Jens Finkhäuser ¯\_(ツ)_/¯ @jens@social.finkhaeuser.de

@halcy Yup! That's how I would do it, essentially. Loses everything from the previous such barrier (or scope start), except for the keep list.

@halcy Yup. But blocks are already associated with scopes, so the inverse of this. That'll get very confusing.

@WAHa_06x36 @halcy Related to this: in-language versioning of interfaces, functions, etc. Bonus points if that get translated to shared object export maps.

@halcy It's odd, but not wrong. Tying this to a block feels weird, however. I'd use a kind of barrier statement like a label.

@chjara @aety Had a JS developer have trouble with the notion that UUID fit into 16 Bytes, because no matter how they remove the dashes and count the rest, they'll always remain longer.

@June 5.

@thegibson Which doesn't have me opposed to reading banned stuff anyway. But rather than saying "they're afraid of what you may learn", maybe go with "there be dragons here - for you or for them, and the trick lies in figuring out which it is" or something along those lines.

(Which, incidentally, is also a warning in the foreword to the Satanic Bible. That just adds a layer of deliciousness here that I couldn't resist sprinkling on.)

@thegibson You know that Mein Kampf was banned in Gemrany (until fairly recently), right?

The problem I have with this way of phrasing things that it sort of assumes all things you may learn have equal value; actually this suggests that forbidden things may have extra value due to being forbidden.

There *is* value in learning terrible things. No denying that.

But there is also that thing where exposing a particular crowd to a particular idea can lead to some truly horrible stuff.

@openrisk It probably is due to this, yes.

@openrisk Popularity and success *are* important, because they provide long-term stability (in principle). I'm not arguing for finding the most niche shit and building your life on that.

I just don't like how monocultures crowd out other feasible approaches, which may even have some benefits the popular thing will never have.

I have no solution, really.

@openrisk And this isn't about moving away for the sake of moving away, but for their own benefit, whatever it may be.

I see so many people jumping on JavaScript framework of the week because it promises to solve some problems, and then they work with it and discover there are new problems. And then the cycle repeats with the next framework.

What does it take for people to stop and make a checklist of their acceptance criteria, and measure projects by how they fit into those?

@openrisk ... of criteria. For example, my set of criteria includes "must be able to understand everything easily", which is why I am very much behind the one-tool, one-job philosophy, and don't love systemd as a result.

But if your criteria include things like fastest boot time or the smartness in managing service dependencies, then systemd clearly wins.

The point, though, is that once there are winners by popularity, it's very hard to convince people to move away.

@openrisk Mhh. Maybe I was a bit obscure. I tried to fit a thought process into a toot, after all.

No, this is nothing against FOSS. It's about chasing the biggest thing. Like, I was discussing Linux and systemd the other day, and I ended up asking myself why Linux is so immensely popular when there are BSDs? Why systemd is so widely supported when there are objectively better systems.

Interjection: "objectively" doesn't mean the judgement applies universally, but according to a specific set..

Some kind of success is vital to FOSS. If nobody uses or knows about it, it's as good as dead.

At the other end of the spectrum, we tend to go for the most popular solution.

I know why it's happening. "So many other people can't be wrong" is a convenient shortcut with a high success rate.

But it also leads to a kind of social Darwinism, "might makes right" measured in popularity.

This is how suboptimal things become standards, monopolize the problem space, and kill progress.

@humanetech @jon Hmm. Open Source was successful at that (over Free Software). I'm not sure what it would look like, but tapping into making people feel good about giving stuff is apparently possible, and people will be willing to let someone else capture the results to some extent.

@jon @humanetech ... tack ActivityPub on e.g. a blog than it is for RSS, because you have that whole subscription management to deal with on the side that produces content. That can probably be improved upon.

Still, I'm more hopeful for the fediverse or some iteration of it to survive a big company.

@jon @humanetech ... and then when it got killed off, a lot of the blogosphere went with it because it became inaccessible (there is still the old reader).

I'm sure there are/were other issues. I'm just focusing on what stood out to me.

Anyway, the big bonus point of ActivityPub here is that there isn't a need for consuming software that can be controlled in a similar manner. Publishing and consuming is both built in.

The downside it's a little harder to...

@jon @humanetech I don't know. Conceptually, the fediverse is like the blogosphere. It's not one big place that someone controls, it's pockets that everyone creates for themselves. They just happen to overlap because the tech permits it, and one big instance isn't going to take that away. Worst case they stop being interoperable and we're back at hellsite vs. fedi.

What kind of killed the blogosphere is google reader. It was so good that everyone used it, ...