Skip to content

Patent Example with LRA

Jump to bottom
Amelia Eiras edited this page May 12, 2021 · 5 revisions

DISCLAIMER: The information provided in this document does not, and is not intended to, constitute legal advice; instead, all information available inside are for general informational purposes only. Readers of this document should contact their legal counsel to obtain advice with respect to any particular legal matter.

I thought it might be good to use Long Running Actions (LRA) as a use case for our Patent Protection options.

WAY TOO SIMPLISTIC PATENT OVERVIEW

Here is a recap of my overly simplistic description of legal terms thrown around software:

  • Patent: ownership of an idea or concept of a thing that could be
  • IP: ownership of the actual created thing made real
  • Trademark: ownership of the name you gave the created thing

Examples

  • Patent: conceptual things like algorithms, techniques, etc
  • IP: material things like source code, binaries, docs, specification text, etc
  • Trademark: branding and identity things like names, logos, slogans

In the game of legal rock-paper-scissors, patents always win. It basically allows you to argue, "I don't care that you wrote it from scratch (IP) or used a different name (trademark), I own the idea (patent)." The primary patent fear is that we put a lot of work into generating IP (a specification) around an idea (potentially patented) and someone implements that specification and is then sued (potentially by a contributor) for using that idea without permission.

We collect patent permissions (grants) from contributors through the various contributor agreements and also through successful Progress Reviews and Release Reviews votes, which are public and intentionally cast by the entities that are most likely to own patents. When we distribute that permission to implementors is what the two options are about.

Long Running Actions (LRA) was a particularly long running spec as Emily very cleverly joked. Here's how each option would play out with that spec:

OPTION A: The Eclipse Foundation Intelectual Property Policy Page 4 called Implementation Patent License

LRA was started in May 2018 and people begin implementing it. Any contributor who may have contributed some of their own patented ideas could potentially file a lawsuit against anyone implementing LRA up until the first Progress Review. Per the EFSP a specification cannot go longer than one year without a Progress Review or Release Review, so as LRA was not ready for final would have been forced to complete a successful Progress Review no later than May of 2019. Following the successful completion of that Progress Review, people implementing LRA and distributing that implementation for feedback can feel safer they won't be sued by another participant for patent infringement. At least not over anything contributed up until May 2019. When work continues and new potentially patented ideas are contributed, there is some potential for an infringement lawsuit over those new ideas. Again the maximum allowed time without a review is one year, so potential to be sued over the post-May-2019 ideas would go away again in May 2020 when another Progress Review is forced. After the successful May 2020 Progress Review vote, all implementors would be protected for the full two-years worth of potentially patented ideas.

Failed specification scenario: If the specification died after the May 2020 Progress Review and never reached a final, anyone who leveraged the LRA spec is still legally ok. Of course any ideas contributed to LRA after the May 2020 Progress Review would not have the benefit of a vote to "lock in" permissions on those ideas and people would be wise to only leverage the LRA specification as it existed after the last Progress Review. On that note, we are allowed to have Progress Reviews more frequently than one year and can have them as we feel necessary; perhaps after a rather large contribution of new ideas people want to implement and evaluate.

In short, it is legally safe from a patent perspective to implement any specification that has gone through a successful Progress Review or Release Review.

Pros

  • If we want to encourage people to implement specifications like LMR that are in development, collect feedback from users and contribute that feedback before a final (and not risk getting sued for patent infringement by another contributor), then this is would be a good option

Cons

  • We don't get to use risk getting sued for patent infringement by another contributor as motivation for being LRA TCK compliant. We'd have to rely on other mechanisms such as ability to use the MicroProfile certified logo (in the context of a future platform spec), consumer pressure, etc.

  • If you contributed a patented idea to LRA, you don't have the ability to say "people should only get permission to use this if they're fully LRA compliant."

OPTION B: The Eclipse Foundation Intelectual Property Policy Page 4 called Compatible Patent License.

LRA was started in May 2018 and people begin implementing it. Any contributor who may have contributed some of their own patented ideas could potentially file a lawsuit against anyone implementing LRA up until the specification has passed a Release Review and the implementation has passed the corresponding LRA TCK. The Progress Reviews in 2019 and 2020 would still provide patent permission "lock in" from the contributors to the Foundation and Working Group itself, but that would not flow to the LRA implementors (even if they were in the working group) and they would still have risk of a patent lawsuit from any contributor for the full three years that LRA was under development and continuing on afterwards until such time they were compliant with the LRA TCK.

Failed specification scenario: If the LRA specification died after the May 2020 Progress Review there and never reached a final anyone who leveraged the LRA spec was already at risk of lawsuit and that risk would remain perpetually until they removed all LRA-related code from their product.

In short, it is legally safe from a patent perspective to implement any specification that has gone through a successful Release Review and for which you are fully LRA TCK compliant.

Pros

  • We can use risk getting sued for patent infringement by another contributor as motivation for being LRA TCK compliant. This would be in addition to other mechanisms as ability to use the MicroProfile certified logo (in the context of a future platform spec), consumer pressure, etc.

  • If you contributed a patented idea, you have the ability to say "people should only get permission to use this if they're fully LRA compliant" and take legal action against anyone implementing the specification during develop phase and additionally after Release Review if they do not pass the LRA TCK.

Cons

  • As "another contributor" is likely to be also a competitor there can be reasonable fear that could discourage early implementations. There can be additional fear if there is doubt the specification can reach a final release.

  • Even if you start implementing after a Release Review there there may be a period of several months before you are fully compliant, during which there is still risk of getting sued for patent infringement by another contributor.

This was longer than I thought it would be, but I guess that's only fitting the LRA theme :)

I hope there was at least some use to this. As long as this email is, it is still not complete. Others are encouraged to point that out by writing the rest :) If there are major flaws, that is good too. All feedback would be seen as a positive contribution and progress.

-David

Clone this wiki locally