copilot might be the straw-on-camel's-back thing that makes me think very not-my-usual-thing thoughts about whether publishing source code was ever actually a good idea at all.

maybe the future of communitarian software development is going to be passing things around on a sneakernet where we make sure you're human first and if we find out you let a corporation or a robot see it, you're a narc.

sure, we'll be hunted like dogs through the wastelands outside the city walls, but at least our text editors won't be trying to stab us in the back and i bet we still get better drugs.


I think the current career trend that I find most disturbing is not copilot, but the notion of migrating all development proper into the cloud via browser-based IDEs.

@lordbowlich it all feels very of a piece to me: a sort of total enclosure and dependence on impossibly complex systems that both determine the course of your work and feed on it to sustain their totality.

@brennen a long while back, an internet famous person, who I can't quite recall at the moment, left FOSS because user demands have gotten ridiculous. They instead reverted to a friends-only license where their code is only available to good friends who have reasonable expectations of development priorities and available time.

I've been operating on that model for a while now, via unlisted repos on and it's pretty nice. it's also why most of my repos have the disclaimer I put on the bottom of

@brennen regardless of copilot and the like, the foss community has gotten really garbage about demanding instant responses and professional grade support from folks with unrelated day jobs and differing personal priorities.

@sungo @brennen

possibly unwelcome, haphazardly timed, pedantic reminder that "publish to an entitled world in a way most permissive and convenient to downstream" is a very "open" value

none of the usual licenses require that at all, though

even the GPL requires giving source only if you gave/sold someone a binary

it's like, we already have freedom-compatible options but people seem hell-bent on ignoring them or running them down instead of just ... using them?

it's just a constant whipsaw between free riding and Gollum grasping and ... I don't know why.

@idlestate @brennen I'm not going to get into a licensing discussion beyond this but the AGPL requires public publishing of source code and modifications. GPLv3 has similar provisions. But those folks have a different definition of "open" than the open source initiative.

The garbage of the FOSS community that I was referencing has nothing to do with licensing though. I'm talking about the folks that show up in the issues or bug emails for projects and demand instant and personalized support and then get nasty when they don't receive it. It's entitlement at the highest level. Folks who scream that the documentation is slightly wrong or out of date and have no intention of helping or fixing the product. Or yes, you should be using their license and their preferred frameworks or language or wtf.

Being an open source developer in 2022, particularly as a leader, is a hostile garbage experience and I recommend that no one get involved.


"only if you gave/sold someone a binary" for GPL ("convey" but still)

it sure as hell doesn't require public bugtrackers and all that shit. that's a common model, sure, but that's it.

say what you will about the sqlite3/fossil dev's religious & social views, but they defy the open-as-in-entitlement take, for example. but people still ship it because it's free enough.

I think the way they put it is: it's open source, not open contribution

I mention licensing because that's the only practical qualifier I recognize as FOSS/not FOSS

people whine stuff like "but but but COMMUNITY" and I'm over here like there was DECUS and all sorts of other "community" user groups well before FOSS.

a lot of what people want to call FOSS is feel-good, we love our customers, take a moment to give us your feedback marketing BS that isn't unique to FOSS at all.


@idlestate I hear you on the legal technicalities. But you know what the sqlite community gets all the fucking time? tweets, blog posts, emails about "you should use github! you're a terrible person because you don't." etc etc. That development community is just well practiced in shrugging it off.

But this exact conversation is why I want no part of FOSS in 2022. I can't say "The FOSS community can be total garbage sometimes" without immediately having to define what FOSS means and talk about licensing. Instead of talking about the bullshit entitlement, harassment, bigotry, and misogyny that runs rampant in the FOSS universe, we're talking about licensing and legalities?

And if what you were really getting at is that sharing code amongst your friends is still technically "open", I don't think anyone was worried about that. At the point where I've walked away from the FOSS universe, the FOSS universe's thoughts on my licensing (if any) is of no import.



I am trying to talk about the entitlement.

I'm trying to say it was deliberately cultivated with FOSS mostly as the vehicle rather than the wellspring


@sungo @brennen

anyway, my favorite convention-defying approach is PyMOL, which does publish source, and with a non-reciprocal license.

but they don't generally publish binaries

other people do, but in so doing they're at least invested enough to figure out how to run builds

@brennen my current hypothesis is that co-pilot will follow a trajectory similar to that of IBM's Watson (but feel free to remind me of this when we're in the unemployment line 😛)

Sign in to participate in the conversation is brennen's single-user Mastodon instance. This instance runs on, and is thus bound by's ToS, which bar instances dedicated to racism, Nazi shit, transphobia, misogyny, incitement to violence, and the rest of the usual litany of horrors.