/tech/ - Technology

>>30355
A computer is capable of displaying a tree structure. Why should you maintain the tree in your mind when the computer is perfectly capable of showing it for you? (Indeed, to varying extents it already does.)

All real world objects have this property too. We build mental models of drawers, filing cabinets, etc, and we expect them to remain consistent - all without arbitrarily making the physical object disappear and keeping a mental model in our mind. Even blind people don't do that.

I may develop on this in the morning, but I'd summarise that it might even be worth thinking of as a matter of game design (not gamification) or interior design. The computer is a digital environment, and within an environment people can make some really very clever inferences - beat a game in half an a-press or whatever - if you design it well, with clear rules and useful feedback. If you design it badly, it's rather the same as designing a cup with the handle on the inside rather than the outside in the real world - you can drink, but why? A terminal does not usually give the necessary feedback to the user for them to treat the computer in this way. (Terminal based games, oddly, an exception)

Think of disparate file movements in terms of this metaphor - if you find a pair of trousers, a fork, a pan, some old papers, and various other junk lying around your physical environment, how do you deal with them? Would you be reassured or unnerved if I was to say some magic words and make them disappear - perhaps to their proper places, perhaps to Hades?

>>30356
>Why should you maintain the tree in your mind when the computer is perfectly capable of showing it for you?
If you've ever used one of those menus where opening a directory shows all its children, you should know how terrible they are. Selecting a file multiple layers deep always floods your view with unrelated crap. The program from the clip actually existed on IRIX and has a fairly modern version with a gtk port https://fsv.sourceforge.net/screenshots/ I agree it can help the user visualize files, i used it once or twice to that effect in the past, but it's even more cumbersome than a folding tree menu for regular use.

Why use arabic numerals, when we could just write tally marks, or perhaps a hybrid system like babylonian numerals? Because these symbols afford a higher-degree of abstraction, that allows us to grasp and express ideas more efficiently and a file path is one such abstraction. The things you need to do to interact with regular files and directories are precisely those you already care about: create, copy, move, remove, rename, link, feed to another program, etc; less actions than are displayed under the windows right click menu i'm certain.
>a pair of trousers, a fork, a pan, some old papers, and various other junk lying around
Sounds like the setup to an adventure game (>ᴗ•) Humans are great at adapting to conventions, provided they need to and aren't coddled by half-baked metaphors or being asked the same questions over and over again (that's essentially what menus are). Insisting on treating program and system state as various types of clutter only hinders an actual understanding of it.
>Terminal based games, oddly, an exception
You mean roguelikes? Technically nethack gives you exactly the same amount of feedback as rm&co relative to what it does. Rm only tells you if it fails, otherwise success is assumed. Nethack has to relay certain variable aspects of the game world, like FOV ar status messages, yet the principle behind it is the same, only show what the user can't already know. Moving onto a fountain doesn't ask you if you want to quaff from it, there are no visual flourishes to your actions, nethack doesn't even display thPost too long. Click here to view the full text.

>>30361
A more general reply follows, but hitting a few points quickly: I think folding tree menus are fine, as are ordinary "display the contents of this folder, its file path, and - if entered from another directory - a back button, plus a folding tree on the side" type file menus. both show you the menu in a context while - thanks to the file tree and back button - doing nothing to hamper understanding of the overall structure. The windows right-click menu, in different order, contains every menu option you list and more, provided you treat "open with…" as a stand-in for "feed to another program". Nethack doesn't have to output any feedback (you could just keep dying randomly with no explanation!), it does so because doing so is good design. You raise arabic numerals - but why use base ten as our default? if we counted using binary, you could finger count up to 1023…

Success being assumed is terrible UI design. It costs almost nothing to output "x deleted" - and in the cases where it does cost something (such as in a script that deletes a lot), it costs almost nothing to add a flag telling it to shut up. i assume, to you, this appears terribly redundant (why tell the user the computer's done doing what they asked?) and perhaps even in violation of the unix philosophy (rule 2!!!) - but there's value in redundancies. planes have two pilots for a reason. (a graphical interface doesn't need to pop up "file deleted", you can see it disappear in response to your actions without having to type "ls" and look for it)
i'm not assuming bad faith here, but i'm not sure i can productively convince you of my point. (which, going way back, is really about treating the user with contempt for using the computer "wrong", rather than UI design itself.) there's just too wide a gulf in assumptions about the default - it's like trying to convince a native japanese speaker that most people find it easier to work out the meaning of a sentence based on word-order rather than by marking everything with particles. (although i think this is only true thanks to English and Chinese tilting the statistics, whereas a preference for the convention of graphical interfaces is - i suspect - closer to "human nature". I mean, sticking with games, there's a reason there are more graphical games than non-graphical games! and, for all the terminal use involved, linuxPost too long. Click here to view the full text.

>>30362
>is it really necessary to explain that users like systems to be forgiving, and that a digital system is usually capable of undoing them?
This is moreso a problem with systems than with interface design. Interfaces can only plaster over the ugly truth that the filesystems currently in use don't have proper versioning or an equivalent to garbage collection. I remember reading a blog post about the inadequacy of the "save" button in OSX and how it lures users into a false sense of security directly leading to data loss. The plan9 CWFS actually fixes some of this at a system level, by requiring you to commit desired changes to disk and leaving any one of these snapshots accessible indefinitely.
>ask me to justify each little detail, each piece of feedback, each option displayed on screen instead of held in memory
I understand this type of approach isn't the most useful in practice (as neither is hiring a ux design team is for a free software developer), but i want people to analyze software in these terms and strive to design efficient interfaces with the least amount of visual clutter (just like the gnome team wants devs to follow their interface guidelines).
>build websites and tools that draw people in
You seem to equate this with copying state of the art or apples design circa 1993. Developers should strive to find the interface most adequate to their problem domain, making it halfway palatable to a competent user would be the next watermark.
>why use base ten as our default?
Because it's a proven convention, that people are capable of intuitively understanding. The main issue i see is that within the capitalist framework of producing worker drones at the minimum of required competence, people aren't afforded the time to properly understand things. It's the type of policy which equates learning computers with sitting schoolchildren in front of microsoft office for a single period.
>I think folding tree menus
Maybe this is a difference in our workflows. When i'm sitting at the wine file explorer to select a file from one of my game directories, i always need to pass through the 'vg' directory. This isn't a problem with cd, as a game is something you typically know the name oPost too long. Click here to view the full text.

>Open source operating systems are software communism…building alternatives?
Silicon Valley anarchists are like the original anarchists who used hippie ass Christian theology to resist primitive accumulation…and like them, their "praxis" is immediately recuperated as part of capitalism. Utopian idealism is pre-Marxist pseudoscience from people who didn't understand the power of the proletariat (or actively denied it in favor of their individualist petite bourgeois narcissism)

>>30355
>What is even the usecase for moving a lot of individual files to different directories?
>>30356
>A computer is capable of displaying a tree structure. Why should you maintain the tree in your mind?
Twin Peaks fireman voice: "it is happening again"
<Lumpers and splitters are opposing factions in any academic discipline that has to place individual examples into rigorously defined categories.
https://en.wikipedia.org/wiki/Lumpers_and_splitters

>>30388
It's the same issue mentioned already, assume that they fixed the conversion rules and a browser implemented it, suddenly a lot of sites that used to rely on the old behaviour would be broken and as a consequence users would use other browsers where their favourite sites still worked.

>>30389
In 1997? The standard came about because there was already incompatibilities between browsers…

>>30390
In 1997, yes.

>>30391
Okay, a nicer hill than any other to die on lmao

>>30373
>technology/abstraction intended to facilitate computer programming for people with no formal computer science knowledge is actually full of unintended caveats which complicates programming rather than simplify it
How many times has this happened before

>>21346
So you don't know how Monero works. Unlike bitcoin, transactions cannot be tracked back to you. When you are controlling a node, you can wipe/not save all the logs you want. Nobody is forcing you to keep logs on yourself lol

bohemia/cannabia might be exit scamming

localmonero is dead

Archetyp got taken why do these morons host shit in Germany?????

Also Europol put out the most cringey fucking page of all lmao

>>30187
hosting shit on a .su domain hosted on an abandoned russian warehouse is not viable for sites that big

>>30370
HugSQL looks like a DSL rather than an ORM so it's going to avoid the usual ORM pitfalls,looks cool btw,didn't know about it

>>30364 (me)
>>30366
Well, ORMs tend to tightly couple your application code to your database schema, which is my reason for avoiding them entirely. You need to keep track of all your migrations and even then if something breaks, it might permanently ruin your database. This is why I try to use the most generic interface possible when interacting with databases with the least amount of abstractions, which in this case is SQL. As long as you don't do anything stupid, raw SQL is fine. But as another anon already mentioned, you're building a hobby project and it'll probably work just fine even if you don't write raw SQL queries.

>>30374
In this application there was a loose coupling between the table UserInDB and the rest of the application. Even though the data is ultimately stored in UserInDB what's passed around are User and UserCreate objects. Ended up really liking this structure, but only understood why after you mentioned this.

class UserCreate(BaseModel):
    username: str
    email: str | None = None
    full_name: str | None = None
    password: str

class User(SQLModel):
    id: uuid.UUID = Field(default_factory=uuid.uuid4, primary_key=True)
    username: str = Field(unique=True)
    email: str | None = Field(default=None)
    full_name: str | None = Field(default=None)
    disabled: bool = Field(default=False)

class UserInDB(User, table=True):
    hashed_password: str

async def get_user(session: AsyncSession, username: str) -> User | None:
    user_in_db = await get_user_in_db(session, username)
    return User(**user_in_db.model_dump(exclude={"hashed_password"})) if user_in_db else None

async def get_user_in_db(session: AsyncSession, username: str) -> UserInDB | None:
    result = await session.execute(select(UserInDB).where(UserInDB.username == username))
    return result.scalar_one_or_none()

>>30377
nope, still dim.

>>30374
Am talking too much and thinking too little; for unknown reasons.
I'm disappointed, and cooked.

A query builder lets you conditionally chain parts of queries.
This is unlike PugSQL, HugSQL, or pg-promise.

You're perfectly free to translate from the ORM objects to other objects.
So the coupling to the database schema shouldn't really be an issue.

Really the only substantial issue can understand at my level is missing features in the ORM.
This is like for example the missing TRIGGER functionality for Async SQLAlchemy.

>>30330
>but they want to be pb again!
And most proletarians just want to win the lottery and not have to work again, it’s kind of the job communist parties to educate and organise people towards a much more likely goal.
And let’s face it, that’s what the seething is about ITT, that educating and organising involves talking to people who might earn more than yourself and presumably that being impossible as they start vomiting blood at the idea of sympathising with people who went to university or had a spare $500 to lose on crypto five years ago.

>>30235
>Ideologically, programmers align with Estonia and Tallinn
nani???

>>30238
>>30234
>meaningless nitpick
>owning capital
how some of you manage to be so smug despite not reading a single book never ceases to impress me

I think a lot of tech workers are in the process of being proletarianized, or will soon be.
If you got in late you probably still don't own shit anyways, I certainly don't. I'll likely be crushed by debt I didn't want for the rest of my life.
What I'm concerned about is that the current batch of Leftists are incredibly hostile to anyone that "hasn't been" or is only a recent proletarian struggler. As the century marches on, more and more people from educated professions and the labor aristocracy are going to get proled against their will, some of these individuals will be able to be swayed over, but not if we petulantly make an enemy out of every single individual that at some point belonged to a different class. These people NEED to be captured.

This kind of ongoing hostility to anyone who has at some point had it better than you reeks of reactionary bullshit to me and is honestly kind of christcuck-esque in its identifying/constructing an identity around suffering.

>>30333
>Ideologically, programmers align with Estonia and Tallinn
<flat tax rate.
<digital democracy.
<social problems have technical solutions.
Something like this anyway.

>>30220
>You may just as well argue that "linux" is a bloated OS because you collect the entire system a single giant unit.
A microkernel is certainly more suckless/minimal than a monolithic kernel.

>>30221
Typically the suckless people do argue that the Linux kernel is bloated, but want a small monolithic kernel like plan9.

>>30345
Well I can't imagine that a microkernel is actually minimal nor keeping it simple stupid, since otherwise we'd not have to keep calling the bastard GNU/Linux because GNU would be using the Hurd.

>>30347
But a microkernel is more aligned with the UNIX philosophy.
I have heard that the issue with Hurd is that they are perfectionists, not that they can't get a microkernel working. Doesn't harmonyOS (huawei OS) have a microkernel?

>>30220
>It's absolutely terrible
What do you mean by this? The daemontools approach to service supervision is unambiguously superior to cgroups monitoring and other hacks systemd uses (see http://cr.yp.to/daemontools.html). Runit sits on the lower end of daemontools derivatives, nosh is another one that can run systemd unit files http://jdebp.uk./Softwares/nosh/worked-example.html

>>30056
If you have any kind of formal employment, financial or medical history, it's pointless to consider the privacy benefits open-source software. Unless, the activities you're trying to hide are done exclusively on an encrypted p2p network, but that too could bite you in the ass someday.

>>30052
>>30057
Lay down the serious hat for a bit.

>>30050
>>30062
Lunduke has been covered for being a petit bourgeois piece of shit already, see: >>28088

No idea about the other guy, but I think Luke Smith is actually on a trajectory to de-chuddify. The barrier is his viewership/fanbase.

OP I'm confused as shit too, he constantly rails about the negative consequences of profit-motivated design, but then simultaneously he stinks of lolbertism. I can only conclude he experiences a lot of dissonance

>>30050
Luke is a terminal contrarian who repeatedly overcorrects. I don't have a materialist critique of him.
> DEI is bad so thus catholicism must be good. > Consumerism is bad so all corporate jobs are horrible
Etc.
>>30119
I was in a similar situation as Luke and I ended up as a leftist because I realized leftist ideologies are the ones who actually want to reward hard work and bettering your community. Right wing ideologies only do this incidentally.

>>30062
>A lot see themselves as cool digital gangsters, as if pirates were historically criminals for crimes sake instead of radically democratic compared to the empires of their day.
Pirates were a kind of proletarian dictatorship…but lets just say most of them were gross Johnny Depp guys and not based lesbian pirate queens!
These PMC redditors are pirates in the sense of being downwardly-mobile petite bourgeois who emerge during a crisis in capitalism and instead of being low paid servants of the ruling class, choose to be thuggish parasites who steal the fruits from farm and factory laborers. Same thing as happened during the rise and fall of the British empire

Apparently being critical of artificial intelligence is good for you.

Artists are going to pretend this is about AI art and not clearly text parsers.

There's brain scans now:
https://arxiv.org/abs/2506.08872

i just dont believe that deterioration is permanent and also why should we be alarmed that mind-numbing office work is mind-numbing

>>27265
patient zero

Not really /tech/ but still /space/.
Lunar eclipse happens tonight, will be visible in the Americas.

Yet another giant leap on the road to the Moon and Mars. The Chinese will never catch up now!

I might be stupid but I still don't understand what problem starship (or even falcon9) is trying to solve

>>30027
Isn't the purpose to soak up government gibs while maximizing profit (offering a cheaper service and doing things no one else can) and inspiring-following a vision of futurism. If the vision of the future were different you'd probably get it.

Another day, another stunning SpaceX success…

>>30155
>Keeping track of terminals
you have to keep track of what the terminal does, or of all fork calls for that matter, not the terminals themselves

>>30157
I have a very specific scheme in mind, which i'd hoped would be aided by something like inheritable x properties, but since they don't exist a kludge it is:
/etc/profile runs a script writing the window id to /run/foo/$$. If a new window is spawned, get the process id with _NET_WM_PID (fails with 9term, so much for icccm compliance), then walk up the process tree using ps until encountering a process in /run/foo containing the window id with which to set the correct geometry.

>>30158
Wait, i think i've found a near-perfect solution. Terminals mimicing xterm will set the WINDOWID environment variable, so for every non-terminal child process you could get the parent terminal with an SUID script like:

#!/bin/sh
strings /proc/"$1"/environ | grep '^WINDOWID=' | cut -d= -f 2-

>>30152
That's cool and I don't want to sound like a dick but have you considered using 9wm or Wio? They are Rio-like WMs for Linux.

>>30174
In rio only the plan9 terminal (which gets its geometry passed in plan9port) can be created this way. As you can see from the xterms, which don't accept pixel coordinates, my scheme works for every program. Also notice the sloppy focus in the video, what you don't see are the 3x3x5 virtual desktops, so there's that. My next step is programming keybindings that align with my emacs muscle memory to work, but the scripting language is finicky and a module would clearly be overkill.