing you to find documents linking to a web page via AltaVista.

I had started at one point, a small post on drop.org (A), which led me to an article on O'Reilly Network (B), which intrigued me to do a search on Google (C), which led me to a historical article about WebEnhancement (D) and eventually to CritSuite (E), which brought together the idea of backlinks and annotation in one, user controlled, application, thus completing the circle.

Surely it would've been easier for (B) to have pointed me directly to (E) rather than having to go through, (C) and (D). A simple search by the author would have provided the relevant links.

This is the idea behind 'backlinking'; providing a route directly through the web following a well trodden and, hopefully, relevant path. If I had known about annotation as well I could've updated the information in the article for the author providing a link to CritSuite, or other annotation tools I found about once I had done another search on Google. I could even have written an article on annotation and added that to the original article, thus marrying the two ideas together. All this without touching the original article. I would've been providing more of a choice of paths from (B), thus helping to facilitate a choice of movement within the web, rather than the surfer only having the choices given by the author of the article.

------------------

Below are detailed three annotation systems which I have found and two, basic, ways of 'finding' 'backlinks':

CritSuite seems to be the oldest system for annotating web pages. As well as allowing you to add comments to any web page, you can also retrieve backlinks via AltaVista. CritSuite is extremely easy to use and the only downside is that you have to surf the web through the CritSuite web site.

Another, similar, server-side annotation system in use is the Annotation Engine. Although this system doesn't look as good as CritSuite it could still provide an interesting way of adding comments to any webpage.

A more flexible annotation system has been implemented by the W3C. This system, Annotea, is already built into the Amaya browser. You connect to an Annotea server using an annotation client, implementations for IE and Mozilla are available, as well as javascript for use in bookmarklets, and then retrieve or post the relevant comments for the webpage from the server. I have tested this out using IE and Mozilla and have not had much success in using either implementation. Amaya is easier to use with annotations but is not really suitable to be used as an actual web browser. Annotea also allows you to run your own annotation server and provides the information to write your own client.

Although it is only CritSuite which provides backlinks as well as annotation functionality it is very easy to add client side backlinks to your web browser using bookmarklets - javascript links which you save as either a 'Favourite' or 'Bookmark', in Internet Explorer or Netscape respectively. An example is given below:

Right click on the following link and save it to your 'Favourites' or 'Bookmarks':
Google Backlinks

Then, whenever you are surfing the web and you want to see which pages link to the page you are viewing, goto the saved 'Google Backlinks', a new window will pop up showing you the Google search results.

At Link Feedback a way is given in which anyone can provide a basic 'backlinking' method on their website by adding some javascript to their web page. Each time someone visits the page their referrer is logged in a database and the last ten referrers are displayed on the visited web page. Unlike Disenchanted this is not filtered and so includes links from search engines and e-mails.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Google
o drop.org
o invented
o article
o post
o article [2]
o search
o WebEnhance ment
o CritSuite
o path
o AltaVista
o Annotation Engine
o Annotea
o Amaya
o IE
o Mozilla
o bookmarkle ts
o server
o client
o Google Backlinks
o Link Feedback
o Also by dazzle


Display: Sort:
Backlinks and Annotations | 31 comments (19 topical, 12 editorial, 0 hidden)
I don't really understand (2.66 / 3) (#8)
by iwnbap on Thu May 16, 2002 at 09:51:23 PM EST

Are you proposing to create the transitive closure of a link? So if A -> B and B -> C than A -> C? (where -> defines a "links to" relation). It sounds like a good idea; but surely the way to do this is to change servers/http so that they make these references automatically.

Re: I don't really understand (none / 0) (#20)
by Vs on Fri May 17, 2002 at 02:41:10 PM EST

No, it's more like adding B->A, with added value as you might want to comment on why you did link back.

So you could go C->B->A, to reuse your example. Of course, if you think it's important, you'd create the immediate link from C->A.
--
Where are the immoderate submissions?
[ Parent ]

ack! efficency (4.25 / 4) (#9)
by Matt Oneiros on Thu May 16, 2002 at 10:06:31 PM EST

It sounds far too efficient

seriously.

The only reason I keep using the internet is because it's the easiest way to inadvertantly slip from the path of a->z and end up saying "screw the fricking z! I'm going straight for 289b"

the total and utter unmoderatedness/randomness of the internet is what makes it so damn good. The fact that you can jump on the subject of buying a tatoo gun then next thing you know you're reading a parody page in reguard to a beowulf cluster of furbies... it just tickles my fickle and utterly randome fancies.

but you've got some good ideas

Lobstery is not real
signed the cow
when stating that life is merely an illusion
and that what you love is all that's real

Meta (none / 0) (#11)
by Holloway on Fri May 17, 2002 at 03:47:03 AM EST

The long dead Teethmag.com had a type of link where random words were linked internally to other pages. It might have been called Bubba*

A browser plugin annoting system Third Voice was big a few years ago (but they're now dead). It was predated by something from at least '96.

There was webpage IRC where channels were urls, you spoke in speech bubbles, and you could drag an icon of yourself around the page. You could leave messages and some pages (like, say, Yahoo's homepage) were littered with arguments.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

interesting (4.00 / 1) (#13)
by tps12 on Fri May 17, 2002 at 11:49:15 AM EST

I kind of like the idea of combining, e.g., IRC and web browsing. It seems like there's a lot of potential for unifying the different protocols and applications on the Internet. Although I can't think of a single other useful example. :(

Anyway, it just reminds me of BBS's and Hotline and stuff, where you have different services all "under one roof."

I know why I like it. It's a classic UI design principle, which is designing the interface to meet the users' needs and expectations, and not the underlying program's. So controls that are logically related should be grouped, even if they are functionally very distant, and vice versa.

Likewise, Internet users shouldn't have to be thinking "I'm looking at a web page" or "I'm transferring files" or "I'm talking to other people." Those are implementation details, and just distract the user from the real topic they are interested in (warez and porn).

[ Parent ]

similar to this IM.. (5.00 / 1) (#15)
by dazzle on Fri May 17, 2002 at 11:56:25 AM EST

There's an IM called Odigo which shows you people, using Odigo, who are looking at the same web page as you and allows you to chat to them. You can also leave 'notes' on the web page for others using Odigo to read. It is also compatible with other IM systems like AOL and MSN.

---
the internet: a global network of small minded people


[ Parent ]
side note.. (none / 0) (#16)
by dazzle on Fri May 17, 2002 at 12:03:24 PM EST

Ideas similar to ones expressed in this article have been theorised or implemented in wikiwebs:
IAwiki
Meatball

---
the internet: a global network of small minded people


Other linking models (none / 0) (#17)
by shivers on Fri May 17, 2002 at 01:19:08 PM EST

There are many other hyperlinking models different to that used by the web - most of which provide backlinking as a byproduct of their implementation. One commong way is to have databases of links, holding source page+location and destination page+location.

There are a few advantages - easier to manage, easier to search, never any broken links. But implementations are a little more complex and there can be scalability problems.

The university I went to had their own version which was quite successful in its own right (Microcosm TNG).

I guess its a bit esoteric given the kind of hyperlinking we're all used to - but interesting nevertheless.

Well I be damned, it's Xanadu! (5.00 / 3) (#18)
by romanpoet on Fri May 17, 2002 at 01:35:23 PM EST

Serious geekpoints to anyone who remembers a man named Ted Nelson [Info], A PhD in Philosophy from Swarthmore before being introduced to computers by some friends at MIT. He developed a very interesting network model called Xanadu. [Xanadu.com] Back in the 60s we was proposing ideas very similiar to that which you propose. However, a Wired article from 1995 [article] basically destoyed his career and his prospects of ever bringing his Xanadu network into reality.

Give a look at his network model, http://www.xanadu.com/xuTheModel/.

Alternative and theoretical network protocols and models have always been a favorite interest of mine and his work is some of the most interesting out stuff there.



Yes, it is (none / 0) (#21)
by jazzido on Fri May 17, 2002 at 03:15:06 PM EST

'Backlinks' can be seen as a special case of the idea of 'bidirectional links', one of the main concepts in which Xanadu is based.

Nelson's model, IMO, is brilliant, but as the Wired article says, is the longest running case of vaporware ever seen. In addition, many of his ideas, such as the 'transpointing windows' are pretty esoteric.

--
"Patriotism is the last resource of scoundrels" (Samuel Johnson)

[ Parent ]

No it's not (none / 0) (#25)
by arthurpsmith on Sat May 18, 2002 at 02:09:13 PM EST

There was something called Hyper-G that came out shortly after the WWW that allowed separation of the "links" from the content - obviously in that case forward and backward links are essentially equivalent, and this is pretty close to that model. Xanadu was based on a whole concept of authorship and ownership of chunks of text (that would be "transcluded" in others, and paid for by complex credit mechanisms) that is very foreign to the way the web has turned out. On the other hand, Tim Berners-Lee did give Ted Nelson some credit for inventing hyperlinking, so there is a connection.

The Wired article didn't do in Nelson; the simplicity of the Web had already made his complex hyperlinking model obsolete. There are a huge variety of linking models out there (for example some old SGML standards used by literary text purists to track multiple versions of a manuscript) and some of these are being re-invented again with XML. Nelson certainly deserves a lot of credit for coming up with and evangelizing these ideas very early; but they're really not terribly relevant to the real world any more.

Energy - our most critical problem; the solution may be in space.


[ Parent ]
Annotations are OLD (none / 0) (#19)
by Maniac on Fri May 17, 2002 at 01:40:35 PM EST

To quote from NCSA's archive...
http://archive.ncsa.uiuc.edu/SDG/Software/Mosaic/Docs/old-whats-new/whats-new-06 93.html
[If the link above is broken (should end in 0693.html) - it is Auto Format's fault... :-)]

 June 19, 1993
    A new version of the public group/community annotation server running on hoohoo.ncsa.uiuc.edu:8001 has been put in place. Now, backlinks to the annotated documents are automatically provided in new annotations. (For information on the public annotation server, see here.)

    A discussion of some design issues in group/community annotation systems, including the collision of WWW annotation systems and news, is now here; currently there are discussions of a Usenet approach to global annotations, and notification methods for tracking annotation activity.

I also did a quick search w/ google with the phrase mosaic web browser annotation and found several interesting reports on this kind of material.


backlinks + freenet (none / 0) (#23)
by nadaroot on Fri May 17, 2002 at 09:00:28 PM EST

All of the systems mentioned use centralized storage. This has a tendency to not be available when you need it. Servers go down and businesses leave the scene. The solution would be to store the backlinks on a distributed, redundant system such as Freenet, etc.

Google, Alexa, (5.00 / 1) (#24)
by jbond on Sat May 18, 2002 at 04:18:41 AM EST

Both the Google and Alexa toolbars provide you with a button for "Related pages" and "Backward Links". But my experience is that this tends to only work effectively for home pages. Maybe what's needed from both of them (As Alexa uses Google now as it's search engine)is a script that finds all links to the site and excludes links within the site, while grouping links from a single site together. And while we're at it, stripping out the links from aggregators. This is not too hard to do on your own server as server side processing of the referrer log, but it's tricky (probably not impossible) to do it generally as a bookmarklet. Anyone up for taking Dazzle's bookmarklet and taking it to the next stage?

Vannevar Bush and the MEMEX (none / 0) (#26)
by netjunki on Sun May 19, 2002 at 02:13:29 AM EST

dazzle's comment about "relevant path[s]" reminded me of an article by Vannevar Bush that I read entitled "As We May Think" that appeared in Atlantic Monthly in the summer of 1945. In it Bush proposes the construction of a device to help facilitate the sharing of knowledge between scientific fields. One of the features of this was the concept of "trails" or paths through a collection of information. So the reader could follow the same path that led someone else to an idea. Ted Nelson and Doug Englebart were both influenced by Bush's work.

...and backlink again... (none / 0) (#27)
by dazzle on Sun May 19, 2002 at 07:14:06 AM EST

Which in turn is similar to a Glass Bead Game mentioned in the book of the same name by Hermann Hesse.

From a Wired article:

I'm certainly not the first to think of it. From Hesse's Glass Bead Game, to Vannevar Bush's memex, to Ted Nelson's Xanadu, poets and scientists have dreamed of building a universal compendium of human knowledge. Now that we have the tools to construct what the librarians of Alexandria could only imagine, what stands in our way?


---
the internet: a global network of small minded people


[ Parent ]
An older annotation and weblog site is.. (none / 0) (#28)
by johwsun on Mon May 27, 2002 at 02:42:16 AM EST

www.hypernews.org is older that crit.

It started as an annotation service, then turned into a weblog. Now it is a dead site, no one talk there. It died simply because it did not incorporate readers ratings.

Another backlink server (none / 0) (#29)
by cdbackslash on Wed May 29, 2002 at 02:25:42 PM EST

See the brand-new SackLinks page at http://www.mindsack.com. Centralized database, stores all referrers with hit counts, shows the top ten on your page, easy to surf around inside. Uses four lines of JavaScript on your page, but you can change your referrer if your page comes up inside a frameset. Does not exclude hits from search engines, so you can see the goofy sh!t people are really looking for when they hit your site.

weblog -> annotation gateway (none / 0) (#31)
by n4th4n on Fri Oct 11, 2002 at 05:39:21 PM EST

If I provide a way for weblog entries to be translated into annotations, then you can look at my entries as you surf the sites they apply to. If you can subscribe to the annotations of people whose opinion you respect, then the annotations you see as you surf the web become a valuable "third voice" (oops). As a weblog author, I can write entries designed to be viewed as annotations as well as in the context of my weblog, and that subtly changes my context. I'm super interested in this idea, to the point of creating an annotea gateway to my weblog (maintained with b2). Anyone else who's interested feel free to drop me a line.

Backlinks and Annotations | 31 comments (19 topical, 12 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!