This week’s Judge Spotlight focuses on [Andrew "Bunnie" Huang]. If you haven’t heard of him you need to pay more attention. His hacker cred goes way back to the original Xbox, which he reverse engineered and laid bare its security flaws. Maintaining his hacker spirit he went on to design and hack the Chumby. More recently he took on the challenge of developing and Open laptop called Novena. All of this while continuing to explore and experiment with all kinds of electronics, posting about his adventures for those of us that care about an electronics ecosystem that doesn’t shut out the user from tinkering with the hardware. Join us after the break for our conversation with The Hackaday Prize judge [Bunnie Huang].
You will have eternal hacker fame for reverse-engineering the hardware security on the original Xbox. What were you doing in your life at the time and how did you settle on that piece of hardware for the challenge?
At the time, I was completing my PhD dissertation on computer architecture at MIT. My advisor encouraged our research group to study the current crop of video game consoles to see what we could learn about how they achieve such high performance at a low price. As such, the Xbox was one of the three main consoles at the time and as a result I set upon reverse engineering it.
You wrote a book called Hacking the Xbox that described your adventure. Did you have any concerns about the repercussions of making that knowledge public and what pushed you to follow through?
Yes, of course I had concerns. At the time, the DMCA was just a couple years old and the as-of-yet untested legislation stood as a major impediment to our freedom to research and to tinker. MIT’s institutional counsel even sent me a letter repudiating their involvement with my hacking activity, possibly in part because they saw a lot of legal risk in aiding the disclosure of my findings. It was convenient for them that the actual implementation of the hack was done on my personal Xbox using my own resources, and largely during a winter break period called IAP.
What pushed me to follow through? Hacking, and the freedom to hack, is an important part of me. I grew up with this freedom, and new legislation stood to take it away. I suppose as a result, I had nothing to lose — whether I stepped up or backed down, either way I could lose an important freedom. And I’d rather go down with a fight.
Can you describe your role in Chumby?
I was responsible for the design, manufacturing and operations of the consumer hardware half of the business.
Chumby surely holds the record as the hackable device which gained the widest public acceptance. What do you think of that part of Chumby and is there a good argument for increasing the number of hackable devices available to the average consumer?
Hackability is something that only a small fraction of the population actually takes advantage of; however, I think there is a certain peace of mind that a larger portion of users get knowing they have the *option* to hack and fully understand their technology. There is something vaguely disconcerting when you become so reliant upon black box technologies. So, sharing the designs and plans with your customers gives them back a sense of agency that I believe is meaningful.
Even though few people exercised their option to hack, I was pleased at the kinds of applications our hardware found. It ended up being used in applications as diverse as a braille terminal for the blind, to the controller for a walking robot, to a console for controlling A/V functions installed at a college campus. We couldn’t have predicted these hacks and the greatest pleasure of producing Chumby was always reading about the clever things people would do when you enable them to hack.
At the end of the day, I’d say the openness and hackability of Chumby had a neutral impact on the business end of things; it didn’t drive sales, but it also didn’t hurt it. But it did create a very loyal customer base and I’m extremely pleased to see that one of our other original founders has rebooted the Chumby servers and there are still lots of enthusiastic users who are delighted by its reincarnation.
The Novena Open Hardware Laptop has two points that stick out in our minds: it’s hackable and free of “black boxes” (like binary blobs on a video card). Did you have both in mind from the beginning of the project?
Yes, of course. Hackability and depth of openness were two major goals of the project. We took special pains, for example, to source a wifi card that is blob-free; the wifi card isn’t the cheapest or best performing one, but it also doesn’t require a blob.
Although, I do have to make a correction to your question: we went as open as we could, but that does still mean that individual sub-components still contain their firmware. The SSD and microSD memory card, for example, still contain the load of firmware permanently burned in there by the original manufacturer. On the other hand, firmware that is “burned in” to a device and not typically visible to the user is not considered a show-stopper by the standards of the broader Linux community.
Furthermore, there are some components which can accept a proprietary blob, which would cause some things to run faster, but they are not required to boot or to function well. For example, the decoding of video can be accelerated using a proprietary DSP built into the CPU, but we don’t include that blob in our distribution; instead, we opt for software decode running on the ARM CPU. Also, the 3D graphics engine is the subject of an on-going reverse engineering effort that we’ve partially funded from campaign proceeds, and with any luck by the time we ship we’ll have an open-source 3D-accelerated desktop environment.
Powerful, portable computing hardware that is extensible is obviously useful to any project that needs custom hardware as they don’t have to start from square one. Do you foresee changes or iterations in Novena’s future that will gain it wider adoption like the Chumby experienced?
Possibly. A large part of Novena’s future will depend of course on how the silicon that powers our machines evolve. If Freescale will do a new, more powerful processor with the same level of openness, I would be very enthusiastic to upgrade my personal laptop by building a new motherboard for it. There’s also some hope that there will be some other open SoC designs coming out in the future, which can give us more options in terms of cost and feature sets that can make the system more affordable to end users.