During the patent examination process, I find that examiners rarely use prior art products as references. Instead the focus is on literature and thus, in the case of products, on product documentation such as user manuals, FAQs, etc. I find this to be problematic because product documentation does not and cannot cover every subtle aspect of all the features and functions of a given product. Therefore, reliance on non-patent product literature is covering only a subset of the prior art associated with a given product, whereas the product itself can be used to demonstrate anticipation of claims in a patent application under examination.
There have been occasions when I have been analyzing a given patent and can find a prior art product that seems to anticipate the patent’s claims. In these cases the product’s literature does not sufficiently describe or demonstrate the functionality covered in the patent while the product itself appears to.
This seems to me to be a fundamental problem with the examination process, and therefore a potential issue with a granted patent’s presumption of validity. Of course, examiners do not have access to many prior art products, particularly those that are costly, rare, large, etc. However, when a patent is asserted, the accused party will often go much further than would an examiner in looking not so much for literature as for products to use defensively in order to raise questions of validity. Certainly one of the first places I look in validity searches is for products, as I know that the examiner has likely not done so. This is despite 35 USC 102(b) :”A person shall be entitled to a patent unless …the invention was patented or described in a printed publication in this or a foreign country or in public use or on sale in this country, more than one year prior to the date of the application for patent in the United States…”
In fact, disclosure and examination forms such as the USPTO’s Information Disclosure Statement (IDS) (form PTO/SB/08a) (submitted by the patent practitioner) and the Notice of References Cited (form PTO-892) (provided by the examiner) only include table fields explicitly denoted to be used for documents. How does one then go about citing a product?
What are the readers’ thoughts on the concentrated patent examiner reliance on literature and not on existing products when searching for and analyzing prior art? And how do practitioners list a product in an IDS? What do examiners think about this lack of product citation? Is it indeed an issue? I look forward to your comments.
The Patent Analyst 
Most of the functional parts of software are kept from the public eye. Although I can type in a window and get a desired result, I do not know the background that processed the input to get output. Without the reference being “enabling” it is not a good reference.
Second, to anticipate the software must do the steps claimed. If the software does A, B, C, E, and the claims do A, B, D, then the software doesn’t anticipate the claims. It may be a good start for obviousness but not anticipation.
Finally, some software comes with confidentiality provisions in the agreement and may never have been disclosed publically. Especially, expensive or custom software frequently cannot be copied, shared, opened, or modified by anyone other than the creating company. Without public disclosure, it may not provide art.
I admit that past programs are a great place to look, especially if you can get information about the steps taken and the date it became publically available.
It is true that determining what steps are performed in software can often be difficult. Though for user interface types of functionality, it is usually straight-forward. Otherwise, a packet capture utility such as Wireshark can come in handy to capture and analyze incoming/outgoing traffic.
However, in the case that infringement of an invention would not easily be detected or determined, then of course it could be worthwhile considering maintaining the invention as a trade secret. As a former coworker of mine often said (not sure if he was quoting someone else), “That which comes before anticipates. That which comes after infringes.” The full implementation of the “after” can be more “easily” delineated through e.g. source code discovery as part of litigation proceedings. Good luck getting the source code of the “before” implementation!
The advantage provided by documents and especially patents is certainty of the dates involved. Much of deciding if prior are can be relied on is deciding if it really is “prior.” A patent clearly shows the publication date and the filing date, right on its face. A published non-patent document may be as useful as a patent but there can be a lot of gray area. When is a document from the “June 1996” issue of a publication actually available to the public? How about something from the “Winter 2001” issue? A few days or weeks can make a big difference, and if you can nail down the date of public availability with certainty, then you know whether you can use that document to invalidate (or reject) a claim.
Products as prior art are certainly effective, but again, you need some way to verify the date of public availablity. For a cell phone, “the date we submitted all the circuit diagrams and other required info to the FCC” is a pretty good date. The submission is a public record, with a date-received indication from the FCC. On the other hand, a product whose samples were provided to customers “last spring” is going to be tough to rely on as prior art.
I also completely agree that products are an oft-overlooked source of prior art. One other challenge with products is – how do you search for them/become aware of their existence in the first place? Google used to have a product known as Google Catalog Search, I believe, that provided access to old product catalogs, but has now been closed (booksearch.blogspot.com/2009/01/farewell-google-catalog-search.html).
I work for a website that is focused in part on collecting information about patent searching in wiki pages. We host a number of wiki pages on best practices in patent searching, at http://www.intellogist.com/wiki/Category:Best_Practices . It would be fascinating to have a Best Practices for Product Search article for our collection – I’d love to collaborate on such a thing (or anyone can start it for themselves after registering on the wiki!). If you’re interested, please let me know!
I agree that product searching is not easy and especially those where process of making or functioning of a particular device/article/composition is needed. GNPD (global new products database) is a good place to look at for preliminary searches but this is limited to FMCG/CPG products mostly. No idea about softwares and electronic devices.