/tech/ - Technology★

If I knew anything about working on kernals, I'd volunteer. As is, I don't even fully know how to use the ecosystem in the unix shell.

It will take SJW genocide. The gnu rises.

>What would it take?

Literally thousands of high skilled devs.

>>975731

Are you saying maintaining a mildly diverging fork requires "thousands of high skilled devs". That doesn't sound right.

>>975731

>list of the responsibilities

>Literally thousands of high skilled devs.

Did you drop out of high school, by chance?

>>975732

It's defeatist bullshit. It just takes a few people with enough autism and time to maintain a fork.

>>975721 (OP)

At this point, I'm gonna stay put and see what happens next. As far as I know, Linus may wind up rescinding the CoC right after I switch over to OpenBSD. That or amend it to keep contributors from pulling their code out of the kernel.

>>975739

>Lulz

Hi FBI!

>>975746

Linux Libre is a de-blobbed fork. While a nice thing to have, it doesn't explicitly reject the CoC, afaik.

Also, Linux in the process of being de-blobbed gradually. Which would eventually render Linux-Libre pointless.

>>why wont we fork it?

thats what they always say

and they are never heard again

>>975782

>Also, Linux in the process of being de-blobbed gradually.

i question this assertion

>>975721 (OP)

Sorry, OP, “I’m not versed enough in these topics to do it myself”. :^)

>>975796

I second this. The facts are incontrovertibly indicating otherwise over time. Probably a glow shill post.

>>975774

Hello glowstick-kun. LKML.org != lulz.

https://archive.is/vJCtJ

>>975721 (OP)

A few dozen talented full-time kernel developers or a few hundred talented voluntarian kernel developers. Note that talented kernel developers are six-figure expensive, not just in silicon valley but anywhere in the world.

>. build a reputation for this project being the true legitimate continuation of The Linux Meritocratic Project,

Costs a lot of $$$

>. serve -reproducible- binaries for all distributions,

$$$$

>. make it politically unsubvertible.

$$$

>. There's probably a lot of Linux contributors looking for a way out.

I don't think so. What makes you think that?

>Can we figure out the list of skills / roles required?

Skills:

- Excellent knowledge of C

- Good knowledge of driver development

- Ability to read and understand other peope's code

- Resistance to bullshittery

Roles:

- Driver development

- Non-driver kernel development

- Code review, QA, security checks

- Running development infrastructure (easiest and least important part)

- PR, advertising just kidding

Codes have to go through a peer reviewed process before they get approved anyways.

>>975823

A few dozen talented full-time kernel developers or a few hundred talented voluntarian kernel developers. Note that talented kernel developers are six-figure expensive, not just in silicon valley but anywhere in the world.

>

>>Can we figure out the list of skills / roles required?

>

>Skills:

>

>- Excellent knowledge of C

>

>- Good knowledge of driver development

>

>- Ability to read and understand other peope's code

>

>Roles:

>

>- Driver development

>

>- Non-driver kernel development

>

>- Code review, QA, security checks

>

>- Running development infrastructure (easiest and least important part)

How is all that necessary just to keep the repo up to date with upstream?

It would be nice to get a lot of progress done on this fork but we don't need that to begin with. As long as we keep it up to date, it should be good enough.

>>. build a reputation for this project being the true legitimate continuation of The Linux Meritocratic Project,

>

>Costs a lot of $$$

How? How do other open-source not-massively-funded projects build their reputations? I might be wrong, but I doubt Apache foundation, for example, is investing a lot on PR. Doesn't their reputation rest mostly on their projects working well?

>>. serve -reproducible- binaries for all distributions,

>

>$$$$

Hosting fees will be low to begin with and pay themselves once there's enough people interested in keeping the project alive.

>>. make it politically unsubvertible.

>

>$$$

How? How would that take money? If the rules are clear from the start and aren't subject to these types of tactics, it sounds to me like it should be enough.

>>. There's probably a lot of Linux contributors looking for a way out.

>

>I don't think so. What makes you think that?

Read the comments on the CoC drama anywhere where there are open comments. No matter how left-leaning the platform is, most people are on our side what makes you thing autistic devs are going to be less on our side than the general public?

Bumping for interest. If I know these SJWs well enough the most they will do is change names and terms into something more "inclusive" to fit their agenda.

Whatever these assholes do, I assume it wouldn't be too hard to revert. All it takes is someone to initiate a fork.

I'd do it myself, however I'm running on very low-end hardware that can't work with big files. I'd be very happy to contribute in any way I can nonetheless. Right now, forking the project seems like the best idea so far.

I encourage anons to start learning C and kernel development if they haven't already. Most of us are NEET basement dwelling neckbeards with a lot of free time so why not try to make this a thing?

Being a lazy piece of shit isn't an excuse.

>>975897

Fair enough.

This is a good thread but I'm pretty sure the real contributors are going to do something similar soon... r-r-right???

>>975893

Having a desktop focused fork would be nice. It shouldn't take too long to learn about general architecture of the kernel to learn what can be removed (starting with CoC, then other bullshit that has no use on desktop like Hyper-V clipboard support...). Then after code has been removed we build custom pottering-free distribution and we have a perfect system. Somebody should create a logo and a name.

>>975904

I propose we try to put our fork on as many repositories as we can so that people have a choice just like when picking between linux, linux-lts, linux-libre, etc.

As for creating a fork designed for desktops; I can see the appeal but what about users who use Linux on their servers? Why'd we leave them to get assraped by the SJWs? And creating two different forks and maintaining them wouldn't be optimal. The most we're doing here is sharing around ideas, so I imagine maintaining two forks would be hell.

>>975721 (OP)

>I'm not versed enough in these topics to do it myself. I'm just some guy

Hey, hey, hey, you're not "just some guy!" You're an Idea Guy™, and we never have enough of those.

I'd love to help. In fact,

I'LL

MAKE

THE

NEW

LOGO

>>975844

>Read the comments on the CoC drama anywhere where there are open comments. No matter how left-leaning the platform is, most people are on our side what makes you thing autistic devs are going to be less on our side than the general public?

LKML has been remarkably silent on this topic. To me it seems like most actual kernel devs either don't care or can't speak up due to pressure from their employers. Either way they won't support a fork.

>>>. build a reputation for this project being the true legitimate continuation of The Linux Meritocratic Project,

>>Costs a lot of $$$

>How? How do other open-source not-massively-funded projects build their reputations?

>>>. make it politically unsubvertible.

>>$$$

>How? How would that take money?

Don't forget you're fighting a massive propaganda machine backed by jew money. Coraline won't be happy about your fork.

>>>. serve -reproducible- binaries for all distributions,

>>$$$$

>Hosting fees will be low to begin with and pay themselves once there's enough people interested in keeping the project alive.

Hosting is cheap as fuck, but getting linux to build reproducibly in the first place sounds like quite an amount of work. I don't know, maybe it already does.

>>[...]

>How is all that necessary just to keep the repo up to date with upstream?

>It would be nice to get a lot of progress done on this fork but we don't need that to begin with. As long as we keep it up to date, it should be good enough.

Just keeping up to date requires no people at all. A cronjob can do that. But we want to find and keep out bullshit and add code by contributors who are banned from the main project, don't we? So we have to review the code that gets added to the kernel. If we assume almost all of the code that gets accepted into the kernel is good, we can do with a few dozen voluntarian talented kernel devs, but good luck finding them.

>>975904

Derailment: The Post

>>975909

It's also not the point of this thread. The point of this thread is to keep Linux as it is and offer a non-compromised alternative that keeps merit at the center of things.

>>975913

Fuck off, irony fag. This thread is proving so far that there aren't good arguments against us doing this. Despite what it would look like when reading the other threads.

>>975922

Not the anon you were talking to, but:

>Coraline won't be happy

What's that failure of nature gonna do about it? Throw a temper tantrum on Twitter? The only people who will take it seriously are the same people we're trying to isolate from the project.

As for reviewing the code, finding politically related changes isn't gonna be difficult, however because now you can't disagree or say no to someone's shit code or else you're a nazi, there's likely going to be code that will cause issues, or cause a performance drop. This is pretty much the only thing maintainers of this fork would have to actively look for.

>>975929

>What's that failure of nature gonna do about it? Throw a temper tantrum on Twitter?

She wouldn't have gotten as far without active support from major tech companies. I'm trying to say that she (or any woman, let alone blue haired ones) is not the driving force behind the CoCs.

Oh so it looks like I found the designated stormweenie thread.

I too hate women and minorities, do I fit in yet? heh

>>975909

Of course any desktop kernel would be more than capable of running server software, I'm talking about removing stuff like BPF https://www.memsql.com/blog/bpf-linux-performance/for which you probably have no use of, unless you're some large company with massive datacenters (notice /* Copyright (c) 2018 Facebook */ in source) in which case you are most likely already pozzed. Unsurprisingly exploits for this exist https://www.exploit-db.com/exploits/45058/

>>975933

Careful. Your reddit is showing.

>>975932

>she

>>975937

These kinds of features should be optional but still in the source and you should be able to add flags that ignore those parts during compilation.

>>975937

>removing stuff like BPF

>which you probably have no use of

I'd like to keep seccomp-bpf.

>>975940

>>she

my bad

Have a code of conduct that just says "No discussion allowed except directly pertaining to the project. No politics of any kind".

>>975937

Careful. Your fascism is showing.

>>975933

The problem/danger with raiding 8ch is you get exposed to the views and arguments of the posters.

You may very well find that one day you will come back, but this time you agree with many of the posters. I am speaking from experience.

The main problem is that the sort of autists that work on Linux are too easily bullied by women, even "women" (trannies) still have power over them. It's like that pic of the tech company where 7/8 women do shit like work on inclusiveness, diversity, anti-bullying etc and the one guy just writes code.

>>975922

>LKML has been remarkably silent on this topic. To me it seems like most actual kernel devs either don't care or can't speak up due to pressure from their employers. Either way they won't support a fork.

Again. Most people are on our side. Devs are more autistic than most people.

Those who don't speak because of pressure from their employers would probably be very inclined to contribute anonymously to a fork. You'd be surprise how much people dislike not being able to express their opinions.

>Don't forget you're fighting a massive propaganda machine backed by jew money. Coraline won't be happy about your fork.

And yet bad press can only hurt you so much in a time when the press itself is seen as a bad actor. There's little they can do if you stand by your positions. Most people are broken down by gaslighting tactics. Most of us see through that.

Everyone on this thread seems so invested in making this idea seem unfeasible, but with such flimsy arguments.

>Hosting is cheap as fuck, but getting linux to build reproducibly in the first place sounds like quite an amount of work. I don't know, maybe it already does.

I assumed making them reproducible would be simple enough. It's not the main point, though. The main point is being able to distribute it for easy consumption.

>Just keeping up to date requires no people at all. A cronjob can do that.

Then let's start with that. At worst we're going to have a kernel that's just as bad as Linux. We can figure how to build on top of that progressively.

>But we want to find and keep out bullshit and add code by contributors who are banned from the main project, don't we? So we have to review the code that gets added to the kernel. If we assume almost all of the code that gets accepted into the kernel is good, we can do with a few dozen voluntarian talented kernel devs, but good luck finding them.

This can be done in incremental steps as more people join the project. In the meantime we could probably rely on long testing periods to make sure bugs aren't merged into releases, can't we?

>>975721 (OP)

Its in the drivers you git, Intel driver code is completely closed off in a propriety format. Forking OS code doesn't matter any more, they're in the hardware.

>>975952

>equating being fascist with rejecting authoritarianism

My brain is full of fuck

>>975989

left wing authoritarianism == morally good

right wing authoritarianism == morally bad

>>976006

>inb4 mao stalin and 420 pol pot were not real leftists

>>975803

If I was MikeeUSA/Gregory Smith, I'd kill myself, Logan.

How do you plan to achieve these goals especially this one:

>. make it politically unsubvertible.

>>976077

That's the point of this thread. To discuss that.

Having a strict "no personal stuff" or "your feelings don't matter" is one possibility.

>>975782

Doesn't Stallman support GNU/Linux libre? Wouldn't rms reject CoCk because its not free-as-in-freedom?

Then again maybe he wants linux to crash and burn so he can try and snag up some devs to push Turd

>>975782

>it doesn't explicitly reject the CoC, afaik.

RMS rejects the CoC and there for a whole shit ton of the FsF does.

Theres a good chance they won't adopt the CoC simply because RMS is behind it.

>>976115

So RMS is next on the list. ESR will probably get in a fatal car accident in the next month as well. gg no re

>>976093

>>976115

>>976127

Doesn't GNOME already have safe spaces for women. I can't find it now, but I remember seeing that in 2015.

>>976077

Have a code of conduct, that says "The only allowed discussion is about things that directly relate to the project. Any and everything else is not allowed."

>>976135

But anon, the problematic use of the terms "slave" and "master" is directly related to this project.

>>976176

USER WAS BANNED FOR THIS POST

>>975721 (OP)

>Make your case by making a list of the responsibilities a bunch of us couldn't possibly cover. Or STFU.

As you said:

>I'm not versed enough in these topics to do it myself.

C developers are well paid and with plenty of available work, you won't easily find people that can and want to help.

Average anons here don't have the required skills, you can't stack multiple meh programmers to make a good programmer.

If the entirety of the fork was changing a handful of license files, that could be done, but the moment anything technical starts to change you'll have trouble: there will be at least a few poeple calling the change pozzed/bloated/backdoored and then you get scope creep up to 11.

>>975721 (OP)

You know what would be funny? if Linus came back in like a month and forked the Linux kernel himself, then convinced all the main contributors to switch to his fork.

>>975721 (OP)

>As long as we keep this fork shitpost-free and focused on meritocracy, it seems to me that quite a few should be glad to jump the shark.

This, if someone actually starts it please do keep it shitpost-free, even if that means redditors will join in, help should be welcomed.

>inb4 nice LARP

Hey I know most of us are LARPers/webdevs/nodevs, but I also know we have some actual smart and old people here, I think we'd have a better chance than /g/.

>>976348

L-linus?, what if they blackmail you again?

>>976348

>funny

Try 'heroic'.

>>976363

I would like to think that I'd take the initiative if I wasn't some spic living in Zimbabwe 2 that might get shot tomorrow.

>>976348

>he thinks Linus is alive

Why are we forking and not them?

>>976345

>C developers are well paid and with plenty of available work, you won't easily find people that can and want to help.

And yet a lot work for free on open source projects. They not only have enough professional motivation to work on this project but political motivation too. To counter these infiltrators destroying their communities.

>Average anons here don't have the required skills,

The key part of that being "average"

>you can't stack multiple meh programmers to make a good programmer.

That's only partially true. With enough time and testing you do end up reaching essentially the same code.

I guess the part where the experts are irreplaceable is in decision-making. But I still haven't seen anyone make a compelling case for why getting professional devs to join would be impossible.

Can you make your best case for it?

>If the entirety of the fork was changing a handful of license files, that could be done,

That's the starting point. And that should get the project going for a while. I doubt they're going to insert Microsoft binary blobs or Pocket integration into the kernel right away. That should give us time to build something.

>but the moment anything technical starts to change you'll have trouble: there will be at least a few poeple calling the change pozzed/bloated/backdoored and then you get scope creep up to 11.

I'm well aware. That's what I'm trying to figure out. How to get people to join, and how to manage things to avoid these pitfalls.

What do you think would be needed? How many experts / benevolent dictators or, hierarchical structures, or written-in-stone principles would be necessary to prevent the ship from drifting away?

>>976423

delet

>>976428

<you can't stack multiple meh programmers to make a good programmer.

>That's only partially true. With enough time and testing you do end up reaching essentially the same code.

Confirmed for never having programmed something serious. Also your post is shit and wastes 50% of its vertical space with whitespace.

Sure, we alll know that monkey in front of the typewriter, but writing good, manageable and ideally lightweight C code is fucking hard. Especially when it gets close to the hardware. You can't just throw your average FizzBuzzer at this. Also conceptually knowing how one could reach some certain goal and actually achieving it by producing compiling code that does exactly that are two different things.

These times, way too many people just talk about programming instead of actually discussing the fucking CODE.

I'd just like to interject for a moment. "Fork" is problematic terminology that goes against the CoC, because forks are dangerous and potentially triggering. Same with "branch". Please use "spoon" and "river" from now on.

>>976368

You have a better cover than Finish American Linus.

>>976115

>RMS rejects the CoC and there for a whole shit ton of the FsF does.

That is, until RMS is dealt with and Matthew Garrett takes over, and we all know how that will go over.

>>975980

Or get a gym membership and

BECOME THE CHAD

>>975721 (OP)

Or we could skip most of those steps and just start contributing to a *BSD (that isn't FreeBSD) so they're able to grow a bigger userbase and usurp Linux's place. They already have a compiling kernel, they already have a reputation, and they already have a contributor-base.

>>975956

Solution:

Work on sex bots and artificial wombs so female 1.0 can finally be depreciated after the "feminism" bug has proved resistant to fixes.

>>976591

Feminism is a (((backdoor)))

>>976428

>And yet a lot work for free on open source projects.

Because they want to. The broader the appeal of your project, the more you attract.

>They not only have enough professional motivation to work on this project but political motivation too.

Are you hoping for Linux devs to switch over to the fork?

It's not likely to happen, whatever social game is used to keep them working on the main fork will be used to keep them off your fork.

>The key part of that being "average"

8chan is not a big site, not enough anons to find a sufficient number significantly above average.

>With enough time and testing you do end up reaching essentially the same code.

There is not enough time, because you must keep up with the main branch.

Also I doubt tests and time alone can fix lack of coding skills, less readable code can still do the same thing when ran.

>But I still haven't seen anyone make a compelling case for why getting professional devs to join would be impossible.

It's very hard and unlikely, not impossible.

You need to find someone willing to spend time on mostly menial work, on an unpopular project, with the media painting anyone involved as a bad goy, all to keep up with a more popular project.

>And that should get the project going for a while.

Isn't removing the CoC a single change to a single file?

If changing distro name and such takes a significant time, the project is doomed from the start.

>What do you think would be needed?

Divine intervention.

Any governance system tried has failed, and I can't see a realistic way out.

If the next jump to another OS, I hope it's not a fork.. or even GPL. Sounds like a perfect time to give BSD it's rightful place. lol. But something gear towards everyone, not OpenBSD.

I love Linux

https://github.com/rianhunter/wasmjit

>Linux

>enormous amount of cruft, experimental stubs, yet stubborn refusal to nail down driver interfaces so you have to constantly maintain your code to work with the newest revs or risk being dropped for brevity

>even though the kernel appears, like an amoeba, to double in size every 48 hours

>any person who wrote something or something that something's derived from can revoke their license retroactively, essentially cutting out industry and leaving a thriving but zero-sum hobbyist market

>current day commits are STRUCT and that's about it, no new features of consequence for years (check the Linux kernel, it's the most conservative these days and the BSDs are pushing experimental bounds along important lines

It's the sunset of an era and it just shows that GNU is the cuck license.

ITT /pol/thetic retards think up "fork the kernel!" but sadly will amount to nothing because they are useless lazy faggots that would rather spend time "making arguments" for a new/forked/shooped kernal rather than actually writing a kernel. And thus nothing of consequence happens.