Microsoft Kills IE6 Smart Tags
June 28, 2001 4:59 AM   Subscribe

Microsoft Kills IE6 Smart Tags "External feedback" was one of the factors that led the company to remove the feature, although he indicated it could be resurrected in later versions. Perhaps the constant barrage of complaints worked on this one? They've got it working, now they are back peddling. I for one am glad, "No Squiggly Lines".
posted by benjh (60 comments total)
 
good riddance
posted by lotsofno at 5:12 AM on June 28, 2001


Once again the MeFi mob gets results?
posted by jpoulos at 5:17 AM on June 28, 2001


if you think it's going away, you don't know microsoft. "a form people can live with", my ass.
posted by quonsar at 5:19 AM on June 28, 2001


In this recent brouhaha, why was it almost completely overlooked that SmartTags - far from "going away," as some seem now to believe - are, in fact, already here. The code for them is in Office XP, which is or will be installed on some insanely high number of Windows OS machines. The UI "switch" that turned them on in the browser may be hidden, but the functionality is - and will still be - there. I wouldn't get too jubilant too soon, kids. It'll only be a matter of time before you'll be pointing your Microsoft browser on your Microsoft PC to some Microsoft website and be presented with some Microsoft Special Feature that links individual words to other Microsoft content (presumably, content you'll pay for using Microsoft HailStorm). And the implementation of that same "special feature" on other sites that have to compete or at least stay current with Microsoft will be fast and furious...
posted by m.polo at 5:40 AM on June 28, 2001


Yeah, I agree that they're probably not "gone". The fact is, they already exist, and it'll be possible to use them in the future anyway.

They're removing it from IE6, but there's nothing saying that they've killed the feature entirely. It's still in XP. This announcement is just meant to extinguish the fires somewhat. It's a PR move.
posted by Succa at 6:21 AM on June 28, 2001


Actually, correction, this article states that MS is removing Smart Tags from Windows XP itself, not just from IE6.

Well, fine. But I'm still not convinced.
posted by Succa at 6:34 AM on June 28, 2001


Yeah, good work "mefi mob".. oh and every single other person on the web. I actually expected certain other people to start patting themselves on the back for this, but rest assured: there were thousands and thousands of people complaining.

Anyhow, I wouldn't worry too much about the whole "might be back in 6 months" thing, if MS tries to put this back in IE, the same thing will happen, and they'll probably have to back off then too.
posted by beefula at 6:45 AM on June 28, 2001


Microsoft has decided to exclude Smart Tags--a technology that could alter the Web surfing habits of millions of consumers--from the version of Windows XP that will ship later this year.

What I'm wondering is that Microsoft's next OS, Blackcomb, is highly dependent on SmartTags. That makes me agree that, yah, it's coming back. But maybe, just maybe, they'll make the lines more... uh, aesthetically pleasing this time.

Ha!
posted by dequinix at 6:49 AM on June 28, 2001


Yes, it'll read the stylesheet and link attributes of the page to make their links blend in. ;)
posted by solistrato at 7:22 AM on June 28, 2001


Thank you, Bill. This whole story proves that developers still have some power. MS does listen to the web development community. At least a little.
posted by ubique at 7:47 AM on June 28, 2001


Chalk another one up to the paranoids. I've talked (well, emailed) with people involved in the development of the Smart Tags implementation in IE6. Maybe they're naive, but they honestly thought the ability to download pieces of code that extend hyperlinking in pages that you're viewing would be something people would use and like.

I, for one, thought it was an interesting way to extend hyperlinking. (Yes, please, forward me your "MS shill" hate mail...) They hadn't anticipated that there would be an uproar from optional extensions that enabled pervasive hyperlinking.

I think they presented the feature poorly, but that it was a great idea that did a good job of indicating future directions for the web. My only objections were from a technology standpoint, I thought it should have been easier to create an XML file that would describe patterns to be matched. As it was, most existing Smart Tags were just list-based matching, which DeepLeap already did rather effectively.

I wonder who some people would want to create our future pervasive hyperlinking systems. It seemed to me Microsoft was in the best position and had a relatively credible (if, as ever, overly proprietary) implementation. But it appears to have fallen victim to the accumulated ill will towards Microsoft's past behaviors.

There are probably ways to make an ActiveX control that would have linked to the IE HTML parsing engine and talked to a DeepLeap-esque system. But I doubt anyone will make them, now.
posted by anildash at 7:58 AM on June 28, 2001


What you're missing is that they are leaving Smart Tags in Office XP, which has a bigger footprint into individual people's world.

Putting it in Windows isn't really a gain for anything, since there is no real content that it would be able to function over. Keeping it at the application level, they can now tie their applications to the internet without Explorer.

See, the whole thing about Smart Tags being in Explorer was that it would change content of web pages and hi-jack competitor's sites without their knowledge.

With it in Office, however, they hi-jack a larger space - personal and professional documents ranging from Powerpoint presentations to Excel documents, with Word being the big one.

No while you're reading a document, a smart tag will direct you to a Microsoft owned or partnered web site to drive revenue off of some individual's document.

See, it's easier to co-opt the individual than it is major companies' web sites. I think it could make a good class-action lawsuit, but it would be hard to prove and get moving.
posted by rich at 8:04 AM on June 28, 2001


(Oh, and I think the idea behind Smart Tags is great, but having a standard default link back to Microsoft's wants and desires is where the problem begins)
posted by rich at 8:06 AM on June 28, 2001


Agreed Anil. The thing that Winer, Slashdot, etc. seem to not understand is that the feature was off by default. Ho would it be any different than NBCi's Quickclick? The only determinant can be that this was Microsoft, ergo it must be evil.
posted by owillis at 8:09 AM on June 28, 2001


Anil, all they had to do was to make it opt-in rather than opt-out and a lot of people would have had no problems. It's not the technology per se that's the problem, it's the fact that the 75% of the pages on the web that are authored by people with minimal HTML skills would have been altered willy-nilly.

And, at a slightly deeper level, a lot of us are sick of Microsoft's "Do What I Mean" (or "Do What Some MS Programmers Think I Might Have Meant," but DWSMSPTIMHM just isn't as catchy somehow) approach to interface design. If you get sick of URLs magically turning into hyperlinks in Office, and numbers suddenly becoming dates in Excel, you probably don't appreciate your losing control of your web surfing too.
posted by rodii at 8:18 AM on June 28, 2001


if MS tries to put this back in IE, the same thing will happen

Actually, I doubt it. Next time it won't be shocking, and the reaction will be more like "oh, THIS again?" than "how dare you even contemplate such a defilement of my sacred HTML?".

I'm more annoyed by the reaction than the feature. Of COURSE the user agent should have the ability to present a page in whatever way it likes. Of COURSE it should be able to spontaneously redesign the page on the fly in whatever way the browser designer thought would be useful. Content vs. presentation, and all that, and the browser is all about Presentation. Philosophically speaking, smart tags are a great idea - an example of the sort of user-interface tweaking the browsers should have been experimenting with all along.

Still... given the fact that content-vs-presentation is pretty much dead, along with every browser but MSIE, and the fact that Microsoft is the new Standard Oil, maybe it was worth getting worked up about.

-Mars
posted by Mars Saxman at 8:36 AM on June 28, 2001


Smart Tags will be back on MSN the butterfly Network. Microsoft Never reduces code size. they issue patches.

When the MSN overtakes AOL for the vast majority of folks who believe that those are the internet, smart tags will be back in all its color.

They didn't say they were going to disable the My Images 'feature'. This feature just helps folks save a right mouse click to save images. Thanks Bill!!
posted by headlemur at 8:53 AM on June 28, 2001


The people who raised the stink to get Smart Tags removed in the first place were the ones who wouldn't have used the feature anyway, and hadn't used the feature when they began complaining about it.

Web designers hated the idea of having someone else dictate what their sites should look like, and solved the problem by dictating to their own visitors what software they should be able to use.

I think this turned into a war between the wrong armies.
posted by wenham at 9:20 AM on June 28, 2001


They didn't say they were going to disable the My Images 'feature'. This feature just helps folks save a right mouse click to save images. Thanks Bill!!

Which, of course, one certainly can't do now. Come on; having opt-out SmartTags puts the burden of work on the developer, but it's just plain foolishness to get your knickers in a bunch about the thought that someone might actually want to save content! To their hard drive! Off the Web!
posted by snarkout at 9:26 AM on June 28, 2001


Smart Tags are stupid.
posted by 120degrees at 10:23 AM on June 28, 2001


Web designers hated the idea of having someone else dictate what their sites should look like, and solved the problem by dictating to their own visitors what software they should be able to use.

I think this turned into a war between the wrong armies.


Wenham, you said much more eloquently what I've been trying to spit out all along. Thanks.
posted by anildash at 10:26 AM on June 28, 2001


Web designers hated the idea of having someone else dictate what their sites should look like...

Okay, why are you trying to turn this into a web designer thing? Designers are paid to make pages look a certain way. Why is that such a horrible, horrible thing?

I've never understood this whole "well, the user wants to see the entire world in Courier New with no graphics" bullshit argument. Since when is color, formatting, layout, typography, etc. so horrible? Why is it such a crime to format a page?

And now you've completely distracted me from the main point of the problem with Smart Tags, in that with Smart Tags, Microsoft would bend common English words to their own purposes for marketing.

And what boggles me about Microsoft defenders, most of whom skew to the "right" and consider themselves big capitalism = freedom people, is that they don't see that Microsoft is attempting, with their .Net vaporware and Windows XXX, to turn a decentralized, democratic medium into a centralized, oligarchical medium. Where the great criss-cross and chaos of the Web becomes nothing more than a constant surveillance to make sure you're not using a pirated version of Microsoft's software. Where you have to report to Microsoft.com anytime you want to use your own computer.

I'm ranting. It's fine. I'm considering switching to Linux.

Oh, and if you think Smart Tags are all neat and stuff, try envisioning Metafilter with every other word having a purple squiggly underline. Have fun!
posted by solistrato at 10:35 AM on June 28, 2001


For some reason, I keep thinking that this Smart Tags debacle is yet another reason why style and content need to be separated. Hear me out.

If you have indeed parsed things out to CSS and HTML, once someone visits your site they can feel free to change the appearance of the page to their liking. 24 point Verdana for body copy? Go ahead. 200% zoom in Opera? Sure! Custom CSS? Okay, have fun! But the content doesn't change.

Now, Smart Tags change the appearance - but arguably not as much as they muck with the content. I don't want my page to have a link unless I tell it to have a link. Do I care about the color of that link, and if it blinks in 10 pica Script? To a point, sure. But I also accept that users should have control over the content, in the end.

Smart Tags take the control over content and put it back in the hands of Microsoft. That's what I don't like about Smart Tags.
posted by hijinx at 10:45 AM on June 28, 2001


Designers are paid to make pages look a certain way. Why is that such a horrible, horrible thing?

And now you've completely distracted me from the main point of the problem with Smart Tags, in that with Smart Tags, Microsoft would bend common English words to their own purposes for marketing.

Visitors want to switch-off whatever gets in their way.

If Smart Tags grow to be nothing more than an injection of advertisements, then users will switch them off. If the decimation of the Banner Advertising market hasn't already made it blatantly obvious, the visitors seem to be very comfortable with their ability to willfully reject advertising no matter how hard it's pushed on them.

Don't fear that Smart Tags will "bend common English words to their own purposes for marketing" because as soon as that happens, Smart Tags will become useless. They will be switched off by the users themselves because they've become a distraction.

This is intricately related to the web design issue, because users tend to tollerate a designer's choice of presentation only when it works. As soon as the design becomes a distraction - because the type is too small, or the colors are wrong, or whatever - then the user will deal with the problem by either switching off the presentation, or leaving the site.

I think designers need to get to grips with the fact that if visitors prefer to read everything in "Courier New with no graphics" then perhaps their design skills are flawed, perhaps the money they were paid to make pages look a certain way was wasted.

Likewise, if visitors are going out of their way to switch on Smart Tags, even attempting to over-ride the META tag which is supposed to switch them off, then perhaps your visitors are trying to tell you something. And perhaps you should listen to them.
posted by wenham at 11:17 AM on June 28, 2001


BTW, I added the Meta tag to prevent smart tags from showing up on any MetaFilter page the other day.
posted by mathowie at 11:58 AM on June 28, 2001


Hyperlinks are not design elements, just as footnotes and references are not design elements. They are, within the terms of the medium, typographical elements, which is something different.

Most people here grok the Web enough to know the typographical nuances of the hyperlink: not just where you link, but what you tag. It's what astonished me about Mosaic in 1993; it's what underwrote Suck's arch know-all stance; it makes a difference.

You know that sinking feeling you get when you take a novel or a play out of the college library, to find that some infantile fuckwit has marked out the "key points" in yellow highlighter. It makes the text unreadable; it destroys its integrity. That's because the act of making marginalia (link shamelessly stolen from our factotum and tag-blocker) is personal and intimate, and just as important as the content: it's a kind of territorial marker, a trigger for memories.

There are ways of providing relevant content on a just-in-time basis: the MIT Remembrance Agent would make for a great browser add-on, because it works through its analysis of personal data that you voluntarily provide. But Smart Tags was the web equivalent of a literature student with a copy of Hamlet, Cliff's Notes and a highlighter pen.
posted by holgate at 1:09 PM on June 28, 2001


I look forward to other attempts to provide the user with content augmenting tools. Sometimes there aren't enough footnotes with the original, which is what has made "The Annoted XXXXX" books popular and necessary.

Take The Annotated Alice in Wonderland, for example, as a book which is now almost necessary because the Victorian cultural references and language of the original story are foreign to modern readers.

Or Asimov's Annotated Gilbert and Sullivan, for the same reason.

And where the cultural and language differences of the past have compelled annotation, the over-specialization and burgeoning vocabularies of modern professions make it necessary for material written today. I could really honestly use something that links a medical or engineering term to a glossary entry - especially when it never occurred to the original author to do it himself. I want the Smart Tag module that does this. I want that feature.

I've even been working on a glossary system for my own web site, trying to implement the mechanism on the server side, and realizing all along that the most elegant place to put it was on the client side instead.


But to soothe the paranoia of content producers, the difference between a library book which someone else has insensitively marked up and Microsoft's Smart Tags is a wide and significant difference that has already been discussed in other forums.



The difference is that the reader can switch Smart Tag annotations off.



In fact, one of the concessions Microsoft made before yanking the whole feature was to leave them switched off by default. And something I've said before is that your readers will switch them off if they become annoying and counter productive. The collapse of the advertising market is historical evidence that readers will do this.

Smart Tags can be customized by anyone with development skills if the API is published. A voluntarily downloadable module can solve the problem imposed by the physical limitations of print and the legal limitations of copyright: we don't have to wait for someone to obtain the rights and put up the expense of producing "The Annotated Alice in Wonderland" or "The Annotated MSN.Com" or "The Annotated MetaFilter".

The original remains uncontaminated and its copyright unviolated, but the reader has the freedom to apply third party annotations at her discretion. There is nothing illegal or inethical about that.

The visitor will have the option and the means to include and exclude whatever annotations she wants. If the visitor doesn't want to go Where Microsoft Wants Her To Go Today, then she won't. And Microsoft can't make her. Paranoid visions of every word having purple dotted underlines is both ill-informed and betrays a lack of faith.

I think the reaction to eliminate Smart Tags completely was stupid. But more importantly I think the protesters lost an opportunity, because Smart Tags are a double-edged sword.
posted by wenham at 1:49 PM on June 28, 2001


You know that sinking feeling you get when you take a novel or a play out of the college library, to find that some infantile fuckwit has marked out the "key points" in yellow highlighter. It makes the text unreadable; it destroys its integrity.

Hardly. In fact, if the annotation is done well, it can be a great asset to study. This is why, when you buy a used textbook for a college course, you search the available copies to find one that seems to have useful notes and highlighting already in it. Other examples are legion: annotated copies of Shakespeare that (gasp!) come from the publisher with someone else's notes in them; copies of the Bible with the Words Of Our LORD In Red.

The only reason it's bad to mark up a library book is that the book is used by others who might not appreciate your markings. But on the Web, each person gets their very own copy of your page. They can mark it up to their heart's content, and they can (if they want) use any automated tools they like to help them. Why do we insist on trying to limit the Web on what's good for books?
posted by kindall at 2:13 PM on June 28, 2001


Why do we insist on trying to limit the Web on what's good for books?

Because scribbles in the margins don't automatically link to stock quotes from a competing company.

Because text marked up with a highlighter pen doesn't automatically link to MSDN Search.

Because companies are used to paying thosuands, if not millions of dollars to create and manage an identity, controlling everything down to the typeface of their stationery. And they won't stand for anything defacing it. Ask any designer who's done identity and brand work for a high-paying client. They are simply not going to stand for the sudden appearance of purple lines that indicate a link to what could possibly be the competition.

Why do people try to reduce the Web to "it's just text", as if it's just scrap paper or a cocktail napkin, and think that invasively messing with the content is OK? We're not talking about something harmless like the Pornolizer or the Jivenator here, people.
posted by scottandrew at 3:16 PM on June 28, 2001


Ask any designer who's done identity and brand work for a high-paying client. They are simply not going to stand for the sudden appearance of purple lines that indicate a link to what could possibly be the competition.

Snarf! So "smart tags" are bad because they're bad for other corporations! Oh, this is rich!
posted by kindall at 3:23 PM on June 28, 2001


What if an author doesn't want his content annotated?

Then that author can simply refrain from publishing his content. That is the only choice he currently has; no author can prevent you from marking up your copy of his book. Nor should they be able to prevent you from marking up your copy of their Web page.
posted by kindall at 3:25 PM on June 28, 2001


Because companies are used to paying thosuands, if not millions of dollars to create and manage an identity, controlling everything down to the typeface of their stationery. And they won't stand for anything defacing it.


Then maybe the web is not for them.

The control of presentation and content on the web is in the hands of the reader, a fact of the technology. Just like the control of presentation and content in print is in the hands of the designer, a fact of that technology.

But if businesses, large and small, insist on having a presense on the web then they should get used to the transfer of control that comes with it.

They can't have their cake and eat it too. They have to compromise as well.

The users of the web have already been compromising by tollerating the design and messages of companies that think they know their readers best, when in fact they don't. Now we, at last, have the ability to put fine-grain control over the presentation and content in the hands of the user - where it was technically infeasable to do so before - why is everyone so surprised when users actually start exercising that?

(Or tool vendors try to empower them.)

While a business might be scared that Microsoft is going to insert links to the competition in their Internet back yard, they should be even more scared if it's a consumer advocacy group, instead.

I said that Smart Tags are a double edged sword and the reason is that it looks like there isn't anything stopping you, or Dave Winer, or anybody else from writing their own Smart Tags to annotate Microsoft's own web pages with their own tools.

If you're influential enough to raise a threatening protest, then you should also be influential enough to encourage your readers to download your Smart Tags.

If Microsoft tries to stop this by using a secret back door, then the fact will quickly become obvious and that is where our protest efforts should focus on.

Microsoft is famous for their Embrace and Extend tacticts. If you haven't already noticed, this method is extremely successful and their competitors' denial-based efforts (deny their right to bundle the browser, deny their efforts to produce an incompatible Java VM, deny their entry into streaming media and so-on) have been profound failures.

The opportunity to give Microsoft a dose of their own medicine has just been passed up.
posted by wenham at 3:43 PM on June 28, 2001


Kindall, if you as the reader wants to mark it up, that's fine. This is not the same.

Smart Tags are the equivalent of the printing company adding annotations to a novel because the book is printed on their press, so they have the right to make whatever additions they see fit. And if someone pays them to put their ad into every book they print, oh well. The author is just SOL.
posted by Dreama at 3:46 PM on June 28, 2001


What if an author doesn't want his content annotated? What if an author expects a user to actually use his own noodle and figure out the words and the references in a particular webpage on his own?

That's the author's problem.

If the reader choses to use a tool that solves a problem that the author tried to impose, then that's the reader's right. The author can either get with it, or go back to print.

This is the web. This is the reader's playground just as much as the author's. You better get used to it.


But since I control the content, I also control the method in which you find out

Forgive me for being absolutely blunt, but this is ignorant and nieve.

You only control the delivery of the content. You do not control anything else.

You do not control the horizontal. You do not control the vertical. You do not control the font, the color, the background or the mood music that accompanies it.

I, the reader, can over-ride all of these. It's a technological fact.

And you can't even reliably detect the presense of the tools I'll use to do this with. Because I can make my browser lie to your server or your Javascript detection scripts if you are stupid enough to try and go to war with me.

I am the reader, and at long last, I am in control.
posted by wenham at 3:52 PM on June 28, 2001


Hell's bells, I've decided I'm not finished yet.

Because scribbles in the margins don't automatically link to stock quotes from a competing company. .. Because text marked up with a highlighter pen doesn't automatically link to MSDN Search.

I think this has more to do with the fact that the medium of paper makes such things impossible than with any notion that it wouldn't be a good idea.

Sheesh. In the Web, we have a medium that pulverizes the limitations of paper. In fact, currently the limitations of the Web are so broad that when compared to paper, the Web's limitations might as well not even exist. And all we can think about is trying to put these limitations back in place!

I guess Richard Bach said it best... "Argue for your limitations, and sure enough, they're yours."

The last thing I would want is to see Microsoft taking the words "Francis Macomber" and linking to some Cliff's Notes precis of Hemingway. Because such a dumbed down bastardization of Macomber really wouldn't do the story justice.

Yeah, God forbid people should decide for themselves how they want to learn! Why, that'd be freedom, and we can't have that, can we?

Smart Tags are the equivalent of the printing company adding annotations to a novel because the book is printed on their press, so they have the right to make whatever additions they see fit.

The browser is the user's agent. In fact, that's what the HTTP header that identifies your type of browser is called: User-Agent. The browser acts at the behest of the user. Nobody thinks, if you use Microsoft Word's in-context spelling checker to fix a misspelled word, that Microsoft had anything to do with making that correction. You did it using the tools provided by Microsoft.

I can click one button and see a computer-generated summary of any Web page (a "Cliff's Notes" if you will). I can use a filter to remove advertising from Web pages I view. And if I want to rely on Microsoft to provide additional information on every page I view -- well, that seems pretty foolish to me, but if I want to do it, it should be my prerogative. it is, after all, my copy of the page being marked up.

Almost everyone here is a Web reader to a much greater extent than they're a Web publisher. I can't fathom any regular Web reader arguing against having the right to read Web pages exactly the way they want to.
posted by kindall at 4:02 PM on June 28, 2001


Smart Tags are the equivalent of the printing company adding annotations to a novel because the book is printed on their press,

I don't believe this is true at all.

If we were to try and stretch your metaphor to describe what's really happening, Smart Tags are more like a book store owner who gives you a template that you lay over the book.

In reality the metaphor doesn't fit comfortably. The browser is not really analogous to the printing press because the choice of browser is independent with the choice of the content - something which isn't true for print. A book is printed by whoever bought the rights to it.

Remember that Smat Tags were to be disabled by default, meaning that the reader must consciously switch them on.

This is a case of the reader marking it up. But in this particular implementation of the same concept, the reader just happens to be chosing the annotations someone else has already written.

ed feels that the reader can get the information elsewhere, but he's missing the whole point of why the Internet became popular in the first place: it makes access to information easier. And annotation tools such as Smart Tags are just attempts at improving the user interface.

If the authors - with their vast creative talent - were to write for the annotation tools as well, then Smart Tags and its cousins will become useful tools.

If those annotations are advertisements, then perhaps the reader genuinely wants them. Can't blame them, nor should you stop them. But I feel that if advertising is all that Smart Tags are ever used for, then visitors will just switch them off.

I am under no illusions whatsoever that readers will put up with a tool that just adds more commercial noise. Therefore, the tool can only possibly evolve towards positive value. If it tries to go in the other direction, it will be killed by the users themselves.
posted by wenham at 4:06 PM on June 28, 2001


Smart Tags as applied to an excerpt of Jerry Kindall's Who Am I page:

Ed, that's cute, but you're being juevenille.
posted by wenham at 4:09 PM on June 28, 2001


Hey, I laughed! Those links were very Suck-esque. If a version of Smart Tags were developed by the Sucksters, I think we'd all install that...
posted by kindall at 4:16 PM on June 28, 2001


it will be a cold day in hell before I see any of my work distorted by a corporation for their own twisted profitable means

Try this morally-equivalent version: "It will be a cold day in hell before I allow my spelling to be corrected by a corporation for their own twisted profit motives."

Looks pretty silly, doesn't it?

Microsoft does not add "smart tags" to Web pages. The user adds the tags by clicking a button in the preferences. To say otherwise would be like saying "Microsoft ignores Web style sheets" because you can turn off CSS support in Internet Explorer. No, the user ignores the style sheets.

it doesn't make any sense to look at a web page in any other way than the manner in which the author has deemed it to be seen. Why the hell do you think there's been such a hue and cry for web standards?

Web standards are increasingly about separating content from presentation, specifically so that the content can be displayed in different ways -- so these two sentences are actually at odds with each other.
posted by kindall at 4:30 PM on June 28, 2001


Sure it does, ed - maybe the author wants it seen in 9 point text, and I think that's too small. So I instruct the browser to enlarge the text, and whoops - out the window goes the designer's vision! Oh, so sorry, but at least now I can read the damned thing.

Microsoft would give you the one link ultimatum. It would destroy the sense of discovery that comes with the Internet

The entire point of hyperlinks is that you can set up a digital link between two related entities. There is a relationship between the words Victoria, B.C. and the web site your browser loads when you click on the link. That is why those words have been linked to that URL at all. That is the point of having a web in the first place - allowing us to link related things together.

Google and other search engines only exist because it's so hard to find things without them, and this is only true because web designers have, generally speaking, done a poor job of linking things together. Search engines are crap - they're stupid, they miss most of the web, and they give you all kinds of wrong answers. They're a poor replacement for hyperlinks, and the only reason we use them is that the current implementation of hyperlinking on the web is so much worse.

What Microsoft has proposed here is simply client-side hyperlinking. Instead of requiring the server side to specify linkages between things, the client makes its own guesses. In theory, the author ought to know better, and server-side links ought to be more accurate - but the evidence that this has not worked too well is right in your face every time you get lost and have to head for a search engine. I don't see how client-side linking can possibly make things any worse.

-Mars
posted by Mars Saxman at 4:43 PM on June 28, 2001


Would we even be having this conversation now if the Mozilla team had proposed the Smart Tags concept instead of Microsoft?
posted by darukaru at 5:02 PM on June 28, 2001


Actually, I just figured out my own question. Instead of 'OMG MICROSOFT IS TRYING TO CONQUER INTERNET', it would be 'More Mozilla bloat? Sheesh, ship it already!'
posted by darukaru at 5:05 PM on June 28, 2001


I said that Smart Tags are a double edged sword and the reason is that it looks like there isn't anything stopping you, or Dave Winer, or anybody else from writing their own Smart Tags to annotate Microsoft's own web pages with their own tools.

Except that Everybody Else's Smart Tags don't have access to Windows Update. And Windows Update, like Hailstorm and .NET, is where the power lies, just as the Windows desktop was once where the power lay. In fact, there's nothing to stop MS from re-imposing its own Tag responders as part of its update policy.

In theory, the author ought to know better, and server-side links ought to be more accurate - but the evidence that this has not worked too well is right in your face every time you get lost and have to head for a search engine. I don't see how client-side linking can possibly make things any worse.

There is no one-to-one relationship between a particular word or phrase and a hyperlink. There never will be. This is not AOL with its fucking keywords. There is no Platonic Ideal Link. And if it ever gets that way, I will unplug my modem.

Say you write an academic paper. You include footnotes for citations and references. Then the journal that publishes it produces an extra edition that sticks in its own citations, without prior consent. That kind of thing changes the nature of the piece.

Why the hell do writers love the web? Because it cuts out the notion of mediation: of editors who cut out paragraphs so that they can get an advertisement on the page; of printers who mess up the italicisation, of all that shit.

Just read some of the early Suck pieces and look at how they use links: creatively, ironically, knowingly.

Would we even be having this conversation now if the Mozilla team had proposed the Smart Tags concept instead of Microsoft?

Yes. Possibly with less vehemence, because Mozilla isn't engaging in a business model built around the provision of networked services. But yes. The power to make links rests with the author. Once marginalia leaves the margins, you might as well give up.

"When I use a word," Humpty Dumpty said in a rather a scornful tone, "it means just what I choose it to mean--neither more nor less.

"The question is," said Alice, "whether you can make words mean different things."

"The question is," said Humpty Dumpty, "which is to be master--that's all."
posted by holgate at 5:30 PM on June 28, 2001


And the user doesn't control the nature of the Smart Tags. Microsoft does.
The user does, however, wield the ultimate control over Smart Tags. He can turn them off (or on, since they would have defaulted to off). The page designer also wields this control. He can disable Smart Tags on his own pages.

spoon feed
subtle and diabolical
distort
vigilant and judicious
dictatorial pixel-based decisions
bland, corporate-approved, designed-for-the-complacent-yuppie links
Use loaded words much?

imagine that all of the content at Suck had been systematically replaced
Has there ever been any evidence that Smart Tags would modify previously existing hyperlinks? You're twisting the Suck example to serve your own purposes, namely demonizing Microsoft.
posted by darukaru at 7:29 PM on June 28, 2001


Try this morally-equivalent version: "It will be a cold day in hell before I allow my spelling to be corrected by a corporation for their own twisted profit motives."

Looks pretty silly, doesn't it?

Not if you're James Joyce and the book is Finnegan's Wake.
posted by davidgentle at 7:36 PM on June 28, 2001


It just occurred to me: isn't it bizarre that the IE group even contemplated this cack-handed, Office-esque implementation of Smart Tags, when another group of Microsoft employees, in association with Dreamworks, is running the astonishingly web-literate AI promo-game-websites?

I know that MS is a huge and disparate corporation, and that the two groups have about as much in common as Bill Gates' and Steve Ballmer's suit measurements, but the browser people really ought to study what's going on with that project if they want to know the real dynamics of the web as a diverse, distributed and hyperlinked medium, rather than a nasty melding of Powerpoint and MSN.
posted by holgate at 7:55 PM on June 28, 2001


There's one little thing that the vast majority of you who would defend smart tags seem to forget:

The users.

The vast majority of internet users wouldn't know a smart tag from a smart bomb, much less how to defuse either. They see a link, they click. They assume that links are put there by the authors of the site to further inform them.

If, in the middle of a vitriolic against M$, smart tags happen to tag the word "Microsoft" and point it to M$'s "innovation" page, it has vast potential to confuse the reader. Why would this guy link to a such a glowing description of Microsoft in the middle of this rant? Is he insane?

Annotations are best done in the context of the content in question. Smart tags can not have any sense of context. They will simply link hither and thither all willy nilly to bob knows what, which, given that most smart tags will be M$ supplied will probably be pro-MS propaganda despite the fact that my site may be a linux advocacy site. And if they link the word linux to redhat's site, I'll get even madder. I'm a slackware user.

Links are part of the content not part of the design.

One other thing the lot of you are forgetting is that if you wish to publish an annotated version of a book still protected under copyright, then you must acquire the permission of the copyright holder. If the copyright holder does not want you annotating his (or her) book, then you're SOL. Microsoft does not, now, nor will it ever, have permission to annotate my words. Period.

{jaded}
posted by jaded at 9:11 PM on June 28, 2001


The act of linking is as much as communicative skill as writing itself; sure, most writing on the web is sophomoric, but that is a failing of the author, and it could prove to be a reason why a user would choose once site over another -- I've preferred using Insound over other CD sites because of their linking to the AMG database. It's why people keep quoting Suck as an example because they excelled at this art. And I recall Zeldman making a similar point about his usage of humor in his anchors' title attributes.

When this editorial voice is interrupted and modified by a third party the way Smart Tags do, it disrupts the intended experience. Changing style sheets doesn't affect editorial voice, only visual presentation. In fact, all of the other browser tricks Microsoft or Netscape have put in to push users to partner sites -- such as custom toolbars/sidebars, location bar searches, and what not -- exist outside that magic box where the content is solely what came off that organization's servers. If Smart Tags were handled in a similar manner, that would be great. But they've crossed the line (or the box, to stick with the metaphor) by putting their content in our documents.

Has there ever been any evidence that Smart Tags would modify previously existing hyperlinks?

Has there ever been any evidence that they wouldn't? Let's face it, we're dealing with pre-release software that's still feature incomplete. The only thing we really know is that they're working on it. And when Microsoft is involved, there's plenty of reasons to question the integrity of this feature.
posted by teradome at 10:25 PM on June 28, 2001


Try this morally-equivalent version: "It will be a cold day in hell before I allow my spelling to be corrected by a corporation for their own twisted profit motives." Looks pretty silly, doesn't it?

Not if you're James Joyce and the book is Finnegan's Wake.

In the context of Word's spelling checker, what you just said doesn't make any sense at all. You're saying users of Word should object to the spelling checker if they're James Joyce? Or what?

Changing style sheets doesn't affect editorial voice, only visual presentation.

So if a user set up style sheets that set, say, all images not to display at all, that wouldn't affect your editorial voice? (The text-centrism of this view should be obvious.) What if users set up a style sheet to suppress the display of certain of your text (say, that text enclosed by the SMALL tag, which you are in the habit of using for ironic asides)? That wouldn't affect your editorial voice?

the user doesn't control the nature of the Smart Tags.

Existence isn't part of a thing's nature? Sheesh, what are they teaching in philosophy courses these days?

There's one little thing that the vast majority of you who would defend smart tags seem to forget: The users. The vast majority of internet users wouldn't know a smart tag from a smart bomb, ...

Guess you'd better make all your links blue underlined then, you might confuse the users! Amazing how designers can generally disdain usability advice, yet call upon it when it suits their rhetorical needs. (Not you specifically, perhaps, but a lot of Nielsen-bashers are probably gleefully using this very argument.) I think this is why Microsoft made the links so godawful ugly, no? Anyhow, if the user doesn't know what a smart tag is after explicitly turning on the feature, then what more can we do for them?

My point through all this is not that the Smart Tags feature is right or wrong, but that it's hardly as black and white as people seem to think. The vast majority of the objection seems to me to be related to Microsoft's involvement, notwithstanding a few brave souls who claim they'd object to the feature even if another company came up with it. They might, or they might not -- we'll never know, because it happens that Microsoft did.

It's not possible to have a book where every word is a link taking you to a dictionary definition. But if it were, do you think book publishers would link to any dictionary other than their own? And do you think the author of the book would be able to persuade Random House to use the American Heritage Dictionary (published by Houghton Mifflin) instead of Random House's own Webster's? Fat, fat chance, don't you think?

Of course the analogy gets blurry. There's nothing that's exactly like the Web. That's kind of the whole point, isn't it? One of the most interesting features of the Web is that it allows readers to participate. Some creators think the only way readers should be able to participate is by publishing their own Web sites. But the very concept is predicated on an old-media mindset and will eventually be pushed aside. The new medium does away with all the old limitations and we point to that fact gleefullly when it does away with restrictions we don't like, but when we like the old way of doing things, we argue vehemently for them, as if they were a moral imperative. What a sad bunch of revolutionaries we make!

We've already seen baby-steps in the blurring of the lines -- Netscape's "What's Related," Third Voice, user style sheets. There have been other attempts. Some of these have been more successful than others, and we will see more. "Smart Tags" won't stick, for now, but eventually they or something like them will arrive. (Microsoft will go back to the drawing board and come up with something less intrusive, you can be sure.) Yes, you will have a little less control as a publisher, but you will have much more control as a reader. And since everyone consumes more information than they produce, this will be a net benefit for the Web.
posted by kindall at 4:57 AM on June 29, 2001


The vast majority of internet users wouldn't know a smart tag from a smart bomb, much less how to defuse either.

How would you know?

There has not been sufficient time or exposure to really know how a representative member of your audience will react to Smart Tags.

The user interface is clearly different for Smart Tags than normal hyperlinks. Without usability testing, you don't know what your readers will mistake them for.
posted by wenham at 8:39 AM on June 29, 2001


Type in "Francis Macomber" into Google and it will be pretty clear that you can find what you want to the depth at which you want to

But a Smart Tag would destroy the process outlined above.

Good riddance.

It occurs to me that you're dilligently ignoring a point I made to you earlier, which is that Smart Tags are an attempt to evolve the user interface. The search engine is a user interface improvement on the library catalog system. Smart Tags are an attempt to improve the user interface of a search engine.

A false assumption that has killed avenues of information in the past is that the user's time is free - that it doesn't matter if it takes them hours to do all the supplementary research necessary to understand an essay or a user interface.

By suggesting that your reader must go to a library and look for Hemingway, or go to a search engine and look for "Francis Macomber", you're placing a burden on the reader that she may not care to take.

She can either solve that problem by leaving your site, having never understood what you were writing about, or use an annotation service to at least muddle through it.

If you don't care whether she muddles through it or not, then why should she care if you object to a third party's annotations? Clearly, the contempt will flow both ways.
posted by wenham at 8:57 AM on June 29, 2001


So if a user set up style sheets that set, say, all images not to display at all, that wouldn't affect your editorial voice?

No, not if you've used proper ALTs, TITLEs, and LONGDESCs on those images. That was the original concept behind HTML (that is, before the corporations stepped in and started mucking with it).

What if users set up a style sheet to suppress the display of certain of your text (say, that text enclosed by the SMALL tag, which you are in the habit of using for ironic asides)?

Well, now I think you're stretching your argument a bit too far. Who would do this? Besides, the user would actively be removing content, and would therefore be aware of the consequences.

In any case, you're stepping into the issue behind the style-vs-content battles -- The writing must speak the irony, not the tags -- if it doesn't, it's bad writing. As for the use of SMALL itself, tags like that don't carry any meaning in them, not like CODE or EM or CITE do; again, it's an authoring mistake. If you want italicized quotes, you use Q tags in the body and Q { font-style: italic; } in your CSS. The browser will, no matter what, know that the text you've tagged up is a quotation, and not just some styled text whose meaning is unknown.
posted by teradome at 9:08 AM on June 29, 2001


So if a user set up style sheets that set, say, all images not to display at all, that wouldn't affect your editorial voice?"

No, not if you've used proper ALTs, TITLEs, and LONGDESCs on those images.

An ALT attribute that says "Picture of the Mona Lisa" has the same editorial impact as the Mona Lisa itself, to be sure.
posted by kindall at 10:56 AM on June 29, 2001


Now that I've installed the IE 6 preview and tried Smart Tags for myself, I have an even better understanding of how misdirected the quantity of arguments against them are.

1) The feature is switched off by default. I have to go to the Advanced settings in Internet Options and find the "Enable Smart Tags" check box to turn them on.

2) Internet Explorer 6 does not include any Smart Tags at all.

None.

Zip.

Diddly.

You must explicitly seek out, download and install a set of Smart Tags provided by either Microsoft or a third party. Right now, the only ones that seem worth installing are the "everyday web" set from MSN.

3a) The Meta tag to prevent parsing of Smart Tags works. In MetaFilter, for example, you won't see any Smart Tags, even for words it's supposed to recognize.

3b) The Windows registry modification published by Dave Winer also works. I've made the change and now Smart Tags work on MetaFilter, Scripting News, and all other "Smart Tag hostile" web sites.

3c) I'm leaving the registry work-around switched on. That's my personal decision as a reader. It remains to be seen whether this is going to provoke any web sites into going to war against me.

4a) The interface for Smart Tags is wholly consistent across all web sites. Once the reader understands what's a Smart Tag and what isn't, there doesn't appear to be any danger whatsoever of mistaking them for conventional hyperlinks.

4b) The web page you must go to in order to download the "everyday web" set of Smart Tags has a large and prominent animated image of Smart Tags in use. A potential user can't help but see exactly what Smart Tags look like before even using them.

5) The "Smart Tags for Everyday Web" provided by MSN are a yawn fest. They recognize a few major companies, colleges and sports teams. The tags attached to them are nothing more than pointers to MSN properties and the respective company/organization's homepages. I don't consider them to be compelling content, and definately not worth my effort of downloading and installing them for. Frankly, I'm disappointed with how lame they are.

6) There are a few third parties providing Smart Tags that appear to be geared towards Office XP, not Internet Explorer. But then, some of them are very industry specific (insurance, for example). It would be promising if, say, a medical organization were to author a set of Smart Tags that highlight names of drugs and offer links to pages that describe what they do, their side effects, and what drug interaction dangers there are. I would download and install those in a heartbeat.

(I'm fairly sure that nobody will be bothered if every instance of "sildenafil citrate" was Smart Tagged and the option of links included a page describing the dangers of blue-green colorblindness sometimes induced by this drug)

7) Smart Tags do not tag any word which has already been hyperlinked conventionally. If you write Microsoft as a hyperlink, then Smart Tags will leave that Hyperlink alone.


Considering the lengths and the mulitple times I've had to explicitly, consciously, deliberately go out of my way to turn Smart Tags on (or even have any Smart Tags to turn on in the first place), I think that all paranoid visions of Microsoft imposed contamination of the web are completely bogus. Ed's lampoon of Kindall's web site isn't in danger of happening unless the reader consciously searches for, downloads and installs those Smart Tags - if anyone even bothers to author a set so ridiculous in the first place.

There cannot possibly be any threat from Microsoft to someone's copyright or integrity of content when the barriers to installation and the implementation of the User Interface make it undeniably the user's conscious choice to have Smart Tags in the first place.

This is not a problem with Microsoft marking up your web page without your permission. That idea is horse manure.

This is a case of Microsoft providing a tool, and the reader chosing to mark up their own copy of your web page in the privacy of their own home or office.
posted by wenham at 12:10 PM on June 29, 2001


To address your point, obviously you don't enjoy the fun of scavanger hunts.

By what logic do you conclude this?
posted by wenham at 1:20 PM on June 29, 2001


An ALT attribute that says "Picture of the Mona Lisa" has the same editorial impact as the Mona Lisa itself, to be sure.

Yes, a picture is worth a thousand words -- but since LONGDESC attributes are URLs, then you can put all those thousand words into an external file, which a browser for the blind could read to a user, or a regular browser could substitute inline as a link.

Is the painting itself compromised here? Of course, it's a painting. It needs to be an image. I could, of course, extend this argument further and say, since I cannot experience an artist's true color and brushstrokes on a 72 dpi Macintosh screen, that the integrity of any painting on the web is already compromised.

But that's besides the point -- we're arguing about accessibility here, and not Smart Tags, so why don't we let the rest of the folks talk about Microsoft (or take it into email if you want to go further)?
posted by teradome at 3:17 PM on June 29, 2001


A strong piece of writing involves reading what lies between the words and finding new meanings upon multiple reads.

I've certainly seen some computer documentation that was written according to this principle, but I wouldn't have called it "strong."
posted by kindall at 1:02 AM on June 30, 2001


It occurs to me that creation of Smart Tag datasets could itself become a new literary form. For example, I know quite a lot about CDMA. I have a CDMA FAQ on my web site. It includes a glossary.

I could also use my bounded expertise in this to create a set of smart tags which, for instance, looked for terms like "long code" and included a brief glossary entry for it plus links to relevant lengthier descriptions of them.

Equally, some of the web-loggers who simply collect links now could instead create Smart Tag datasets. If they encounter what they think is a good article about some subject, they'd add it to the dataset attached to appropriate keywords.

I can see a great deal of potential here. How about a non-sequiter dataset where the links have nothing whatever to do with the words they key? Using one would sort of be like an easter-egg hunt.

Well, enough speculation. I have some comments here, in general.

First, there are some ethical issues where discussion is pointless because practical matters have made them moot, and author-control over presentation of web pages is one of them. It doesn't matter whether authors or page designers want control because they don't have it and cannot ever get it. I can use a tool like Proxomitron (available for free) and override nearly anything about a web page. I was using it for a while to do exactly that, but it's a bit buggy. However, the concept is valid; it is (in Unixy terms) more or less like having AWK sitting in your HTTP stream, filtering the material going both ways. It is, for instance, quite capable of changing CSS (since it's just characters in a row). If I wanted MetaFilter to display with black letters on an orange background, I could have it. If I wanted to force a specific font onto a page which forces another, I can do it. If I want to suppress specific pictures, I can.

And no-one can prevent that. They can't even detect that it's been done.

From this discussion, it appears that the specific characteristics of Smart Tags is not the real question being discussed. What this is really about is control, and it is an example of content creators attempting to take models from other media and force the web into them. In the early days of television this happened; the creators tried to take what they knew about radio and make television act the same way. So for the first few years, television looked a lot like radio with pictures.

It was only later that people realized that television, for better or worse, was something completely different. Now TV and radio content bear no resemblance to each other.

The people who believe that the author must have control are trying to make the web into a magazine without paper. It doesn't work that way. The web is something completely new, and it's not going to fit into that mold.

The critical difference in the web is that it is truly interactive, and the reader does have control over the experience, both in gross and in detail. With a magazine the only choice the reader has is to read or not to read, but if she reads then she reads what was presented. That isn't how the web will work. It's been like that to some extent for the last couple of years, just as early television was radio-with-pictures for the first couple of years, but change is inevitable as the users become more sophisticated (which will inevitably happen).

On some level it's always been possible for users to alter the content which was presented to them, but those means have always been crude. You can use a black marker to blot out words or phrases. You can use a razor blade to eliminate sections of pages or entire pages. You can use the mute button or a trip to the bathroom to ignore sections of TV shows.

But recently technology has provided more sophisticated tools to the customers in, for instance, TV (e.g. TIVO) and when this has happened the consumers of that material have grasped it eagerly and used it heavily. TIVO, in particular, is being used to skip advertising. The only reason users haven't done much modification of material in the past is simply that the tools weren't there.

But this has led content creators to believe that the ability to control the content is some sort of God-given right, rather than simply being an accident of the media technology. Users have always wanted to have some control over the material.

With the web, they have it and they're not going to give it up. Content creators can try to exercise control, and if they are too heavy-handed about it then the readers will send them a message by avoiding the material. The ultimate consumer control over material is in ratings.

Here's an example: I use a browser plugin called iMarkup. I use it heavily. One version of this gives people the ability to export the markups they've created so that others can load the datasets and set them. This includes not merely the ability to put sticky notes of different shapes onto pages with embedded content, but also the ability to directly graphically modify the page. Using this tool, I can put stickies on a page (which can contain text), or add graphics and text, or directly annotate the text itself. (The annotations pop up on a mouse hover.) I can give these things to other people (I could, for instance, provide them as downloads on my own web page, or send them via email). And creators of web content can't even detect that this is happening. I don't happen to give my iMarkup changes to anyone else but in the 9 months or so I've been using it I've created several thousand of them.

People who are successful in a medium are the ones who embrace all aspects of the medium and try to express themselves within both its capabilities and its limitations. The best of them turn the limitations of a medium into advantages. The most important difference between the web and other media is the much greater extent to which the reader is in charge. In one sense this is a limitation; in another sense it is extremely liberating.

Trying to turn the clock back is futile.
posted by Steven Den Beste at 10:29 AM on June 30, 2001


I might mention that I don't use iMarkup routinely to paint moustaches on people; I was just using that as an example of what the tool can do. In fact, I don't use its graphical capabilities at all.
posted by Steven Den Beste at 10:36 AM on June 30, 2001


There's a difference in the control ideology of corporate websites that don't provide external links (of which I've had to build far too many) and personal websites that encourage comments, citation, in-jokes, textual promiscuity. As I said earlier, the feedback loop between the A.I. game-site designers and the groups responding to the game offers a reminder of the true possiblities of the medium.

(As with many web technologies, those content providers who've exerted the greatest control over distribution and reception are the porn sites.)

I'm waiting for a just-in-time information service that empowers the reader. I don't think Smart Tags does it, at least in its current incarnation. It's clumsy. It doesn't yet grok the web's complexity as a writerly text. If it did, I wouldn't be criticising it.

Now, if there were a way for readers to generate Smart Tags that took account of their personal interests -- for instance, by running Bayesian searches on pages, that compared them with your email archives or bookmarks, and flagged them accordingly, then that'd be a step forward. (Autonomy's Kenjin -- no longer supported, but downloadable from here -- is actually a basic example of this, and it doesn't impose itself on the page itself.)

For the moment, Smart Tags are an office-application tool yoked into service for Microsoft's dream of an AOLish, keyworded, corporatised web. But there's still plenty of time to corrupt that ideal, just as Marc Andreessen's IMG tag horrifed the SGML purists and transformed the web.
posted by holgate at 11:52 AM on June 30, 2001


« Older Aw, what a cute kitty...   |   This page Newer »


This thread has been archived and is closed to new comments