Uncool: USPTO Breaks Millions Of Patent URLs Without Public Notice
Posted March 26, 2007, in Clients, Clock Tower Law Group, Erik's Favorites, Patent Law, Privacy, Trademark Law, Weblogs, Worst Practices by @ErikJHeels (permalink: http://erikjheels.com/?p=733)
Static URLs? We don't need no stinkin' static URLs!
The United States Patent and Trademark Office (USPTO) handles both patents and trademarks, but the patent side of the house lags behind the trademark side of the house in implementing user-friendly technology. Over the weekend, the USPTO updated the URL format that it uses for it's public PAIR system <http://portal.uspto.gov/external/portal/pair> (PAIR stands for Patent Application Information Retrieval).
PAIR is used by patent applicants, patent practitioners, and the public at large to check the status of pending and issued patent applications. The public PAIR portion is for applications that have been published or issued, the private PAIR system is for the rest (and requires a special encryption key to access, which is available to patent practitioners).
My patent and trademark law firm, Clock Tower Law Group <http://www.clocktowerlaw.com/>, routinely sends static URLs to clients so that they can monitor the progress of their own applications. We also monitor all of these URLs. One of the reasons that we proactively monitor URLs for our clients is because we believe that it is better to prevent problems before they get out of hand. For example, one of our client's trademarks was accidentally assigned to Viacom due to a USPTO error. If we had not been monitoring our client's trademarks, then this problem may not have been discovered.
Trademark URLs have a very simple format (based on the trademark application number).
For example, here is the URL for one of our client's trademarks (the KAYAK.COM trademark):
<http://tarr.uspto.gov/servlet/tarr?regser=serial&entry=78640286>
And here is the URL for one of our own trademarks (the CLOCK TOWER LAW GROUP logo trademark):
<http://tarr.uspto.gov/servlet/tarr?regser=serial&entry=78614746>
The Trademark Office's TARR URLS (TARR stands for Trademark Application and Registration Retrieval - the USPTO loves its acronyms) are very helpful. They include the current owner, current attorney, current status, and a brief prosecution history. Perfect, really. And they are constantly updated.
The Patent Office's PAIR URLs, on the other hand, are ugly, long, and difficult to parse. But up until today, they were at least functional. Here is the URL for one of the patents we prosecuted (for Inceptor, now owned by Idearc):
Hint: the above URL does not work.
Note the "09456777" in that URL, which is the patent application number for this application, which issued as United States Patent 7062707 <http://www.erikjheels.com/?p=648>.
The irony is that if you search for the application and bookmark one of its URLs (such as the file history tab), then you will get a URL of the above format. You can bookmark the URL for that tab, and another tab, and both will work, but only for the duration of your browser session. If you quit your browser and restart, then those URLs will not work. Until today, it did not work like this.
It gets worse. The PAIR system is setting a cookie on your system, which includes session data. The cookie expires at the end of the session, but the URL appears to depend on the session data in the cookie. I would argue that there is no reason to set the cookie in the first place. There is certainly no reason to have the a static URL depend upon it, and to have the URL expire at the end of your browser session, but that's how it now works. This also appears to be contrary to the USPTO's own privacy policy <http://www.uspto.gov/web/doc/privact.htm>, which discloses the use of cookies for other USPTO systems, but not for public PAIR.
Despite the general ugliness of the PAIR URLs, they used to work just fine. You could bookmark them, email them to your clients, or use a URL monitoring service <http://lifehacker.com/software/web-utilities/download-of-the-day-infoic-157706.php> to monitor their status.
It's bad enough that neither the Patent Office nor the Trademark Office offer ATOM or RSS feeds for trademark status (TARR) and patent status (PAIR), but it's unforgivable that the Patent Office broke millions of static URLs over the weekend without notice (or apparent reason).
If you are as upset about this as I am, then please send email to the USPTO Electronic Business Center at ebc@uspto.gov and urge them to restore static URLs, to simplify and standardize their formats, and not to change the URLs without adequate public notice and comment.
You could also urge the USPTO to add ATOM or RSS feeds, but one step at a time.
In the spirit of full disclosure, Kayak and Inceptor are/were clients of Clock Tower Law Group and are mentioned in this article.
[...] Erik J. Heels: Uncool: USPTO Breaks Millions Of Patent URLs Without Public Notice [...]
See also:
http://inventblog.com/2007/03/create-static-urls-of-patenttrademark-pages-for-clientsthey-are-probably-broken-now.html
http://www.patentlyo.com/patent/2007/03/uspto_public_pa.html
http://www.rethinkip.com/archives/urls_gone_bad.html
http://lawreview.justia.com/2007/03/us_patent_office_removes_abili.html
See also:
http://techdirt.com/articles/20070328/132314.shtml
You said, "The public PAIR portion is for applications that have bee unpublished or issued …"
I think you mean, "The public PAIR portion is for applications that have been published or issued …"
Greetings Anon,
Thanks - I just fixed it.
Regards,
Erik
While you're at it, here's another typo:
"For example, here is the URL for our of one of our client's trademarks"
Should be:
"For example, here is the URL for one of our client's trademarks"
Greetings JKovacs,
Thanks - fixed.
Regards,
Erik
Typo:
"Here is an the URL for one of the patents we prosecuted" — get rid of "an"
Another annoying (and contradictory) issue with the public PAIR site is: the amount of unrelated info present. While on the one hand the PAIR gods may cite resource issues such as disk space, servers, and bandwidth, on the other hand they burden the public PAIR site with a 62-to-1 ratio (approximately) between unrelated/unwanted/un-needed info to what's really needed by their customers. In other words: why does the USPTO PAIR site force me to download an approximately 62KB web page so I can look at approximately 1KB of bibliographic information? They are the first ones who should conserve resources by designing and implementing a low-overhead info retrieval site!
Their HTML and JavaScript code is a mess. Here's an example from todays public PAIR–these are the parameters sent "over the wire" to the USPTO so your query can be processed–it's ridiculous that this many should be required–also note the duplicated "lang=DINO" key/value pair!
loadImages=LoadAllImages
searchType=
isSubmitted=isSubmitted
selectedApplication=undefined
dosnumber=NUMBER
historyNumber=
isHistorySelected=
submitButtonClicked=submitButtonClicked
AppSearchType=appId
AppSearchType=
selecteddosnum=NUMBER
lang=DINO
testHidden=appId
selectedSearchOption=
public_selectedSearchOption=pair_applicationSearchoption
is_pair_new_search=appId
pair_control_no_search=
paramForXmlDownload=
appSelectedType=appId
dosnum=NUMBER
lang=DINO
dosnumsele=++++…+++++
Contrast this with a "Publication Number Search" from http://www.uspto.gov/patft/index.html:
TERM1=20010000044
Sect1=PTO1
Sect2=HITOFF
d=PG01
p=1
u=/netahtml/PTO/srchnum.html
r=0
f=S
l=50
Updated numbers:
62KB HTML page PLUS 48,525 bytes for images (.gif); JavaScript (.js); and Cascading Style Sheets (.css) =~ 109KB. What a waste!
If only the PAIR portal were implemented by the same programmers who did the PatFT and AppFT interface at http://www.uspto.gov/patft/index.html. Sigh.
I usually proof my posts with a text-to-speech program to check for typos. Due to the urgency of this issue, I didn't do that this time. Thanks all for your help.
Thanks, Kerry, for the link to Patent Lens–very interesting treatment of IP.
I noticed that Patent Lens provides an even simpler mechanism for linking to PAIR via redirection:
http://www.patentlens.net/cambia-api/pairlink.html?filing_number=11503243
http://www.patentlens.net/cambia-api/pairlink.html?publication_number=US20070067865
Just substitute the numbers in these examples with your own. Note the presence of "US" before the publication number and its absence before the filing number. Interestingly, I was able to use both 10-digit and 11-digit publication numbers (e.g., US20070067865 and US2007067865).
Message from:
PAIR
Internal Server Error
The server encountered an internal error or misconfiguration and was unable to complete your request.
Please contact the server administrator, usptoinfo@uspto.gov and inform them of the time the error occurred, and anything you might have done that may have caused the error.
More information about this error may be available in the server error log.
————————————–
IBM_HTTP_Server/6.0.2.15 Apache/2.0.47 (Unix) Server at portal.uspto.gov Port 80
Greetings JKovacs ,
TDR has also been down for about the last 4 hours. But it's after 3:30 pm ET, so I'm guessing it won't be fixed until morning.
Regards,
Erik
BTW, the links have changed yet again, so our static links no longer work. We're continuing to look for another solution.
Greetings Kerry,
I contacted the ABA yesterday but they were no help.
The USPTO is now putting session data garbage in the URL as well. Check out this URL (same example as my original post):
http://portal.uspto.gov/external/portal/!ut/p/kcxml/04_Sj9SPykssy0xPLMnMz0vM0Y_QjzKLN4gPMATJgFieA
fqRqCLGpugijnABX4_83FT9IKBEpDlQxNDCRz8qJzU9MblSP1jfWz9AvyA3NDSi3NsRAHxEBJg!/delta/base64xml/
L0lJSk03dWlDU1lKSi9vQXd3QUFNWWdBQ0VJUWhDRUVJaEZLQSEvNEZHZ2RZbktKMEZSb1hmckNIZGgvN18w
XzE4TC8zL3NhLmdldEJpYg!!?selectedTab=fileHistorytab&isSubmitted=isSubmitted&dosnum=09456777&pu
blic_selectedSearchOption=#7_0_18L
Nice. I somehow think that blogging is the only way this problem will be fixed.
Regards,
Erik
Hi all,
I recently contacted EBC by phone to protest. They just said they were aware of the issue and that they didn't expect any changes in the near future.
So I came up with a php script that will fetch the desired information with wget from an application number. The principle is to first load the search form page, save the session cookie and retrieve the randomly generated "URL key" from that page. A session unique URL is then generated from the key and the application number, and wget is used on that URL, passing on the previously saved cookie.
I haven't fully tested this yet, but my first trials work. I assume this script can be adapted to create a redirection to the correct link (but I haven't worked on that because I didn't have that need).
Enjoy!
[...] (which are now hidden from search engines with non-static URLs in the public PAIR system). Static URLs for patent documents were unceremoniously eliminated by the USPTO in May 2007. (I protested. Some other bloggers blogged. But the patent bar in general appeared not to [...]
[...] commentsVerizon FiOS TV Problems - 32 commentsHow To Set Up VNC Over SSH On Windows XP - 29 commentsUncool: USPTO Breaks Millions Of Patent URLs Without Public Notice - 19 commentsAmerican Idol Top 6 Recap - 18 commentsHow To Print OS X To XP - 16 commentsAre Blogs [...]
Word on the street is that the USPTO is considering privatizing the access to PAIR, probably selling it to Thomson & Thomson or some other money grubbing scum buckets. So the taxpayer will have to pay twice to prepare then retrieve public information. This bureaucracy is OUT OF CONTROL
Update 2007-01-02: PAIR is in the process of implementing a Captcha challenge to "…prevent disruptive use by automated programs."
HA! There are captcha mines in China, India and other countries that can be employed to circumvent that.
The real issue is as I mentioned in my post dated March 29th, 2007: the inefficiency of the PAIR site itself, namely its 65-to-100 to one ratio of peripheral information to that which is desired.
I have uploaded a "first look" at the PAIR captcha scheme here:
http://i12.tinypic.com/82050et.jpg
Here's what one blogger wrote about captchas:
"Of course, it’s no help to the clever counter-attack strategy of bots getting people to solve CAPTCHAs in exchange for access to porn"
So, can one conclude that the USPTO is implicitly contributing to porn access? Heh heh.
PAIR is back up and running… with Recaptcha.
PAIR–via Recaptcha–is employing a proprietary MPEG audio layer 3 CODEC that precludes some users (like me) who choose not to pay for it (directly or indirectly).
Let's see how long it takes for someone to figure out how to programmatically get around their latest hurdle.
What exactly is the USPTO's publicly stated reason for making it more difficult to obtain this information?
John; Just go to the PAIR site and you'll find this:
"To continue, you are required to enter the verification code as shown in the box below. This step helps prevent disruptive use by automated programs. For information on PAIR Usage Policy, visit http://www.uspto.gov/ebc/index.html."
[...] - 43 commentsGuest Blogger - 34 commentsHow To Set Up VNC Over SSH On Windows XP - 32 commentsUncool: USPTO Breaks Millions Of Patent URLs Without Public Notice - 27 commentsHow To Print OS X To XP - 19 commentsChinese Domain Names, Shanghai IDC, and Gardiner [...]