09/18/02 The WORD [appropriate use of the "C" word]: Until searchers are given the power of coherent logical Internet infrastructure, they're going to have to adopt some crude, sloppy, "gorilla methods" in the war against the evil menace--the chaos pimps. Last night a combination of factors got me thinking about words, synonyms, and the capacity at most search engines to exclude content that contains a certain "word". I started speculating that maybe most Spam content reveals its true colors by the use of some common word, or words that could be excluded from search results. So I tried an experiment.. I started with the phrase "search industry" and added the word "Internet" this combination returned 19,400 results. As "search engine optimization" is the vernacular of the snake oil sellers relevant to this subject I thought excluding [-]SEO would take a good chunk out of these 19,400 results. Surprisingly the search returned 19,200 results --I still can't believe only 200 of the references contain the letters SEO... But moving on, here is a list of the next six words I tried and the results returned: directory-- 16,800 Marketing-- 10,600 [added] 2002 -- 9,550 Consulting-- 748 recruitment-- 701 solutions-- 642 The word "consulting" was clearly the big winner in this "bad word" contest and if you combine it with the second place finisher "marketing" you pretty much have the two word synonym for "Buy some worthless BS from me." These Spam Cutter words are I think justifiable weapons of war as they do kill a lot of bad guys with little collateral damage. My first SpamScraper bookmarklet. [does a search on selected text subtracting sites that include the words marketing, consulting, optimisation, services, promotions.-- Improved models coming soon] Seems to me there is lots of room for research and cottage industry innovation in this whole area of Spam-scraping using "exclude these words" bombs. The really neat thing is if google ever indexes this page I will be a victim of my own friendly fire.

09/17/02 ICANN-ot Understand: This article implies that the government intends a status quo extension of Icann's mission and authority. At the very least the government should make no long-term commitment until it receives the $800,000 opinion regarding DNS reform it purchased from the National Academy of Sciences (which I keep bringing up with cynical regularity) RSD [Regularly Scheduled Disappointment]: This will probably be the most used tagline on this Blog. I received a reply from Tim BL. Viewed from someone else's perspective it likely would be seen as a positive, helpful reply-- but from my "tired of all this talk and no action" perspective-- I was probably going to be disappointed with any reply that did not include the statement "I understand that your better idea is an entirely sound and harmless extension of existing infrastructure and I will see to it that it is implemented asap" Understandably he did not have time for detail, and this sentence sums up his "first impression": "the idea of defining some interesting common terms is useful, but the centralized database is not going to work. People hate central control and insisting on more registration they will not like at all" This "central control" rhetoric is a little too close to a "cheap shot" for me to feel obligated to take it seriously so I will reply with inkind simplicity. I don't see any mobs in the streets burning phone books or library card catalogs. "People" don't hate the efficiencies of centralization-- liars' and cheaters hate the organization and standards that centralization brings because it puts them out of business. What most regular, "real people" using the Internet really hate is the Spam e-mail and Spam Web pages a profiteering and unaccountable "navigation and infrastructure industry" has been allowed to impose on us all in the cover of chaos. In truth I don't think the idea of living in a world with a "whatis" database scares' Tim BL. Understandably his own vision of the road to a better future "The Semantic Web" is his priority and I am, and should be, grateful for the opinion he offered regarding my alternative. Some opposing perspectives regarding The Semantic Web that are almost as incomprehensible as the idea itself: Catching the Dreams Semantic What?

09/16/02X marks the spot [a spot of catch-up]: Did a little looking at the least of the worst in the standalone (independent xGo/xODP) web directory category. Being kind, let's remember that it could take sailing ships years to map a few hundred miles of coastline. Being realistic, the people working these directories should start deciding who should get eaten first when things get rough on the lifeboat. I've advised most directory starters, that they are likely wasting their time if they cannot secure access to a rich "site submission" stream. The fact that the ODP so grotesquely, squanders that valuable asset should be reason for every Internet lover to thoroughly resent Aol's management and google's support of the ODP. The other problem, is that most of the control freaks managing these directories make Captain Bligh seem like a fair guy. In the past, I have made an effort to make helpful contributions, as I would like ODP to face some real competition--unfortunately you have to blow too much smiley-face-smoke to make even the simplest, most logical changes. "Just for the fun of it" I made a modified version of my "select and search" Bookmarklet script for the three directories. I'd suggest you not select more than one word as content at these directories doesn't run very deep. Select some text and: SearchGoGuides SearchJoeAnt SearchXODP

09/15/02Lost Horizon [stop looking in your wallet]: From one of the replies to the National Academy of Sciences' request for public comment regarding DNS reform: "The National Academy of Sciences estimates that 80 percent of the information on the Web has a spatial component, and yet we have very limited ability to search for information based upon location. Such mature capabilities would have enormous benefits in the area of location-based services, opening new global and local opportunities for science, commerce, education, and enterprises, particularly in the developing world." The rest of Martin Reddy's statement proposes a solution that I think is unnecessarily complicated-- but the fact that somebody else is talking about this problem is a hopeful sign. As part of my "whatis" proposal web sites (or content pages) that have a regionally relevant component would, as a matter of routine, have that information made a part of the descriptive keywords ( longitude/latitude, postal code, city, etc). that would represent the site/contents database identity. In a "Whatis" world, "where" something is is a part of "what" something is. It is a curious fact that GPS technology in mobile applications has become commonplace. People are driving around right now with the aid of electronic "smart maps" built in to their dashboards that tell them when and which way to turn to get to almost anyone's driveway. To the Internet's embarrassment, it doesn't seem to know that anything smaller than a state even exists. It's like a super computer not being able to add 2+2 and represents a level of dysfunction (or unapplied functionality) that is awfully hard to just accept as acceptable.

09/14/02Always thinking [wishfully]: I've been thinking about the problem of google's "you link, you vote" algorithm. As yesterday's Blog entry demonstrates sometimes a link isn't intended as a complementary expression. None of yesterday's links deserve anyone's vote, and they shouldn't be getting an increase in rank merely because I do my readers the courtesy of providing an active link (to examples of some junk). One solution I suppose would be for me to disguise "bad links" with some JavaScript, but I really don't like having to make my pages unnecessarily script dependent or complicated. I'm thinking this is a problem of google’s creation and they should provide the solution. Perhaps a simple page tag or link property <google=novote> that would enable content providers to offer links to competitive or even hated content, without paying the price of -- in very real terms -- damaging their own relative page rank. How about a novote.txt file like robots.txt telling google what pages not to count votes on? A thing I'm not sure about is whether google bot recognizes web addresses in plain text. To be on the safe side I've decided to start spelling out the Dot in dotcom. He likes it [yes it happens]: I really do like Bookmarklets. The control they offer over how content is displayed, used, and in general processed is easily acquired and substantial. I think the fact that thay make it possible to relatively easily refine and customize off-the-shelf web browser function can only have a positive effect on innovation and productivity. Now for the half-empty stuff: Why are there so few good programmers (or scripters) actively producing better bookmarklets? The best known Internet reference on the subject of bookmarklets (bookmarkletsDotcom) hasn't done any substantial updating in over two years. Fact is, the standardizing of the "dom" and the absence of a real incompatible browser war, has provided opportunity for improvement. The search bookmarklets on the left menu [click here if you don't see it] for example work on most framed pages (as long as all frames contain pages from the same domain) this makes them a big improvement over what is offered on most of the high traffic, but basically archived, bookmarklet sites. Go ahead and try it (as a link), select some text and just click one of the search links- results provided in a new window. The "x" stands for "exact phrase", and the "++" means "all the words, anyorder". As I see it higher screen resolutions have provided a lot of "white space" at many websites, and I think more sites should be using that's space to provide visitors the power and convenience of "on site/on page" bookmarklet tools. Even though I'm not a very good programmer, as time permits I will be adding more bookmarklets (unique & new variations) they are free for your own use-- just bookmark the links and apply the favorite to any (compatible) page you're viewing If you'd like to provide them as links on your site (just copy the link address) a thank-you link to this site would be appreciated. P.S.-- If anyone knows a practical work around for the "one domain only" frames security BS limitation, please let me know, what you know.

09/13/02Personal Note: I had a *#&%fn tooth pulled today. Funny how when they are out of your head they look so strong, healthy and harmless. GooFle: I did a google/Fast search on the phrase "Internet navigation". Google returned 28,000 results and did so much faster than "fast" (alltheweb). Unfortunately faster junk isn't better junk. Both engines return the same top three results. The #1 "Internet navigation" reference is: This really content thin Rice University "can you big boys and girls say search engine" baby page. It's hard to imagine how out of 28,000 references this little pile of dated, stagnant Internet dust could have the highest "page rank." Ironically or perhaps suspiciously the site doesn't even provide a link to Fast. Why would the Fast algorithm give a search engine links page such a high ranking when it doesn't even link to its site? Answer: Because the fast algorithm has only one qualifier "copy google". The #2 "Internet navigation" reference: yesenglishonline.com -- "Linking you to a wealth of free English-language resources around the world around the clock." I'm not sure but I don't think that sentence is even good English --maybe free is a little too inexpensive. I am sure that there's no readnable rxcuse for this sites high rank on this phrase. #3: internav.com -- The ad at the bottom of the page says all that need be said "Received 10,000 guaranteed visitors for only 29.95" -- sadly it might be worth paying, as they obviously know how to play Foogle. Googles fourth pic was: A really super cool site called Galaxy that provides like 20 Super links to really neat stuff like AOLdotcom. Fasts fourth: DoTheWebGuide.com another American Cheesy directory site. From this great site you can get to [every other word in the English language]Webguide.com, even SUPERwebguide.com. And of course like all great web navigation web sites there is an ad for Zappos. The google algorithm must have a preference for real "vintage" content as the fifth highest ranked site of 28,000 was "Last updated January 5, 1998" I shall retire to bedlum... -Dickens

09/12/02 GOD -1 [The name game]: I noticed on another Blog somewhere a reference to the fact that a google search on the name "Tim" returns a profile page of Tim Berners-Lee as the number one result. The ranking seems appropriate as this ultimate Internet guy's name is everywhere. I had read a lot of Mr. TB Lee's stuff in the past and thinking about some of the compatible aspects of our ideas on Internet navigation, I decided to try another e-mail contact. (last one over a year ago) Nothing ventured, nothing gained as they say. In this 1997 article Mr. Lee almost suggests the idea of establishing a "common vocabulary" for the limited purpose of providing consistent content description and enhanced Internet navigation-- which I have been advocating. This archived [here] sitepoint discussion is in my opinion a very informative and entertaining hashing out of the issue. [of course Google and Fast have not indexed the content even though their botts visit regularly] RSD [Regularly Scheduled Disappointment]: Okay it has only been a week, but I did think this Blog would provoke comment from someone. Truth is, the home page has been getting pitifully little traffic over the last six months. The fact that I'm no longer willing to waste time posting on censored SEO message boards is probably largely responsible. I did post a remark at XODP regarding site enhancements but that site is probably getting even less traffic than this one.

Archived Pages Week1 09/06/02 - 09/11/02 Week2 09/12/02 - 09/18/02