[ / / / / / / / / / / / / / ] [ dir / agatha / ausneets / general / htg / kpop / leftpol / sw / zenpol ][Options][ watchlist ]

/tech/ - Technology

You can now write text to your AI-generated image at https://aiproto.com It is currently free to use for Proto members.
Name
Email
Subject
Comment *
File
Select/drop/paste files here
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

[–]

 No.878414>>878431 >>878445 [Watch Thread][Show All Posts]

At the heart of the software, the implementation is the application of the ideas that stem from the programmers imagination, with many different approaches reach to the same conclusion at the end. However, the path that is taken to reach that conclusion is sometimes bumpier than the others and sometimes, it is outright wrong. This is why program bugs exist, because not all solutions are born equal, and as a consequence, they yield different amount of efficiency and correctness.

At this point, we can pretty much agree that incorrectness of the programs stems from the flaws in the implementations, right? So we can now discuss about a rather popular, but buggy program. Or should I say a suite of programs? KDE.

Now, KDE has many neat features. It has a fucking taskbar for starters, in contrast of its closest competitor, GNOME. It has a much better design than any other desktop environments I have seen. It has a proper file manager. It has fancy promises. But underneath them, there lies an ugly truth, that KDE is a buggy mess.

KDE will crash on you when you least expect it. KDE will completely restart the computer when you are adding another torrent to the most stable torrent client out there, qBittorrent. KDE will try to index encrypted files in your home directory and dedicate an entire CPU core for this very cause. KDE will make you rip it out of your system and never touch it again.

Until this point, I have typed out my personal rant about KDE right there, most of it centered around the usability of this software. But why? I could have very well easily set up a blog for myself and take this entire pile of shit right here, and put it there. I would, but it doesn't stop there.

You see, KDE exhibits some characteristics of poor programming practices right in front of the users face. But what about the background? All those processes, taking up millions, sometimes billions of bits in the volatile memory, what about them? If the very structure of this program is broken, so should its smaller parts, right? What happens then? Surely, the poor programming practices that piled up shit code over the years do have a consequence, and that is insecurity.

You see, improper software development puts the confidentiality, integrity and security of the user data on the risk. It just doesn't stop at problems that you can fix by firing up another tty, or copy-pasting some commands from a wiki that most of the user won't be able to find in the first place. No.

This is why software like KDE will never be the default desktop environment for Ubuntu. This is why RedHat chooses to focus on the GNOME instead. Companies that aimed at distributing GNU/Linux systems has a pretty good reason to choose GNOME, because its alternative can't even help itself.

All in all, all incorrectly implemented software should be taken out with the trash, and no amount of widespread adoption should undermine this practice.

 No.878424

>all incorrectly implemented software

That's all software, except for some trivial software and an incredibly small amount of verified software. Writing correct software is so hard that it's almost never worth it.

That said, GNOME werx. It has a surprising amount of features, it just places them tastefully so they don't get in your way but they're still easy to find when you go looking. The only really inexcusable thing is that they got rid of the system tray, but that's easily fixed with Ubuntu's AppIndicator Support extension.


 No.878426>>878427

Why don't you look at software designed to be minimal such as Plan 9 OS? Why must you insist on staying with the software that you consider inefficient and incorrect?


 No.878427>>878435

>>878426

Good software doesn't just need to be good at being what it's trying to be, it also needs to be trying to be something good in the first place.


 No.878431

>>878414 (OP)

gnome is just as shit


 No.878435

>>878427

I thought that's what KDE and Gnome are doing right now. Are you telling me that it's possible to reach KDE's and Gnome's level of functionality while not being a "buggy mess"? I don't think it's possible to do that while being reasonably timely in releases.


 No.878442

All software should be written in a language that enforces the "one-way-to-do-it" paradigm like python or rust.


 No.878445>>878450 >>878486 >>878532

File (hide): 705731a46f5ed61⋯.png (10.75 KB, 392x225, 392:225, seL4.png) (h) (u)

>>878414 (OP)

>all incorrectly implemented software should be taken out with the trash

Indeed. poorly made software can also be a massive security issue.

It is because of this that you should take a look at seL4!

>>878163

seL4 is the world's first operating-system kernel with an end-to-end proof of implementation correctness and security enforcement.

This project is a mathematically-verified bug-free microkernel system that supports x86 and ARM, and aims to be general-purpose.

Website:

https://sel4.systems/

Whitepaper:

https://sel4.systems/Info/Docs/GD-NICTA-whitepaper.pdf

Genode OS, an OS framework that supports seL4 and other kernels:

https://genode.org/index

seL4 is already being used and researched by the military for avionics and helmet displays.


 No.878450

>>878445

>mathematically-verified bug-free

No its mathematically proved that the implementation they wrote corresponds with a particular specification that they wrote. It does not mean that the specification is free of bugs.


 No.878486

>>878445

L4 is explicitly designed to be minimal. This means that its scope is well defined and it can feasibly achieve the minimum level of complexity. Most other software systems are not quite minimal in scope. There is significant complexity in combining pieces of general software to form a sophisticated general system.


 No.878510>>878512 >>878555

Focusing mainly on software implementation and licenses isn't enough. Nowadays Red Hat, freedesktop, Google et al. are centralizing control of GNU/Linux, and of FOSS in general. The writing's on the wall but RMS and his zealots refuse to see the big elephant. Ironic, considering his bitching about ebil corporations. As long as it's "free software", nothing else really matters.


 No.878512

>>878510

>are centralizing control

Ah yes you mean like the feudal hierarchy that is Linus's Linux. Its a NEW thing. Not like all the old corporate Unixs.


 No.878532

>>878445

that’s all fine and dandy, but does it play my animus?


 No.878555

>>878510

We can see a big elephant and that elephant will not cause damage to us because of the power of the GPL. Corporations aren't inherently evil, we don't oppose business on the basis that they make a profit, we oppose business that profit through an unjust form of control. The GPL means users can never be controlled and users will always have the four freedoms of free software.




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
13 replies | 1 images | Page ?
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / agatha / ausneets / general / htg / kpop / leftpol / sw / zenpol ][ watchlist ]