Profile cover photo
Profile photo
Martijn Oostdijk
Martijn's posts

Post has attachment
New version of our app is available from Play.

Post has attachment
Martijn Oostdijk was tagged in Martijn Oostdijk's album.

Post has attachment
Seminar Privacy in the Internet of Things 24 November in Utrecht. Een COMMIT seminar georganiseerd door Innovalor. Over grip krijgen op het gebruik van persoonlijke gegevens door anderen. Sprekers: Maarten den Braber - quantified self, Linda Kool van Rathenau – eCoaching,  Jaap Henk Hoepman – PI Lab – privacy design, ob Hulsebosch – InnoValor – privacy policies, Joris Janssen – Sense – platforms voor privacy enforcement. Meer informatie en inschrijven:

Post has shared content
Trustworthy Computing is dead, long live Trustworthy Computing!

The news came through yesterday that as part of a series of layoffs and reorganizations, Microsoft had shuttered Trustworthy Computing as a standalone business unit.  In the midst of the lamentations on Twitter, John Lambert assured us that “all your TwC is still here. SDL, operational security, pentest, MSRC, Bluehat are just under a new roof,” but the change is still worthy of note.

We don’t do a lot of reflection on the past in the security industry, unless it’s about failure.  But the creation of Trustworthy Computing a dozen years ago as a standalone unit and the top engineering priority at what was the most important software company in the world has had a monumental impact on the landscape of information security.

The full story is of course bigger than I can go into here, but the computer security profession pre-TwC was in what I’ll call the “l0pht/@stake era”.  This was roughly equivalent to the time of the Spanish Conquistadores in Mexico and Peru. Giant empires of software and developers could be laid low by a few elites wielding offensive technology (and unwittingly unleashing the occasional plague) that seemed like magic to the hapless defenders. We tried to keep the hackers out, but without really understanding how their methods worked, instead just digging deeper moats and building higher walls.

When I was trying to make information security my full time specialization in that era, the CISSP training that was the best the mainstream had to offer was cargo-cult stuff about firewalls, system audit, and password policies. Information security was regarded by ISC2 as an insufficiently demanding discipline on its own that a certified professional obviously should also be proficient at things like disaster recovery, physical security and choosing the right kind of data center fire suppression systems.  Meanwhile, I can’t remember seeing a single mention of writing secure code or software testing – and I read all the recommended materials, amounting to thousands of pages, not just the official study guide.

Trustworthy Computing was one of the most important forces in changing all of that.  It was the first large-scale private sector effort to approach information security as both a science and a systematic engineering discipline, to do serious root cause analysis of incidents, and to use that knowledge to redesign the processes by which software was created and maintained. And an important and under-appreciated part of the work was that Microsoft in those years brought in, as employees, contractors or Bluehat guests, every talented hacker, security engineer and pentester they could get their hands on for a two-way dialog.  They institutionalized everything they could learn from them, and they taught them in turn the systematic approaches, tools and methods they’d developed.   They didn’t keep what they were learning secret for competitive advantage – they shouted it from the rooftops to everyone who wanted to learn.

That Trustworthy Computing diaspora today constitutes a big part of the core of the modern information security industry.  Veterans of TwC are security leaders in at Yahoo, Google, PayPal, Facebook, Adobe, VMWare and dozens of other companies.   From the hapless, hopeless position the industry found ourselves in a dozen years ago, we’re today starting to stand up credible defenses against nation-state level attackers. And while the heavyweight SDL processes of five years ago have been streamlined even at Microsoft, every security program today has some of the DNA of Trustworthy Computing in it and thinks about the job it exists to do in a different way because of it.

To the visionaries who led this charge for a dozen years, thank you.

Post has shared content

Post has shared content
This chart shows the world’s internet usage shifting to smartphones:
Animated Photo

Post has shared content
My educated guess right now: iPhone6 will have NFC. And it will have a secure element, despite Apple having their own "secure enclave".

I don't think Apple really wanted to use NFC. They could perhaps have built a superior experience with iBeacons alone. The problem for Apple is that NFC capable payment terminals are rolling out on a massive scale worldwide, and I think their market share is too small to push an Apple-proprietary alternative, at least outside the US.

The market in the US is special; NFC payments are hardly a thing here, and even those stores that have NFC-capable terminals like Best Buy still turn off NFC functionality (why is a long story...). But if you go outside the US, it becomes more clear that NFC is a standard that is hard to avoid. Canada, the UK, Poland, Turkey, Australia and Japan are just some examples of countries where contactless payment is big, and growing very fast.

So Apple really had two options:
1) Exclusively push their own proprietary solution based on iBeacons. It would probably be awesome, but the problem is that it would require merchants to buy hardware capable of supporting Apple's proprietary tech. Perhaps Apple could pull that off in the US, but it's hard to see merchants investing in it outside the US where their market share is in the 10-15% range. And let's not forget that merchants deploying new payment hardware is not an easy thing: it requires buying and installing hardware, training employees, maintenance and a lot more. Apple would have to come up with an insane value add to justify it - though I'm sure they're working on it ;)

2) Use NFC technology; perhaps not the user experience they were looking for, but good (and rapidly increasing) world-wide acceptance, and the possibility to complement it with their iBeacon-based solution over time.

My only remaining question is whether they will limit NFC functionality to countries where it's already big, and disable it in countries like the US, so they can push their own solution there instead. It would be a weird move, but somehow also Apple-like :-)

Finally, why does Apple need a secure element when they have a "secure enclave"? Because secure elements can be completely powered by the field that is generated by an NFC reader, even when the battery of your phone is dead. Secure Enclave, being embedded in the power hungry A7/A8 CPU, could never be powered that way. Having a secure element would allow Apple to support transit use cases, where "omg my battery is dead how do I exit the subway after I entered with my phone" is a real thing to worry about :-)

Of course I could be wrong; I think not shipping NFC this year would be a big risk for Apple, but they're known to do such things ;-)

Post has attachment
Just pushed version 2.0.8 to Play. Minor bug fixes and some new pre-filters for OCR.

Post has shared content
These are really exciting days for Android, we're extending and improving the platform in so many different ways that it's hard to keep track of it all! 

For all you NFC lovers, here's Andres from our NFC team talking about some of the cool new NFC features in the upcoming L release! And yes, we do have even more cool stuff brewing :-)

Post has shared content
So fingerprint authentication is coming to Android? 
Wait while more posts are being loaded