Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-4041-

NaCl enabled in the Chrome Beta

Name: Anonymous 2011-08-12 2:57

Native Client (NaCl) has been shipping with Google Chrome for a while now, but it wasn't enabled by default. It has just been enabled in the Beta channel which means it'll hit the stable channel in about 6 weeks or so.

For those that don't know, NaCl is a plugin that runs native code in the browser. For instance, C/C++ code compiled against NaCl will run sandboxed in the browser, and the Pepper library allows your program and the browser to communicate.

I have a bunch of little games I wrote in C that I'd like to port. Last I heard, NaCl had an SDL port, but most of my games are written with Allegro so I'm hoping someone ports that soon if they hadn't already.

http://chrome.blogspot.com/2011/08/building-better-web-apps-with-new.html

Name: Anonymous 2011-08-12 3:00

Chrome offers table salt?

Name: Anonymous 2011-08-12 3:35

Auditing a simple sandboxed VM's security is already hard enough, so the hell with it, why not allow native code to run and expose any and all hardware faults and privilege escalations to fucking browser code. Brilliant.

Name: Anonymous 2011-08-12 7:35

The main advantage of Web scripts is portability across different machines, but you lose it once you use native code. So what's the point of native code Web scripts? You get the worst of both worlds!

Also,
seamlessly
capabilities
leverage
expertise

Name: Anonymous 2011-08-12 10:14

ActiveX was such a brilliant idea it deserved to be re-invented!

Name: Anonymous 2011-08-12 14:41

>>5
pff get with the times, gramps

Name: Anonymous 2011-08-12 14:41

>>4
You can still have your web apps written in pure JavaScript and use NaCl to handle the bottle necks. Take for instance a photo editor written in JavaScript. Some of them will send the data over the network and process the photo server side. You could embed a binary written in C/NaCl to process the photo client-side and write a hook that will use that if the browser supports it or fall back on the old method if it doesn't.

The entire app doesn't have to be written it native code, but native code can be leveraged to resolve performance bottlenecks on systems that support it. This in combination with WebGL, which Chrome has had for a month or two and which speeds up graphic processing in the browser 1000 fold, could lead to some really killer browser-based games, which Google is really pushing for.

Name: Anonymous 2011-08-12 14:47

Fuck NaCL. LLVMCl is the future.

Name: Anonymous 2011-08-12 14:48

Are >>3 and me the only people in the world who see the amazing security issues with this?

Name: Anonymous 2011-08-12 15:05

>>9
Yes, they found more than 10 bugs in their security contest. And it's good enough metric to claim that it is not safe.

Best part:

http://code.google.com/p/nativeclient/issues/detail?id=70

works ok

http://code.google.com/p/nativeclient/issues/detail?id=49

error 403. WOW! Typical security by obscurity! I'm feeling so safe!

Name: Anonymous 2011-08-12 15:19

>>9
No.

http://en.wikipedia.org/wiki/Pentium_F00F_bug

Although I really wish Intel (and AMD) would stop selling CPUs before they had all the errata dissolved.

Name: Anonymous 2011-08-12 15:55

fuck you cose i have no cloue about what u are talking

Name: Anonymous 2011-08-12 16:02

What about sodium chloride?

Name: Anonymous 2011-08-12 16:14

Seriously? This just opens up a whole can of worms, both in security and cross-browser/system compatibility problems. As if we didn't already have enough shit to deal with thanks to IE.

Firefox FTW folks.

Name: Anonymous 2011-08-12 16:21

>>14
I like the first line of your post.

Name: Anonymous 2011-08-12 16:34

why the fuck would anyone complain about the rise of python in the browser???

fuck /prog/

Name: Anonymous 2011-08-12 16:53

>>10
OMG, a security issue that was fixed two years prior to launch. Everybody to the bunker!

Name: Anonymous 2011-08-12 17:33

>>17
FIX THIS!
f0 0f c7 c8

Name: Anonymous 2011-08-12 18:15

>>16
Why would anyone want extra Python in their browser? You can already run it (if you know what you're doing) and Python is kind of a shitty language imho

Name: Anonymous 2011-08-12 18:31

Native Client? Can someone tell me how C/C++ code compiled to VM byte-code running in a VM in a browser can be considered native, because I just don't see it?

It's the same as Google's "Native" SDK for Android... it's not actually native, running straight on the metal... it's just compiling your C/C++ to their VM byte code.

Name: Anonymous 2011-08-12 19:06

>>20
This is actually compiled to x86.

Name: Anonymous 2011-08-12 19:09

>>20
It's not byte-code and it's not running in a VM. It's an actual binary that Chrome pulls from the server and runs on your machine, and yes, you must compile a separate binary for each platform you wish to support.

Name: Anonymous 2011-08-12 21:19

>>21,22
Isn't this ActiveX all over again?

Name: Anonymous 2011-08-12 21:23

>>23
DON'T HELP THEM!!!

Name: Anonymous 2011-08-12 21:28

>>20,23
http://code.google.com/games/technology-nacl.html
Native Client is a sandboxing system. It runs code in a virtual environment where all OS calls are intercepted by the NaCl runtime. This has two benefits. First, it enhances security by preventing untrusted code from making dangerous use of the operating system. Second, because OS calls are virtualized, NaCl code is OS-independent. You can run the same binary executable on MacOS, Linux, and Windows.
But syscall virtualization by itself wouldn't be as secure as Javascript, because clever hackers can always find ways to exit the sandbox. NaCl's real contribution is a software verification system that scans each executable module before it runs. The verifier imposes a set of constraints on the program that prevent the code from exiting the sandbox. This security comes at a relatively small performance price, with NaCl code generally running at about 95% the speed of equivalent compiled code.

Name: Anonymous 2011-08-12 21:42

DOSBox has been ported to NaCl.

http://www.naclbox.com/

Name: Anonymous 2011-08-13 3:08

software verification system that scans each executable module before it runs
http://en.wikipedia.org/wiki/Halting_problem

Name: Anonymous 2011-08-13 3:19

>>27
No, it's theoretically doable, but the human factor is too much of a risk. Untrusted native code simply shouldn't be allowed to run.

Name: Anonymous 2011-08-13 7:48

>>25
Watch out for low-flying pigs.

Name: Anonymous 2011-08-13 8:11

>>25
But syscall virtualization by itself wouldn't be as secure as Javascript, because clever hackers can always find ways to exit the sandbox.
I don't see how, except if there's a security flaw in the kernel.

NaCl's real contribution is a software verification system that scans each executable module before it runs. The verifier imposes a set of constraints on the program that prevent the code from exiting the sandbox. This security comes at a relatively small performance price, with NaCl code generally running at about 95% the speed of equivalent compiled code.
I suppose indirect jumps, indirect calls, and jumps or calls in the middle of an instruction are forbidden. Code making extensive use of callbacks is probably slowed down much more than 5% then, and can even be impossible to compile. I hope I'm wrong.

Name: Anonymous 2011-08-13 9:18

>>30
They'll find a way. If people can punch through VM's like Virtual PC and Virtualbox, they'll do it to this as well.

Name: Anonymous 2011-08-13 9:45

>>10
403
Google, are you even trying?

Name: AnonEEEMouse 2011-08-13 9:49

It seems so.

Anyways, /prog/ can suck my big fat greasy nigger cock.

Name: Anonymous 2011-08-13 10:52

>>33
Back to /b/.

Name: Anonymous 2011-08-13 11:32

Why would you bother with Google code? It's shit.

Name: Anonymous 2011-08-13 12:10

>>34
fuck you faggot

Name: REPLCOCKS 2011-08-13 14:18

(loop (print (eval (read))))
Everyone can go straight to /b/!

Name: Anonymous 2011-08-13 16:50

nackles

Name: Anonymous 2011-08-13 21:44

>>36
And here comes the imageboard paladin jerking off and spraying shit everywhere, his brain too cooked from imageboard retardery to even notice.

Name: Anonymous 2011-08-14 2:09

>>39
fuck off and die, fagstorm

Name: Anonymous 2011-08-14 2:10

>>40
polecat /kebabs/, you ``faggot.''

Name: Anonymous 2011-08-14 2:12

>>41
you cant win, faggot. so pack up your shit and leave, nigger fagstorm

Name: Anonymous 2011-08-14 2:14

>>42
We, (as in The Sussman and myself), have already won. You, sir, are the one who has lost most spectacularly. Good day.

Name: Anonymous 2011-08-14 4:04

>>43
oh yeah? prove it, faglips.

Name: Anonymous 2011-08-14 4:24

>>44
Back to /b/, please.

Name: Anonymous 2011-08-14 5:06

>>45
yeah, that really helps your cause. now I'm going to shitpost TWICE as much. great job, fagstorm

Name: Anonymous 2011-08-14 5:51

>>43
you are morbidly obese

Don't change these.
Name: Email:
Entire Thread Thread List