Articles about EFS:
General discussion of EFS. EFS is a system which will save the Patent Office a lot of money. Patent applications submitted via EFS save the Patent Office from a lot of scanning and retyping. The filing receipt and indeed the issued patent will all be based on the information provided electronically to the Patent Office in the EFS filing. Filing receipts will be more often correct on the first try, saving the Patent Office some of the cost of generating and mailing corrected filing receipts. Patents won't need as many certificates of correction and it will cost the Patent Office less to issue the patents, because the typing will have been done by someone other than Patent Office personnel.
From the practitioner's point of view, the only pressing reason why you must learn EFS is that if you are filing a redacted specification the Patent Office won't take it on paper, but will only take it via EFS. The usual reason for filing a redacted specification is that you have filed a US application and later performed foreign filings which contained less information than was in the original US application.
The other motivations for learning how to do EFS are perhaps less pressing. It will save you a trip to the Post Office and might let you get today as a filing date even after the Post Office has closed for the day. And it should result in an issued patent that contains fewer errors than a patent that is typeset by the Patent Office from a paper filing.
How EFS works. The EFS system makes use of a standard called XML (extensible markup language). You prepare the patent application coded in XML, and then you send the XML document to the Patent Office via software called ePave. ePave gives you an electronic filing receipt with a filing date and application serial number.
Coding your patent application as XML can be done in any of several ways. The Patent Office provides software called PASAT which, used together with Microsoft Word, permits you to use Word to create the XML document. The Patent Office also provides a WordPefect XML Template which, used together with WordPerfect, permits you to use WordPerfect to create the XML document.
One approach is to write the patent application using the word processor that you usually use, and then copy and paste the paragraphs of the application from your regular word processor into Word/PASAT or from your regular word processor into WordPerfect with the XML Template. Another approach is to write the application from scratch in one of the two authoring systems (Word/PASAT or WordPerfect with the XML Template).
What is XML, exactly? Sooner or later, anyone who is reading the EFS materials will wonder what exactly XML is. The EFS materials make it seem like some mysterious thing that no mere mortal could possibly understand. The clear message of the EFS materials is that without PASAT or the WordPerfect XML system, there is no way you could possibly generate a patent application that is in XML form. But you can surely tell from the way I am introducing this topic that I feel XML is not complicated or mysterious at all.
XML is nothing more nor less then a methodical insertion, into your document, of "XML tags". Each tag is denoted by angle brackets. The tags are nearly always provided in pairs, rather like parentheses or square brackets. The second tag in a pair will start with "/". Whenever you see a particular XML tag that starts something, you can expect that later you will see an XML tag that ends the something. In the XML format defined by the US Patent Office each patent claim, for example, starts with <claim> and ends with </claim>. Tags can be nested. For example, the patent specification starts with <specification> and ends with </specification>. Between those two tags are the entirety of the specification, including lots of other tags. There are tags that define the title of the invention (<title-of-invention> and </title-of-invention>). The title goes between those two tags. There are tags that define the background of the invention (<background-of-invention> and </background-of-invention>). The background of the invention goes between those two tags.
Those who have done HTML coding for web sites will immediately feel quite familiar with XML. In HTML, as in XML, the tags are denoted by "<". In HTML, as in XML, the tags usually come in pairs, with the second tag of each pair denoted by a "/" in the tag. In HTML, as in XML, if you were to misspell a tag, or delete one tag of a pair without deleting its mate, you would end up with a defective file. In HTML, as in XML, if you were to put a tag in the wrong place (e.g. a <title-of-invention> tag between the tags <claim> and </claim>, you would end up with a defective file.
Within your law firm, if you are trying to decide which person should be
the first to suffer through learning how to do EFS, you should consider
choosing someone who has done HTML coding for web sites. Such a person will be
right at home with the tags that appear in EFS XML documents.
Randy L. Milbert has created a couple of introductory tutorials that may be helpful to brand-new EFS users. You can see:
How do you submit an inventor's Declaration with filing, if the specification is changed by XML formatting after the inventor has reviewed it? It is drilled into every practitioner ever since studying for the patent bar exam that a patent application may not be changed after the inventor has reviewed the application and has signed the Oath or Declaration. If even one word of the application changes, the practitioner who wants to avoid being disbarred will immediately shred the Declaration so as to avoid breaking the law by submitting a Declaration that relates to a version of the application differing from that which was actually filed with the Patent Office.
This raises the question of how a practitioner can possibly include an inventor's Declaration with an EFS-filed application. The inventor reviews the specification, signs the Declaration, and then ... and then the practitioner adds XML tags using PASAT or the WordPerfect authoring tool, and processes the application further by entering it into ePave (thus converting it into a ZIP file). And then sends the application to the Patent Office, having transformed it several times from the format in which the inventor had reviewed the application.
Mr. Ross Dannenberg of Banner & Witcoff, Ltd. asked the Patent Office as to whether the reformatting performed as a part of EFS filing require that a new Declaration be signed, and here is the response he received:
Mr. Dannenburg, I checked with our system developers, and they agree that a FAQ on this topic would be useful. We started the process for developing the FAQ. I'll be writing the response to that issue in due course. The response will indicate that the USPTO considers the minor re-arranging of the parts of the application to not be a change in the content of the application. The applicant will be careful to avoid adding any new matter to the application in the process, except for headings and other organizational indicia needed to fit the EFS requirements for submission. Best regards, Jay Lucas Sr., Legal Advisor, USPTO
Ways to generate TIF files. Michael Dryja (mike@dryjapat.com) says one way to generate TIF files is to use a "TIFF printer driver." This way, you "print" to a TIFF file from any program. He says he uses the one by Informatik, available at http://www.informatik.com/tiffwork.html. He says his files are around 40K in size.
David Hendrickson & Associates, a patent drawing firm, has prepared a very helpful tutorial describing their experiences using several approaches for generating TIFF files for use with EFS.
Here's how I have generated TIF files. I use a Paperport Vx scanner which is connected to a computer. I feed the figures into the scanner, and it creates files which are visible in the Paperport application which is running on the computer. I can then export these files as TIF files. The Paperport software permits the user to configure the export to a desired resolution (such as the 300 dots per inch required by the Patent Office) and to a desired compression (such as the Group 4 compression). Unfortunately, I find that the moment I insert such a figure into my PASAT/Word patent application, it leads to a problem that PASAT/Word will crash the moment I try to export XML or preview the application in a browser.
The only way I have been able to overcome this is to open Wang Imaging (which comes standard with Windows NT). I use Wang Imaging to read in the TIF file, and then I write out the TIF file again. While there is no reason why this should make any difference, it does. Inserting a post-Wang TIF image into my PASAT/Word patent application does not give rise to the export problem. It seems that PASAT/Word has very finicky taste and will not accept every TIF image, even if the TIF image fully complies with the stated requirements for TIF images.
Another way to make TIFF files. Here's another way to make your drawings into TIFF files. Get a free account with j2 Global Communications. They will assign you a fax number that is personal to you. Faxes sent to that fax number will appear in your email inbox as TIFF files. The TIFF files will, unfortunately, fail in two different ways to comply with the USPTO's picky requirements -- they will be 204 dots per inch while the USPTO requires 300; they will be multipage TIFF files while the USPTO requires that each page be its own TIFF file. Thus you will be stuck using some other software (such as the Wang Imaging that comes with Windows) to read in the TIFF files and write them out as single pages with 300 dots per inch.
One way that the j2 approach can be particularly helpful is when the client is going to fax the figures to you. Normally what you would do is to receive the fax from the client and run it through an image scanner, which would degrade the quality of the figures. If you have the client send the faxes to your j2 fax number, you can skip a scanning step and thus preserve the image quality.
Testing the characteristics of a TIF image file. A Patent Office person kindly provided a program to me called Tiffinfo which can be used to determine the characteristics of a TIF image file. It runs from the DOS command prompt, and takes as an argument a file name (optionally including path) of a TIF file. Here is a typical report:
C:\EFS>tiffinfo d:\sensorsolutions\p001\fig15.tif d:\sensorsolutions\p001\fig15.tif: Warning, unknown field with tag 32934 (0x80a6) ignored. TIFF Directory at offset 0x8 Subfile Type: (0 = 0x0) Image Width: 2300 Image Length: 1296 Resolution: 300, 300 pixels/inch Bits/Sample: 1 Compression Scheme: CCITT Group 4 Photometric Interpretation: min-is-white FillOrder: msb-to-lsb Software: "Oi/GFS, writer v00.06.00P, (c) Wang Labs, Inc. 1990, 1991" Orientation: row 0 top, col 0 lhs Samples/Pixel: 1 Rows/Strip: 110 Planar Configuration: single image plane Group 4 Options: (0 = 0x0) C:\EFS>If you have trouble with PASAT crashing when you insert a TIF image and then try to export as XML (or view in a viewer), then use this utility to study the inner workings of the offending TIF file and provide the results in a posting to the discussion group. This should help users to work out what there is about some TIF files that would cause PASAT to crash.
Tips and tricks for PASAT. For many users, the Patent Application Specification Authoring Tool (PASAT) isn't easy to set up or use.
PASAT crashes and loses all your work. I happened upon another bug in PASAT. Suppose you create a patent specification in PASAT. The smart thing to do, of course, is to save your work frequently so that if PASAT crashes, you will only lose the work that you have done since the last time you saved your work. For example, PASAT often crashes when you try to export XML or preview in XML.
So, suppose you wish to save your work in PASAT. You click "file" and "save" or "save as", and then you type in the path to which the file is to be saved. And suppose that you type in a path that points to a subdirectory that does not exist. This could happen, for example, if by mistake you misspell the name of the subdirectory. What will happen next is that PASAT will crash, and will lose all your work. Of course, what PASAT would do if well designed is display an error message and then allow you to try again, without crashing and without losing all your work.
The way to protect yourself against this type of crash is to make sure that you never type in a subdirectory name -- only "browse" to it by pointing and clicking in the subdirectory browsing screen.
The other problem is that during a PASAT editing session, you will want to save your work repeatedly. And with a newly created specification, each time you try to "save," PASAT will force you to mouse around and retype everything each time to wish to "save." PASAT apparently forgets everything about where you saved the file, each time you save it.
The one way to solve this problem that PASAT forgets where you saved the file and keeps asking, again and again, where you want to save it, is this -- when you have saved the file once, close it and reopen it. PASAT will then remember where the file is, and when you repeatedly save it again (as is prudent to do), PASAT will not force you to retype everything about the file's location again and again.
Getting your Microsoft Word to work.The first problem is that you can't just use any old Microsoft Word. You have to have Microsoft Word 97 and it must be patched with two "service releases" from Microsoft.
You might get lucky and find out that your Microsoft Word already has one or both of the service releases. To find out, open Word and click on "help" and "about". If you see "SR-1" then you have Service Release 1, and all you need to do is install Service Release 2. If you see "SR-2" then you have Service Release 2, and you can start using PASAT. If you don't see SR-1 or SR-2 then you need to install Service Release 1 followed by Service Release 2.
You will find the following articles to be of interest:
Installing Service Release 2 often fails. It fails so often that Microsoft has a discussion on this subject, entitled SR-2 Patch Fails Even Though SR-1 Has Been Installed. Quite often the only way to get to Service Release 2 is to uninstall Microsoft Word completely, and then reinstall it, followed by Service Release 1 and Service Release 2. Nothing worked for me, so I finally gave up and called Microsoft tech support at 1 (425) 635-7056. I was pleasantly surprised that they did not charge me for the call and actually got me to a functioning Word with the two service releases.More PASAT tips. A reader reported a problem:
I still haven't been able to get PASAT to work. Any idea why (when I load PASAT) I get a Microsoft Forms error box saying: --Could not load some objects because they are not available on this machine.-- I get the error box twice, and then Microsoft Word loads. I don't see "New Specification" option under "File", but I do see the options of "Load Specification", "Save Specification", "Import XML Specification" and "Export XML Specification." I uninstalled Word 97 and then installed SR-1 and SR-2 and then reinstalled PASAT. Still didn't work.Barry Gaiman suggested: "You need to install the Windows imaging component." This made the error message go away, but the reader was still unable to see the option of "New specification." Julie R. Schwartz then suggested the following:
Here is something to try to make sure all your tools options are accessible. Exit out of PASAT and in normal WORD under Tools, Templates and Add-In, make sure all "efs applications.dot", "S4TempTemplate0.dot", and "S4TempTemplate1.dot" are checked. Go back to PASAT and see if that helped. I had the same problem, and I believe PASAT needed a "tool bar" fix.Using PASAT. To use PASAT, you must first get your Microsoft Word updated to SR-2. Then click on the PASAT icon. Microsoft Word will then load. You can start authoring your document, following the instructions in the PASAT Authoring Manual.
Inserting Dependent Claim References. I have had great troubles trying to insert dependent claim references. From the documentation, the impression given is that you would insert a "dependent claim reference" into each dependent claim, identifying the claim from which it depends. The idea is that if the independent claim were to get renumbered, the dependent claim would automatically get its reference updated to the new number that is on the independent claim.
I must say that my impression is that this portion of the PASAT user manual is extremely poorly written. I wonder if others have the same impression?
Example. Suppose claim 10 depends from claim 9. So we insert a "dependent claim reference" into claim 10, and as a result, claim 10 says "the system of claim <dependent-claim-reference>stuff</dependent-claim-reference> wherein the framus is connected to the clevis." The idea is that the "stuff" follows claim 9, no matter what new claim number it may some day receive. If claim 9 some day becomes claim 11 (because two new claims were added above it) then PASAT would be clever enough to (a) renumber claim 10 to be claim 12 and (b) keep track that claim 12 now depends from claim 11, and will be printed henceforth as depending from claim 11.
The problem for me is that sometimes when I follow the instructions to enter a dependent claim reference, I get an intelligible claim number such as "<dependent-claim-reference>5</dependent-claim-reference>", and others times when I follow the instructions, I get "<dependent-claim-reference>[Claim Reference]</dependent-claim-reference>". A USPTO person admitted that she also gets inconsistent results like this when she tries to insert dependent claim references. She says don't bother to try to insert the references. She says just give up and hand-type the claim numbers into the dependent claims.
Right-clicking. Maybe right-clicking is hidden away somewhere in the PASAT documentation, but if it is there I cannot find it. What I suggest you do, whenever you are inserting things into your patent specification, is simply position the cursor at the place where you are planning to insert something, and right-click there. The right-click will bring up a menu of things that you are permitted to do there. If the thing you want to do (e.g. insert a figure or a new claim or a dependent-claim-reference) doesn't appear on the menu, then you can't do it there. Try positioning the cursor somewhere else and right-click again, and maybe you will be offered the menu choice that you are looking for.
Viewing tags. I recommend that whenever you are using PASAT, you should keep tags in view. To turn on tags, click "view" and click "view tags". This will allow you to see the XML tags. You can then watch closely to see what happens, for example, when you insert another claim or insert a title. As you gain experience with PASAT, you can anticipate where is the right place to insert something new. For example, you do not insert a new claim between the tags <claim> and </claim>. You insert a new claim after a </claim> tag and before the next <claim> tag.
If you always keep tags in view, you can more intelligently figure out where to right-click to insert a new thing. There is no sense inserting a new claim within an existing claim, for example, and the tags and right-clicking will make this clear to you.
Exporting XML. As you know, part of the electronic filing system is that you (a) use one of the authoring systems to generate an XML file and (b) use the ePave software to send the XML file to the Patent Office. If you are using PASAT, then for this to work, of course you need to be able to generate an XML file. The way PASAT works is that it initially creates a file with an extension "s4w". Then after you have gotten everything done, and are finished revising the document, you click on "file" and "export XML specification" and PASAT is supposed to generate the XML file.
I tried authoring a very simple patent specification and exporting it as XML, and it worked fine. Then I tried a real patent specification, one with real claims and figures, and ended up with an "s4w" file that crashes Microsoft Word every time I try to export the XML specification. Have others had such experiences?
Michael Dryja offers the following comment on pasting claims into PASAT/Word:
I use Word's paragraph-numbering and cross-reference tools to automatically number claims, and cross-reference claims for dependent claims. If I try to paste these into PASAT, I get gobbledy-gook. I first paste them into NotePad, then repaste them into a new document in Word, and then paste them into PASAT, and they're all converted fine.Dealing with the symbol editor. A member of the EFS discussion list says:
We are in the process of preparing our first application for filing using the PASAT software and are encountering a problem. After highlighting the text, that does contain formulas, and pasting it into the "detailed description" section of the template, the view symbol manager says it doesn't recognize certain symbols and ask if you want to run the symbol manager. If you select yes, your workstation freezes. This has been duplicated on more than one machine. If you select no, the Pasat module complains about the degree symbol. I have tried many iterations of inserting the symbol but the results are the same. In addition, I have saved the Word Perfect file as a Rich Text file and then pasted into PASAT to no avail.Here is my comment in response:
This is the problem where you paste into PASAT, and it encounters an unrecognized individual character symbol. The one I run into all the time is the opening or closing ("curly") double quotation marks or opening or closing ("curly") single quotation marks. PASAT plays dumb and claims not to know what those symbols are. Of course what one wishes is that PASAT would say to itself "I know perfectly well these are curly quotes, and I will change them to plain quotes." That is not, of course, what happens. PASAT simply throws up its hands and says "would you like me to bring up the symbol manager so we can decide what to do with these mysterious symbols that I have never seen before?" And if you are gullible enough to say "yes", it promptly crashes and loses all your work. As you correctly note, the only way to proceed is to say "no" to the symbol manager, and then go back and straighten out the unknown characters one by one by starting the symbol editor later. In many cases you can actually find the "correct" symbol, such as the vertical-line type of quotation marks, and paste them in. And yes, I just checked and it turns out there is a degree symbol in the symbol editor.List member Rich Tripp mentions:
This issue and others are addressed in two separate areas of the USPTO's website: http://www.uspto.gov/ebc/efs/faq/troubleshooting.htm and http://www.uspto.gov/ebc/efs/solutions/solutionindex.htmHere is an article on the problem of error messages that talk about expired macros.
Tips and tricks for the WordPerfect system
The WordPerfect XML system only works if you have WordPerfect 9, they
say. I don't have Wordperfect 9. I only have 7 and 8. If anybody has any
experiences to report with the WordPerfect system, please let me know.
Getting your PAIR cryptographic certificate. It is no easy trick getting the Patent Office to issue you a PAIR cryptographic certificate. Mine took a couple of months, and other people have reported asking for their certificate again and again for months before it finally arrived. You have to be nice but persistent. I suggest you call up the patent electronic business center weekly to ask where your certificate is.
Once you receive the certificate, use the USPTO Direct software to test it. See if you are able to gain access to the PAIR reports for your pending patent applications. See our discussions at http://www.patents.com/pair on how to use PAIR and how to get your cryptographic certificate.
Getting your downloaded copy of ePave. At least one user has reported difficulty downloading the ePave software. The USPTO people tell me that they decided to try to make it difficult for people outside the US to get their hands on the ePave software. The reason, of course, is that the ePave software does encryption, and encryption is the same as nuclear weapons, it is illegal to let it out of the country. So the USPTO checks your IP address and does a reverse domain name lookup (also called in-addr-arpa) to map the IP address to a domain name, and they look to see if the mapped domain name appears to be in the US. And if there is no in-addr-arpa record for your IP address, they will refuse to let you have the ePave file. (This basically means your system administrator forgot to set up your in-addr-arpa record. In such a case, ask your system administrator to get your in-addr-arpa record set up.) Or if the USPTO can't manage to do a Whois on the domain name that results from the in-addr-arpa lookup, then they will refuse to let you have the ePave file. So you should ask your system administrator to tell you what domain name is mapped from your IP address, and you should check to see if it is in Whois. I mention this because the user who inquired had an email address that ends in "dot mil", and many Whois programs can't figure out how to do a Whois on a dot mil.
The usual trick that works fine, whenever one is having trouble downloading encryption software (PGP, ePave, 128-bit Netscape, etc.) is to download it from an AOL or Compuserve account. Such accounts always have functioning in-addr-arpa records for their IP addresses, and they always "look" as if they are in the US. I assume that anyone outside the US who wants to obtain ePave will manage to do it by using an AOL or Compuserve account.
Tips from Anton J. Hopen. Anton J. Hopen has prepared a page detailing some tips for use with ePave and a page with some tips and tricks. See also Windows Server 2003 Generates ePAVE Validation Error.
Error message: host environment does not support date formatting.
Once you have successfully filed a patent application using EFS, you
will want to be able to print out what you filed. You can see a
discussion of a way to print out what you filed.
Getting USPTO Direct to work. Some users, especially those with firewall protection for their local area networks, have trouble getting USPTO Direct to work.
See this article which talks about troubleshooting USPTO Direct and dealing with firewalls that interfere with USPTO/Direct (and thus interfere with ePave).
Getting USPTO Direct working on a second computer. We have
prepared a short article on the subject of getting USPTO
Direct to work on a second computer.
This page is http://www.patents.com/efs , revised January 25, 2012.