Sabitlenmiş Tweet
sol🏴☠️
1.4K posts
@sol_plunder @chaserxy @statebox This is really cool. I haven’t seen literate programming in a while, what a great concept. Gonna have to spend some time poking at this (and by extention you when I get stuck lol)
English
@antonlivaja @chaserxy @statebox And a language toolchain for that, written entirely in PLAN Assembly: GitHub.com/sol-plunder/re…
English
@antonlivaja @chaserxy @statebox GitHub.com/xocore-tech/PL…
Atm the code is bigger because we are focused on scaling/perf/persist, but a minimal subset needed to run everything is truly tiny.
English
@antonlivaja @chaserxy @statebox This is awesome, btw. We need more stuff like this.
Everything should be bootstrapped from source!
English
@antonlivaja @chaserxy @statebox I believe my system can improve on the bootstrapping builds approach.
Instead of: minimal asm -> subset of C -> GNU Mes -> real c compiler
You do: minimal asm -> PLAN -> Reaver -> full C compiler + linker
Rough estimate is that PLAN is smaller than a mini C compiler.
English
@antonlivaja @chaserxy @statebox Hi! I've never heard of stageX. Is it related to the "Bootstrapping Builds" idea used by Guix via GNU Mes?
English
@Ngnghm @satisfiesvalues @msimoni > Free markets beat centralized systems, every time.
Not true, centralized systems have dominated human organization throughout almost all of history.
Even today, free markets are an illusion. A mask which gives legitimacy to hidden central structures.
English
@Ngnghm @satisfiesvalues @msimoni This is probably the actual core of all of the paradigm debates: paradigms are INHERENTLY political and tribal.
SW systems are literally political systems for abstract entities (objects/processes/functions/etc).
Human interactions with these systems mirror those structures.
English
Static types catch errors early and that's great—but they also catch non-errors early, preventing you from writing the software you want—and that's terrible.
Those who only tell you about one side of the tradeoff, or claim the other side is universally negligible—are dishonest.
English
English
English
@Ngnghm @satisfiesvalues @msimoni > the masses of lesser-brained people can still handle larger and more complex systems than you ever can thanks to OO.
The end result is horrible.
English
@Ngnghm @satisfiesvalues @msimoni Very interesting comment.
I guess my whole thesis is that competent people should get together and create much smaller, simpler ecosystems to replace these large, complex systems.
Our dependency on this complexity makes us the slaves of its managers, and we should cut ties.
English
@sol_plunder @satisfiesvalues @msimoni My brief experience with the Haskell ecosystem is that DLL was far from solved: the feature or bug fix you need is in the latest version of library A, but it needs the latest version of library B, meanwhile you're using library C that can only deal with an earlier version of B.
English
@Ngnghm @satisfiesvalues @msimoni Stackage solved DLL hell for the Haskell ecosystem, as discussed on the thread.
Just have an organization that coordinates ecosystem-wide release. Works extremely well.
Big proprietary ecosystems do the same thing, and they have the best software.
English
@sol_plunder @satisfiesvalues @msimoni As to the designing nice big parameterized systems: you'll find that it requires a lot of coordination, and makes each small upgrade incompatible, triggering DLL hell. OO Fixes That. OFT.
English
@hassundotX The purpose of this is to create a complete panopticon.
The only solution here is to fully exit from any states that require this, and to pull out of any technological systems which comply with these requirements.
And, ultimately, to destroy the humans who are pushing this.
English
How digital identity verification should be done, but with certificates to ensure the data is not set arbitrarily.
They do not NEED all your private info from a driving license/passport and the gov won't protect you from its abuse
Pirat_Nation 🔴@Pirat_Nation
The main program that starts and manages almost everything on modern Linux computers, “systemd”, recently added an optional "birthDate" field to its user database records. This stores a user's full birth date so apps can check age, for example, to comply with new age-verification laws in places like California, Colorado, and Brazil. It is not automatic age checking. Systemd only saves the date. Only admins can set or change it, but the user and some sandboxed apps can read it.
English
English
English
@spandrell4 @Air_Peasant Wrong conclusion, I think.
The baby in this case is a big pile of hard to maintain code. You have to actually reject all code built by hostile actors, and only use code you can maintain without them.
The anti-systemd people were right.
English
@Ngnghm @satisfiesvalues @msimoni @ngnghm what do you think about actor model systems and their relationship to OO?
Gerbil scheme has actors, iiuc?
In a pure FP system, actors are probably the closest thing you can get to OO, since they are stateful, support messages/late-binding w/o mutation or value identity.
English
@Ngnghm @satisfiesvalues @msimoni Unless you include things like the actor model in your conception of OO, in which case I agree.
Type systems are brittle and don't compose with each other, which is why we will have typed languages over a unityped core in PLAN.
English
@Ngnghm To what extent is it true that dependant types are opt-in?
Is this just a problem that comes from over-use / mis-use of dependant types, or a problem that is inherent to any language that even supports them?
I have never used DT in anger, but they seem nice, superficially.
English
My experience with less-than-dependent types is that of "The Bear and the Gardener": an idiot friend who does heavy lifting for you, but will kill you trying to eliminate a bug.
My experience with dependent types is that they are too much work for an overly rigid result.
English