Cnaiür

[Cnaiür @MarcBeirne]

@JackalOakheart @Seanchuckle I think when someone vagueposts against her it's because she is not as pro trans as the British centre/left, although i think since the cass report she's less an outlier as the consensus has shifted towards her https://t.co/SOE0Nekt2i

[Cnaiür @MarcBeirne]

@ToastingTweets @tenobrus @Dragonogon It's some of the best data i guess, it probably goes out through mailing list so probably not related to whether you're still using SO.

[Cnaiür @MarcBeirne]

@ToastingTweets @tenobrus @Dragonogon https://t.co/lcyxFrgUpn

[Cnaiür @MarcBeirne]

@tenobrus @Dragonogon Vibecoding has a different definition than "written with ai"

[Cnaiür @MarcBeirne]

there's a well known inevitable pattern:
(config as data) -> (add templates/limited programming) -> (have bad probably TC languages) -> (give up and use a real language) -> (extract patterns) -> (config as data)

but the better path imo is
(config as data) -> (generate it with real language) -> stop

[Cnaiür @MarcBeirne]

@wispem_wantex LLMs are going to spin the dial to some new equilibrium, a shallow read is package availability must matter less. I don't get the beef with containers, if they didn't exist we'd have to invent them, as what micro-vms? Language-level packaging?

[Cnaiür @MarcBeirne]

@wispem_wantex Your argument seems to imply more that its hard to reject any part other than the language in isolation. Theres also a skills/culture/organizational element, microservices via conways imply ~ org nihilism/dev fungibility

[Cnaiür @MarcBeirne]

@johndrewmarkley This is discussed in Elements of Clojure: using an unknown or made up word allows you to attach meaning to a concept without clashing with the meaning of a real word which might lead to misinterpretation, at the cost of losing the "shortcut" to understanding of existing words

[Cnaiür @MarcBeirne]

@KarenFuchs16 @johndrewmarkley Thanks, i meant dead in the sense relevant for this which is "meaning of words is not changing"

[Cnaiür @MarcBeirne]

@AlexanderTw33ts @alexandr_wang Inask of it's a high agency idea or destroying the commons? Its a good idea sir...

[Cnaiür @MarcBeirne]

For sure but you are ultimately reviewing /some form of code/ even if not the "real" code. UI has less verification requirement than business logic in general I think, if its to be interected by a lower-rights user since the blast radius is limited. I just think if you get lisp its for legibility not efficiency, time will tell.

[Cnaiür @MarcBeirne]

@Ngnghm but evaluation eagerness the other way around

[Cnaiür @MarcBeirne]

@nateberkopec If you think 20% less code isn't relevant

[Cnaiür @MarcBeirne]

@solraig Innate in the sense of unexamined/unconscious not "natural"

[Cnaiür @MarcBeirne]

@solraig It's because of our innate convention about which side is the plane to consider as"default" prioritizes (out of screen, top of screen, right of screen). So when you consider the orientation fixed from top of screen it switched and that's a default

[Cnaiür @MarcBeirne]

@davidzmorris Narrator it was both

[Cnaiür @MarcBeirne]

It's not necessarily true, you're assuming noone else would have come up with pagerank. Monopolies create value but they also destroy it, notably through suppressing unfavourable innovations. A classic example is standard oil suppressing cracking. So we get pagerank months early and lose whatever ideas/cos Google strangled in the crib.

[Cnaiür @MarcBeirne]

@msimoni We could try for rl#6

[Cnaiür @MarcBeirne]

@msimoni Can you explain it? First class continuations are more or less equivalent to reifying the stack? To continue an exception i guess you need something like that but not too be able to copy the stack which you need to pass around multiple first class continuations?