The Wall Street Journal Lies

The Wall Street Journal Lies when it says, “New Windows XP Feature Can Re-Edit Others’ Sites“.

I know Mossberg’s usual slant is a of a layman approaching technology, but his misrepresentation of what Smart Tags are, and how they work, is ludicrous and paranoid, and seems fueled by the kind of ridiculous anti-Microsoft rhetoric that is usually ascribed to fanatics who think faith and emotion should be an element in a technology discussion.

The truth is, Smart Tags work like this: There are two elements, a recognizer and an action. The recognizer “reads” your Word or Excel document, looking for patterns of letters or numbers that it understands to be a certain kind of information. Here in the U.S. a classic example is the 212-555-1212 layout of numbers which signifies a telephone number that the recognizer can process.

The second part is the action, which takes the data that the recognizer has read and does something useful with it. In the most common example used, that phone number that was recognized would be passed to Outlook and used to create a new address book entry.

That’s Smart Tags, in a nutshell. So where did Mossberg get his crazy theory from? Well, if a person has installed Word XP, Excel XP, Office XP, or Internet Explorer 6, he has the option of enabling Smart Tags in web documents, but they’d pretty much only show up in a page that was created by saving a Word or Excel document as HTML.

[Addendum: This is based on the assumption that IE6 continues to ship without any recognizers or actions. I’ve been sent some info that leaves this assumption ambiguous. Right now the Smart Tag Development Kit says, "Text is not recognized automatically in Internet Explorer 5.0 and 5.5. However, specific custom HTML tags can be added to the HTML source file to enable recognition of text for which you want to have smart tags."]

If that set of criteria were met, then the recognizers present on that person’s system would determine what text had a Tag attached to it. If there were appropriate recognizers, then they would hand the Tagged information to actions. Those actions could be provided with the software, or more likely by the company for which that copy of Office XP had been installed. The classic enterprise use of this technology would be to enable quick and easy lookup of information relevant to their work. (Part numbers would be understood by the recognizer, which would pass them to an action that would retrieve inventory information and quantities in stock from the intranet.)

It is quite possible, even probable, that, at some point in the future, various Microsoft properties with valuable information to offer (MSNBC news, Encarta maps, Expedia flight information, etc.) will provide recognizers and actions for people to install. They haven’t yet. At that point, people could choose to install them and add Tags to their documents. Then those documents could be saved as web pages and their friends who also had this same combination of software could visit those Office-created HTML pages.

And yes, Walt, this would result in your nightmare situation of people being able to easily look up stock quotes on MSNBC from web pages that mention a ticker symbol. Call in the Feds.

On a semi-related note, I’ve been working on a Smart Tag for Google. The actions part is done, but I’ve been struggling with the recognizer for a few weeks, off and on. I’ll post it when it’s done.

Maybe a better way to explain all this would be to let people know that Smart Tags are content- and context-sensitive Bookmarklets for Office. You can see some of the few tags that already exist at

[Discuss](javascript:viewComments(3968661)) Smart Tags. ( echo commentCount(3968661); ?>)