Update: two more recent webinars took place and recordings are now available.
Hello folks. Today I presented a webinar on the risks of using the USPTO’s DOCX e-filing system. The recording is now available and you can see it here. The duration is 1 hour and 35 minutes. You can download the presentation materials here.
Keep in mind there will be another live webinar tomorrow on mostly the same subject matter. Click here for more information or to register.
Hi Carl,
Thank you very much for the yesterday’s webinar. Very interesting and useful.
Regards,
Olga
It is indeed a “remarkable” idea to provide the filer with a message digest not of the document as received from the filer but of some edited version. Possibly the USPTO no longer knows the message digest’s purpose (namely, that the filer can prove that the USPTO possessed the file at some point in time by filing evidence that the USPTO calculated the digest from it (this incidentally also appears to require that the digest is time-stamped and signed with the USPTO’s private key; otherwise the filer could try to make a filing receipt in photoshop). However, as sheet 22 shows, exactly the same can weird procedure can be used by the USPTO when receiving PDF files. So, this is not inextricably linked to the use of DOCX, isn’t it?
The reason a SHA-512 hash can’t be “reverse engineered” is that it doesn’t contain all of the information in the original document, and thus doesn’t unambiguously point back to the original. There are in fact a vast number of other “original” documents that would produce the very same hash. It’s extremely unlikely that any of these other potential originals would be anything but gibberish, and that’s why we trust that identical hashes are, for all practical purposes, proof of the identity of the originals.
Interesting question: what if you generate a hash of a trusted pdf of your application, and submit it as a “miscellaneous communication” along with the application papers, certifying that “This is the SHA-512 hash of a pdf of the specification filed herewith”? The PTO has no rational reason (yeah, I know…) not to believe you if you later submit the pdf and point to it to support an amendment or correction.
Thank you for commenting. Indeed why not simply paste the SHA-512 hash of the trusted PDF into the body of the DOCX patent application with a bit of explanation?
Pasting the hash into the document will alter the hash of the corresponding pdf – and I question whether it would count as a “disclosure” of anything. That’s why I suggested submitting it as a separate item. (Calling it a “preliminary amendment” might be a way of adding it to the specification without compromising the validity of the hash.)