This is part of The Pile, a partial archive of some open source mailing lists and newsgroups.
From: peter.boersma@_REMOVE_THIS_satama.com (Peter Boersma) Newsgroups: comp.human-factors,alt.hypertext,comp.infosystems.www.authoring.site-design Subject: Re: Theory: "information density" Date: Mon, 22 Jan 2001 12:54:02 GMT On Sun, 21 Jan 2001 04:19:01 -0600, jorn@mcs.com (Jorn Barger) wrote: >My main web-design page-- http://www.robotwisdom.com/web/ --currently >uses the expression 'content-centered' as the umbrella-concept under >which all my design ideas try to find shelter... > >But "content-centered" _sounds_ completely tautological, so I'd like to >find a replacement that's a little more self-explanatory. And one idea >might be to build on the idea of 'information density'... [snip onion-model description] >This is what I mean by content-centered. Any page that offers content >(ie, all pages, hopefully) should try to _optimise_ that content, which >normally means raising the information density (and info quality, too, I >guess). Hi-density/hi-quality information includes: > [snip list] >A hi-density page should include all of these, arranged clearly and >compactly. Instead, if you analysed the info-density of the average >page, you'd probably get something like: [snip percentages] A series of similar indexes, a tool to measure them, and a large number of examples can be found at: The Rating Game http://stein.cshl.org/~lstein/rater/ with an explanantion (from 1997!) at: Sifting The Wheat From The Chaff http://www.webtechniques.com/archives/1997/05/webm/ === From: jorn@mcs.com (Jorn Barger) Newsgroups: comp.human-factors,alt.hypertext,comp.infosystems.www.authoring.site-design Subject: Re: Theory: "information density" Date: Mon, 22 Jan 2001 09:16:00 -0600 Jerry Muelver <jerry@hytext.com> wrote: > robotwisdom only scored 94.6. Looks > like Jorn's got his work cut out for him to get that > information density up to snuff. the page i rated is my joyce overview: http://www.robotwisdom.com/jaj/ and it's my showpiece of applied info-density theory, so i recommend you look at it before sneering. (yeesh) === From: Stylewriter <scambled@eggs.net> Newsgroups: comp.human-factors,alt.hypertext,comp.infosystems.www.authoring.site-design Subject: Re: Theory: "information density" Date: Mon, 22 Jan 2001 23:49:12 GMT I don't follow any of the meandering, off topic, replies to date, John. So I'll just barge straight in and be as clear as your goodself. In article <1enkfx1.1sv1pjr1eg6ksgN%jorn@mcs.com>, jorn@mcs.com says... ____> My main web-design page-- http://www.robotwisdom.com/web/ --currently ____> uses the expression 'content-centered' as the umbrella-concept under ____> which all my design ideas try to find shelter... ____> ____> But "content-centered" _sounds_ completely tautological, so I'd like to ____> find a replacement that's a little more self-explanatory. And one idea ____> might be to build on the idea of 'information density'... It seems clear to me that the 'umbrella-concept ' under which your site operates is minimalistic and structureless information'. ____> ____> Nothing is more frustrating for me as a web-surfer than being forced to ____> step thru a 'hierarchical' table of contents: several pages in a row, ____> each of which offers just its own very-short table of contents for the ____> next layer down. I've called this the 'stripped grapes' model because ____> it offers just a skeleton with all food-value removed, when it could ____> (and should) be redesigned to enrich the information density .... Absolutely agree. Three licks shold get any visitor to any element of content they seek. ____> .....as much as ____> possible-- This is where you start to lose it .... ____> by collapsing all the layers into a single page, including a ____> summary of the content of all the info-pages, and 'promoting' samples of ____> the best content from each page to the top layer. ____> ____> This is what I mean by content-centered. Any page that offers content ____> (ie, all pages, hopefully) should try to _optimise_ that content, which ____> normally means raising the information density (and info quality, too, I ____> guess). The problem with your style is that you are not optimising anything. What you are doing to stripping the content of all structure and focus. Human communication is not the same as computer or robot communication - as much as you may wish that it were. Human communication depends on shape, form, focus, hierarchy and digestibility. Your theory and site lack of these elements. It may be efficient in communciations information quality down a telephone line - but it is not efficient as communicating information to a human being. ____> Hi-density/hi-quality information includes: ____> ____> - clear summaries ____> - simple lists and tables and timelines ____> - images that really illustrate ____> - vivid pullquotes ____> - links that are carefully chosen and clearly described ____> - forms you can use to get quick answers ____> ____> A hi-density page should include all of these, arranged clearly and ____> compactly. Your site lacks summaries, simplicity. There are NO images despite your definition. A link such as 'service like Atomz.com ' cannot be described as clearly described. The problem with your approach is that it is as inappropriate and pointless as the graphic artist gone mad approach. One end of the spectrum is obsessed with everything but the content - while the other end is obsesse with everything but the presentation. Both approaches are doomed to dismal failure in the face of human communication dynamics which demand a balanced, stuctures and focussed approach for any kind of success to result. Few visitors to a site are so single minded on obtaining the information on 'that' site that they want to do nothing other than read 5,000 information-packed words on the relevant topic. The vast majority of site visitors are seeking bite sized chunks of information that may or may not be present in the site being visited. They do not wish to read through 5,000 words to come to the conclusion that the chunk they seek is not after all there. And most of the rest of the visitors are not seeking any specific piece of information, per se. They simply wish to get a 'flavour' of the content. In the realms of the commercial site, your approach is a suicidal 'trip' for any prospective business. In the realms of the reference site, your approach is guarenteed, imho, to bore the pants of any inquirer. Regards, Stylewriter === From: joshsegall@my-deja.com Newsgroups: comp.human-factors,alt.hypertext,comp.infosystems.www.authoring.site-design Subject: Re: Theory: "information density" Date: Tue, 23 Jan 2001 00:42:10 GMT In article <1enkfx1.1sv1pjr1eg6ksgN%jorn@mcs.com>, jorn@mcs.com (Jorn Barger) wrote: > My main web-design page-- http://www.robotwisdom.com/web/ --currently > uses the expression 'content-centered' as the umbrella-concept under > which all my design ideas try to find shelter... > > But "content-centered" _sounds_ completely tautological, so I'd like to > find a replacement that's a little more self- explanatory. And one idea > might be to build on the idea of 'information density'... > [snip snip] > This is what I mean by content-centered. Any page that offers content > (ie, all pages, hopefully) should try to _optimise_ that content, which > normally means raising the information density (and info quality, too, I > guess). Hi-density/hi-quality information includes: > > - clear summaries > - simple lists and tables and timelines > - images that really illustrate > - vivid pullquotes > - links that are carefully chosen and clearly described > - forms you can use to get quick answers > > A hi-density page should include all of these, arranged clearly and > compactly. Instead, if you analysed the info- density of the average > page, you'd probably get something like: > > 10k of info graphics > 30k of navigation graphics > 40k of decorative graphics > > 1k of content-related links > 10k of navigation links > 40k of 'filler' links (eg motivated by e- commerce, not info-content) > > 5k of content text > > So you're waiting for 136k to load to get 16k of info. > > (And rendering-time also has to be factored into the info-density > calculation: every added layer of tables (especially) further dilutes > your score. Also breaking up text over multiple pages.) > [another snip] All of this smack's of Jakob Nielsen's ideas on usability on the 'net (and his book "Designing Web Usability). I have some disagreements with his theories as well, but he does balance "information density" with usability. I mean, what if I removed all the paragraph breaks, graphics, and white space from a document? The whole page would be one clutter of text: tons of information, but hardly usable. As opposed to reducing the file size (which Nielsen also does, across the board), there's also the "screen" size. Perhaps no more than 10% of the screen should be devoted to navigation, for instance. Of course, a 640x120 navigation bar is only 10% of a 1024x768 screen, but is over 25% of 640x480. So good luck there... I highly recommend reading up on usability as a whole, as opposed to narrowly defining success in terms of "information density." -Josh Segall ===