Karl Fogel

Home page: http://red-bean.com/kfogel/
Fediverse:
- <@kfogel@kfogel.org> (I normally tweetidentitoot from here)
- <@kfogel@rants.org> (my blog, which is also Fediverse-enabled)

Last Notes

2026-05-18 05:14:04 UTC

- reply

@nprofile…50mt Another take: An industrial regulation created by publishers to curtail competition, descended from and modeled on the state censorship law that it replaced, shouldn't be our framework for figuring out the ethics of LLMs nor, that matter, the ethics of anything else.
There are plenty of critical things one can say about LLMs, but the fact that training them involves copyright violation just puts them in the same company as, say, jazz musicians. It's not a point against them; it might even be a point in their favor.

2026-03-10 15:54:29 UTC

Don't look a gift horse in the mouth while changing it in mid-stream after the barn door has closed.
Just trust me on this one.

2026-03-03 20:44:07 UTC

I'm trying to better understand the non-FOSS source-available licenses, and who uses them and why. If you have opinions and feel like sharing them, I'm all ears!
I'm aware of:
Commons Clause
Business Source License (BUSL)
Functional Source License
Server Side Public License (SSPL) (arguably in some ways a kind of super-strict free software license, but that's just my opinion and I know most of you don't agree with me it's okay we can be friends)
Fair Core License (see fcl.dev)
Microsoft's three:
Limited Public License
Limited Reciprocal License
Reference Source License
Elastic License (did anyone other than Elastic use it? Does even Elastic still use it?)
There are a few others listed here but I think I've hit the major ones above.
Did I miss anything big? And does anyone know of any notable uses of any of these (other than the uses listed on the Wikipedia page above)? Especially, if there's some flourishing software project that gets incoming contributions using one of them, that would be interesting to know about.

2026-01-26 16:20:02 UTC

So: governance diagrams for open source projects? Does anyone know of any particularly good ones?
I'm not a particularly visual person myself, for better or worse, so I tend not to turn to such diagrams first. But others have said that they're very helpful, so I'm trying to have some good examples to draw on!

2026-01-19 20:26:50 UTC

- reply

@nprofile…4tva That's what I can't stop thinking about.

2026-01-19 16:02:43 UTC

Real life is analog.

2026-01-08 05:17:54 UTC

- reply

@nprofile…p7yv "Rosamund" or "Rosamond"?

2025-08-22 15:47:15 UTC

Just for next week, everyone switch to pronouncing "logic" with a hard "g" and "login" with a soft "g".

2025-06-11 22:18:37 UTC

Me, four and a half years ago, toward the end of Trump's first term:
It's time for my Increasingly Frequent Reminder:
He will put the military in a bind and force them to choose. He will go there. Why not try? No cost to him if he fails. The only question is what "national emergency" to concoct in order to maximize his chances of success.
(on birdsite: https://x.com/kfogel/status/1330011144795336706)

2025-04-24 03:02:18 UTC

Even good organizations make mistakes sometimes. The OSI should do exactly what this petition says and release the original full results of its 2025 election. The petition explains why very clearly; I've signed it, and I urge you to do the same.
I support the OSI's mission wholeheartedly. I served a three-year term on its Board of Directors, about a decade ago -- that was the Board that transformed the OSI into a membership organization that holds member-driven elections, so I take deviations from that very seriously. (More recently -- full disclosure -- my company has done research consulting for OSI.)
The reason the OSI should publish the unaltered election results is that is the only way such elections should ever be run is with full transparency. That's what voters and candidates both expect and deserve. There can be no qualifiers, conditionals, or hedging when it comes to how an election is done: you either do it right or, in the rare cases where there are truly exceptional circumstances, you disclose them and explain precisely why they required a change in procedure. In this case, there were no circumstances that really required a change in procedure, and the OSI then changed the procedure post facto anyway without explaining why.
One may agree or disagree with the particular platform of two of the disqualified candidates, but that platform is not the issue here. The issue is that the OSI held an election and then neither honored nor published the actual results.
The OSI is clearly aware of this: their updated blog post about the results makes some reference to those in-flight procedural changes. You can read it yourself and compare it with the factual allegations in the petition. I do believe you should take the latter as firm findings of fact: I've known candidate @nprofile…yat9 for many years and I cannot imagine him ever knowingly misrepresenting or misreporting a particular sequence of events that he was a witness to (and note that the OSI's post does not contradict the petition -- it merely presents things in a different light).
As I said at the start, even good organizations make mistakes sometimes. This was such a mistake, and I hope they'll quickly fix it in the most direct way possible.
I believe it is still the case that these elections are technically only advisory anyway. Under the bylaws, the Board ultimately decides who serves. Thus what's been happening in OSI elections all along is the Board has simply chosen to rubber-stamp, and thus officialize, the results of the member-driven elections. This is a fine arrangement; if the Board wants to decline to rubber-stamp a particular election result, it should announce that decision and explain why. But by refusing to acknowledge that that's what they're doing, and by not publishing the actual results, they're just temporarily silencing a signal that is unlikely to go away.
My opinion is that it would be much better for them to let the results stand, get those candidates onto the Board (if they indeed won), and have a healthy debate over the proposed reforms and all the other matters that that signal was signaling so loudly about. But that's just my opinion; if the Board feels that the presence of those candidates would be bad for the organization, then they don't have to seat them. But they should own that decision publicly and not hide behind false proceduralism.

2025-04-17 19:54:07 UTC

Can we all agree that the "I agree to use electronic records and signatures." checkbox on DocuSign and other similar services represents an obvious bootstrapping problem?

2025-03-12 03:28:18 UTC

- reply

@nprofile…yat9 Some of the reactions in there are... I don't even know where to start. Bernie Sanders ran for the Democratic Party's Presidential nomination without joining the Democratic Party. (IIRC, on a second run later he did join the party -- because <drum roll> they'd instituted a new rule requiring it.)
I mean, what you're doing isn't conceptually hard. If someone wants to oppose your candidacy on substantive grounds, then fine, but this weird, incantatory, nominalist officialism-that-isn't-even-official does not fly.

2024-12-27 02:55:36 UTC

- reply

@nprofile…w8ng Ok, but I still didn't know what memo I missed. (Of course, given that I've pretty clearly signaled that I don't have an open mind regarding what that memo is apparently saying, perhaps it doesn't matter that I remain otherwise unfamiliar with it!)

2024-12-27 02:32:07 UTC

- reply

@nprofile…w8ng Wha-hut? Did I miss a memo? Is someone proposing that Wikipedia (to which I donate, and intend to continue donating, monthly) is not a good cause?