FITUG e.V. Förderverein Informationstechnik und Gesellschaft

Vermischtes

------- Forwarded message follows ------- Date sent: Thu, 17 Feb 2000 04:47:28 -0600 Send reply to: patent-l@ftplaw.wuacc.edu From: Gregory Aharonian <srctran@world.std.com> To: Multiple recipients of list <patent-l@ftplaw.wuacc.edu> Subject: PATNEWS: More on PTO's banning URLs in patent specifications

!20000216 More on PTO's banning URLs in patent specifications

But first, Computerworld reports that politicians in the State of Virginia passed something even more crappy than the average software patent, the Uniform Computer Information Transactions Act (UCITA), which would let vendors disable software if they believe users violate a license, and other contempts for the customer. I wonder what the ratio of purchasing legislation is for the federal and state levels. More on the passage of UCITA in Virginia at: http://www.computerworld.com/home/print.nsf/CWFlash/000215ECDA.

Next, the current issue of Technology Review has an article on software patents by Seth Shulman, who was been writing about the craziness of this problem as well: http://www.techreview.com/articles/ma00/shulman.htm Seth's article is online, at the URL above. Technology Review is having an online forum on IP at http://www.techreview.com/forums.

Finally, last week the EE&CS Department at the University of California Berkeley had a two day conference on future information technologies, the latest goodies from their labs. Program papers and keynote presentations in movie form are at http://buffy.eecs.berkeley.edu/IRO/00.conf.html

(Note: the use of URLs above in no way is an attempt to encourage people to do so similarly in their patent applications).

====================

Yesterday I sent out a PATNEWS on a new PTO policy to prohibit URLs and executable Web code in patent specifications. I received a lot of email about this PATNEWS, which surprised me because this naive policy is about as naive as many other computer-related policies PTO management dreams up. Anyways, what follows are some of the comments I received, and my replies.

====

According to the memo, the policy went into effect February 1, 2000 and won't be official from some time. Apparently until then, you can ignore any requests from the examiners to follow the policy, at least based on one comment from a lawyer: To the extent a Patent Office Rule makes a change in stated policy but the new policy is not available to the public, the change is not enforcable under administrative law. See 5 U.S.C. 553. See also Nabisco v. Mossinghoff, an unreported decision from the district of New Jersey of about 1984. And no, I didn't do a deal with the PTO to publish the policy to make it public :-).

Some comments reflecting multiple replies, mostly from lawyers. First, many agreed with the PTO and argued against me when I argued that URLs should be allowable in patent specifications (beyond the irrelevant distinction made in the policy statement to only reject URLs in <>, with the PTO apparently not knowing that programs like Word will automatically turn even partial URLs like "www.uspto.gov" in text files into active hyperlinks that more normally would require the specific form in the memo such as <A HREF="http://www.uspto.gov">. The main objection was the transitory and unverifiable nature of Web links. As one programmer commented: Part of their problems is their reliance on the notoriously security-plagued Microsoft product line, like MSWeird, which is a favorite target of virus hackers, that examiners have to use for their Office Actions.

Unfortunately, in the non-patent prior art that PTO management continues to ignore dealing with in a serious way, URLs are becoming are popular reference mechanism, especially in papers and reports on the Internet. Look at the footnotes and bibliographics of papers coming out of the annual World Wide Web Conferences (most of which are never cited) - lots of URLs used as references. If such references are good enough for those skilled in the art, they should be good enough for the PTO. As one reader agrees: How does the USPTO plan to deal with non-patent prior art which happens to be indexed as a URL? Patent applicants are under an obligation to disclose to the USPTO any prior art of which they are aware. What if this prior art is referred to only as a URL? Should the applicant not disclose it? Or should the USPTO not publish it?

Solution? Require any URLs in patent specifications to be printed out and submitted in IDSs, and/or have Congress fund the Library of Congress to set up a facility to archive and hash-timestamped Web pages for future retreival and verification. Or as one reader suggested, a new profit opportunity for the PTO: An acceptable alternative would be to submit the electronic version of the pertinent information and PTO then can create their own repository of such necessary relevent concerned URLs. And working with the Library of Congress could open up channels to get LoC with the PTO's prior art problems.

One other thing - paper documents can be just as transitory as URLs. Consider old software user manuals, which few people archive, including the software company itself. Many are submitted in as non-patent prior art, and in some cases are the only publicly available copy left after many years. All someone has to do is check out the filewrapper, remove the manual (or any other obscure art in the filewrapper), and check the filewrapper back in. The reference to the manual on the front page is now just as useless as a URL that has been deleted. Don't think this is a problem? Here is some gossip from a private searcher who hangs out in the PTO public search facilities: I was discussing exactly this problem with a professional document handler yesterday. It seems that, more and more, all the documents are missing from patented files. He went into great detail as to the problems he has had getting the non-patent art from files. From my own experience, it may take up to three days [or longer] to get a patented file. Plus, "Abandoned Files" STILL seems to be a "black-hole". Apparently the document handling professionals are not coming up with the documents and others are blaming THEM, perhaps completely unfairly.

Sure the transitory nature of the average Web page is much more probable than for a paper document, but they do share the same problem whose solution is not transitory PTO policies, but a revamping of how the PTO handles non-patent information in general. Never happen, but it is the solution.

One lawyers took exception to the PTO's URL policy:

First of all, mere recitation of a URL is not an incorporation by reference any more than the mere recitation of an article is an incorporation by reference. The recitation must evince some attempt to incorporate in order to be considered an incorporation by reference. If mere recitation is sufficient for a URL, then it will be for an article.

Second, if the constitutional purpose of the patent system is to promote the dissemination of technical information, how can this URL policy be said to promote the purpose it is supposed to be serving? Maybe the USPTO needs to review it's reason for existence.

Dealing with URLs will distract the PTO from its existential mission of delivering as many patents as possible to its large corporate customers. A comment from a corporate patent lawyer, including a criticism of the CAFC:

This PTO notice does indeed seem strange to me, IF it really does requires removal of technical publication web site addresses from patent specifications [as opposed to PTO Form 1449 citation forms]. [E.g., I recently found a publication that was only published on a web site, and must therefor cite it by its web site address.]

I cannot see any logical difference between this PTO notice, preventing citing an internet publishers address in the spec, as compared to a PTO notice that would improperly demand the removal from a specification of the cited mailing address of a Technical Library in which a thesis is cataloged (which is clearly a prior art publication under established case law), OR the identification of the publisher of a hardcover book, OR the citation in the specification of the address of a supplier of a catalogued or trademarked precursor chemical [which is REQUIRED by two particularly egregious "best mode" panel decisions of the CAFC ].

Nor can the PTO take the position that something published on a commonly accessed web site and easily found by a search engine is not a "publication" or not publicly "known" under the patent laws. Even if the publication date is indeterminate, if it was publicly downloaded before the patent application filing date it is a 35 USC 102(a) citable reference.

The fact that web site contents may later change is no different than the fact that chemical products only identifiable by trademarks may also change, or that technical papers handed out at technical conferences may not be readily available later on either.

P.S. I hope the PTO is not planning on relying on the two (note the dissent) patent-practice-clueless CAFC panel majority members of a very recent CAFC decision on "incorporation by reference", who obviously never even read the REST of the very MPEP Section 608.01(p) cited as "authority", or they would have noted that this Same MPEP Section, at I.A.2., ALLOWS cited publications text to be inserted into the specification, without a new matter objection, even for essential claim support.

with a similar comment from another lawyer:

How does inclusion of a URL in a patent application or an issued patent differ from identifying any essential component included in the best mode of the invention by a trade name or by the manufacturer's name. Finding the trade name of the manufacturer's name in the patent, one could then look them up in Thomas Register. This type of craziness is a result of being within the Beltway and having a Beltway, governmental bureaucratic mentality. Note that the Federal Circuit recently rejected such a Beltway, governmental bureaucratic mentality in Atmel Corp v. Info. Storage Devices, Inc. Likely this latest craziness, recall that the USPTO opposed patenting software for 10 to 20 years, will ultimately get straightened around, hopefully not in 20 years.

One lawyer made the following observation on why something must be done:

I had pointed out in a recent post to misc.int-property that the PTO web server for full text patents was passing any HTML tags in the specification as raw HTML to the user's browser. This feature would enable hit tracking by the use of embedded image tags to, for example, a single pixel image stored on your server since that image would be requested from your server each time the patent was viewed on the PTO server.

Apparently the gist of that message was passed along, and was garbled to "URLs" instead of HTML tags generally.

There is a prominent note on the PTO web server to the effect that they don't collect or retain search or hit information. However, if you'd like to know the IP addresses of whoever might be looking at your patents in the full text database, there appears to be a rather simple method for doing that.

Interestingly, the full-text patent database serves exactly what it says - the full text.

And if that text happens to include any HTML or Javascript, well, the PTO full-text database merrily serves that too... and your browser, or anyone else's, will parse and/or execute the code that is included in the text of the patent.

Among the amusing sites to see in the full-text patent server HTML zoo, there are lots of image tags, font formatting tags, a couple of people who've had their patents cut-off in mid-stream by tags indicating the end of the body or the end of the file, at least one pull down menu, a submit button, a java applet window, and assorted form fields. One of my favorite ones is the guy who managed to get a working "mailto" link included in the text of his patent.

For example, scroll through the full text of U.S. Pat. 5,978,817, and you'll get a pretty good idea of what I mean.

http://164.195.100.11/netacgi/nph-Parser?TERM1=5%2CC978%2C817

Better yet, scroll to the bottom of U.S. Pat. No. 5,930,474

In general, it appears that putting the code into a table in the patent text works better than including it in the main text, since it is less likely to have extraneous font format tags inserted in inconvenient places. It appears that most of the unintentional working examples of HTML code in patent text use relative addressing for the image tags instead of absolute addressing (i.e. including the complete URL). I haven't found any pictures yet, but I've only dug into it for about 20 minutes. So, instead of an actual image, what is normally displayed is a 10x10 pixel "missing image" icon (in an IE browser). However, if you'd like to get the picture of your choice (say, a single pixel) into the full-text database, you should include a complete URL to your server in your sample HTML. Then, all you need to do is to keep an eye on the HTTP referrer log for that server, and you will collect the IP addresses of any companies or law firms that have retrieved the text of your patent from the PTO server.

I'm not sure whether the PTO server permits server-side includes, but if it does, then you can use an include to insert text from your server into the text delivered to the browser simply by having the include tag in the text of your patent, along with the URL of your text file. Imagine sparing yourself the headache of re-issue by retaining the ability to alter the claims (at least as provided by the full-text database) at your leisure.

And if you think that's amusing, just wait until the corporate lawyers start doing that to their SEC documents, which are similarly served in full-text format by the SEC's Edgar server - real-time financial, um, corrections.

John Berryhill Ph.D. J.D.

Yes a problem, as another reader points out:

Well, to be honest, I don't think Mr. Kurnin is as clueless as you, Greg. Did you miss the point of his memo? It's a good one. The PTO is making these patents accessible to the public through the Internet and a bunch of wise-asses are taking advantage of the PTO's efforts by embedding links to their sites in the specification. It doesn't strike me as mismanagement to shut that nonsense down. Some idiot could embed a link to some porn site or, who knows, even your site.

But there are solutions that still permit HTML coding and similar types of coding to appear in patent specifications without such problems (such as having the PTO create an archive - and charge a fee to put references into it). Occasionally the PTO has to bend to the technology, and not the other way. Think of embedded coding in Web pages as a potential virus. Well, we have lots patents and products on virus scanners - why not adopt them to analyze patent specifications to remap such potential references to a special Web server at the PTO? Then again, the PTO has yet to write and use a computer program to veryify the antecedent basis of all words in a patent claim (a trivial programming task). So the policy is clueless. There are PTO fees for claims over 20 claims - why not a PTO fee for URLs as well, with the money used to fund an archiving service inside the PTO? Any URL reference in a patent specification could be used to download the Web page to PTO computers, analyzed for potential liabilities and altered to remove such liabilities without changing the information disclosure aspect of the URL. Similar techniques can be used for embedded coding.

So I stick to my original criticism. URLs and HTML/Java/etc. have legitimate uses in patent specifications (just as they do in journal articles and other publication forms) that should be allowed under the patent law. If there are problems, the solution isn't some naive knee-jerk prohibition by PTO management, but along with other non-patent prior art problems the PTO is having, an opportunity to modify existing PTO systems with new features.

Greg Aharonian Internet Patent News Service ------------------------------------------------------------- Private reply: Gregory Aharonian <srctran@world.std.com> Public replies: patent-l@ftplaw.wuacc.edu Technical questions: Paul Arrigo, zzarri@washburn.edu List owners: James Wadley, zzwadl@washburn.edu John Christensen, zzchri@washburn.edu To sign off, send message to: listserv@ftplaw.wuacc.edu Message merely says: unsubscribe patent-l An Experimental and Free Service of WashLawWEB: http://lawlib.wuacc.edu Terms and Conditions: http://lawlib.wuacc.edu/liability.htm ------- End of forwarded message -------

Zurück