~b @torresjrjr@qoto.org

"How Austral’s Linear Type Checker Works"
by Fernando Borretti
29 January, 2023
https://borretti.me/article/how-australs-linear-type-checker-works

Whenever ed's version bumps, I feel the familiar solid ground wobble a little.

"My approach to rigorous resource management in C"
by Sylvain Gauthier
2023-01-06 19:09
https://sgauthier.fr/blog/rigourous_c_coding.html

@outsidecontext
You can ask @vladh, the maintainer. I'd probably say the "file format" section, though he could maybe make more generic headings.

@outsidecontext
Nice, looks great and idiomatic for a new Hare project. Another one for the ecosystem.

See also https://sr.ht/~vladh/hare-project-library/

@blainsmith
Great!

Though I'm surprised you've used `delim: u8` instead of `delim: rune`, and your subsequent use of bytes::Index(). You're discarding Hare's strings support? Particularly in your reader.

@blainsmith
Will you support alternative delimiters? I'd want to handle (superior ) TSV files too.

Everything I wish I knew when learning C
Tom M
2022-11-18
https://tmewett.com/c-tips/

Maybe it would have been smart to add a mandatory "null" or "n/a" option to the Question<>Response ActivityPub interaction.

https://www.w3.org/TR/activitystreams-vocabulary/#dfn-question

Introducing Austral: A Systems Language with Linear Types and Capabilities
by Fernando Borretti
28 December, 2022
https://borretti.me/article/introducing-austral

@ben @J12t
I like to think I'm aware of those phenomemon, and that we still have hope. Oh well.

I also see you have experience with Mozilla and software engineering. And your work on elections and voting.works looks great, btw!

> I strongly recommend reading the Master Switch, by Tim Wu.

Thank you!

@ben @J12t
> some vague opposition to anything that isn't small, artisanal non profit, spec-pure implementation of the fediverse.

Yes, I believe it's important the Fediverse players are largely non-profit and spec-pure, with slow, careful improvements to ActivityPub and the network.

Fee-based models are OK though. I pay for my email server, and would for my fedi.

> That's naive.

I think it's right, and a good ideal to aim at. It's also largely the current form of the Fediverse, so I disagree.

I won't support another "kinda federated twitter but in practice usuable and centralised and dictated by Mozilla proxies".

> The gravitational force of a growing fediverse will cause all of the things you mention.

That's not a certainty. And if it was, I'd rather it happen slowly, and I think Firefox integration would accelerate those problems.

> A large player like Mozilla that understands the open web is a very good way to prevent the worst of natural tendencies and network effects.

I'm not convinced. And what about Chrome being inspired to EEE?

I don't want large players. There's no need for large players. There's no reason for Firefox to adopt such a responsibility, especially when they can make things worse, intentionally or by accident.

I trust random FLOSS volunteers, because they mostly have the same interests and vision for the network. And that's not naive, I argue.

@captainepoch
True.

From personal experience, I've had people look at the two servers and said "Mastodon looks more official and put together". Pleroma has that kind of Y2K web design. So most people are drawn to Mastodon, even if they're equally capable.

Of course, Pleroma ships with the Mastodon frontend, but many don't know that, and a novice won't understand that.

@captainepoch
We can still greatly benefit from more server software. There's way too much Mastodon. Reinforcing interoperability with multiple implementations is good.

@J12t @ben

https://qoto.org/@torresjrjr/109607707185525264

@captainepoch

> @torresjrjr I agree, we already have too many clients for the Fediverse.

And not enough server software!

@captainepoch @ben @J12t
Oh there are so many more problems.

Privacy, anonymity, security, centralisation of the Fediverse, the proliferation of the Mastodon API and quirks in place of ActivityPub, inspiring proprietary browsers to do the same in order to embrace-extend-extinguish, the inevitable commercialisation of the theoretical Mozilla Fediverse instance, inevitable breaking changes when Firefox introduces their own "helpful" features...

Firefox is a bloated, controlled-opposition browser. And it needs to remain at the very least a browser. Considering some kind of fedi integration is rediculous.

Terrible idea and article (no offense).