Slashdot Mirror
Amazon Patents Including a String at End of a URL
theodp writes "On Tuesday, Amazon search subsidiary A9.com was awarded U.S. patent no. 7,287,042 for 'including a search string at the end of a URL without any special formatting.' In the Summary of the Invention, it's explained that 'a user wishing to search for 'San Francisco Hotels' may do by simply accessing the URL www.domain_name/San Francisco Hotels, where domain_name is a domain name associated with the web site system.' Here's the flowchart that helped cinch the deal."
The php website has done this for ages when searching functions. I am sure they have been doing it before 2004.
eg.
http://www.php.net/stupid%20patents
yep, I have never ever seen this one before.
http://en.wikipedia.org/wiki/Prior_art
http://en.wikipedia.org/wiki/SomeStupidRandomSearchTerm
Im a gamer, not a grammer major. This post is full of spelling and grammer mistakes.
I'm a number, not a free man!
Similar implementations have already been done.
/search/blah...
/invoice/# or invoice/customer/[name or number] or search for customers using similar techniques.
.htaccess file that does a rewrite to a single catch-all if the requested URL does not exist. The app can then parse the request and infer what the user really wants, whether it's an action of a controller, a query or similar.
With Ruby on Rails, it uses a similar technique for discovering actions. It even has facilities for creating custom URL maps so what would normally come across as ?search=blah would get converted into
del.icio.us uses that for tag search (ie: http://del.icio.us/username/blah).
For my internal invoicing system that I wrote in PHP (but never finished), you could search for invoices by going to
The trick involves a
Although I've never seen this specifically applied to search (a la google), it's been used for filtering with tags (like del.icio.us).
stupid software patents.
...spike
Ewwwwww, coconut...
This slide from a talk delivered in January 2003 describes the same idea of searching by URL content (listed under "Interesting Uses"). I don't remember being particularly surprised by the idea at the time, so I'm sure there's considerably older prior art, but this was the first thing that sprang to mind.
(Ignore the date on the top right, which always shows today -- the talk's date of January 22, 2003 is listed on the PHP talk index.)
Only, wikipedia search for the string in the URL is an option that is one click away.
http://en.wikipedia.org/wiki/Ain't%20it%20true
If you ask me I'd use the wikipedia way, or the good old search box.
Because if you're typing into the address box in a browser, you're likely to have autocompletion. That means you're likely to start a search whenever you want to get back at the site, bad for the search engine.
Also your searches are accessible through your browsing history - as for all searches through get requests I think.
Having said that, this patent differs from the prior art of wikipedia by simply doing an additional step automatically. Where's the innovation, USPTO guys?
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
Did they just patented mod_rewrite??? Tue Aug 24 06:55:44 1999 UTC (8 years, 2 months ago) baby! http://svn.apache.org/viewvc/httpd/httpd/trunk/modules/mappers/mod_rewrite.c?revision=83751&view=markup&pathrev=573831
When you apply for a patent, that's the day the prior art becomes effective. So if wikipedia did it after they filed, then that prior art would not count. Not saying it is not a stupid patent, but just wanted to point out, as a general rule, these things can take 5+ years to become live, so sometimes prior art comes around after a company starts using the patent-pending technology and others copy it.
Your ad here. Ask me how!
Did you test that link? It does work (after a redirect).
Prov 9:8 Do not rebuke mockers or they will hate you; rebuke the wise and they will love you.
Filed: August 23, 2004 http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=7,287,042.PN.&OS=PN/7,287,042&RS=PN/7,287,042
TODO create witty sig.
Fsck those Amazon idiots.
:) Amazon seems to make a living patenting obvious ideas, makes you wonder why they never patent anything REALLY original.
I'm patenting some_words.topical_search_engine_domain_name.com
I'm pretty sure I had that search-idea a few years back though. Even if it fails the prior-art test, it's pretty friggin' obvious to anybody who has ever used mod_rewrite.
My own website has two mechanisms very much like this patent and has had so for quite some time now; "file.html" requests are parsed by mod_rewrite, then send as a parameter to a PHP where the named page is loaded (and integrated in a template).
My photoalbum can take any name for a photograph, then finds the closest match to that filename, based on a number of search methods and shows it.
Atleast Amazon can't patent THOSE methods now, since I've published them
Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
Don't forget one I use frequently: http://www.php.net/patentssuck/
Do not meddle in the affairs of sysadmins, for they are subtle, and quick to anger.
it wroks bteter if yuo tyr nad palce teh fsirt and lsat ltetres in teh rghit psotiion but taht is nto psosbile for slmal wrods
Right. So innovative that I only put a CPAN Perl module (CGI::PathInfo) up for that kind of crap, oh, SEVEN YEARS ago.
It's called PATH_INFO. I've been using it since 1996 to implement hierarchical indexes and implied searches for internal websites where I've worked. Most people have never heard of it, because they haven't read the CGI spec, they've just cargo-culted something from the examples directory or, worse, copied a CGI from someone who didn't understand CGI, either.
The structure of a CGI URL in NCSA HTTPD and Apache is:
(#anchors aren't passed to the server, they're used in the browser only.)
So, you can use /path_info if you like, instead of ?query=string. It makes it nicer if you can represent something hierarchically, like .../toolindex/autoconf/2.53/ gives you the meta-info page on how and why autoconf 2.53 got on to our servers. But .../toolindex/autoconf/ just tells you what versions are there.
Combine with ScriptAlias / or SetHandler, and you can do it from the root of the server.
It's a little more work if you want to allow some known paths to go to regular static pages and not the CGI-or-equivalent.
http://en.wikipedia.org/wiki/Prior_art
Actually, those are page titles, not search terms... you can't just go to "http:/en/wikipedia.org/wiki/search terms", you have to go to "http://en.wikipedia.org/wiki/Special:Search?search=search+terms&go=Go".
There probably ARE examples of prior art, but Wikipedia isn't one of them.
I did this in 1999 as a part of software I wrote. I created a facility to map actions to URL strings (and yes, this includes the empty string), and one of the examples I gave was taking a freeform string and searching for it.
I am sure there are many, many, other cases where people mapped 404 to a search, which is the same thing.
In short, not only is this obvious, it is defeated by prior art.
I was doing this in 1996.
I had just joined a startup company, "Hells's Kitchen Systems", or "hks.net". We were an e-commerce startup. Our main product was CCVS, a credit-card processing system for Linux and other versions of Unix. But our first product was a shopping mall written in PHP. Not a simple store, but a mall -- it could contain multiple stores, each of which had multiple departments, each of which had a variety of products.
So, the web content was driven from database searches. But we did not want it to look like that was the case -- we wanted it to look like a family of hand-crafted web sites.
So we did exactly what's described. We appended strings to the end of URLs, and parsed the URLs and used them to search in order to build the pages. People would go through an ordering process, and an order was composed and faxed to the warehouse so it could be fulfilled. It was meant to be a cheap way to get any company that could take catalog orders onto the web without forcing them to change their business processes too much.
It was originally written in PHP/FI2, and then ported to PHP3.
Two different stores that used the system made it into production and were up for years. I am going to wrack my brian to try to remember their names, and if I can, I'll find them on the wayback machine so I can point to them. I bet a bunch of my comments made it into the delivered HTML, and so we might be able to actually prove my claim.
"URL of the form www.domain_name/search_string, where domain_name is a domain name of the web server system" Jassy, et al. needs to read the RFCs! There are nice, reserved domains for uses such as this: example.com, example.net, and example.org. This is very handy when writing documents of this type and everyone should use it. http://www.ietf.org/rfc/rfc2606.txt
This absolutely is an obvious use of mod_rewrite, and is something that we did about 5 years ago to make an e-commerce site we developed here in the UK more Google-friendly. So, instead of having URLs of the form /product.do?id=12345, we made them of the form /product/12345 and had a mod_rewrite rule in Apache rewrite that internally so Resin saw it as the first form. It wasn't exactly new back then, either.
I've not read the patent fully, but if the Slashdot summary is accurate then it's utterly ridiculous.
It's official. Most of you are morons.
From what I can see, the patent covers urls like http://www.answers.com/search but not http://en.wikipedia.org/wiki/Search .
From the patent abstract: "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".
Your example, "http://en.wikipedia.org/some random search term" works for me...after a 5 second redirect.
If you ever work with a patent lawyer, the first thing they say is "you have to report any prior art that you happen know about, but please don't search for any yourself." They also tend to say things like "just because it is obvious to *you* doesn't mean it is. Please don't worry about whether it is obvious. Let us do that".
Couple this with an economy that treats patents as valuable assets, and it's pretty obvious where these things come from. You can't really blame the companies too much for this. It's like there was a government office that offered free money for the first person who yelled the word of the day. Wouldn't you be in their yelling?
The problem is the patent office, which allows this bullshit.