I appreciate that the subject line does not narrow things down very much, given a seemingly limitless variety and number of ways that the USPTO has done wrong-headed things with Patent Center over the past five years of its development. But here is a recent breathtakingly wrong-headed move by the USPTO that deserves particular comment — abruptly imposing upon all applicants an extremely disruptive and wholly unnecessary roadblock to the uploading of PDF files. It makes Patent Center, which was already not easy to use given its many bugs, even harder to use.
Let’s start with a brief description of how things were in the past, and what just changed.
Shortly after September 11, 2001, there was “the anthrax scare”. Government agencies worried that incoming postal mail might be laced with anthrax. The USPTO had failed the previous year in its efforts to roll out the ePave e-filing system (which collected patent applications in character-rich computer-readable form). So with the failure of ePave, the USPTO was stuck in 2001 continuing to receive its half a million patent applications per year on paper.
(First side note: our firm was a beta tester of ePave, and there was a stretch of several months during the beta testing of ePave in which our firm, all by itself, had filed more than one-fourth of all of the patent applications that anybody filed in ePave. This statistic tells you that our firm was one of the most loyal beta-testers of that system, and also hints at how few patent applications actually got filed during the beta testing of ePave.)
(Second side note: toward the end of the beta testing period for ePave, I was disappointed to learn that the way that these character-rich computer-readable patent applications actually made their way from ePave into the USPTO systems was, and I am not making this up, a human being would print out the ePave patent applications onto paper, and would then feed the paper applications into the same USPTO workflow as for patent applications received by postal delivery. They were scanned into image-based PDFs and loaded into USPTO systems as flattened images. All of the computer-readable characters in the ePave submissions were irrevocably discarded.)
(Third side note: most patent applications that users file electronically in EFS-Web and Patent Center are filed in PDF formats that are text-rich. Such PDF files are the result of “printing to PDF” from a word processor, and all of the characters from the word processor file are preserved inside the PDF file, where they can be later extracted by suitable computer software. But, and I am not making this up, the USPTO systems in both EFS-Web and Patent Center discard all of the computer-readable characters inside the text-rich PDF files. Both EFS-Web and Patent Center are designed so as to “flatten” any text-rich PDF into mere TIF images. The TIF images are later subjected to OCR processing to try to recognize the characters for use in downstream USPTO systems, for example for the 18-month pub and for the issued patent.)
As I say, ePave failed, and then 9/11 happened, and the USPTO needed to somehow divert the half a million patent applications a year that were being filed, away from paper filing, and into some electronic e-filing path. And so EFS-Web was put into service in 2002. (And yes, our firm was one of the most loyal beta-testers of EFS-Web. And yes, this is when our firm created and launched the EFS-Web listserv, for users of EFS-Web.)
When users began using EFS-Web in 2002, users had no choice but to learn how to modify perfectly acceptable Windows file names to satisfy unnecessarily picky USPTO systems. File names in Windows were (and are, to this day) permitted to contain (for example) spaces and multiple periods, but EFS-Web would puke on an uploaded file if its name were to contain a space or a spurious period. A variety of characters that are not alphabetic letters or numeric digits, but that are perfectly acceptable in a Windows file name, would likewise cause EFS-Web to puke on the file name if present anywhere in the file name.
More than two decades have passed (from the year 2002 to the present year 2023) during which users who are getting ready to upload a file from Windows into EFS-Web have had it engrained into their routine to scrutinize the name of the file carefully to see if anything about the perfectly acceptable Windows file name would make EFS-Web puke, and would methodically edit the file name to eliminate any spaces or spurious periods or other unacceptable non-alphanumeric characters.
Importantly to this discussion, everybody knows that Windows file names are not case-sensitive. A Windows file called “PaTeNt-ApPlIcAtIoN.PDF” is exactly the same as a Windows file called “pAtEnT-aPpLiCaTiOn.pdf”. And for the past twenty years, with EFS-Web, an uploaded file called “PaTeNt-ApPlIcAtIoN.PDF” has been and is exactly the same as an uploaded file called “pAtEnT-aPpLiCaTiOn.pdf”.
Even more importantly to this discussion, for the past twenty years, with EFS-Web, it has made absolutely no difference whether the user’s uploaded PDF file ends with “pdf” or “PDF”. And this has been exactly what one would expect given that Windows is not case-sensitive about file names.
Now we must discuss Patent Center. The USPTO launched Patent Center in 2018 for alpha testing.
(Fourth side note: In alpha test, US national-phase entries in Patent Center were assigned from a block of application numbers starting at 15/733,001. It ws thus very easy to see exactly how many US national-phase entry filings had been carried out in Patent Center. It turns out that from the start of alpha test in 2018, until the end of 2019, Oppedahl Patent Law Firm LLC filed half of all of the 371 cases that had been filed in Patent Center (blog article). Yes, my firm was a very loyal alpha tester of Patent Center for the USPTO.)
(Fifth side note: In alpha test, design patent applications in Patent Center were assigned from a block of application numbers starting at 29/652,001. It ws thus very easy to see exactly how many design patent application filings had been carried out in Patent Center. It turns out that from the start of alpha test in 2018, until April 2020, Oppedahl Patent Law Firm LLC filed half of all of the design patent applications that had been filed in Patent Center (blog article). Yes, my firm was a very loyal alpha tester of Patent Center for the USPTO.)
More than five years have passed (from the launch of Patent Center in 2018, until Wednesday, November 1, 2023, six days ago), during which users who are getting ready to upload a file from Windows into Patent Center have had it engrained into their routine to scrutinize the name of the file carefully to see if anything about the perfectly acceptable Windows file name would make Patent Center puke, and would methodically edit the file name to eliminate anything that would make Patent Center puke.
The USPTO has tried to convince users to migrate their e-filing activity from EFS-Web over to Patent Center, and by now, just over half of all patent e-filing activity at the USPTO has taken place in Patent Center. (This actually counts as a gobsmacking failure for the USPTO, given that this means that just under half of all patent e-filing activity at the USPTO is still taking place through the older and more reliable EFS-Web, which USPTO plans to shut down tonight at 11:59 PM.) Part of the USPTO’s way of trying to lure users of EFS-Web over to Patent Center has been to tout the (small handful of) ways that e-filing is easier in Patent Center than in EFS-Web. There are three such ways, which the USPTO has repeatedly pointed to in its luring efforts:
-
- Users can upload multiple PDF files at once into Patent Center, while in EFS-Web the user must upload each PDF file one by one.
- In Patent Center, users can insert certain document codes within the file name of the to-be-uploaded PDF file, and sometimes the document code will prompt Patent Center to assign an appropriate document description in an automatic way. (See Tips and Tricks for Users of Patent Center.)
- USPTO has reminded users that Patent Center is more friendly to users than EFS-Web is, with respect to puking on file names. USPTO has repeatedly claimed that Patent Center provides “100%” of the functionality of PAIR and EFS-Web, and indeed for the past five years (until Wednesday, November 1, 2023), it has been true that any to-be-uploaded PDF file whose file name would be acceptable to EFS-Web would likewise be acceptable to Patent Center. Not only that, as USPTO has repeatedly reminded users, Patent Center is actually more friendly than EFS-Web with regard to file names. To give one example, Patent Center will not puke on a space in a file name, while EFS-Web will puke on a space in a file name.
Which brings us to the abrupt and wholly unnecessary software change in Patent Center that the USPTO implemented on Wednesday, November 1, 2023. On that day, the USPTO made a software change that pukes on any uploaded PDF file that … wait for it … has a file name ending in “PDF”. I am not making this up. Starting on Wednesday, November 1, 2023, the user who is getting ready to upload (to Patent Center) a PDF file that has a file name ending in “PDF” encounters the problem that it is necessary to manually rename the PDF file so that the file name ends in “pdf”. I promise you, I am not making this up!
This is the same Patent Center that for the previous five years had somehow worked just fine in cases where the user had uploaded a PDF file having a file name ending in “PDF”. Somehow the downstream USPTO systems had worked just fine for the past five years, receiving PDF files through Patent Center having a file name ending in “PDF”.
And, not to put too fine a point on it, EFS-Web for the past twenty years had somehow worked just fine in cases where the user had uploaded a PDF file having a file name ending in “PDF”. Somehow the downstream USPTO systems had worked just fine for the past twenty years, receiving PDF files through EFS-Web having a file name ending in “PDF”.
But abruptly, and with no advance notice to USPTO’s customers, on November 1, 2023, the USPTO made a software change in Patent Center so that it would puke on any attempt to upload a PDF file having a file name ending in “PDF”.
Meanwhile, we note that the USPTO did not feel the need to make a corresponding change in EFS-Web on November 1, 2023. To this day, EFS-Web will not puke on an uploaded PDF file having a file name ending in “PDF”.
Here is the rather condescending banner that appears today on the loading page for Patent Center:
IMPORTANT Some filers are experiencing a “File naming convention error” when uploading files in Patent Center. The filename extension must be lowercase (e.g. .docx) to be compatible with other USPTO systems. This has always been the rule and is now enforced in Patent Center. Please go to the “File Naming Convention” page for details.
So now let’s try, as best we can, to enumerate the many ways in which the USPTO’s abrupt change on November 1, 2023 was wrong-headed.
It was abrupt. The USPTO gave no warning of this change. There was not, for example, some announcement a month earlier, letting users know that a month from now, Patent Center would start puking on attempts to upload a PDF file having a file name ending in “PDF”. Such an advance announcement, had the USPTO had the courtesy to provide it, would have permitted users to start making modifications to their various upstream systems so that newly created PDF files could be crafted to have file names ending in “pdf” instead of “PDF”. Instead the way that users learned of this software change in Patent Center was by encountering the puking.
Never mind that apparently to the extent that it was somehow important to fiddle with the last three characters of a file name, the USPTO apparently had known about this since the year 2002, given that the USPTO says “this has always been the rule”. So there really is no excuse for the USPTO failing to give, say, a month’s warning on this very disruptive change.
Many user systems automatically use an uppercase “PDF” extension. Yes, many document processing systems and content management systems used by many USPTO customers have, for decades, automatically created any new PDF file with a last three letters of “PDF”. Indeed this is a very natural thing to do given that the name that we give to this type of file is … wait for it … the capitalized initialism “PDF”.
It broke the “100%” claim. USPTO had been claiming, since July of 2023, that Patent Center supposedly provided “100%” of the functionality of PAIR and EFS-Web. (The claim was untrue, blog article, but the falsity of that claim is not the point of the present blog article.) Part of the “100%” functionality is the promise that any to-be-uploaded that would not make EFS-Web puke would thus likewise not make Patent Center puke. And indeed from the year 2018 until November 1, 2023, it was true that any to-be-uploaded that would not make EFS-Web puke would thus likewise not make Patent Center puke. But starting last Wednesday, all of a sudden any user who had the temerity to try to upload a PDF file with a file name ending in “PDF” would find that although EFS-Web would not puke on it, Patent Center would puke on it. This aspect of the “100%” functionality promise got broken on November 1, 2023.
There was no consultation with users. Other patent offices routinely consult with users before making disruptive changes like this. If some change is being considered that would make it harder, rather than easier, to use a patent office system, common courtesy suggests that users ought to be consulted. The USPTO did not consult with users about this very disruptive change.
For those who up until now have been using EFS-Web, this is a stupid time to do something that makes it harder to use Patent Center. The USPTO plans to shut down both PAIR and EFS-Web about fifteen hours from now. Again, as mentioned above, these days almost half of all patent e-filing activity is still being done in EFS-Web. The users who are still using EFS-Web will suddenly find themselves forced to use Patent Center. It would already have been a challenge for such users to be forced to use Patent Center, but now it will be even a more difficult challenge, given the sudden software change that makes Patent Center puke on attempts to upload PDF files that have file names ending in “PDF”, considering that such files do not make EFS-Web puke.
For those who have been mostly already been using Patent Center, this is a stupid time to do something that makes it harder to use Patent Center. The users who have indeed mostly migrated to Patent Center already face many challenges, such as the over one hundred Patent Center bugs that remain outstanding, many for over three years now. Starting tomorrow, Patent Center will be harder to use because the traffic will almost double. (The almost half of all filing activity that has been done in EFS-Web until now will suddenly get forced onto Patent Center, thus almost doubling the load on Patent Center.) And layered on top of this will be the completely unnecessary further burden of renaming one’s PDF files so that the file name ends with “pdf” instead of “PDF”.
The Electronic Business Center will be unnecessarily burdened even more than it already was going to be burdened. It was already going to be an enormous strain on the EBC to deal with the flood of telephone calls from users who, starting on November 8, would be forced to migrate from EFS-Web to Patent Center. The USPTO’s needless imposition on November 1 of the need to rename one’s PDF files to overcome the puking based on “PDF” at the end of the file name has already led to an increase of call volume to the EBC. This software change on November 1 will continue to cause its burden on the EBC during the flood of calls that will happen on November 8.
The banner is condescending. The banner (quoted above) tries to “blame the user” for the puking. The banner talks about how “this has always been the rule”. By this, the USPTO means that supposedly there was some documentation for EFS-Web, in 2002, that said that it was forbidden for a to-be-uploaded PDF file to have a file name ending in “PDF”. And maybe the USPTO is not lying about the existence of such documentation.
But even if the USPTO is not lying about the existence of such documentation, the banner is condescending. USPTO’s own actual conduct, for the past twenty years with EFS-Web, is that a PDF with a file name ending in “PDF” is perfectly acceptable for upload. USPTO’s own actual conduct, for the past five years with Patent Center, is that a PDF with a file name ending in “PDF” is perfectly acceptable for upload. The insulting premise of the statement that “this has always been the rule” is, apparently, that every user of Patent Center in 2023 is charged with:
-
-
- having read the EFS-Web documentation in 2002, and
- having somehow noticed in 2002 the existence of an (unenforced) requirement that a PDF file to be uploaded to EFS-Web is not permitted to have a file name ending in “PDF”, and
- having committed this bit of information about the EFS-Web documentation to memory, not forgetting it during the passage of over two decades, and
- being in possession of a high enough quality crystal ball to be able to predict that without warning, over twenty years later, the USPTO would, in a different system that is not EFS-Web, decide abruptly to start enforcing this requirement that a PDF file is not permitted to have a file name ending in “PDF”.
-
Not only that, the user would be charged with somehow figuring out that a piece of documentation from twenty years ago about what the user could or could not upload into EFS-Web would later be construed by the USPTO as if it had been a piece of documentation for Patent Center, which did not exist twenty years ago. Yes, the user would be charged with somehow being able to predict that some requirement for how to use EFS-Web would later turn out to be a requirement for how to use Patent Center.
The recitation that “this has always been the rule and is now enforced in Patent Center” is at the same time insulting and condescending towards the user.
The November 1 software change cannot possibly actually be necessary. No matter what face-saving excuse the USPTO may offer for making this software change on November 1, the plan fact is that somehow for the past twenty years of EFS-Web, a PDF file having a file name ending in “PDF” has not been a problem for “other USPTO systems”. Somehow, for twenty years of EFS-Web, the half a million patent applications filed by users have managed to work their way through the “other USPTO systems” just fine. Issue fees got paid, patents got granted, and the PDF files having file names ending in “PDF” that had been uploaded in EFS-Web did not cause the system to collapse.
Not only that, but somehow, for five years of Patent Center, the hundreds of thousands of patent applications filed by users in Patent Center have likewise managed to work their way through the “other USPTO systems” just fine. Issue fees got paid, patents got granted, and the PDF files having file names ending in “PDF” that had been uploaded in Patent Center did not cause the system to collapse.
But intead, we as outside observers of the USPTO are being asked to swallow the preposterous notion that suddenly on November 1, 2023, after twenty years of EFS-Web and after five years of Patent Center, things have changed. Suddenly a PDF file with a file name ending in “PDF” will cause the “other USPTO systems” to collapse, starting November 1, 2023. (But only if uploaded in Patent Center, not EFS-Web.) I don’t believe it.
But let’s assume for sake of discussion that it is true. Suppose that it is actually true that an upload of a PDF file with a file name ending in “PDF” would not cause a collapse of “other USPTO systems” if uploaded on October 31, 2023, but supposedly would cause such a collapse if uploaded the next day on November 1, 2023. And we suppose that somehow this change in the consequences of uploading a PDF file having a file name ending in “PDF” from Tuesday October 31 to Wednesday November 1 is a change that applies only to Patent Center, and does not apply to EFS-Web. I don’t believe it, but still for sake of discussion let’s suppose it to be true.
From this it does not follow that the correct next step is to modify the software of Patent Center to make it harder to use. It does not follow that the correct next step is to modify the software of Patent Center to puke on an attempted upload of a PDF file having a file name ending in “PDF”.
No, the correct next step would be to add a line of code in the software of Patent Center to change the offending “PDF” at the end of the user-provided file name to be “pdf”, if indeed this is what is needed to avert the imagined collapse of some downstream systems. Any competent computer programmer trained within the past decade will have learned in school how to use “regular expressions” which provide trivially easy and reliable ways to do such conversions. Many programming languages have built-in “string functions” that can convert an entire string of characters, or selected character positions in a string, from uppercase to lowercase. With either of these approaches (the “regular expression” approach or the built-in “string function” approach), we are talking about a single line of code. We are talking about a software edit that would take at two minutes to key in at a keyboard, but might take less than sixty seconds for someone familiar with the particular project. This kind of fix would have the great advantage of … wait for it … not making Patent Center harder to use.
Seizing upon an irrelevant and long-ignored document as an excuse to do the wrong thing. It is plain to see, from the condescending banner quoted above, what sequence of events must have taken place within the USPTO:
-
- Somebody within the USPTO noticed that every now and then, something would go wrong in some downstream USPTO system because whoever coded that system had not paid enough attention to stuff like upppercase and lower case.
- USPTO people then discussed whether maybe the correct next step would be to fix the programming mistake in that downstream USPTO system so that it would handle correctly the same variations of of file name formats that all of the upstream USPTO systems were already handling correctly.
- USPTO people then said “that would be real work”. USPTO people then said “we don’t like doing real work”. So the idea of fixing this incorrect handling of upper and lower case in the downstream USPTO system that was handling upper and lower case incorrectly, was discarded.
- Some USPTO person said, “wait a minute! Can’t we make this problem go away by shifting the problem to the users?” And sure enough, if we push the problem further and further upstream, so that it is the users who must deal with the problem going forward, then we save the USPTO people from having to do anything at all to fix the original programming mistake.
- But then some other USPTO person said “wait a minute! Isn’t it wrong to shift some internal USPTO system to the users, particularly when we acknowledge that it was USPTO people, not the users, who created this problem?”
- And then some USPTO person went rummaging through documents from twenty years ago, documents that predate Patent Center by fifteen years, and found one document that imposed a requirement in a different system (EFS-Web) that the user was not allowed to use uppercase letters in the last three positions of a file name. The requirement had never been enforced, not ever even once with EFS-Web, for twenty years and to this day. The requirement had never been enforced in Patent Center, for five years. But there it was, in this long-ignored document from twenty years ago. And this, dear reader, is the excuse the USPTO offers for this way of dealing with things. The way to deal with some programming mistake made by USPTO people in some downstream USPTO system is not to actually do the tedious work of correcting the programming mistake. Instead, the way to deal with it is to shift the burden to the users, even though they were not the people who made the programming mistake. And the excuse for doing this burden-shift is the discovery of convenient face-saving language in some twenty-year-old document that has never been enforced.
Do we fix a problem by fixing it once, or by making somebody else deal with the problem repeatedly? We are now able to compare the burden imposed on various stakeholders depending on whether one approach or the other is followed. Let’s compare:
-
- A first approach is to incur a one-time cost. Yes it takes some amount of money and effort to fix whatever mistake some USPTO person made when they failed to pay enough attention to case sensitivity in that downstream USPTO system. But that money cost and effort cost is only a one-time cost. Once the mistake gets corrected, that is the end of it.
- A second approach is to incur a recurring cost, in perpetuity. This is the approach where perhaps a half a million times per year, a USPTO user must spend time fussing with renaming a PDF file so that it ends with “pdf” instead of “PDF”. This cost is incurred over and over again, forever.
So which approach is the correct choice? I think most neutral observers would say that the correct choice is, suck it up and fix the programming mistake (the one-time cost) so as to avoid the wholly unnecessary recurring cost that goes on forever.
But the alert reader may be able to guess where this is going. The decisionmaker here is not a neutral party, striving toward the right answer or the best choice. The decisionmaker here is inside the USPTO. And, not to put too fine a point on it, the one-time cost to fix the problem is a cost that the USPTO would incur, while the perpetual recurring cost is a cost that is incurred outside of the USPTO. The USPTO does not have to pay that perpetual recurring cost.
Ignoring fairness. I have been in retail stores where a sign is posted that says something like “if you break it, you pay for it”. There is some general notion of fairness that I think many of us have learned from our earliest days, that if some bad situation has arisen, the fair thing is that if we know who caused the bad situation, maybe that is who ought to remedy the bad situation.
Where does the blame lie for the programming error in that downstream USPTO system? Are USPTO’s users somehow to blame? I don’t think so. Call me old-fashioned, but if the blame lies somewhere within the USPTO for a programming error in some downstream USPTO system, then I think the fair approach is that the USPTO ought to be the one to bear the cost of fixing it. I do not think the fair approach is to make Patent Center harder to use, and to impose this harder-to-use Patent Center burden (over and over again) on blameless users.
- So, to sum up:
-
- This change that makes Patent Center harder to use was probably not necessary at all.
- Even if this problem about uppercase and lowercase was, supposedly, a problem that really did need fixing, it was dumb to address it by making Patent Center harder to use, when it could have been addressed in another way that would not have made Patent Center harder to use.
- Even if we assume for sake of discussion that there was no choice available to the USPTO other than to make this change that makes Patent Center harder to use, this was absolutely the stupidest time to make this change, a mere week prior to the profoundly disruptive event of shutting down PAIR and EFS-Web.
- Even if we assume for sake of discussion that there was no choice available to the USPTO other than to make this change that makes Patent Center harder to use, there is no excuse for doing it abruptly, without warning. If indeed there was some need, starting in the year 2002, for PDF files to refrain from having file names ending in “PDF”, then the USPTO could have given a warning on, say, October 1, 2023, of the impending change that would require users to change their own internal process flows.
I’m astonished by the wrong-headedness of the PTO’s response. It feels like someone found the “must be lowercase document” and decided the software had to be changed without warning to match the document. No one asked, “Are we sure changing the decades-long behavior is a good idea, particularly at the last minute?”
Just wow
I’m such a dork, but this is making me laugh out loud: the USPTO guidance on file names says you can’t include such classics as “LPT1” in your file name. Sure enough, Patent Center is now enforcing this rule (I never checked before, but I doubt that it was previously).
Seems like some project manager got their hands on the file naming conventions and demanded that a coder follow it to the letter. If in 2023 the presence of “LPT1” in a filename could cause an issue in the bowels of USPTO systems, the mind reels…