Notices by wizzwizz4 (wizzwizz4@fosstodon.org)

@A11yAwareness Based on https://www.w3.org/WAI/curricula/, which might be more accessible.

@paul Frame challenge: you say you are learning web development, but React is, like, the antithesis of web development. (It has some good ideas, so it can be worth learning about the framework, but actually using it is another matter entirely.) I doubt you'll learn best practices from this codebase.

As for an entry point… I don't know about "good", but __SECRET_INTERNALS_DO_NOT_USE_OR_YOU_WILL_BE_FIRED is a funny entry-point.

@ranfdev Is that inherent to the problem domain (e.g. graphics shenanigans), or just the dependency being bad?

@mathew @alcinnz I'm disappointed we can't still do this. If you've got a graphically-intensive rendering process, it used to be that you could use other windows to cover it up, and make it run faster. Nowadays, it's not clear how to write a program that behaves this way.

@alcinnz There's an implementation in https://dan-ball.jp/en/javagame/dust2/, though I prefer the 3D ones.

@jamescooke When is 7832 not 7832?

>>> x = 7832
>>> y = 7832
>>> x is y
False

You're holding is wrong.

@jamescooke It's not safe: that's an undocumented CPython implementation detail, and could change at any time. (Compare Java, where the equivalent behaviour is part of the language specification.) CPython is not the only implementation of Python.

You should not be writing code that relies on this behaviour in the first place. The decimal.Decimal limitation is a valid thing to be annoyed about (though it is documented), but is not working is not.

>>> 7832 is 7832
True

@jamescooke Showing the type was sufficient. After all:

Python 3.9.2 (default, Feb 28 2021, 17:03:44)
[GCC 10.2.1 20210110] on linux
Type "help", "copyright", "credits" or "license()" for more information.
>>> a = pow(2, 30, 1<<30)
>>> type(a)
<class 'int'>
>>> a
0
>>> a is 0
False

`is`-equality is not a necessary condition for type-and-value equality, even for interned types like int and str.

@pluralistic While "effect" can be used as a verb (meaning "to bring into being"), I think you meant "affect" (meaning "to modify, disrupt, or alter").

@pluralistic This isn't Twitter. 😛

But good to know about the blue ticks. Reputation-washing, a rogue employee, or an old script getting re-applied to the database, perhaps?

@alcinnz Yup, exhaustive search gives the following four-element example:

2 4 3 1
1 3 4 2
4 2 1 3
3 1 2 4

This satisfies (a * b) * c = (a * c) * b, but
• 1 * 2 = 4 ≠ 1 = 2 * 1; and
• (1 * 2) * 3 = 2 ≠ 1 = 1 * (2 * 3).

I'm not claiming to know a non-group structure where you can usefully do Diffie–Hellman, but I wouldn't rule it out just yet.

@alcinnz Strictly speaking, you only need the weaker property that you've quoted, which I believe can be satisfied without associativity or commutativity (while keeping our trapdoor function suitably injective – maybe equivalent to invertibility??). It's just, normally we're using a finite cyclic group, which automatically gives us these properties.

@alcinnz A smallest counterexample is, apparently, the "winner of rock paper scissors" operator (over the three-element "rock paper scissors" set), which is commutative but not associative.

I'm not quite convinced (yet) that it's unique up to isomorphism, but Wikipedia implies it is, and who am I to doubt uncited passages in Wikipedia articles about abstract algebra?

Edit: never mind. It's cited. https://doi.org/10.1016/j.ic.2014.08.005

@alcinnz Is Es the same as Es1?

@alcinnz You've written "commutative", but then the definition of associativity. Is that deliberate?

@dalias Mozilla could simply modify the source code of Firefox to do that. Your threat model should be data exfiltration, imo, not remote reconfiguration.

Check the source, to see if tracking protection is applied to extensions. I doubt it would be.

@dalias You should provide http://detectportal.firefox.com (though replacing it with a local version is sensible), and it looks like https://tracking-protection.cdn.mozilla.net is just an S3 bucket full of hashes. The others, though, I could do without.

@funbaker @mangeurdenuage @SuperDicq We shouldn't force him to: I agree strongly that "thought crimes" aren't. But, being a leader comes with responsibilities.

I don't think it's unreasonable to expect him to choose one.

@funbaker @mangeurdenuage @SuperDicq Richard Stallman would, I hope, disagree.

> In anything we publish, and anything we send to strangers (they might redistribute it to the public), we have to show that our views about issues are primarily based on the moral level.

https://www.gnu.org/philosophy/the-moral-and-the-legal.html

His writings are part of why I feel this way about things. I would love to see him address these issues, these arguments – but he just doesn't.

@mangeurdenuage @SuperDicq Given how you phrased that, I expect I'll disagree with your assessment. But, I haven't heard about this. What happened?