Wednesday, February 7th, 2001
CNET documents the annoying trend employed by low-rent websites: spawning new windows when you try to exit. The new windows contain all manner of advertisements, inducements, and crap you are certain to have no interest in.
I’m told porn sites have been doing this for years. I have no direct experience with this, of course.
One workaround is to surf with JavaScript disabled in your browser. This is not a great solution, though, because too many sites rely on JavaScript for basic site navigation.
As I mulled this over, I saw a potential solution.
First, the technical background… The annoying window-spawning behavior is enabled by an event handler called onunload, part of the JavaScript/ECMAScript language. Any JavaScript statements that appear in the onUnload attribute of an HTML document’s <BODY> or <FRAMESET> tags will be executed when the document is “unloaded” — when the browser window is closed, or when the user follows a link to another page. (Experiment here.)
And now, the solution: browser makers should give users a way to suppress this handler. For example, perhaps by pressing a key on the keyboard while closing the window, the browser can be instructed to ignore any onUnload calls. Or there could be a preference setting. There is a precedent for this: page authors can control whether links open in a new window, or target a specific frame, by giving a TARGET attribute to the <A> tag — and yet the user can override this through a variety of keyboard shortcuts or menu selections, e.g. “open this link in a new window” or cmd-click (on MacOS).
There are multiple benefits: less Internet congestion, fewer browser crashes, less disk space wasted by unwanted browser cache files, less meaningless traffic logged by advertisers, higher clickthrough rates on advertising, and overall less-annoying user experience (not that any of the advertisers seem to care about that one).
Tags:
posted to channel: Web
updated: 2004-02-22 22:49:16
Tuesday, January 30th, 2001
Anyone familiar with the web industry knows of the trend of patenting so-called inventions to prevent competition. What’s unusual is that some of these “inventions” should not be patentable: according to James Gleick (in Patently Absurd), an idea cannot be patented. Actual programming code cannot be patented either, although it is protected by copyright. “Software and algorithms used to be unpatentable,” says Gleick. But that appears to have changed, as evidenced by recent well-known patents on ideas: Amazon’s “1-click” ordering method, Priceline’s “name your own price” shopping methodology, and Intouch Group’s method of sampling music.
This last example strikes close to home. I know people who used to work at Intouch. And I know some people who work at the companies Intouch has sued for patent violations: Amazon, Liquid Audio, Listen.com, etc.
Today the legal landscape shifted dramatically. A website called BountyQuest, established to find “prior art” in contested patent issues, has announced 4 winners. “Prior art” is something that proves that a patent should never have been issued, because the “invention” was not original to the patent holder. In the case of Intouch’s patent on music sampling, Perry Leopold was able to prove he had released his music-sampling concept to the public domain years before Intouch existed. This would tend to call into question the validity of Intouch’s patent — and if the patent is overturned, the Intouch lawsuit should disappear quickly.
The core problem seems to be that USPTO analysts are not equipped to find prior art when they issue patents on software algorithms (see Gleick’s article for more on this). I hope BountyQuest’s recent success leads to a fundamental change in the USPTO approval process.
Tags:
posted to channel: Web
updated: 2004-02-22 22:49:16
Friday, January 26th, 2001
Whenever I download MP3s with Napster, my first choice (when I’m not searching for something so obscure that nobody else on the planet, apparently, has encoded it) is to download from the bozo who claims his bandwidth is 14.4k. Invariably, these guys have OC3s — capable of serving up not just a song, or even a CD, but an entire genre of music in about sixty seconds. For unknown reasons every one of those guys thinks that by hiding their monstrous bandwidth behind the “14.4'' label, nobody will figure out that they’re running Napster one hop away from MAE West on a 72-node Beowolf cluster with trunked gig-E connections to a NetApp filer.
But sometimes I’m wrong, and I find a guy who really is trying to squeeze dozens of megabytes of music through a crackly old analog phone line — miles of corroded copper wire strung haphazardly to a shed behind an abandoned gas station in Tucumcari, NM — where he proudly boots into Napster on a hot-rodded 486 (133MHz DX-4!) with 8 megs of RAM, and a genuine 14.4k modem with a street value of about five cents.
Tags:
posted to channel: Web
updated: 2004-02-22 22:49:16
Tuesday, January 23rd, 2001
In the spirit of Disturbing Search Requests, I trace back my referer logs to see who is linking to me. I found one that strikes me as odd.
I’ve received a number of clickthroughs from this site. I visited to see why they were pointing to me… and found the page to be written in French. I asked babelfish to translate, and was dismayed to find my site listed under the heading Dreadful, dirty, and malicious. Worse yet is the 3-word summary of my journal: pub with fart
Pub with fart? Maybe this should be my site’s slogan. Some folks appropriate song lyrics, or marketroid buzz-phrases. But I could be much more original: debris.com: Pub with Fart
I wonder if it’s too late to get in as a DigiScents beta site. I’m quite sure their ScentWare SDK allows me to embed hydrogen sulfide in a web page.
Tags:
posted to channel: Web
updated: 2004-02-22 22:49:16
Thursday, January 11th, 2001
Near the top left corner of this image, from Terraserver, is a picture of the house I grew up in. I am just dying to know who wrote the ‘X’ on there.
Tags:
posted to channel: Web
updated: 2005-07-22 05:35:03
