/tech/ - Technology

Where proprietary software comes to die

Boards | Catalog | Bottom

Drawing x size canvas

Remember to follow the rules

Max file size: 350.00 MB

Max files: 5

Max message length: 4096

(292.67 KB 450x399 RMS.png)
/tech/ - Technology Anonymous 01/03/2016 (Sun) 17:52:28 [Preview] No. 125 [Reply]
Welcome to /tech/, a technology board. This board is for discussion of technology, both software and hardware.

Wiki: https://wiki.installgentoo.com/
IRC: #/tech/, #InfinityNow and #endchan on irc.rizon.net - https://qchat.rizon.net/

NSFW files are only allowed if they are spoilered. Tech support, consumer advice and desktop/ricing threads are all allowed for now.

All rules and policies are open for discussion in this thread.

You can view the logs in a less annoying way at https://8ch.tech/endlog/tech.html (available for all boards).

Related boards:
Edited last time by ring on 06/30/2016 (Thu) 01:50:02.
157 posts and 23 images omitted.

Anonymous 12/18/2016 (Sun) 23:52:58 [Preview] No. 7684 del
(249.95 KB 1200x972 1460919021574-0.jpg)
What if the user could enable all images to be spoilered if they are at work/class/outdoors or keep them unspoiled for at home. Additionally, when posting there could be an option to let spoiler anons know if the image is graphic. This could change the color of the spoiler or something to that effect.

Just a thought to give more options for the sake of compromise. Since an image can be questionable without being graphic.

Anonymous 12/19/2016 (Mon) 01:32:23 [Preview] No. 7688 del
That would be nice. Try suggesting it to StephenLynx.

Anonymous 12/19/2016 (Mon) 01:45:48 [Preview] No. 7689 del
I already suggested odili to implement media hiding and showed him that I implemented it on penumbralynx. Is up to someone add it to endchan's front-end now. You can see it working on http://spacechan.xyz

And the alternate spoiler is not practical. Re-spoiling anything is too fast and one should assume anything spoiler is almost guaranteed to be NSFW or graphical.

Anonymous 12/19/2016 (Mon) 01:46:51 [Preview] No. 7690 del
Tell the staff they would have to manually approve posts.

I am certain that no one would even consider the prospect of sitting and approving posts so the board could be used at all.

anal sex 01/05/2017 (Thu) 22:22:33 [Preview] No. 7788 del
(4.29 KB 255x197 index.jpeg)
How do you do, Stallman?

Anonymous 08/07/2016 (Sun) 16:15:07 [Preview] No. 6120 [Reply]
BSD Thread.

Let's have a thread about the permissive, copyleft style of operating systems known as BSD.

Discussion of Free,Net,Open,Dragonfly,Edge *BSD is welcome!

Linux and FSF users who would rather complain about licencing and that linux is better need not apply.


FreeBSD Quarterly Status Report: https://www.freebsd.org/news/status/report-2016-04-2016-06.html

Vulns found in freebsd-update and portsnap: https://lists.freebsd.org/pipermail/freebsd-security/2016-July/009016.html

OpenSSH 7.3 Released: http://www.openssh.com/txt/release-7.3

Message too long. Click here to view full text.

43 posts and 11 images omitted.

Anonymous 01/15/2017 (Sun) 18:04:21 [Preview] No. 7838 del
I've been toying with openbsd's pledge like functionality for linux.

As Theo said in his original presentation it would be trivial to implement it with (lib)seccomp infrastructure by grouping system calls.

The problem is, while openbsd has only one standard c library, linux distros have more than 4 choices for std libc(Musl, uclibc, dietlibc and glibc. standard conformane and general code quality in that order), making the said grouping interesting subject.

openbsd's pledge grouped system calls in a way that makes sense for them(things they learned while investigating their base utilities) but I took different approach and grouped system calls based on how musl implements posix apis.

hope I can settle things down with stali developers to get actual testing going.

Anonymous 01/16/2017 (Mon) 07:18:12 [Preview] No. 7848 del
LibertyBSD looks like an interesting project


Anonymous 01/16/2017 (Mon) 17:48:16 [Preview] No. 7851 del
(92.24 KB 420x420 nah.png)

if you are not intel employee who can review and compile microcodes DO NOT BUY hardwares that require said microcodes to function properly.

said microcodes DO NOT EXECUTE in kernel privilege and they DO NOT fuck with kernel data structure.

they run INSIDE your shitty hardware. OpenBSD kernel READS said microcode and LOADS it inside your shitty hardware's INTERNAL MEMORY which is then EXECUTED BY your shitty hardware's OWN shitty processor.

non libre LINUX kernel ships with BINARY ONLY drivers that RUNS with KERNEL PRIVILEGE and said microcodes.

OpenBSD kernel DOES NOT ship with BINARY ONLY drivers.

Anonymous 01/17/2017 (Tue) 20:09:01 [Preview] No. 7861 del
What about hardened BSD on laptop?

DIY, Modular, Open source Laptop Anonymous 01/03/2017 (Tue) 09:21:19 [Preview] No. 7768 [Reply]
Laptop form factor open source hardware design, built from off the shelf components for DIY, modular, open source computing.

CPU, RAM, graphics: 1x single board computer
FSB, ports: 2x powered USB hub
Battery: 1x USB powerbank, >5000 mAh, >2A
Input: 1x USB keyboard
Input: 1x USB trackpad
Storage: 1x USB flash drive
Network: 1x USB WiFi dongle
Audio: 1x USB sound card/headset
Display: 1x USB powered HDMI monitor
Case: 3D printed or DIY glued wooden case

-Completely modular, customizable, upgradeable and repairable
-User serviceable by all users
-Universal, modular battery and charger. Can swap battery when depleted

Message too long. Click here to view full text.

12 posts and 1 image omitted.

Anonymous 01/11/2017 (Wed) 11:34:17 [Preview] No. 7819 del
>muh intel

I am typing this very post in shitty Chinese board based chromebook so what are you trying to prove?

that you never worked on arm devices?



>muh usb runs fasta than copper substrate

usb sends command and control data through same line.

You can't have 480 Mbps purely for data. at maximum it would be something between 400 Mbps to 300 Mbps (for usb 2.0) for SINGLE connection at BEST(i.e. it's still theoretical value).

Message too long. Click here to view full text.

Anonymous 01/11/2017 (Wed) 11:50:42 [Preview] No. 7820 del
(30.03 KB 300x250 brighter .png)


>There are several reasons why your device will never be able to use all of this bandwidth. First of all, the USB bus is shared among several users. Even if you are plugged into different ports on the motherboard, you are probably sharing the same host controller as all of the other devices on the bus, so your device is sharing the USB bus bandwidth with all of the other devices.

>Second, USB is a packetized protocol where longer blocks of data are divided into 512-byte packets. Each packet contains a header identifying the packet contents, and a CRC at the end of the packet for data integrity. Each packet also requires an ACK from the other side of the link. Start of Frame (SOF) packets are sent every 125 uSec (microframe) to maintain timing on the bus. The net effect of this is that the theoretical maximum bandwidth of USB is 13 bulk packets per microframe, or 53,248,000 bytes/second. Even this limit is not achievable with current host controllers, which can receive 10 bulk packets/microframe or send 8 bulk packets/microframe.

>500mA – This is the absolute maximum power allowed under the USB spec

don't listen to haters op!

bright future awaits you!

Anonymous 01/12/2017 (Thu) 16:08:24 [Preview] No. 7825 del
There are also x86 SBCs out there. Many of them also have LVDS connector for display, which opens up some options with used panels from laptop PCs (although I'm not sure everything is compatible).

Just look into embedded pico-ITX

I want to build a UMPC with Pentium N 3710 (Braswell) based pico-ITX board. Although this would be relatively expensive it has a much more capable GPU than any of the ARM stuff out there. x86 and real OpenGL/DirectX acceleration should make a lot of neat things simple to achieve.

Also for rapid prototyping of enclosure, forget wood unless you're really good at working with that. Look into Sintra PVC foam board. I've worked with it in the past as enclosures for power supplies and other projects. That was with 6mm thick sheets. Really easy to work with, does not take a lot of tools. You basically would want a T-Square or Triangle and some sharp knife or razor to score with then break it. Anywhere that you need to attach a component through a screw hole you would just drill a very small pilot hole and use a small wood screw. The PVC foam board can be glued together with a wide variety of adhesives but thin cyanoacrylate works better than most. That's just regular super glue.

I'm probably going with 3mm this time considering the UMPC will be small I think that's sufficient, although it will be built in layers that are glued together, thicker than 3mm in many spots.

If done properly this could have excellent results and be achievable without spending a lot of time learning 3D modeling or CAD. Or having to spend all that time designing a 3D model and send it off for printing not quite 100% sure if it's going to all fit together just right.

I'm still looking for a good option for the hinges, friction hinges can be surprisingly expensive.

Anonymous 01/12/2017 (Thu) 16:13:40 [Preview] No. 7826 del
A few companies that manufacture and sell x86 pico-ITX SBCs (and have prices and ordering available on their site):

(Israeli company, but if you're buying Intel that's irrelevant)

(DFI is a Taiwanese company that has always made very high quality motherboards. They used to be famous for excellent gaming motherboards during the socket 939 era, but these days focus on industrial and embedded stuff)

Anonymous 01/17/2017 (Tue) 08:58:36 [Preview] No. 7860 del
Specifications for >>7814
CPU: 1 Ghz A31S ARM Cortex-A7 Quad-Core
GPU: PowerVR SGX544MP2
Display: M155 GAEMS 15.5" LED TN LCD, 1366 x 768
Network: 10/100/1000 Ethernet, WiFi 802.11b/g/n
Battery length: 5 hours (10000 mAh @ 10 W)
Ports: 7x USB 2.0, 2x 3.5mm Audio output, 1x 3.5mm Audio input, 1x HDMI
Weight: <2 kg
Dimensions: 38cm x 24cm x 3cm (folded)
Total cost: $254.99

>I am typing this very post in shitty Chinese board based chromebook so what are you trying to prove?

If you used the linked chipset, then with nonfree wireless and video drivers?

Message too long. Click here to view full text.

(10.70 KB 109x130 gccegg-65.png)
Compilers Anonymous 11/24/2016 (Thu) 12:48:06 [Preview] No. 7468 [Reply]
Can someone, please, "red-pill" me in compilers. Particularly, C compilers.
Why use anything besides GCC? What's the difference between all of them?
12 posts and 4 images omitted.

Anonymous 01/15/2017 (Sun) 08:08:41 [Preview] No. 7832 del
i was reading that gcc sucks at some point.... and that clang is better... i dont know personally why, i used gcc and it was ok, i am now on fbsd and using clang and its okay also... dont know

Anonymous 01/15/2017 (Sun) 09:03:55 [Preview] No. 7833 del
Are we really having this thread? Remember the DEPRECATED meme? That was based off of the anti GCC pro Clang sperglord.

Anonymous 01/15/2017 (Sun) 09:06:34 [Preview] No. 7834 del
(311.56 KB 612x792 1352501355093.png)

Anonymous 01/15/2017 (Sun) 12:52:37 [Preview] No. 7835 del

>typical linux fanboy

nobody said LLVM/clang is better than gcc in this thread so far.
(except for brief mention in >>7469)

General consensus was that if you are not kernel/os developer you may as well shut up and use whatever your distro provided.

What are you trying to prove beside your incompetence?

Anonymous 01/17/2017 (Tue) 07:56:13 [Preview] No. 7859 del
redpill me on chiru.no

crypto fails meta thread Anonymous 01/17/2017 (Tue) 01:31:24 [Preview] No. 7855 [Reply]

curve25519-donna implementation by google is susceptible to timing attack as MSVC runtime's 64bit multiplication is nonconstant time operation in x86 environment.

Anonymous 01/17/2017 (Tue) 01:41:22 [Preview] No. 7856 del

Anonymous 01/17/2017 (Tue) 01:44:03 [Preview] No. 7857 del

openssl command line's enc kdf is WEAK

Anonymous 01/17/2017 (Tue) 02:23:57 [Preview] No. 7858 del

Exposing HashMap's iteration order can cause O(n²) blowup even in innocent-looking code without the presence of an attacker.

In the presence of an attacker, access to the order of a dictionary allows HashDoS-like attacks with only two requests in common scenarios.

(128.61 KB 1200x675 chip.jpg)
Pocket CHIP Anonymous 05/15/2016 (Sun) 17:39:56 [Preview] No. 3593 [Reply]
Does anyone here have experience with the Pocket CHIP, or the CHIP in general? As an option for mobile computing it seems much cheaper than things like the Dragonbox Pyra and OpenPandora, but much more powerful/useful than a chroot on an Android touchscreen device. Is it good? Are there similar computers?
15 posts and 1 image omitted.

Anonymous 01/15/2017 (Sun) 21:30:38 [Preview] No. 7842 del

does 3d acceleration work out of the box?

I heard CHIP team were working with ARM for bringing mali gpu support for mainline linux.

can you share contents of dmesg, lsmod and uname -a?

Anonymous 01/16/2017 (Mon) 05:22:53 [Preview] No. 7846 del
It's mostly a problem that there's almost no existing worthwhile software for linux that uses OpenGL ES

But yes the GLES acceleration works now, it's pretty much limited to Quake 3 and a few other things. After I bought it I did the research and you would be surprised just how few open source engines and games support GLES (on linux, there are quite a few for android)

Anonymous 01/16/2017 (Mon) 17:52:35 [Preview] No. 7852 del

so I heard. tbf it wasn't like we had that much hardwares capable of GLES for general purpose boards (outside of android phones that is).

Did CHIP team ported mainline linux as promised?

Anonymous 01/16/2017 (Mon) 19:39:24 [Preview] No. 7853 del
I'm not sure the latest but I think the main sticking point is NAND drivers

Anonymous 01/17/2017 (Tue) 00:05:42 [Preview] No. 7854 del
Actually this thread has me wanting to mess around with the pocket chip some more. It seems to me that a game that I've compiled on the chip before (Cataclysm DDA, SDL) should work automatically with GLES. But it was completely software rendering last time I tried. Unless the MESA libraries in the repos were compiled without the right options.

(4.89 MB 500x281 1446448543096.webm)
Webm thread Anonymous 05/23/2016 (Mon) 23:08:37 [Preview] No. 3991 [Reply]
Post your /tech/ related webms
1 post and 1 image omitted.

niggerfaggot 06/27/2016 (Mon) 01:09:38 [Preview] No. 5546 del
(283.99 KB 480x360 P2P.webm)

Anonymous 06/30/2016 (Thu) 22:55:28 [Preview] No. 5679 del

Anonymous 01/09/2017 (Mon) 16:00:16 [Preview] No. 7810 del
(3.45 MB 450x360 wana dance.webm)
(3.85 MB 1280x720 watchdogs.webm)
(3.32 MB 296x200 win98.webm)

Anonymous 01/09/2017 (Mon) 16:09:45 [Preview] No. 7811 del
(3.67 MB 854x480 the matrix.webm)
(1.46 MB 324x240 missle.webm)

Anonymous 01/16/2017 (Mon) 08:41:00 [Preview] No. 7849 del

(589.86 KB 770x435 opera.png)
Opera neon Anonymous 01/15/2017 (Sun) 13:34:12 [Preview] No. 7836 [Reply]
https://youtube.com/watch?v=t4MikGVvQOA [Embed]

This looks absolutely horrific

Sage Sage 01/15/2017 (Sun) 16:04:29 [Preview] No. 7837 del
>Blog type post
>Stating your opinion

Anonymous 01/15/2017 (Sun) 18:48:00 [Preview] No. 7840 del
proprietary sfotware is GROSS

Anonymous 01/16/2017 (Mon) 07:05:36 [Preview] No. 7847 del

>le edgy greentext
>implying your opinion of OP

Anonymous 01/16/2017 (Mon) 17:32:04 [Preview] No. 7850 del

(751.73 KB 640x400 gentoo.png)
Why haven't you installed gentoo yet? Anonymous 12/16/2016 (Fri) 17:36:10 [Preview] No. 7636 [Reply]
19 posts and 3 images omitted.

Anonymous 01/04/2017 (Wed) 00:47:31 [Preview] No. 7776 del
because sabotage is marginally better

Anonymous 01/04/2017 (Wed) 01:24:11 [Preview] No. 7779 del
fuck this whole 'rolling' release distributions that change compiler once every motherfucking week just to see how gcc developers fuck up this time.

Anonymous 01/04/2017 (Wed) 01:37:32 [Preview] No. 7780 del
>over a thousand lines of code
oh no

Anonymous 01/04/2017 (Wed) 03:57:34 [Preview] No. 7781 del

but why would you run Gentoo on production server?

good to see someone else running hardened gentoo on Chromebook. Are you using arm based one?

I got Asus rockchip base Chromebook last month and it's working great with musl hardened

Anonymous 01/15/2017 (Sun) 22:29:42 [Preview] No. 7845 del

Tor Anonymous 12/20/2016 (Tue) 16:00:02 [Preview] No. 7698 [Reply]
>Single Onion Services for muh "i don't need anonymity"
>decentralized RNG for muh cryptography wonders
>removing "tap handshake" old stuff in favor of ntor

>meanwhile block bypass still gonna eat your post and say "fuck you" to uncommon browser
15 posts omitted.

Anonymous 12/27/2016 (Tue) 12:05:22 [Preview] No. 7739 del
Actually i did too.

Anonymous 12/27/2016 (Tue) 14:07:24 [Preview] No. 7740 del
I heard that if you girlfriend has a feminine penis then you should eat more vegetables.

HTTPS at hidden services Anonymous 01/01/2017 (Sun) 00:11:24 [Preview] No. 7758 del
>Subject: businessCategory=Private Organization,jurisdictionOfIncorporationCountryName=US,jurisdictionOfIncorporationStateOrProvinceName=Delaware,serialNumber=5019303,STREET=20 Paoli Pike,postalCode=19301,C=US,ST=Pennsylvania,L=Paoli,O=Duck Duck Go\, Inc.,CN=3g2upl4pq6kufc4m.onion


>tfw https made friendly privoxy install incapable of removing browser headers, browser using that proxy sends something "bad" and fuckduckgo responds with "forbidden"

Anonymous 01/15/2017 (Sun) 21:26:13 [Preview] No. 7841 del
The trouble with Tor is the people treating it like a big truck that you just dump something on, when it is really a series of tubes, and when those tunes get filled your Internet can be delayed.

Anonymous 01/15/2017 (Sun) 22:11:22 [Preview] No. 7844 del
(164.55 KB 316x315 Al_Gore_on_Futurama.png)
Tell me more...