My Photo
Blog powered by Typepad

May 2005

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        

« Odeo Pushing Podcasting Ahead | Main | Dossier-Leaking ChoicePoint Tells Public: Tough Luck »

February 25, 2005


Ian Lloyd

Just to clarify, when I refer to opt-in, I mean that the user who is installing/running the toolbar should have to opt-in for the toolbar to display these Google-created links. There is some talk of the need for opting out from the point of view of the web site owner wh may not want links to be displayed (in other words, by using a meta tag or similar). Thought I should clear that one up ...


Hi Ian, you're mostly wrong in what you wrote, but one glaring example...

"And my Treo dramatically alters the way a page looks compressing pages and images."

But does it add links that weren't there to start with? No, it makes the best guess it can to display the page as it was designed on a small device. Not the same thing as the Autolink, once again ..."

Yes, yes it does. It detects phone numbers and turns them into links that I can click on and phone. I wish it would do the same of addresses!


"We all hates Microsoft's Smart Tags idea because it gave more, unearned power to Microsoft. Google doesn't have that same dominance, but it has enough to worry about."

It's not about dominance. It's about monopoly.

We all hated Microsoft's Smart Tags because it was a monopoly illegitimately trying to extend its monopoly.

Jacob Kaplan-Moss

There's all this brouhaha about the Google Toolbar, but I still fail to see the problem. Dan, you wrote that "as users install the toolbar they should be asked if they want features that change content on web pages. There should be an opt-in process, not an opt-out process, for such things."

Fact is, is is an opt-in process. Let's go through the steps I need to follow to use AutoLink:

1. Go to and download the toolbar (that's me opting-in once).
2. Run the toolbar installer (that's a second opt-in).
3. Click the "AutoLink" button (that's opt-in number 3).
4. Click "OK" in the dialog that explains what AutoLink is and what it does.

That's four -- count 'em, four -- times that I, the end user, has to explically and deliberatly agree that I wish to add this feature to my browser.

The fact is that the very nature of the web makes it laughably easy for the end-user to change the content of the pages they view.

For example, how is the Google Toolbar any different from the "Links via Feedster" Firefox plugin that my collegue Adrian Holovaty developed? Or, how is it different from the "Zap Colors" bookmarklet I use to magically turn pages with offensive color schemes into good old black and white? How does it differ from my "linkify" bookmarklet that converts plain-text URLs into clickable links to that URL?

Each is very simply a tool that I get to decide weather or not I want to use. You, as the content producer, have no control over what I do once your HTML hits my computer, and that's all there is to it. This is the way the web works, and has always worked.

David Sims

I also am against Autolink, but for reasons I haven't seen expressed often. Thank you for allowing me to contribute my (rather lengthy) thoughts on the matter:

So much has been discussed about Google’s new “Autolink” feature in its newest BETA Toolbar, but I can only remember one single, and brief, mention regarding Autolink’s potential impact on academic, government, legal, medical, non-profit, and scientific information websites and web-enabled online repositories. All Google’s efforts seem to involve commercial services and commercial benefits at some level, which is understandable for a commercial entity like Google; however, Google seems to ignore the fact that a great many of our society’s benefits are not driven primarily, directly, or even indirectly by monetary considerations.

There are many millions of web pages (probably hundreds of millions) produced by organizations such as mine that have a mandate to produce authentic, authoritative documents. You know, that neat stuff they’ve got “over there” at NASA; that really useful report on licensed contractors at your state Consumer Protection department; the immense archives at the Library of Congress; helpful publications you may unfortunately find yourself needing from the NIH or the CDC on various diseases; those great government tourism sites you log onto every Spring, in anticipation of an exciting Summer getaway; the non-profit foundations and research organizations that provide analysis, research, and resources on a whole rainbow of topics; the really cool articles from science journals that contain the newest in scientific and technological research and discoveries…you get the idea, don’t you?

Whether you are for or against Google’s new Autolink feature (or the earlier Microsoft smart tags) or any other specific third-party software that manipulates content producers’ web pages, it should be discussed whether, as a society, it really is okay to “mix” or “remix” *ALL* varieties of information—the “users want it, so it must be okay” argument espoused by Google and many of its supporters, for whom “user convenience” is the (stated) driving force behind this technology.


For instance, “my” site contains thousands of pages of government statutes, regulations, and legal opinions that carry the weight of law (known as caselaw), and other information that, among other requirements:

(1) must remain intact, as it has been created by a legislature or other mandated government authority;

(2) must remain impartial in its presentation;

(3) must serve all participants within a system (legal, medical, etc.) equally, regardless of who a particular participant is, what political ideology they hold, etc.

Why must it remain intact? —To truthfully serve our “users” and not lead them astray, which results in effective and efficiently-running legal systems, medical practices, and more.

Why must it remain impartial? —To provide the credibility required so that our “users” can come to trust us as a (or *the*) source of reliable, authoritative information and helpful service.

Why must we serve everyone, without preference? —Because we are a *public* agency; our mandate is to serve everyone, without giving preferential treatment to anyone, but rather ours is to provide assistance and service fairly and equitably.

For *any* software to add content (even in the form of links) to the content my agency publishes online distracts users from the true intent and value of the publication of that content and, in the future, could result in potentially unnecessary, lengthy, and costly litigation. One simple example: every single legal opinion on our site is a public document that carries the names and addresses of the attorneys involved, yet linking them all completely distracts from the true (non-commercial) value of the document and could (in the minds of our online customers) imply our agency’s endorsement of particular individuals and/or firms, which is neither our intent nor our right—remember, unlike a private sector commercial entity like Google, we’re mandated to be impartial.

And if future “upgrades” to Autolink come down the pike, other “distractions,” potential “misinformation,” and inadvertant “endorsements” could show up on our pages unannounced. Will our taxpayers, our customers, know the difference? We’d like to think so, but many years of experience leave me very unencouraged in this regard.

In addition, is it fair to citizens of this nation for agencies like ours that serve public, societal interests, and for those of us who work in these organizations, to be forced to spend time (and dollars) on “protecting” ourselves from private-sector encroachments like Autolink? I’m not going to give my personal opinion, but my professional opinion is that it certainly weakens our ability to carry out our missions as public, societal institutions since we now have to “protect” the material as well as simply creating it and making it available. The net result will be less content, and that goes against the inherent strengths of the web to serve our fellow human beings (and save money in these tight-budget days).


Another problem: Google’s own documentation says, “Google may collect information about web pages that you view when you use advanced features such as...Autolink...” Now, some people would say this is spyware...some would say this is just unethical...yet others just see it as a part of “capitalism” and therefore, to them at least, it’s perfectly okay. Whatever you believe, at the very least this tracking of web activity conflicts directly with the privacy policies that many, if not most, government agencies must abide by. My guess is that it also conflicts with the policies of many other types of organizations as well. Legally, I don’t know whether this is a problem for us (I suppose only my Attorney General could tell me for sure).

However, so many “users” won’t see any line dividing our pages from the “extras” that Google gives them, that it may create a problem for academic, government, legal, medical, non-profit, and scientific entities in the “user’s” eyes—they’ll go to one of these highly-textual, information-rich sites with the expectation of (relative) anonymity, but will (through no fault of the sites visited) be tracked.

Now I ask you: is it fair to people who have paid for a website (either through tax dollars or through charitable giving) to be tracked, spyed upon, in return for those dollars? Would you like it? Is it the worst thing that’s happened to humankind? —no…but it’s not nice, and it’s neither ethically nor morally defensible.


I have read in various accounts statements from Google to the effect that Google knows what users want and knows their online behaviour, so they can and should do this. I admire Google’s ability to read their “users” minds—this is truly legendary talent—after nearly 8 years developing a site and speaking with many hundreds of its online customers, I am still amazed at the various viewpoints from which so many intelligent “users” like attorneys, medical doctors, university researchers, etc. come—I myself would *never* be able to make a statement to the effect that “when a user comes to a [whatever type of] number, they’ll of course be ready to leave the site.” I am thoroughly floored that any human corporation, such as Google, can actually know *when* people will leave a site, and further that they have the educational and professional expertise in *every* situation to provide the *most* appropriate links for “users” to follow…really, truly gifted individuals these Google folks!

In the worlds of academic, government, legal, medical, non-profit, and scientific publishing, in so many instances links are endorsements of the legitimacy of information *and of its pertinance in certain circumstances*. I know it might seem like heresy to some reading this, but I actually do believe that the publishers in those communities know better than Google what is legitimate and pertinant in a given legal case, a page on treatment for cancer, the intracasies of nuclear particle behaviour, and so on.

As good as the folks at Google are at what they do (and they *are* good, we all admit that), they are *not* the very EXPERTS creating and publishing the billions of web pages that Google itself indexes.

It is just plain irresponsible for Google (or any other “reputable” organization) to make changes to content where a seemingly innocuous change (from Google’s perspective, in this case) might have an unknown (again, from Google’s perspective) and potentially negative effect upon parties relying on a particular body of online information.

I hope Google will reconsider implementing this and similar technologies; it’s a big mistake for their “users,” contrary to what they themselves may believe. Changes to such material might suit some “users” just fine, but as a responsible society we don’t let kids play with fire just because they want to—even adults cannot be experts in all fields of endeavor and should not, through this type of technology, be made to feel that they have the moral or ethical right to “remix” material for which they simply don’t have the authority, background, education, experience, training, etc.

I don’t consider this attitude to be snobbishness on my part, and I am certainly not trying to put “users” down—I am a “user” too, and am comfortable enough with my own human limitations that I can admit that I myself am not necessarily an expert in medicine, law, science, or any number of other fields—I've got to have respect for those who *do* know about these things, those who have spent lifetimes gaining specific knowledge in specific areas, and to have confidence in their abilities and motives—we have ALL got to do that. If the integrity of the content they produce comes into question, then we will *all* in this society have a problem far bigger than Autolink…and if we cannot respect other people and value them for their gifts and abilities, we will lose our humanity.


One last argument I’ve seen wielded in favor of Autolink-type technologies is the notion that if someone builds a site but doesn’t give you good links, then “we’re here to ‘save the day’ and provide a way out of this (lousy) site.”

That argument seems to me to reward those who don’t build good sites and hurt those who do build good sites, the polar opposite of the service upon which the good Google name and reputation have been built. I frankly don’t understand this viewpoint; it seems to me to undermine the validity of the core service for which the whole world admires Google—it just doesn’t make sense.


I ask Google to remember that an important, sizeable, and significant part of the web consists of material that is beyond the scope of its own professional expertise—this viewpoint requires humility, but I think Google can handle the challenge of respecting the rest of the world’s contributions to the web as the rest have already found respect for Google’s own contributions. This material plays an essential (though often not directly commercial) role in making this a relatively safe, enjoyable, and desirable society in which to live. “Playing around” with this material for “fun and profit” diminishes its true value and is a short-term investment in the web—it *may* make money for Google and its stockholders, it *may* make some “users” happy because they feel a freedom and power that is near unlimited, but it won’t serve society’s deepest and most important human needs.

I respectfully ask Google to remove the Autolink feature immediately, and further to develop no similar features in the future. Such features simply are too editorial in nature and diametrically oppose Google’s existing strengths.


As has been suggested elsewhere, IF Google insists on going through with this disrespectful scheme, then I would *demand* that Google provide a simple, reliable means for content publishers to "opt-out" of any Autolink-type features, something like:

(meta name="GoogleAutolinkPreventParsing" content="TRUE")


In some cases, it really *is* necessary for the benefit of the “users” to preserve document integrity: you not only want, but you NEED, the “correct” statute, the “real” info on cancer treatment, an impartial guide to resources, etc. without the added commercial pressure of the pursuit of monetary gain influencing such important documentation and informational content.

Publishers MUST have the final say in these types of circumstances: they are, after all, the experts in their respective fields and presumably that is why visitors to their sites are going to *their* sites for information. If it were otherwise, we could all just go straight to Google and nobody would need any “external” sites anymore, Google would have “all the answers.”

Please Google, listen to and respect *all* of us who are part of a diverse, healthy, vibrant web.


I think, when your article says "... What Google isn't taking into account is that its market power, and the tendency of users to accept the default -- to eat what's on the plate someone puts in front of them -- will tend to create Google's version of the Web, not the users' version ...", you really should be saying "... What Google IS taking into account ..." I think they know full well that users obviously accept the defaults, and in their opinion that means they're onto a winner.

The comments to this entry are closed.