4 ms·
Patent agent here. At Google we investigated patents and their possible invalidity almost every day. Claim language is obtuse because it's for lawyers, not lay
by AlbertCory 5y ago
Patent agent here. At Google we investigated patents and their possible invalidity almost every day.
Claim language is obtuse because it's for lawyers, not laymen. An engineer can write the description, but only the lawyer can write the claims.
It's immensely complicated, but if Unity does someday assert this patent against someone else, then you all are helping the defendant when they have to defend themselves. At least, I hope the defendants find this discussion.
The defendant's standard tactic is usually to file for IPR, or PTAB reexamination of the patent, which is like a trial without a courtroom. The cost figure I recall is $500,000; maybe it's different now.
If the IPR succeeds, then Unity is SOL. There was a Supreme Court case recently where some plaintiffs tried to get the whole IPR system thrown out. They failed, as far as I know.
Lastly, the whole back-and-forth with the PTO is in the public record, if you know how to find it (and if you think reading claims is difficult, wait 'til you try reading a prosecution history!)
- akersten 5y ago> An engineer can write the description, but only the lawyer can write the claims. If anyone needs a one-sentence rationale why the entire concept of a software patent should be invalid, here it is.
- jlokier 5y agoNot just writing them. Apparently only a lawyer should read the claims too: > Lawyers are not very expensive compared to the consequences of trying to read it yourself (https://news.ycombinator.com/item?id=28451317 https://news.ycombinator.com/item?id=28451317)
- AlbertCory 5y agoThe same is true of any patent, even the hardware patent in Flash of Genius [1] [1] https://www.imdb.com/title/tt1054588/?ref_=nv_sr_srsg_0 https://www.imdb.com/title/tt1054588/?ref_=nv_sr_srsg_0
- jlokier 5y ago> The cost figure I recall is $500,000; maybe it's different now. That's the standard tactical defence fee? I would guess that kind of money is very far above the means of most open-source developer defendants. And wouldn't a larger, closed-source business rationally choose to pay Unity (say) $400,000 extortion fee instead of taking it to court with the associated uncertainty?
- AlbertCory 5y agoThat's not court; that's an IPR [1]. In court you have a jury of ordinary citizens; in an IPR you have experienced patent people. As for defense strategy: you're right, but Unity might want a quick hit. They don't want a lengthy proceeding where they might lose. [1] https://www.uspto.gov/patents/ptab/trials/inter-partes-review https://www.uspto.gov/patents/ptab/trials/inter-partes-revie...
- jlokier 5y agoTo the defendant who doesn't have $500,000 or $400,000 (maybe $1-10k is available), whether it's court or IPR seems like splitting hairs. It doesn't matter what kind of legal proceeding, because they can't afford to participate in any legal proceedings at this level anyway. They also can't afford to pay for more than superficial advice. What I'm getting at is: $500k is more than most people's entire assets, and more than the assets in many small businesses and organisations. With software patents like this one, most infringements are likely done by the "little guy", who can end up threatened when they publish code, or have to operate in fear of that happening, if they know about it. They simply cannot challenge the patent through any legal method that has a fee like $500k. If the rational closed-source company will pay Unity's hypothetical fee because it's less than the IPR fee, and the open-source defendant doesn't have any choice because they can't afford any of the options... who is left that is both motivated and capable to use the IPR procedure? In the case of this particular patent by Unity, the techniques are things people thought of decades ago, but like most "inventions" of this kind, it's only when you have a specific project (typically paid) where they'd come in useful that you typically get around to using it. Nobody has time to document and explain every little "invention of the day" every time they think of a new technique just to feed the prior art corpus. I tried that once. When you've come to see how low the threshold is for patentability in software (and to be honest, in some digital hardware too), it's seems literally impossible to write down every patentable insight as fast as they come once you're on a roll; you'd need a support team just to do the documenting. It's upsetting to think up a technique, flesh it out, not have a use for it for 10 years (you're busy, right?), then find one day when you want to use the fruits of your own figuring out 10 years prior on a project, it cannot be used any more. Even more so when it follows naturally from techniques dev communities have been discussing for years in the field. (A number of us also have stories of publishing a sketch of a technique on some forum, only to find it patented later. It happened to me with a texture-mapping graphics technique in the 90s, where I found a later patent actually cited my Usenet post; in practice that meant I couldn't develop it further because the obvious next steps were blocked.)