3

I may file a patent in the future that relates to a user interface on smartphones. In the meant time, I am building the algorithm stored in a GitHub repository. My question is:

  1. Can I use this GitHub page history to prove I am the original inventor? I.e. if someone else tries to file a patent before me but is shown to have invented it at a later date than when I started, will the GitHub page be enough to grant the patent to me?

  2. I have trouble understanding the grace period. If I publish an art and have not filed for a patent within the year, does that mean anyone else can file the patent, even though the art verifies that I am the original inventor?

  3. What constitutes an art being "published"? If I create a private GitHub page, and one year later change it to public, is the date of publication the date I created the GitHub page, or the date that I set it to public?

2 Answers 2

2

I did a little digging and found this post:

Register prior art, but not wanting a patent

In case it might help anyone else, I'll answer my own questions for archival reasons. If I got anything wrong, any help correcting would be much appreciated.

  1. Yes, GitHub can be used as a published prior art.
  2. If I wrote the code on GitHub, and I was the first that invented it, no one else but me can file the patent. However, I only have 1 year from the publish date to file that patent. After that year passes, no one can file the patent (I assume its under public domain?)
  3. From the URL I linked, it seems like the GitHub page must be publicly available for it to be considered a published prior art. I'm not sure on the specifics if the repository was initially private, then set to public, then set to private again some time later, then public again. I assume the publish date of the prior art is on the first time it was set to public (so the 1 year grace period would start ticking on that time).
1
  • Not to resuscitate an old thread, but it seems doubtful that the repository would be considered prior art if it was not linked to or otherwise searchable based on its subject matter. However, this may be incorrect.
    – user132162
    Aug 13, 2017 at 16:14
-1

Github is apparently not considered prior art by the USPTO.

See the case of Microsoft patenting RANS while it was already readily published and being in use: https://encode.su/threads/3863-RANS-Microsoft-wins-data-encoding-patent?p=74250&viewfull=1#post74250

I am in contact with EFF regarding this patent, they might help somehow, e.g. finding a lawyer who could make it for a lower cost.

Turns out github is not prior art for USPTO! Patent officers often have no computer science education (granting 20 year monopolies!) I don't understand it, couldn't believe it ... but thousands of programmers publicly sharing only source code seems to have zero protection from anybody copy&pasting their code, ideas into a patent! Please share your new ideas also in a different way like arxiv ... What is non-github prior art for rANS with Markov models?

And also

I have talked yesterday with EFF lawyer and was told that, while they try to change it, for now github is not prior art for USPTO. If going to court, it might change - but it is hundreds of thousands of dollars and years. I was clearly told that using github as prior art e.g. for this reexamination makes no sense.

4
  • This is not correct - if you are saying they don’t look a GitHub yes, probably true. They don’t look at thesis in libraries in foreign counties but if it’s indexed by subject it is prior art. And examiners have expertise and get very good at searching in their narrow fields of assignment. Even read to the bottom of the linked post to see that this random does think GitHub can be prior art.
    – George White
    Oct 13, 2022 at 15:39
  • Have you even read this post, or the linked thread? An inventor has contacted the EFF because Microsoft has patented his invention. The EFF has told him they know the USPTO doesn't consider GitHub prior art. This person has put a lot of effort in keeping this invention patent-free, but failed anyway. See en.wikipedia.org/wiki/…
    – ktmf
    Oct 13, 2022 at 17:33
  • The Wikipedia article says google tried to gat a patent in the subject matter topic and failed. Later Microsoft filed an application in the same technological area, got a rejection, overcame the rejection and the application was allowed. (A very normal occurrence) All of this does not make your answer authoritative. And the text at the original link does have “Any other ideas what realistically could be reached here? I know ... github is prior art ... but getting it to the patent system might be tough ...)
    – George White
    Oct 14, 2022 at 2:43
  • Having read the 3rd party submission - it is a long story not an effective 3rd part submission. Should be — “attached is a prior art document dated ABC that shows claim x is not novel because of what the document teaches on page Y.” The Register article says it would be bad for the world for the MS patent to issue. The patent doesn’t add anything important. — if it is not important no one will implement it and no one will infringe.
    – George White
    Oct 14, 2022 at 2:56

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .