Gonna do me some #lawtwitter and live-read this. Let’s go.
Nov 3, 2022 • 9 tweets • 4 min read
Wrote today on the work of Elinor Ostrom (of @Ostrom_Workshop and Nobel fame) and what we can/can’t learn from her in open. tldr: our commons of *code* have never been good matches for her work, but our commons of *developer time* is—and that points at some warning signs.
🧵version: Ostrom uncovered design principles that made commons more resilient over long time periods, by studying 900+ IRL commons. She was not dogmatic about the rules: in different circumstances the humans in a system could/would adjust. And adjust open has!
Nov 1, 2022 • 8 tweets • 3 min read
There’s a lot going on in the RAIL licenses—responding to tech changes; cultural changes; and having the unavoidable apps that most new licenses can only dream about. I tried to put down some of my thoughts on them here 👇🏽
The license is very clearly not Open Source. And that means companies should use models licensed under it with some trepidation; the essential *predictability* of traditional open is missing.
Jun 30, 2021 • 6 tweets • 3 min read
Since lots of people are asking overnight, here’s my top-level take on Copilot and IP. Will collect more comments from overnight in a thread here.
I should preface this by saying that this is all new law, based on analogy to some very old things, and rarely tested in court (because, in programmer terms, there is no CI and each test framework execution is 💸. So lawyers are necessarily making some guesses when discussing it.
Have been thinking a lot about the gap between what people want to be possible, and what is possible, when using the technology of copyright licensing. This paper feels very timely.
The ODbL drafting group was very optimistic about making copyleft possible in data, but we did not think pessimistically enough about the challenges; or perhaps we did but did not communicate them well enough to the client.
Feb 3, 2021 • 12 tweets • 4 min read
Next time I write about ethical licenses, the NC licenses as a proto-ethical license would be a fruitful topic.
Other future blog topics: Do we still expect “mass gathering of innovators and participants” primarily (or even in part) by dint of an open source license? Data clearly says no (median # of participants on both SourceForge and GH: 1), but...
Hot take: the @ml5js license draft is roughly infinitely more interesting than Elastic/Mongo/SSPL.
There are some things I have learned from the furor around SSPL, that inform how I will think in a healthy, coherent way about evaluating new licenses. Blog post tomorrow about this, based on my post from last week: blog.tidelift.com/so-you-want-to…
