The following was written by Robert Scoble

I've taken another look at linking technologies (Microsoft's failed Smart Tags in a beta of IE 6, Google's Autolink, and now GreaseMonkey are examples).

I've changed my mind -- somewhat. I'm now going to take a user-centric stance instead of a don't-tread-on-my-content approach. I'm also admitting I was wrong in calling all linking technologies evil.

So, let's throw out what I've written so far. Let's, instead, attempt to get big companies like Microsoft to take as user-centric an approach as possible when building linking technologies.

Many people, including Cory Doctorow told me that I was on the wrong side of the line by being anti-user (our debate should be up this weekend, Doug Kaye told me -- he apologizes for taking so long to get it up, but he's at O'Reilly's ETech conference right now so can't get them up).

So, how should we judge whether these new linking technologies are really "pro user?" Here's six guidelines:

1) Does the linking technology ship with a default set of linking behaviors?

If it does, bad. If it doesn't, good. Why is shipping with a default set of linking behaviors evil? Because that tells me that the company that is shipping the linking technology is more concerned by its revenue generating potential rather than taking a user-centric approach. From studying past human behavior we know that most people will just live with the defaults.

If you must include a default set of linking behaviors just as a proof of concept, then link to someone other than your company at first. Personally, I'd rather people who get the linking technology be asked to download a "behavior pack" which will be clear about what it does.

For instance, if Microsoft releases a linking technology, then it should let you choose your first "behavior pack." It should take you out to a Web site and let you choose from a variety of packs so you know what you're getting and how it'll behave. If a default is loaded it should set off bells for being "anti user" because then it's clear that the linking technology was designed to take users to advertising or company-specific stuff.

2) Does the linking technology potentially confuse some users when it's turned on?

Now, Microsoft's Smart Tag attempt (picture here) at least tried to use a squiggly line to differentiate itself from regular old A HREF links. I'd like future attempts to go even further.

Why can't you put something that looks like a transparent piece of glass over a word that a linking technology is adding? Why do we need it to look like a link at all?

That way no user would ever be confused. But, if you decide to stick with a link then Google's approach is OK. That forced the user to click a button on every page to see the new links. I just would hate to see what would happen when a company would try to turn that on by default.

I see a ton of potential user confusion coming. It should be clear where the new link is coming from, and, what behavior pack added it.

3) Does the linking technology give Web designers an "opt out?"

I might not want anyone to use a linking technology on my page, for instance. I'm a user, this is my work, and it seems sane to me to tell linking technologies to stay off while visiting this page.

Taking a you-must-display-my-new-links approach seems to be anti-user. Maybe that's just me, though.

4) Can the linking technology be programmed by the user?

This is a key point. It should be #1, actually. If you're gonna argue that linking technologies are "for the user" then you MUST let the user have control of the linking technology.

You MUST let the user remove linking behavior and you MUST let the user ADD linking behaviors. If your linking technology does not have this it CAN NOT BE SEEN AS USER FRIENDLY.

This is why I love Greasemonkey but I don't like other linking technologies currently being tested. Greasemonkey puts me in control. It's user centric.

5) Can the user package up new linking behaviors and distribute those to other users of the linking technology?

Again, one of the key arguments that has gotten me to switch my mind is that linking technologies are potentially user empowering.

Well, if you're gonna make that argument then you need to go all the way to the mat for this one. If I'm a user and I want to create linking behaviors that'll take you, to, say, a favorite charity of mine, then I should be enabled to do so.

If not, then the linking technology HAS TO BE judged as anti-user.

This is why I like Greasemonkey again. Mark Pilgrim, the other day, shipped a behavior pack for Greasemonkey that he called "Butler." I thought that was great. I wish I could do it with other linking technologies currently on the market.

6) For linking technologies to be seen as "user centric" they must be explicitly loaded by the user.

Including them by default with other products or services is treading on dangerous water. This is why Greasemonkey is good and Smart Tags were evil.

7) If there is conflict between links, always display the original link.

That way the integrity of the linking system stays intact. Keeping the integrity of the linking system is very important to users.

8) If linking technologies are fighting over links, either display no links, or let all the linking technologies display links and let the user choose.

It's possible, for instance, to come up with a menu. Look at the Smart Tag implementation again. A link could pull up a box that'd let you see all the linking technologies that are fighting over that link and let you choose.

Again, take as user-centric an approach as possible and make sure you don't confuse anyone as to where the new links are coming from.

Make sense?

What do you think about this set of guidelines? Disagree or agree?

I'm still waiting for a really user-centric linking technology. Greasemonkey is close, though, and made me realize I was wrong in just calling all linking technologies evil.

__________________
Former WebProWorld Admin
IntentionalFoul.com

Some of these suggestions are quite reasonable.

Allow for an opt-out tag that the technology will comply with.

Make the generated links different in appearance, such as the suggested pane of glass, a squiggly line instead of an underline, or even a yellow highlight.

Behavior packs.

These 3 suggestions would be enough for me to embrace link technology.

The Laws and other protections that protect and authors works, are not just to protect the author from plagurinsm or whatever.... they're also to protect the audience from imposters or misrepresentation.

You wrote the articel that i am replying to, I need to believe that YOU wrote it, and you intended for the contenst of it to be what you wrote. I'm depending on that information to be accurate as you intended it to be.

I'm depending on any links in your articel to be links that YOU intended to put there, or that YOU alolowed to be put there on your behalf as a relavent part of the article, and I ned to trust that if I follow thos links thay will take me to places that you intended me to see as part of illustrating the points in your article or providing resources and information that YOU belive I can trust as relaible, accurate and unbiased.

I DON'T need or want middleman shyster website owner advertiser or whoever to stick THEIR links into YOUR article whethor relevant or not. experiance has shown that their motivation is not the furtherance of my education or enlightnment on the subject matter that you are writing about, but their motivation is to sell me something oneway or another, to get me to see their advertisements to buy their products or to get me to see their point of view and ideas on a subject matter that may or may not be a point of view or ideas that you had in mind for me to be waylayed by while I'm trying to read Your article.

Now if YOU want to allow them to place all their ads and links and ideas AT THE BOTTOM of the page AFTER your article, so that AFTER I'm all done reading Your article, I might be free to explore those other ads, links and ideas on my own with the understanding that they are not part of your article and may lead to other articles, ads or ideas that are not your and are not endorsed by you, than that would be fine... but that would make them just another bunch of competeing ads that we already see on websites all the time and and entirely not what "they" had in mind.

Call me a purist, but I don't believe an authors work should "poisoned" or altered in any way not intended by the author, except maybe his Editor in the context of a newspaper or web news article.

If "they" want their viewpoint heard, "they" should write their own article not weasle into yours.

Joe DeMarco
Farnsworth Financial Service LLC
jdemarco@farnsworthfunding.com
http://farnsworthfunding.com
Toll Free 866-222-9979

One aspect of linking (#6 - Loaded by the User) relates specifically to the notion of publish-subscribe. If we viewed link definitions as subscription feeds that each user could opt-in (or opt-out of), it would simplify the debate and decision to use them.

Imagine you like Scoble's blog so much that keywords and blog post titles appeared as links in all your Microsoft Office documents when those same terms were used. Further envision subscribing to this Smart Tags collection in the same way that you subscribe to Scoble's RSS feed.

This is the model Microsoft had in mind when they first released Smart Tags in 1999. Unfortunately the the more pervasive understanding [of syndication] that we have today wasn't present in 1999, and more than a few writers jumped to conclusion that all Smart Tags were evil.

We provide an experimental [free] sand-box for Smart Tag subscriptions at http://myst-technology.com/mysmartchannels.

--bf

One aspect of linking (#6 - Loaded by the User) relates specifically to the notion of publish-subscribe. If we viewed link definitions as subscription feeds that each user could opt-in (or opt-out of), it would simplify the debate and decision to use them.

Imagine you like Scoble's blog so much that keywords and blog post titles appeared as links in all your Microsoft Office documents when those same terms were used. Further envision subscribing to this Smart Tags collection in the same way that you subscribe to Scoble's RSS feed.

This is the model Microsoft had in mind when they first released Smart Tags in 1999. Unfortunately the the more pervasive understanding [of syndication] that we have today wasn't present in 1999, and more than a few writers jumped to conclusion that all Smart Tags were evil.

We provide an experimental [free] sand-box for Smart Tag subscriptions at http://myst-technology.com/mysmartchannels.

Call me a purist, but I don't believe an authors work should "poisoned" or altered in any way not intended by the author, except maybe his Editor in the context of a newspaper or web news article.

This is a bit narrow-minded for me. The benefit of agile technologies (specifically linking systems) is that they can be useful in many ways. Some users find great benefit in automatically creating intersections between new articles being read other information resources that may have already been deemed useful.

Information users should be able to decide (for themselves) when a packet of relevant links (terms intersections) are overlayed on top of new information that they may encounter. The objective in assessing the value of any information is to determine its context. What better way to do this than by letting a Smart Tags system point out where content resonates in your own vernacular. Ideally this is achieved through a publish-subscribe mechanism.

bf --