Flutterby™! : AutoLink

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics


2005-03-03 16:53:15.182443+00 by Dan Lyke 2 comments

A lot of people have been up in arms over the "AutoLink" feature in Google's toolbar. Scoble calls it "evil", while on the other side Dori compares it to printing web pages. Here's my take:

Demanding that a web page be viewed only in a certain way is akin to those asshole movie folks who create DVDs that can't be fast forwarded through. The whole point of HTML was (back in the day) that a given local system could use the hints provided in the markup to render it appropriately to the end-user, whether that was in a text-only mode (many of you probably don't remember text-only modes), or via a voice synthesizer, or whatever.

I have issues with people who take web sites and republish their contents wholesale, changing things like link targets (often to point to "herbal v!@gR@" sites), but if you, in the privacy of your own home, want to do this to my web page, more power too you. Furthermore, if you want to sell people a tool to make their web pages more useful, more power too you.

Just as I actually support those folks who make that DVD player that can take external edit decision lists for a given CD and play an edited version of a movie. If you're not republishing it or calling it your own, you're free to make any re-interpretation of something you wish.

[ related topics: Intellectual Property Privacy Weblogs Movies ]

comments in ascending chronological order (reverse):

#Comment Well, yeah, but there are nuances made: 2005-03-03 20:53:01.266063+00 by: baylink

We tend to assume, when saying "yeah, maybe this might not be so bad", that we're in control of it, which is likely only true for people who are a) geeky b) libertarians. Lots of people have *others* geek their computers, from little old ladies with grandkids, to office workers with IT staffs.

And don't forget webapps.

If I had to support a big web app and your plugin made my job harder (and couldn't be configured to avoid my site, preferably by seeing a meta header or something), I'd have the customer rip it out.

On other topics, RT is nice, but slow (probably mod_perl problems) and LID is still LID.

#Comment Re: made: 2005-03-04 14:05:25.262753+00 by: ebradway

WikiWeb! I guess I missed the SmartTags debate a few years back... But isn't AutoLinking kind of like building a "WikiWeb"? A WWW where every linkable word or phrase is linked?

Of course, the issue has to do with maintaining content purity and presentation. What if AutoLink replaces the links you put in your page?!? How can you tell an AutoLink from a link placed by the page editor?

Personally, I think there are some MAJOR benefits to be had from AutoLinking with the caveat that purity and presentation be preserved. First, I spend alot of time cutting and pasting key ideas into Google to find more info and even more lately, I've been using WikiPedia. I'd love a tool that automated that task, as long as I could fine tune it so it links in the direction I need. For instance, if I'm reading a paper on genetics dealing with isomorphisms, I don't want it linked to the math term isomorphisms. But I have that sort of problem already with spelling and grammar checkers. And the fact that I had to go to the second page of links in Google to find a that reference says that Google would be off the mark!

Second, I find it quite frustrating to be engrossed in a web page, click a link, and end up at a 404 message because the linked page has moved or has been removed. Google's cached pages are a great source for dead links. It would be great if dead links were detected and I was sent to the cached link (with appropriate messages, like the cached pages are currently presented).

But the real issue underlying all of this is the matter of who controls it. This kind of meta-search that must be executed on your web page pretty much has to be done using a tool like Google's search engine. The tool would have to have fast access to the search engine and have enough data on my browsing habits to know that I don't want the first link to "isomorphisms". That's alot to turn over to any company, even as benign as Google.