News for hackers. Stuff that matters.
Re:There's No DRM in JPEG – Let's Keep It That Way by Spivak on Tuesday October 13, 2015 @04:27PM (#10383230)
Wouldn't this DRM require every implementation of the JPEG standard to honor the DRM or am I missing something?
Re:There's No DRM in JPEG – Let's Keep It That Way by TD-Linux on Tuesday October 13, 2015 @04:28PM (#10383242)
Yes it would - but that's how all DRM systems work, plus an attempt to make the image unreadable to "unauthorized" JPEG implementations.
Re:There's No DRM in JPEG – Let's Keep It That Way by bediger4000 on Tuesday October 13, 2015 @04:43PM (#10383346)
How, if some Future JPEG (FJPEG) is a standard, does one make a standard FJPEG image unreadable to unauthorized FJPEG implementations? Either the standard itself has to be kept secret or the FJPEG implementation has to contain some secret, right?

If the standard is kept secret, by maybe charging a tidy sum for it, and getting all buyers to sign an NDA in blood, we'll all have fewer implementations of FJPEG viewers. There would probably be no "open source" implementations of FJPEG. We would all loose, as the market for FJPEG viewers has barriers to entry, and therefore would serve buyers poorly. That's standard, Free Market economics.

If FJPEG viewers have to contain some secret (an algorithm or a key both come to mind as things that have happened in the past) the same sorts of things happen - barriers to entry to sell in the market, higher costs, fewer options to consumers.

Of course the 3rd option is to legally mandate some kind of overall DRM at the operating system level, which would have much the same effects, except that we'd have a narrow choice of Windows or Mac OS, with maybe a larger choice of FJPEG viewers.

All of these options should be repugnant to Free Market believers such as myself. I predict the Republican Congress will laugh this out the door.

Re:There's No DRM in JPEG – Let's Keep It That Way by TD-Linux on Tuesday October 13, 2015 @11:18PM (#10384816)
Yeah, those are the only options. The second option is what Blu-Ray and EME currently do, for example.

But yeah, basically no one is going to use this ever.

Re:There's No DRM in JPEG – Let's Keep It That Way by airza on Tuesday October 13, 2015 @05:02PM (#10383456)
Just like they did with the DMCA, right?
Re:There's No DRM in JPEG – Let's Keep It That Way by bediger4000 on Tuesday October 13, 2015 @06:45PM (#10383990)
I'm sorry, but could you elaborate on this? I'm not understanding the connection with DMCA.
Re:There's No DRM in JPEG – Let's Keep It That Way by mzs on Tuesday October 13, 2015 @05:12PM (#10383515)
Not necessarily, for example a few MSBs of Y, Cr, and Cb DC and AC could be encoded with the current JFIF markers and new ones used the for rest in an encrypted representation. In that manner a current decoder could produce a lower fidelity version still:

http://www.jpeg.org/items/20150910_privacy_security_summary....

Re:There's No DRM in JPEG – Let's Keep It That Way by nickodell on Tuesday October 13, 2015 @04:42PM (#10383342)
No, that would be unenforceable. If your implementation doesn't honor the DRM, they won't provide you with the encryption keys in the first place.
Re:There's No DRM in JPEG – Let's Keep It That Way by MichaelGG on Tuesday October 13, 2015 @05:55PM (#10383761)
Then you end up with a key distro problem. And unless the idea is to do it with hardware-protected keys, it's pretty much DOA for such low-bandwidth, simple-format as a still picture.

Otherwise, as an example, Firefox wouldn't be able to decode these images without some plugin. And that plugin would be just as subject to key-dumping.

Re:There's No DRM in JPEG – Let's Keep It That Way by pdkl95 on Wednesday October 14, 2015 @08:21AM (#10386164)
> subject to key-dumping

Not if you are running a skylake CPU with the new SGX instructions that encrypt RAM.

https://software.intel.com/en-us/blogs/2013/09/26/protecting...

    6, Enable software vendors to deliver trusted applications and updates 

    8. Enable applications to define secure regions of code and data that
       maintain confidentiality even when an attacker has physical control
       of the platform and can conduct direct attacks on memory.
By "trusted", Intel almost certainly means "trusted computing"/"trusted execution environment". It took a decade, but Palladium (NGSCB) is here.

Firefox already caved and ships the EME plugin, so obviously they would cave to an image-DRM plugin as well.

/* this is why it's so incre4dibly important to take a hard-line stance against DRM; it's a lot harder to reverse course once you cave and start accepting some DRM */

Re:There's No DRM in JPEG – Let's Keep It That Way by MichaelGG on Wednesday October 14, 2015 @12:00PM (#10387358)
SGX doesn't provide a way to secure output at the moment. It's awesome for remote attestation. Like, you could run a probably secure Bitcoin tumbler. (Well I'm not sure how the key verification part works.) But it does not provide a way to stop me from pressing PrintScreen.

I agree that they may expand it (MS's waste of engineers during the Vista period in their rush to provide Protected Media Path comes to mind.)

But I'm also saying that web comics could do this today. But I'm guessing that they don't want to sacrifice the UI of just loading images.

Re:There's No DRM in JPEG – Let's Keep It That Way by ekimekim on Tuesday October 13, 2015 @10:57PM (#10384752)
Yes, all DRM systems are inherently impossible to make fully secure with only local data.

However, several other options exist:

* Needing to contact an external server to view an image (of course, you could always take the opportunity to save the now-viewable image then share it)

* Charging a large fee / legal contracts for the information on how to decode (security through obscurity)

* Making circumvention illegal and enforcing this with fines or jail time. This is probably the only option that "actually works" as you've turned the technical problem into a social problem. People would still break it, but you could simply arrest them if they try to help others do the same.

Re:There's No DRM in JPEG – Let's Keep It That Way by MichaelGG on Tuesday October 13, 2015 @11:19PM (#10384821)
I'm not sure these approaches work for images. Certainly web comics could require special clients even now, yet they don't. I'm guessing they want the ability for their pictures to still work normally, but "without saving".
Re:There's No DRM in JPEG – Let's Keep It That Way by ekimekim on Thursday October 15, 2015 @02:53AM (#10391743)
Yes, that none of these systems will ever actually work is the point ;)