Gentlemen: On October 31, you wrote:
I chose to differ with the efficiency of Mr. Larsen's paper. If one were to have a log with but a few entries, it is reasonable to assume that his philosophy might be effective. However, when a log files contains millions of lines, the necessity of special handling becomes self-evident.By the way, there is a useful article which has just appeared at http://www.klarisma.dk/articles/aindex.asp about adding fake parameters to requests in order to pass information to yourself. For example, the author recommends adding parameters to all href's so that you can quickly see which link on a page was clicked. If the page is a static page, the server will ignore the parameters, but they will still get written to the logfile.
Does this information not already exist in a log file? I have two fields in my log file, the requested page and the referring page. If I understand Mr. Larsen correctly, he's adding additional information to the referring field to tell me something that is already there. For example, he suggests that parameters be added to a link to indicate where the link came from. I already get this information. I have only one link on a page to another page. It might make some difference if I were to have links in five places on the same page and you want to track which link is the most effective. However, this skewes the accuracy of the study, since repetition is the dominant factor at work here, not the effectiveness of the link.
By the vary nature of a web-site, it is organized in much the same way as a corner variety store is organized. The milk is always put in a location the farthest from the door to force people to go by all of the less-needed impulse-buy items. If you know that people are coming to your site to access a specific resource, you put into place "distractions" that will "sell" your site or your products while people naviate their way to the "milk."
My site is organized in a hierarchy from the top down to as much as 15 levels deep. It is a simple matter for me to determine where a page came from by the referrer field. I might have 10 different places on a page where that link could have come from but then, my site is a major exception to the norm. My objective is not repetition but relevance. My site is organized in the same way that a dictionary or encyclopedia is organized, with a predictable hierarchy. Entertainment sites or marketing sites will have a different hierarchy where the objective is to "enhance the experience of the visitor" to entice them to stay longer. My site is the opposite, I want readers to get the information they want, quickly and comfortably and be able to move around the site to other information and references quickly.
I can't help thinking that I'm missing something in Mr. Larsen's proposal. However, that he has presented his position is laudable. I present this, not in criticism of his position but to extract that which I might have mis-understood or not read. I look forward to comments that might clarify this.
Ron Woodall
---------------------------------------
Ron Woodall
[EMAIL PROTECTED]
The Compendium of HTML Elements
"your essential web publishing resource"
- available at/disponible à:
http://au.htmlcompendium.org/index.htm (Australia)
http://www.htmlcompendium.org/index.htm (Europe and North America)
+------------------------------------------------------------------------
| This is the analog-help mailing list. To unsubscribe from this
| mailing list, go to
| http://lists.isite.net/listgate/analog-help/unsubscribe.html
|
| List archives are available at
| http://www.mail-archive.com/analog-help@;lists.isite.net/
| http://lists.isite.net/listgate/analog-help/archives/
| http://www.tallylist.com/archives/index.cfm/mlist.7
+------------------------------------------------------------------------
