/tech/ - Technology

>>>/agdg/ - Amateur Game Development General

>>>/cyber/ - Cyberpunk & Science Fiction

>>>/emacs/ - GNU Emacs

>>>/sci/ - Science and Mathematics

>>>/electronics/ - Electronics Engineering

>>>/laboratory/ - STEM Discussion and STEM Shitposting

>>>/hamradio/ - Amateur Radio

>>>/lv/ - Libre Video Games

>>>/netplus/ - Networks and Plus

>>>/prog/ - Programming

>>>/rice/ - Desktop and Phone Ricing

>>>/t/ - Torrents & Trackers

>>>/templeos/ - The 64-Bit Temple Operating System

>>>/vape/ - Vaporizers

>>>/vir/ - Virtual Reality

>>>/wg/ - Wallpapers General

Is this motherboard compatible?

http://www.newegg.com/Product/Product.aspx?Item=N82E16813121515

I have an i5 2400, and an NVIDIA GeForce GTX 450 with PCIe 2.0.

>>548797

>thinkpad

>gaymen

they're netbooks. they're not designed for loads like that.

>>543862

I have no problem with SystemD, just the way they've shoehorned it into everything. People should be free to choose the init system they want.

I personally have no problem with SystemD and have found some of the features it brings to the table useful. This is not for everyone and it should be opt-in, not forced by RedKike.

Gentoo has it right, you are not forced to use it, but the option exists if you want/need it.

>>543872

This is a babby's first distro topic, not a "my autism distro is better than your autism distro!!!!1111 " topic.

OpenSUSE

Kubuntu

>>545844

>fucking md5

>Not GPG or a way stronger checksum. E.g. sha256

Mint

>>547951

Servers are indeed important, but there are a lot more clients than servers by nature, and GNU still has to compete with BSD and to a somewhat lesser extant Windows Server in that field, so even there the overall importance of GNU is dwarfed

GNU and their culture of incompetence I genuinely believe has been one of the factors holding back Linux on desktop, whereas the Linux kernel used in other OS' like Android exells elsewhere

GNUs signature ecosystem was designed by computer engineers and enthusiasts for computer engineers and enthusiasts, people willing to put up with dependency hell to be able to compile their own software and other such shit other OS' like Windows and Android don't have to deal with in exchange for less customizable subsystems. It's not nor will it ever be ready for the average users needs

>>547693

> the earliest version of a company was compiled

I stopped reading here. Linuxfags have literal autism.

>>547977

I'm an average user, quite happy with Xubuntu 14.04 after the clusterfuck that was my Windows 8 experience.

>>547522

Those two concepts aren't remotely similar to one another.

One is observing that productivity hasn't gone up when it seems like it should. One is predicting that the internet will have a comparable impact on our lives as the fax machine.

Hardened kernels are useful in any applications where security is important.

Because most hardening mechanisms exhibit minimal slowdown, and because adding compiler flags and recompiling your kernel isn't particularly difficult, it is advisable to utilize hardening mechanisms wherever possible.

The only cons are slightly increased memory usage and reduced execution speed for some hardening mechanisms.

>>548189

>Container is compromised because of buffer overflow vulnerability

>Attacker escalates privileges because all security measures were relegated to the compartimentalization software

>Get pwnd anyway

You know grsecurity does way more than permission control, right? Fuck, you can use it for containers too.

>>548208

Are you seriously considering coding your own patches? Why not go all the way and code your own kernel from scratch too?

>>548318

>The only cons are slightly increased memory usage and reduced execution speed for some hardening mechanisms.

Someone should tell the Debian maintainers about that. More specifically, the Iceweasel maintainer seems to have little to no idea about hardening.

Thanks for your responses.

What does hardening do? How could I be attacked, by whom, and what does a hardened kernel do to prevent such attacks?

Are we talking a teen hacking the computer, or government agency, how would they do it, and how does a hardened kernel prevent that.

I am using linux-grsec 4.4.6 on Arch, just in case.

>>548818

Hardening is about preventing exploits from working.

For example, the linux kernel exploit that people were shitting themselves about a few weeks(?) ago, to do with a refcount overflow in the keyring functions in the kernel, was completely prevented by grsecurity.

That was prevented by a refcount overflow checking plugin for gcc, used when compiling the grsecurity kernel. If a refcount overflows, it makes the object indestructible. This results in a small memory leak triggered each time the exploit is attempted, but it's better than the alternative of allowing kernel-level access to the machine (...while still having the memory leak).

You can start here for a list of options presented to you when compiling the grsecurity kernel: https://en.wikibooks.org/wiki/Grsecurity/Appendix/Grsecurity_and_PaX_Configuration_Options#Non-executable_pages

That should give you an idea of the type of protections offered. Everything from enforced ASLR, to RWX mprotect banning, to fixing the double-chroot exploit, to adding a sysctl option which lets you ignore all new USB devices.

It has a lot of features and it would take me a long time to list them here. I will say that the RWX mprotect ban is great to prevent some ROP-based exploits (as they can't mark their real payload executable so they have to rely on extensive ROP gadgets which probably don't exist), and the improved ASLR makes remote code execution exploits very difficult even if everything was RWX. On top of that, the "bruteforce prevention" (slowing down a program's re-launching if it crashes a lot) stops people from trying to guess where ASLR has put a specific address.

There is of course a lot more that you'll see if you read the link above.

I think grsecurity is a good choice of kernel. Selinux is a different kettle of fish, and it doesn't actually cover this kind of hardening at all. It's mostly about MAC, but I think if you don't harden your kernel at the very least then tPost too long. Click here to view the full text.

>>548189

SELinux is a SHIT, disproportionately invasive implementation of labeling MAC pushed completely needlessly and inappropriately onto regular users and run-of-the-mill servers by Red Hat, that's only appropriate in formal organizations and at that only because there's no real alternative in that scenario.

99% of desktop PCs (that warrant a MAC) and 95% of servers out there are better off with another MAC like AppArmor or Grsec RBAC.

testing%20my%20broken%20ass%20script

Keep getting an invalid board error.

>>548823

Are you the one creating the 8ch poster?

>>548777

This.

I will use a WRT54GL if I want. As long as I can replace my router and install free firmware on it.

"</tmp/msg.txt"

Pretty sure you can't make a router from a raspberry pi.

Just make a PDA instead.

>>548749

>Literally no excuse

I am /tech/ illiterate. I don't understand the significance of the raspberry pi, let alone how to make one into a router.

>>548841

wrt54g master race

>>548692

About mid last year HW hadn't updated his Arch install for over 9 months if I remember correctly. Better times.

>>548800

Nice couter argument. You totally haven't proven his point.

Shame such nice digits had to be given to a extremely sub par memer who can not even make a proper argument.

>>>/oven/

>>548830

I said

>newer, less tested, potentially buggier packages

it's more than time - it's the other effects that time causes. If a buggy package sits on your system for 6 months without any changes will still be buggy. Older releases have had more time to be bug-tested.

this is pretty simple stuff tbh

>>548670

In my whole year of using Arch, I've had startx break only once.

Pretty sure this was due to me going ape on the keys, too.

>>547885

>almost certainly nonfree firmware

It's still possibly a botnet.

>NSA designed

Trusting the NSA, ever.

I have more respect for the CIA.

Not from Amercia.

>>548077

Crazy secure puts it above both wicked and hella secure on the NSA phone security scale.

>>547850

She looks like Grimes

>>547850

what habbened?

>not using an NSA-designed phone that runs windows

I could just use a regular windows phone and get the same result

There are mac zombies so obessed with Apple products they are now making fake reviews of products that don't even exist and they even go as far as printing out fake packaging and 3d printing a device for the review...

I mean seriously what kind of person does this shit? Just look: https://www.youtube.com/watch?v=6ZNlGbVy67Y

>>548828

>reversing erectile dysfunction

>reversing type II diabetes

oh lordy

also, what's wrong with the first pic? Those temps are a tad on the high side for the load but its not something unusual or non-typical or anything to panic over, I've seen worse tbh

>>548848

>also, what's wrong with the first pic?

Nothing, actually, except that it was a $5000 trash can (Mac Pro 2013).

Very interesting design TBH, and with Windows 8.1 ran like a champ. Just couldn't be upgraded whatsoever unlike the classic Mac Pro.

>>548832

>Once you know what you're doing, they're actually very easy to work on in my opinion,

>sometimes just some ridiculous shit Apple engineers would come up with seemingly only to fuck with technicians.

that's how I felt when I took apart my MBP 5,5 for the first time, ribbon cables, the odd screws, the weird battery connector, etc. the stock paste job on it was terrible, just putting new paste on it dropped the normal idle 10C. didn't change shit about it overheating on load though, such a shame.

>>548856

was everything soldered on? you'd figure at least the GPU and SSDs could be upgradable.

>>546706

Hinduism has become the second largest religion as Indians have completely replaced the white collar workforce in America.

>>546706

you just got ran over by a 2nd generation syrian while walking in a crosswalk

<iframe width="560" height="315" src="https://www.youtube.com/embed/IFe9wiDfb0E" frameborder="0" allowfullscreen></iframe>

Wow I'm a 72yo virgin

>>546720

See you in he(aven)(ll)

>>546762

>since it it's more acceptable to be a slim girl than a slim boy

>no cuteboys to gawk at

Time to produce the new super race.

>>546844

m8, >>547438

Also,

>>547066

>Richard Stallman uploads himself to the waifunet and lives forever

That's just a clone of him, a pretender. Read permutation city.

>>547271

And then?

>>547551

bretty gud fam.

Why are you putting information that you want to keep secure, on a fucking cellphone? Are you an idiot or something?

The confidentiality problem of android apps is a popular CS research subject since 2011, you can search google scholar with keywords like "android privacy leaks" and it will yield a lot of results.

There are some related open-source projects about program analysis that I know of:

https://github.com/pjlantz/droidbox

https://github.com/androguard/androguard

http://wala.sourceforge.net/wiki/index.php/Main_Page

Also there exist some noob friendly products on market, but most are expensive and enterprise targeted.

>>548438

I personally use replicant now and it's really nice to see that there is no play store installing itself some new update without asking you.

Only replicant can say that they have almost solved the problem (waiting for a phone with libre hardware).

But wile we have to use satellites with non libre hardware and software controled by govs no phone will be secure

https://en.wikipedia.org/wiki/Replicant_%28operating_system%29

>In March 2014, Replicant developers found and closed a backdoor present in a wide range of Samsung Galaxy products that allows the baseband processor to read and write the device's storage, sometimes with normal user privileges and sometimes as the root user depending on device model. It is unknown whether Samsung's proprietary firmware for the radio chip can be remotely instructed to use these access features and whether the vulnerability was introduced with legitimate uses in mind.

>have no google products

>install xprivacy and set all apps to request usage of all permissions

I've caught 2 apps that were only using internet to send analytic data or advertising data.

The best solution, though, is to use open source software.

Root your shitbrick and uninstall all the google apps. Now throw it out your window and go pick up an old nokia or something.

>>548759

I remember hearing them saying that it will be important to let people swap out parts of the network stack without compromising basic things, for instance things like what encryption model you want to use, what kind of peer routing system you want (this is where i2p comes in), etc..

I know that kind of model is the goal but I have no idea how close they are to that right now, the developer said something about trying to be like IP and have a thin waist where only some things are standard but everything else could be improved as needed later or swapped out to fit preferences/requirements.

I'd assume once they get message passing working it wouldn't be too hard, you just have to make sure some peers allow all supported methods and pass messages. Like if I didn't care about privacy but wanted to help those that do I could act as a relay for i2p nodes. That's my speculation, maybe they have a more elegant idea in mind that is more resistant/safe.

>>548574

There's really no difference beyond scale. Home routers have built-in DHCP servers too. Many ISPs even leech internet access from bigger ISPs.

>>548759

>Copy everything you like on the Internet to the distributed IPFS network instead

I'm really annoyed how people keep talking about IPFS like it's some kind of bag of holding. It's not. It's basically a very complex stack that just provides a torrent-like CDN solution. It doesn't magically provide any extra availability over traditional client-server internet, it's still same plain old computers serving plain old content. If you don't have anyone to serve content, it doesn't work. If computers that can serve content can do it poorly, it works poorly. You still need traditional servers and datacenters for it to work.

>>548773

It absolutely provides more availability in several ways, the most obvious one being content addressing. With IPFS you request the content, it doesn't matter where it comes from. With traditional client-server you're still at the mercy of a domain and your specific CDN servers.

The other obvious one is "ownership", if you want a video on youtube.com you have to hope the youtube server is up so you can pull it from them, again when things are content addressed it doesn't matter, if you want the video from there and the original content host is no longer online, anyone can fill that role as long as they have the content.

It's strictly better than the traditional model, it only has the potential to be more reliable without sacrificing anything, if you're the only perr hosting it then it's no worse than http while also gaining content addressing. If you use a CDN then again you can just plop down more ipfs nodes to act as them in other parts of the world, no sacrifice, at worst it's the same but it has the potential to be better. For each visitor you have that's another point of redundancy and geographical distribution.

There's more things as well that exist now but one of the future plans is to have an opt in distribution cache much like freenet and perfect dark. So I can say "I'm offering 100GB of disk space" and my client will pull in 100GB of random data from other IPFS peers for redistribution.

The other important thing to consider is that not everyone has the ability to host content on a CDN. In the same way torrents empowers individuals to easily share files this does for site hosting. From the perspective of an individual, hosting a website or whatever should get easier as traffic goes up unlike the traditional client-server model. It makes it actually feasible to have a popular website hosted out of home.

>>547524

Get yourself a business class connection from your local isp the. Set up multiple wifi routers around your house to cover your neighborhood then sell your wifi to your them.

>>548849

>>548852

Seems like your convincing yourself, try new stuff man and stick in the environment.

>>548854

I have aero disabled too. Also it's totally not the xp taskbar, it's completely different because the launchers and tasks are combined.

>>548855

I've tried lots of stuff. I think Windows 7 is the comfiest os/de i've used.

>>548858

Then why do you have this thread?? Its really just heres my opinion I will now defend it.

>>548859

Because I care for linux and I want it to suceed. And in my opinion this is where at least part of it should be moving to.