Deficiencies in the “auxiliary PDF” approach for DOCX filing

There are many reasons why the USPTO’s “auxiliary PDF” approach, which the USPTO hopes would induce reluctant DOCX filers to do DOCX filing, is unacceptable.  (See USPTO blinks a second time on auxiliary PDF with DOCX filing.)  Here are some of the reasons. 

    • The Federal Register notice carefully avoids saying that the uploaded PDF “controls” or is “authoritative”.
    • The USPTO does not come out and expressly promise that the message digest in the ack receipt will match the auxiliary PDF file.
    • The USPTO does not even come out and expressly promise that there will be a message digest in the ack receipt for the auxiliary PDF file.
    • Until now the USPTO has not been storing the auxiliary PDF bit-for-bit in SCORE.  Instead, the auxiliary PDF gets dismembered into TIF images (one per PDF page).  Then the TIF images get downsampled or upsampled from their original resolution to 300 dots per inch, thus losing resolution.  The TIF images also get halftoned, which ensures blurring for any gray scale or color in the PDF.  If and when the filer tries to download what purports to be the auxiliary PDF from IFW, the file downloaded from IFW it will be a different file size than the actual auxiliary PDF that the filer actually uploaded, and it will fail to match the message digest from the ack receipt, and it will be blurred, sometimes beyond readability.
    • It is not at all clear from the Federal Register notice how the filer can escape the legal consequences of having clicked on the adhesion contract that says the D2 docx file (not the PDF) is the controlling document.
    • The Federal Register notice carefully avoids saying how closely the auxiliary PDF has to “match” the D1 DOCX file (or the D2 DOCX file).  Suppose they don’t really match very closely, maybe because the USPTO’s DOCX engine renders things strikingly differently than the filer’s word processor does.  Is the USPTO really committing to issuing a CofC to say that the issued patent is really what the auxiliary PDF said, no matter how much it diverges from the USPTO’s rendering of the D1 or the D2?

7 Replies to “Deficiencies in the “auxiliary PDF” approach for DOCX filing”

  1. “The United States Patent and Trademark Office (USPTO) is continuing to modernize and streamline its patent application systems to support robust and reliable patent rights, speed the issuance of patents, and reduce the costs and barriers of global patent protection. The submission of patent applications in DOCX format facilitates the USPTO’s ongoing efforts.”

    Sure. And the earth is flat.

  2. Indeed it comes down to CoC practice (10 years from now during, or shortly before, litigation). You have to be sure that the USPTO will then allow you to invoke the document that *you* submitted and will correct any conversion errors at their end. By the way, the EPO does not permit such corrections.

  3. Can anybody explain why the PTO goes so far out of their way to mangle the pdf files that applicants submit, as opposed to simply SAVING the damned things?

    1. That’s easy. I answered your question during one of my recent risks-of-DOCX webinars. When the USPTO decided that ePave was a failure, and then decided to start taking PDF patent applications, the USPTO tried to create its own software for this and failed. At that point USPTO licensed EPO’s “Phoenix” software which is what EPO was using to manage its image-based patent filing system. This must have been terribly embarrassing for the USPTO from an institutional point of view because it was all to much like admitting that some other patent office was better at doing something than the USPTO was. Anyway, one thing that was baked into Phoenix at a very basic level was that it only really stored TIF images, one per page. So if the applicant were to e-file (say) a ten-page PDF, the first thing that would happen is the PDF getting dismembered into ten one-page TIF images. Later if the applicant wished to download a file from the (Phoenix renamed as EFS-Web) system as a PDF, the only way to do it was to stitch the TIF images back together on the fly, into a simulacrum of the original PDF.

Leave a Reply

Your email address will not be published. Required fields are marked *