US Patents: Can this be real?! How?!!

I guess many people learned through slashdot that yesterday, Amazon was awarded US patent 7278042, titled Search engine system supporting inclusion of unformatted search string after domain name portion of URL. In case you are still wondering what this patent entails, read the abstract:

A web site system provides functionality for searching a repository of information, such as the World Wide Web, by including a search string at the end of a URL without any special formatting. In one embodiment, when the system receives a request for a URL of the form www.domain_name/char_string, where char_string is a character string that may include spaces and non-alphabetic characters, the system initially determines whether the character string includes a prefix that identifies the URL as a non-search-request URL. If no such prefix is present, the character string is used in its entirely as a search string to execute a search, and the results of the search are returned to the user.

So, basically, this describes an application that would process URLs of the form http://<somedomain>/<search string> by feeding the text value of <search string> into a search engine and return the search results. The flowchart describes the process:



The patent claim includes a few variations:

As will be recognized, various modifications can be made to the method shown in FIG. 1. One such modification is to treat all URLs of the format www.domain_name/char_string as search request URLs. If this approach is used, a subdomain (e.g., www.pages.a9.com) may be provided for accessing other content of the web site. Another variation is to treat certain URLs that lack the full prefix as search requests if such URLs are otherwise invalid. For example, assuming "-/" is used as the only valid prefix, a request for www.a9.com//text may be treated as a request to search for "text" if the URL does not point to a valid subdirectory or object.

Another variation is to use a suffix, rather than a prefix, to identify non-search-request URLs. For instance, a URL of the format www.domain_name/char_string may be treated as a non-search-request URL if and only if char_string ends with a predefined suffix such as "/-" or "/-/"

The patent was filed in 2004. Personally, I have a hard time believing Amazon was the first one to think of this, but even then...I never realized this patent system really is...what will happen next? Is there a patent to interpret subdomains as search strings? Should I file one, just for fun?

Why webdevelopers should use Linux

Recently, I revamped my old Organization Chart javascript widget to a new version. Since I released this code some time ago I received quite a good deal of email requests from people that are actually using it for their company's intranet site.

On the one hand, it makes me very happy to see people are actually using it for professional purposes. On the other hand, at the time I slapped it together in a few hours total, building on some old experience. As it turned out, users were/are having quite some issues to properly customize their organization charts. I even received some code contributions to improve the functionality.

Now, a few days ago, I finally found some time to address the major issues. But something grave changed in the mean while. When I first created and released the widget, I was a windows user, and used to create stuff that would work on Microsoft Internet Explorer, and then modify it to work on Mozilla/Firefox. Now, it has been about a year since I replaced Microsoft Windows for (K)Ubuntu Linux. Meaning, I am without Internet Explorer....or am I ;)

Enter IEs4Linux:



The IE4sLinux project offers the real, genuine Microsoft Windows Internet Explorer as a Wine emulation:



One of the compelling things about the IE4Linux project is that it allows you to run several versions of MS Internet Explorer on the same machine - something I never got to work on my native windows machine.

Obviously, web developers that need to test their pages on several browsers and browser versions are better off on Linux + Wine than native Windows, as Linux + Wine + IEs4Linux allows them to test their pages on several versions of Internet Explorer.

Thanks to the people from IEs4Linux and Wine!

Google Gears Query Tool: bugfixes, new features

Yes, I committed another deadly sin programming for the off-line web :0

A few days ago, I wrote how I didn't test the Google Gears Query Tool with Internet Explorer.

This time, I was kindly informed by Jon Stephens that the query tool didn't check to see whether Google Gears is actually installed. It's fixed now. You should see the following message if the query tool detects that google gears is not yet installed:

Apart from that, I also added a few decorations and features. Most notably, the SQL statement textarea now uses a monospace font, and an SQL statement history was added.

Check it out:

The statement history will simple maintain a stack of successfully executed statements, and put the most recent statements on top. Selecting a statement from the listbox will copy the SQL back to the textarea.