>I think I like the idea, but I can't help wondering if it would have unforeseen consequences.

As I said in a sibling comment, quickie comments on HN should be taken more as mental stimulation and kickoff points for further discussion as opposed to "final bill that has been revised in committee and is going to the floor for a full vote". The details of implementation are certainly critical, and not trivial either! I'm fully in support of thinking through various use cases. But part of why I'm interested in alternate approaches is that they might give us finer grained tools.

>Could this approach undermine the protections afforded by open-source licenses? (IANAL.)

I have actually considered that as well but didn't add it into a quickie comment. If we take the second path of approaches I listed there, then thinking about it all open source software would fall under a special even more permissive class of the tier 3, in that it already has "fair, reasonable and non-discriminatory" licensing for all right? Except that it's also free. The motivation here is the "advancement of the useful arts & sciences" and the public good, so having it be explicit that "if you're releasing under an open source license and thus giving up your standard first, second, and part of your third period of IP rights and monopoly, you're excluded from needing to pay a license fee because you've already enable the public to make derivative works for free for decades when they wouldn't otherwise anyway."

All that said, I'll also ask fwiw if it'd even be that big a deal given the pace of development? I do think it'd be both ideal and justified if OSS had a longer period for free, that's still a square deal to the public IMO. But like, even if an OSS work went out protection (and keep in mind that a motivated community that could raise even a few thousand dollars would be able to just pay for an extra decade no problem, the cost doesn't really ramp up for awhile [which might itself be considered a flaw?]) after 10 years, how much is it worth it that 2016 era OSS (and no changes since remember, it's a constantly rolling window) now could have proprietary works be worth it against 10 year old proprietary software all getting pushed into the public domain far faster? That's worth some contemplation. Maybe requiring that source/assets be provided to the Library of Congress or something and is released at the same time the work loses copyright would be a good balance, having all that available for down the road would be a huge win vs what we've seen up until now.

Anyway, all food for thought is all.

> quickie comments on HN should be taken more as mental stimulation and kickoff points for further discussion

Indeed.

Setting aside variable details like time frames and cost structures which can be debated separately, what I found interesting about your suggestion is it's a mechanism to create an escalating incentive for copyright holders to relinquish copyrights even sooner than the standard copyright period. Currently, no matter what the term length, it costs nothing to sit on a copyright until it expires - so everyone does - even if they never do anything with the copyright. And the copyright exists even if the company goes bankrupt or the copyright holder dies. Thus we end up with zombie copyrights which keep lurking in the dark for works which are almost certainly abandon-ware or orphan-ware simply because our current system defaults to one-and-done granting of "life of the inventor + 70 years" for everything.

Obviously, we should dramatically shorten the standard copyright length but no matter what we shorten it to (10, 15, 20 yrs etc) we should consider requiring some recurring renewal before expiration as a separate idea. Even if it's just paying a small processing fee and sending in simple DIY form, it sets the do-nothing-default to "auto-expire" for things the inventor doesn't care about (and may even have forgotten about). That's a net benefit to society we should evaluate separately from debates about term lengths.

I see your suggestion about automatically escalating the cost of recurring renewal as another separate layer worth considering on its own merits. My guess would be just requiring any recurring renewal would cause around half of all copyrights to auto-expire before reaching their full term - even if the renewal stayed $10. The idea of having recurring renewal costs escalate, regardless of when the escalation kicks in, or how much it escalates, is a mechanism which could achieve even more net positive societal benefits by increasing the incentive to relinquish copyrights sooner.