[ / / / / / / / / / / / / / ] [ dir / agatha2 / b2 / choroy / dempart / freeb / mde / trap / vichan ][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.
Email
Comment *
File
Select/drop/paste files here
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

[–]

 No.1065045>>1065151 >>1065179 >>1066017 >>1066103 >>1066358 [Watch Thread][Show All Posts]

Do you think it’s possible for a piece of software like Linux to ever reach a final version and never need an update again and everyone would be satisfied with it? Or is kernel/OS development an endless game of cat and mouse whenever new security holes and bugs are introduced as more and more libraries are being added and more and more hardware is getting supported ad infinitum

 No.1065047>>1065049 >>1065052 >>1065093 >>1065185 >>1066407

"If the son of Adam were to possess two valleys of riches. he would long for the third one." [Muslim 1048]

In conclusion: No. It will simply die or meld into something else at some point. Humanity cannot help but continue progression whatever the cost, such is our affliction.


 No.1065049>>1065063

>>1065047

You could’ve told me that without shoehorning your Sandnigger trash at me


 No.1065050>>1065053 >>1065061 >>1065081

You answered your own question. Even if you end up with some magical, 100% bug and exploit-free kernel, there will always be new hardware to support.


 No.1065052

>>1065047

>Mud trying to do the Christian thing and post passages from Jewish Adventures III: Revenge of Mohammad

Thats really sad and desperate


 No.1065053>>1065062

File (hide): dfae9d403402b21⋯.jpg (104.13 KB, 570x570, 1:1, ighgydshdy.jpg) (h) (u)

>>1065050

When will we stop developing new hardware?


 No.1065054>>1065060

A program in general can certainly reach a final stage, even though this will be difficult/impossible in something like C. Linux can't be finished because of the driver question, but maybe this would be different if drivers weren't such an Anything Goes land and there were some (sane, nobody needs another ACPI) standards.


 No.1065060>>1066358

>>1065054

The solution is obviously microkernels. separate the kernel, drivers, and firmware so each component is as mutually exclusive as possible


 No.1065061

>>1065050

Also, what if we set final versions for hardware? Then the issue would just be ironing out security bugs. Unfortunately security bugs exist aplenty in hardware nowadays since we "softwarized" hardware development


 No.1065062>>1065066 >>1065644

>>1065053

Not until industrial civilization dies out.


 No.1065063>>1065064 >>1065172

>>1065049

I posted it because the fact you asked the question shows that you don't understand humanity. Would you be this mad had I quoted the Unabomber?


 No.1065064>>1065081 >>1065100

>>1065063

>Would you be this mad had I quoted the Unabomber?

No I wouldn't because the Unabomber wasn't a desert rat camel jockey


 No.1065066

File (hide): 00a5d9362aad8c4⋯.jpg (75.25 KB, 500x375, 4:3, 0jrtsh5.jpg) (h) (u)

>>1065062

What if the hardware we get stuck with has security bugs a la Meltdown and Spectre? Are we just fucked forever then?


 No.1065081>>1065083 >>1065091 >>1065092 >>1065097 >>1065102 >>1065105 >>1066466

>>1065050

Ah but once you've got the version compatible with your hardware you'd never need to update again.

>>1065064

If you like alcohol (al-kuḥl) and use algebra (al-jabr) you are indebted to Muslims.


 No.1065083>>1065088

>>1065081

>Islam is a race

lmao go ask the Persians


 No.1065088>>1065090 >>1066466

>>1065083

The descendants of the Persians are now all over the world. Your Muslim neighbor has ancestors who invented Algebra. Your Muslim dentist has ancestors who invented your Arabic numerals. Everywhere you look, a touch of Islamic ingenuity.


 No.1065090

>>1065088

>Islam is a race

The retard keeps on going


 No.1065091

>>1065081

Alcohol is harram and predates the invention of Islam


 No.1065092

File (hide): d2c8fc0ff0e0825⋯.jpg (111.11 KB, 668x623, 668:623, 0f39aaffc6a6584aac85540b77….jpg) (h) (u)

>>1065081

>This retard thinks Muslims invented alcohol

>This retard thinks alcohol didn't exist before 600 AD


 No.1065093

>>1065047

>Humanity cannot help but continue progression whatever the cost

what progress do you speak about nigger?

we have degradation in every area of life:

>computers and software getting worse and more botnet

>cars designed to fail after warranty. also botnet in cars

>women brainwashed by jews

>richest jews getting richer, common people poorer

where is the progress you speak about?


 No.1065097

>>1065081

Indebted? Why don't we have a parade for the first fucker to suck on a goat tit, while we're at it?


 No.1065100>>1065103

>>1065064

Birthrates lmao.


 No.1065102

File (hide): 2c217db04093a51⋯.jpg (26.13 KB, 398x399, 398:399, 1469819189887-0.jpg) (h) (u)

>>1065081

Giving something its common name doesn't mean you invented it, faggot. Both alcohol and algebra predate Islam by thousands of years. If I call you a goathumper and everyone follows suit, does that mean I created you?


 No.1065103

>>1065100

Those are nosediving in the Arab world too. The only places on Earth not projected to have decreasing birth rates is South Sahara Africa


 No.1065105

File (hide): e8a79c3fb7ac150⋯.jpg (9.98 KB, 226x225, 226:225, froge.jpg) (h) (u)

>>1065081

>If you like alcohol (al-kuḥl) and use algebra (al-jabr) you are indebted to Muslims.

Fantastic bait. Really just excellent bait, this thread will probably never recover. Here, have a free (You).


 No.1065117>>1065148 >>1065156 >>1066358

The linux kernel, and web browsers, will never have a final version because then all the programmers whose livelyhood that it depends on would be out of a job. So they program shit code and add bugs or new features like renaming all the functions for no techincal reason beyond making uneccessary work just to keep their jobs. Progress on improving the linux kernel's base stopped at around version 4.18. For examples see the kernel commits or try implementing new features or look at out of tree forks doing so on top of the kernel for post 4.19. Now its very little bug fixing, drivers getting added, and mostly make work. Most software is like this when corperations get their hands on it. Take chromeium and firefox/forks. They keep adding shit features like renaming everything or always on javascript because if they fixed all the bugs and stabilized, improved secured and made performant, a select set of features like, i dunno, PARSING HTML THE WHOLE FUCKING POINT OF A BROWSER they would be out of a job. Lynx and other text browsers parse html better the chromium and firefox as a example to such.


 No.1065148

>>1065117

Also remember that wither we want to admit it or not, a lot of software development is developed by for-profit companies and most commercial software development is marketing-driven (OUR WEB BROWSER CAN RENDER SHIT THE COMPETITION CAN’T!) so this puts a burden on FOSS developers who just want shit to perform basic functions but are also under pressure to actually gain adaptation among normalfags outside the FOSS community


 No.1065151

>>1065045 (OP)

Only if it had been written by lord terry.


 No.1065156>>1065159 >>1065160 >>1066026

>>1065117

If you feel that later versions of kernels and web browsers are so meaningless, why don't you keep only the last meaningful version that you wish? You should never upgrade from that version and hopefully you will have zero problems in the future.


 No.1065159>>1065163

>>1065156

Thats the problem. If you just want security based updates even thats not enough because either superficial shit piggybacks off of security updates or vice vera where security updates have to be apart of superficial shit like new web browser versions so he just can't


 No.1065160

>>1065156

Security updates and hardware support. On browsers its fesable to backport security updates like forks of firefox do all the time. For the linux kernel not so much, just look at grsec for a great example. The dev of grsec is a crazy dedicated autist who undid decades of damage and still had updates on occasion to fix the shitshow of the kernel for newer hardware supported by newer versions. Grsec is long gone now though since the lead dev is dead.


 No.1065163

>>1065159

Web browsers are much easier to backport updates for compared to kernel updates. The problem you described of small updates hiding security fixes, in my experience, seems to only apply to the kernel. Most browser bugs are caused by newer features that you just don't need to browser the internet while avoiding the botnet. If you want botnet like goybook or something like that then you are in for heaps and heaps of shit from modern javascript malware.


 No.1065172

>>1065063

>Would you be this mad had I quoted the Unabomber?

No, because people are starting to admit that Kaczynski is a bright guy who got some things very right. Muhammad got nothing right, he was just some merchant who got religious hallucinations out of epileptic seizures.


 No.1065179>>1065180 >>1065190

>>1065045 (OP)

kernel development can only ever stop if either an eternal universal standard hardware interface is developed (unlikely, but a man can hope) or hardware development stops.


 No.1065180

>>1065179

As far as general purpose machines go we already have UEFI. For OEM devices that have their own proprietary interface it doesn’t matter because in that case the burden of adapting the software to their proprietary interface falls on them only and is not expected to be developed further


 No.1065185>>1065202 >>1065316

>>1065047

>even muslims are posting on /tech/

wew, the moderation in here truly respects freedums :)


 No.1065190>>1065203

>>1065179

Every new OS now begins in a VM, most "machines" of any sort now run in a VM. VM performance is therefore what matters. Only a matter of time before some crazy bastard makes a bare metal machine based on the virtio devices. And it some of it is really a shim in ring -1 who would know, and would it really matter?

Bang, universal hardware standard.


 No.1065202>>1065316

>>1065185

IIR he’s literally one faggot that sometimes posts mud propaganda on b2 and stands out like a sore thumb


 No.1065203

>>1065190

Aren’t you literally just describing hypervisors?


 No.1065205

These attempts to turn /tech/ against /pol/ without lulz are remedial


 No.1065316>>1065700

>>1065185

I've been here since 2014 lmao

>>1065202

I've never browsed /b2/, you do realise there's an entire board for Muslims on here right.


 No.1065644

>>1065062

what happens if we hit a limit with what is possible in tech and the laws of the universe?

then would hurd finally be finished?


 No.1065700>>1066027

>>1065316

>you do realise there's an entire board for Muslims on here right.

Its not even in the top 50, meaning theres less then 50 ISPs lurking it. Stop lying and stop kidding yourself.


 No.1066017

>>1065045 (OP)

that would mean that its perfect and its never going to happen when its so big. there will always be bugs in it and they will keep adding support for new hardware. a "final version" means that its dead.


 No.1066026

>>1065156

i dont really update kernels that often. only if there are serious vulnerabilities or bugs that affect the things that im using. most of the changes in the kernel is to things that i dont even use.


 No.1066027

File (hide): ec24a37589d52d9⋯.jpg (42.21 KB, 272x272, 1:1, giggle.jpg) (h) (u)

>>1065700

>He thinks boardspeed matters


 No.1066060>>1066090 >>1066358 >>1066977 >>1067179

File (hide): a601e87784045d7⋯.jpg (35.04 KB, 660x440, 3:2, Unix-2.jpg) (h) (u)

At this point, it's hard to care when Linux is still not 100% backwards compatible with Unix.

You'd think after more than 30 years, they'd manage to reverse engineer it all the way legally!

I'd probably be surprised how they're still profiting from Unix when almost nobody uses it today except a few Java Shrills, AT&T Buffoons and a good chunk of the HP-CompaQ Cocks (Solaris, HP-UX and AIX).

MacOS doesn't count since they're based off a modified BSD that only got Unix-certified because they made it proprietary later (look up OpenDarwin).

Even every Supercomputer in the world stopped using Unix in favour of something they don't have to pay for tens of thousands a year.

Can anybody fill me in on why does a OS with such great potential still kick back on anything that's not Android (besides the people)?


 No.1066080

>it's hard to care when Linux is still not 100% backwards compatible with Unix.

What did he mean by this?


 No.1066090>>1066092

>>1066060

sauce on the second image?


 No.1066092>>1066094

>>1066090

The OS the 500 fastest Supercomputers in the world use by year.


 No.1066094>>1066095

>>1066092

i said the sauce, not the filename

<where the fuck does that info comes from


 No.1066095>>1066100


 No.1066100>>1066977

>>1066095

newfag much


 No.1066103>>1066358

>>1065045 (OP)

Linux will never reach a final version because computing will never reach a final version.

Computing is a dark art that feeds off human knowledge. As we change, so will computers.


 No.1066358>>1067099 >>1067103 >>1067109 >>1067181

>>1065045 (OP)

These do exist, but you need a simple interface. An operating system can have many times more features than UNIX while having a smaller kernel simply because less occurs in kernel mode. UNIX weenies have trouble understanding modular design. This seems to have come from the origins of UNIX. Originally, UNIX had no DLLs, shared memory, or threads and ran one executable in one process in one address space. The only way to communicate with another process was by pipes or temporary files. The only way to share code between processes was by putting it in the kernel. Other operating systems have dynamic libraries where code could be shared outside the kernel.

Those graphs also show that Linux is getting worse and worse. If you would have told computer scientists in the 70s that a kernel would cost 2.2 billion euros, have more lines of code than bytes of disk space on real machines, and need 15,600 developers, but still can't do what Multics did in the 60s, they would be less likely to believe you than telling them that you're from the future.

>>1065060

Microkernels are a solution and probably the only way a kernel can have a final version. They have small enough interfaces and separate all code related to devices into other processes, so when all the bugs are gone and it's optimized, there is nothing left to do.

>>1065117

The business model is based on spreading Linux and making it hard to move away from Linux, in order to get more developers and more money focused on Linux and away from real solutions to problems. That's why they don't want you to use a different free operating system. That's why the OOM killer is still there. C sucks and allows them to create "new" software that is just bug fixes without having to add or fix actual features.

>>1066060

Why does proprietary software have "Live Free or Die" as a slogan? Revisionist weenies are trying to claim Stallman's work as the "freedom" of UNIX, which is like Microsoft saying WINE and ReactOS are proof that Microsoft has always been a supporter of free software. Stallman chose to clone UNIX because it was popular, not because it was good or because he liked it.

https://en.wikipedia.org/wiki/Coherent_(operating_system)

>Coherent was not Unix; the Mark Williams Company had no rights to either the Unix trademark or the AT&T/Bell Labs source code. In the early years of its existence, MWC received a visit from an AT&T delegation looking to determine whether MWC was infringing on AT&T Unix property. The delegation included Dennis Ritchie, who concluded that "it was very hard to believe that Coherent and its basic applications were not created without considerable study of the OS code and details of its applications."

Expect "delegations" showing up at your company if you don't pay those licensing fees.

>>1066103

The solution is to design the OS for the hardware and its uses and not to have all this hacky bullshit like building I/O on top of a system call originally meant for controlling typewriters on a PDP-11. Phones don't need UNIX file systems, commands, and shells. They should be more like a modern version of old Mac OS where those program icons really are files instead of faking it. Many things need no OS at all. Routers don't need to run a UNIX clone, just simple software. Game consoles can use simple menus and maybe a few system calls or BIOS, not a huge OS. This makes even less sense when the companies have control over the hardware and don't have to support all the possible kinds of PCs. Maybe their "programmers" just don't realize code can run without an OS? That would explain a lot.

    Shite painted pink is just pink shite.  Why didn't FSF
try to invent a free operating system that was good?
Answer: their political agenda is 10^6 times more
important to them than good engineering.

I doubt that, given that RMS had no particular unix
experience when he started FSF (which may explain a lot).
Given a background of ITS and Lisp Machines, I suspect unix
was chosen as the OS of choice because:

1) Useful tools could be developed and used before the
OS was ready.

2) There was sufficient public domain and portable
pieces of unix to serve as a place to start.

3) replacing unix would be "better for the world" than
replacing ITS.

RMS spent a lot of time before FSF trying to do basically
similar things with lisp machines. No one noticed, no one
cared, and there aren't any more lisp machines anyway...


 No.1066407

>>1065047

Real halal hours in here today


 No.1066466

>>1065081

>>1065088

Your Arab world is fucked sideways thanks to Islam and the cia, Islam is nothing but worthless cancer to the middle east.


 No.1066977>>1067179 >>1067417

>>1066060

>You'd think after more than 30 years, they'd manage to reverse engineer it all the way legally!

Both Stallman and Linus have expressed distaste for various aspects of Unix and POSIX standards. While things like "posix me harder" exist, there's likely no effort to extend compliance because there's no practical reason to do so. If they're not compliant, it's intentional.

Furthermore this isn't something anyone should strive for anyway. Ironically, reading the design docs for Plan9 (written after decades of real world experience by the Unix authors themselves) highlights all the shortcomings of Unix better than any of its detractors ever could. And even today, those same people are talking about the shortcomings of Plan9, asking people to put effort into OS research again. Because of course an OS designed in the 60's and 90's are not going to be considered perfect today after all the changes in how we interact with various technologies.

Microkernel devs do appear to be trying today. Some of that stuff looks promising and sane because it builds off of real world experience over multiple decades in the same way. They have the ability to build a project out of frustration rather than speculation. To do what they want to do directly at the system api level rather than extending or hacking existing things to function in ways they where never intended to. (Re-)Implementing Unix and retrofitting things on to it will never be a worthwhile effort as far as I can tell. Emulating electronic typewritters is not sane in 20XX. No matter how much you want to say it's diverged.

>>1066100

Not sure what site you think you're on.


 No.1067099>>1067417

>>1066358

Linux does everything Multics did in the 60's. prove me wrong.


 No.1067103>>1067417

>>1066358

protip: the vast majority of Linux are hardware drivers


 No.1067109>>1067417

>>1066358

You can have true UNIX compatibility without tying your entire OS to UNIX paradigms. That’s pretty much what Apple did thanks to using a microkernel. And modern Darwin systems are superficially derived from BSD at best since Apple rewrote a lot of shit after acquiring Next. Under the surface it resembles UNIX and it got official UNIX certified but Apple was able to do that without really being UNIX. That’s the beauty of microkernels. If Apple really wanted to they can remove the UNIX core shit and just stick with their proprietary APIs and libraries


 No.1067179>>1067417

>>1066060

Literally no one cares about muh Unix compatibility except Enunchs weenies.

>>1066977

>Both Stallman and Linus have expressed distaste for various aspects of Unix and POSIX standards. While things like "posix me harder" exist, there's likely no effort to extend compliance because there's no practical reason to do so. If they're not compliant, it's intentional.

This is correct. Many people still fail to realize that Stallman made GNU similar to Unix from a functional point of view simply because Unix was popular at the time, and thus it would have made the switch easier. He himself said it. He also never used Unix, "not even for a minute", before starting GNU. This choice was never for technical reasons, as the real goal was, and has always been, spreading free software.

It's not like the world is in a race to clone Unix, like some people here believe.


 No.1067181>>1067417

>>1066358

Alright, I'm willing to bite: what are some comprehensive, reliable sources on Multics, especially some that highlight objective advantages it had over Unix?


 No.1067417>>1067441

>>1066977

>Ironically, reading the design docs for Plan9 (written after decades of real world experience by the Unix authors themselves) highlights all the shortcomings of Unix better than any of its detractors ever could. And even today, those same people are talking about the shortcomings of Plan9, asking people to put effort into OS research again.

Plan9 is crap. If it wasn't for the names associated with it, it would be considered on the same level as some hobby UNIX clone. It's even more disappointing considering it was made after all that money AT&T got from licensing fees, so they should have been able to hire real OS developers, but no, it's the same old crap with some badly implemented features removed instead of redone properly.

https://en.wikipedia.org/wiki/Open_Systems_Interconnection

>The NCC (UK) publication 'Why Distributed Computing' which came from considerable research into future configurations for computer systems, resulted in the UK presenting the case for an international standards committee to cover this area at the ISO meeting in Sydney in March 1977.

https://en.wikipedia.org/wiki/V_(operating_system)

>The V operating system (sometimes written V-System) is a discontinued microkernel operating system that was developed by faculty and students in the distributed systems group at Stanford University from 1981 to 1988, led by Professors David Cheriton and Keith A. Lantz.

If OSI is from the 70s and distributed microkernels were developed in the 80s, why do all these weenies point to Plan 9, which does it worse than any real distributed OS? It's the same reason weenies always bring up C++ and Java when complaining about OOP, and not CLOS.

>>1067099

Read this. It has no segmentation or rings, other than the bare minimum needed to work around them in x86, while Multics is based entirely on segments and rings.

https://multicians.org/exec-env.html

>>1067103

That just means Linux sucks. Mainframes in the 60s already had drivers distributed separately from the OS, so it doesn't even need a microkernel. Microkernels are just a way to provide additional protection and separation.

>>1067109

I posted a link to an OS written in Modula-2, which was tiny. It might have even fit on a floppy. They ended up "converting" C programs to run on it, which made it into UNIX with a tiny Modula-2 subsystem. This is exactly what happened with microkernels.

>>1067179

UNIX weenies care more about UNIX than free software. They would rather pay licensing fees to use C and UNIX than get something better for free. That's why Stallman cloned UNIX, and it sucks. Ironically, if he cloned something good, none of the UNIX weenie "programmers" would want to use it, which means the quality of free software would be better because it would be made by competent people.

>>1067181

Read some links on this site. One of the big advantages is segmented virtual memory. In other words, all files are mapped into the address space and can be accessed directly with CPU instructions.

https://multicians.org/

https://multicians.org/multics-vm.html

https://opensource.com/article/18/10/linux-data-streams

>Think about how this program would have to work if we could not pipe the data stream from one command to the next. The first command would perform its task on the data and then the output from that command would need to be saved in a file. The next command would have to read the stream of data from the intermediate file and perform its modification of the data stream, sending its own output to a new, temporary data file. The third command would have to take its data from the second temporary data file and perform its own manipulation of the data stream and then store the resulting data stream in yet another temporary file. At each step, the data file names would have to be transferred from one command to the next in some way.

>I cannot even stand to think about that because it is so complex. Remember: Simplicity rocks!

This is a great example of UNIX brain damage. A real OS wouldn't use temporary files for this at all. It would just call functions in libraries, which is much simpler and needs no system calls or task switches or address spaces at all. UNIX weenies don't want simplicity, they want UNIX. Even worse, pipes can only send sequences of bytes so you have to serialize and deserialize data in both directions, which you wouldn't have to do with real IPC. It all makes sense when you realize pipes were a hack over teletypes, the 70s equivalent to controlling a program by scripting keypresses and mouse movements.

       Raise your hand if you remember when file systems
had version numbers.

Don't. The paranoiac weenies in charge of Unix
proselytizing will shoot you dead. They don't like
people who know the truth.

Heck, I remember when the filesystem was mapped into the
address space! I even re<BANG!>


 No.1067441>>1067507

>>1067417

>being vague about why you hate Plan 8 again

Of course, you've made a fool of yourself every time you tried shitting on it in detail so you dug up two Wikipedia links and called it a day.


 No.1067507

>>1067441

> people still respond to the unix hater




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
65 replies | 7 images | Page ?
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / agatha2 / b2 / choroy / dempart / freeb / mde / trap / vichan ][ watchlist ]