8 replies to this topic

[kensplace]

Not sure if this has been mentioned before, but if not...

Amazon has been granted a patent for passing parameters as part of the url, without a query string
(ie instead of example.com?search=dvd you use mod rewrite to use example.com/dvd )

Surely this falls under the its obvious to people in the field, has prior art, is not unique and probably more things....

I cannot believe the patent office is granting patents like this, but they are, the big companies are slowly patenting pretty much everything (often where they should never be allowed to).

link to slashdot about it

[eKstreme]

Filed 3 years ago, so not that long ago.

I read the claims and skimmed the rest. There is nothing in there that qualifies as inventive. To take Apache as the embodiment of the web server, they're using a feature of that software that's designed for this purpose.

Which begs the question, which moron granted that patent?

Unless Bill (or anyone else!) can tell me what I'm missing.

Pierre

[bragadocchio]

Amazon has been granted a patent for passing parameters as part of the url, without a query string
(ie instead of example.com?search=dvd you use mod rewrite to use example.com/dvd )

It's not rewriting to a specific URL, but rather taking the string of words after the slash, and using them to perform a search.

It's technology that Amazon is presently using, and you can see it by typing (or pasting) the following into your browser address bar, with the spaces between the words at the end:

a9.com/newark delaware pizza

or

a9.com/cre8asite forums marketing 101

There are a number of folks in the slashdot thread who claim to have been doing this for years. It's mentioned that php.net has been for at least 7 years.

paste this into an address bar:

php.net/list of functions

You get an error message when you do. But if you type something that is in their online manual, you do get redirected:

php.net/function exists

So, I don't think that php.net is doing quite the same thing.

[kensplace]

Amazon appears to be redirecting also, the server headers for an example above give

#1 Server Response: http://a9.com/cre8asite forums marketing 101
HTTP Status Code: HTTP/1.1 301 Moved Permanently
Date: Wed, 28 Nov 2007 22:49:12 GMT
Expires: Thu, 01 Jan 1970 00:00:00 GMT
Pragma: no-cache
Set-Cookie: a9locale=en_US; Domain=.a9.com; Path=/
Location: http://lite.a9.com/a...query=cre8asite forums marketing 101&search_source=WEB
Content-Type: text/html;charset=UTF-8
Content-Length: 0
Vary: Accept-Encoding,User-Agent
Connection: close
Redirect Target: http://lite.a9.com/a...query=cre8asite forums marketing 101&search_source=WEB

#2 Server Response: http://lite.a9.com/a...query=cre8asite forums marketing 101&search_source=WEB
HTTP Status Code: HTTP/1.1 200 OK
Date: Wed, 28 Nov 2007 22:49:12 GMT
Set-Cookie: a9locale=en_US; Domain=.a9.com; Path=/
Expires: Thu, 01 Jan 1970 00:00:00 GMT
Pragma: no-cache
Content-Type: text/html;charset=UTF-8
Content-Length: 7919
Vary: Accept-Encoding,User-Agent
Connection: close

ie..

#2 Server Response: http://lite.a9.com/a...query=cre8asite forums marketing

It must be pulling the data in from there, and displaying it on the main site.

I would assume (without trying it) that php.net could easily get around the problem illustrated earlier by using the feature documented in the apache 1.2 docs, available since 1996 (apache docs

'noescape|NE' (no URI escaping of output)
This flag keeps mod_rewrite from applying the usual URI escaping rules to the result of a rewrite. Ordinarily, special characters (such as '%', '$', ';', and so on) will be escaped into their hexcode equivalents ('%25', '%24', and '%3B', respectively); this flag prevents this from being done.

as its the space in the url that was being escaped into the hex character that stopped it finding the exact term.


There is bound to be more ways to do it than this also....

After all, all we are talking about at the end of the day is getting the url the user typed and converting it into a parameter - how its done is fairly irrelavant, its the fact it just does not seem a novel concept, nor one that is new....

I cant see how they can patent something 'in the manual' so to speak, does that mean people can just grab the docs for a language and patent all possible uses for something designed specifically to do those uses?

[eKstreme]

Well, clean URLs are not new, and using space as a separator is... well... the law. It's English grammar!

Basically, this bit of Apache .htaccess will do what the patent claims:

RewriteEngine On
RewriteBase /
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule . /index.php [L]

Basically it says if you get a URL that doesn't map to file or folder, route it to index.php.

What index.php does is then the search functionality the patent talks about. To take Wordpress as an example, the search would identify one post (if the request was for a permalink), a set of posts (e.g. a category or archive page), or a page from the database of posts and pages.

To take a bigger example: del.icio.us/tag/seo tools. The space gets encoded to its hex format, but, again, that's the standard (in this case HTTP).

I still fail to see anything inventive, and if so, we have prior art in the case of delicious. According the great Oracle of the Web (Wikipedia, who else?), the site was founded 4 years ago, which puts this a few months (about a year) ahead of the patent filing.

Pierre

[bragadocchio]

Is delicious doing a search, as in what a search engine might do, or are they attempting to match a tag, and if there is no tag, then they just show a "no items"

Is that a difference that should make a difference? I don't know.

This really isn't a complicated invention/process. It may not be all that "non obvious."

One of the rules usually followed for granting a patent is that it be:

New
Useful
Non obvious

Now I don't see any of the major search engines doing this, and I'm questioning the usefulness of it myself. A lot of the slashdot commentary questions whether it is really new or nonobvious.

A lot of patents don't describe something earthshakingly new, but rather a variation of something that no one else seems to be doing, that has some amount of uniqueness to it, and is useful.

[eKstreme]

I'm trying to avoid debating obviousness because we would need to define what are the skills and know-how of someone skilled in the art. That's two large tubs of worms, hence me attacking it on the grounds of newness and trying to find prior art.

For usefulness, I think it's very useful as a concept but the encoding associated with it makes it a very stupid implementation on usability grounds. Imagine a poster that contains a URL with the spaces included; let's take this excerpt:

"Go to http://a9.com/hotels san fransisco for the latest hotel deals in San Fransisco ."

Where does the URL end and the text start? The other thing is: who on earth actually types in the address bar to search? Everyone either uses the browser's built-in search box or types in the Google or Yahoo search boxes of their home page. On very rare occasions, I edit URLs and once in a blue moon, I type in the most likely URL on Wikipedia to save a search. In short: not functional in its current implementation. It's useful though to think about delimiters in URLs, but that's already set in the HTTP protocol.

Sorry, I think it's a stupid patent to have been granted.

[A.N.Onym]

As a matter of fact, If you type "g cre8asite marketing 101" in the address bar of Opera, or just "cre8asite marketing 101" in Firefox, you'll get SERPs from Google. I've been doing it for quite some time and it is faster than first going to the SE and then searching.

While the concept is the same, it is the site that is doing it, not the browser, in the case of Amazon. I am not sure how it is unique, but it is useful if many people know about it and use it

In fact, I wanted to do the same for Yahoo (it is really slow to use it itself), so maybe if a browser can do that, it'd be nice. Now, I guess Yahoo will have to buy the rights to use this technology, though.

[kensplace]

If that patent stays undisputed then there is something seriously wrong with patents....

I wonder how many other similar type patents there are, where they are in effect patenting the obvious use of a tool designed to do the thing the patent is granted on. Worse, its not just a tool, its the generic concept of taking a url and doing what you want with it.... There is nothing unique here that I can think of.
If they had found some clever way of doing something that no one else was capable of doing, or could not work out how to do in a few minutes (or less) then maybe - but thats not the case here.

Its like patenting putting in nails in wood with a hammer, thereby forcing anyone who wants to use a hammer to pay up.... Its a obvious use for the hammer....

I cant see how they can stop people from taking the url and doing what they want to do with it in regards to splitting it into parameters etc and deciding where to go from their, its plain crazy if they can patent simple concepts like this.

Someone will probably be granting patents on what parameters you can pass to functions next, or how you can use string manipulation functions to manipulate strings..... If they are not already that is....

Edited by kensplace, 29 November 2007 - 04:25 AM.