|
Linking 1-2-3
December 7, 1998
If you've been following Today's Links for the past month you've seen the results of my work to compile and select links on several different topics about the Web.
It's an interesting process to track down links using several different methods and to see how it fits in with how sites are built and content published.
Fishing Holes
I don't have a news ticker that pushes me all the content that makes up Today's Links. So it's a mix of luck, experience and finding some new toys… a lot like fishing.
There are a few regular spots where I usually start off at to get a feel for what stories are developing. Two of those regular spots are News.Com and ZDNet News. Both sites provide a unique service to let their readers know about stories that are still being finished off before the complete story is published.
In most cases, I won't even link in the first story I see on a subject until I see what coverage other sites are providing. Industry Standard has an interesting exercise with Media Grok where they take a few stories and look at what kind of coverage different sites devoted and how they approached the story.
On a more limited basis, you can also find larger sites playing the linking game with News.Com, Wired, ZDNet, and other sites all devoting some space for links to stories from other sites. They seem to understand the value as well.
Lists of lists
Mailing lists. The reason I'm swamped with hundreds of e-mails a week. But there are some lists where I usually read to find an occasional interesting link or story. You'll be deleting a lot of messages on a high traffic list, but if there are some well-connected people subscribed then it should be worth the time.
One alternative is A List Apart that is unique with a moderated digest of new threads and responses on everything to do with the Web. Every few days there will be a digest of selected new topics and responses to previous topics. It's not real time, but that can be a good thing.
Also in e-mail, I subscribe to newsletters and a few sites to get notified by e-mail when there's new content available. Though, I tend to visit sites at routine times to get new stories linked in to Today's Links before I even get a notification message from the site.
How Do I Link This?
Some sites just like to make it hard to link to an article with content being posted to generic pages that will surely be replaced in a few days.
Most sites generate content and direct it into a page with a unique name with an index somewhere else that lists the stories. But there are several sites where content is featured on generic pages that are constantly replaced with new content. So they'll have a todaystories.html file that is constantly updated with the newest content and it's hit and miss if you can find an archive when the content is replaced.
There are two sites that are at least making some attempts to accommodate linking that will continue to work a week later. First, AdAge provides content on generic pages but at the time of publication they also make the archives available. And to go the extra mile, they split up the content into individual stories even though it is presented as a single compilation of stories on the generic page.
Another beautiful example is Peter Merholz's site peterme.com where he posts pieces on the front page but dutifully includes a link to the permanent residence at the same time.
I talked about getting notified when sites update new content by e-mail, but I also subscribe to sites with Internet Explorer 4.0 to tell me when a page is updated. Well with sites starting to drop static pages, IE4's method for subscribing to a site of checking time stamps on an HTML file seems to be a fading solution.
Microcontent
I usually link in stories with the title as the only hypertext anchor along with the publication name up front. So I can understand what Jakob Nielsen was talking about with his Alertbox article on microcontent.
In certain situations, I know that if I just post the title as a hyperlink that readers will have a small mystery on what the story is about. So I'll include some additional details after the hyperlink: author's name, brief description or comments.
Even knowing that I selected the links might be of some comfort to readers, but it's something that I want to keep an eye on when I link in to Today's Links.
JIT-SE
A good acronym, just in time search engines (JIT-SE) which was coined by Dave Winer are search engines that search selected sites and create an index on a timely basis. Two JIT-SEs are offshoots of regular search engines with NewsTracker from Excite and NewBot over at HotWired's HotBot.
NewsTracker beats NewBot hands down with the number of sites it compiles for its JIT-SE and the different topics covered. But like regular search engines, JIT-SEs aren't perfect. On NewsTracker, when I do a search for WebTV, I'll always see pages from CBS Sportsline. Why? Well there's some JavaScript on all the CBS Sportsline pages that mentions WebTV.
It isn't a new problem to JIT-SEs or regular search engines, because you can see NewsTracker's problems documented in the DaveNet on JIT-SE from July 1997. The problem is only magnified when you jump from the limited sources that a JIT-SE looks after to the regular search engines ploughing through the Web.
Something that I've been playing with recently are compilation sites that flirt with the role of a JIT-SE. Two that I know of are NewsHub and Newslinx. Newslinx, which was purchased in November by Mecklermedia's Internet.Com, seems to be compiled by hand. While you can see NewsHub bring in new headlines every 15 minutes.
Like JIT-SEs, it's an important consideration to see what publications these compilation sites are covering. So you don't miss out on stories not being indexed by these sites.
That's all?
I hope you start to use some of the tools discussed like NewsTracker, because it really does solve some
problems as a JIT-SE. I have it hooked into IE4 so in my address bar I just type "nt webtv"
and my search gets passed to NewsTracker and I see the 1000 hits from CBS Sportsline.
While discussion on how improve search engines and JIT-SE continues, maybe you can learn about
microcontent and consider how you can improve a site your working on.
Off to find some more links for Today's Links!
Later,
Lawrence
|
