Hacker News new | comments | show | ask | jobs | submit login

Go after "Railroad Baron Money." That is to say, to invest in a fundamentally new kind of very useful infrastructure, like a subset network of computers where DRM actually works. If not absolutely, then well enough in practice. That would mean it would take something like 2 to 3 years for people to jailbreak new hardware, and almost a year to break revisions to old devices that fix an earlier jailbreak. (We're getting close to this level of security for some organizations.)

The trick is to be one of the Barons when networks are becoming widespread, and not an early innovator who gets in the history books but dies penniless. (Again fits with the railroad analogy.)

(And yes, trusted computation [DRM] as it is practiced now is bad, in the way that many things only possessed by only the powerful are also bad.)

Yeah, some kind of trustworthy computing would be around a 4-5 on the scale. Maybe 6.

> Yeah, some kind of trustworthy computing would be around a 4-5 on the scale. Maybe 6.

A great sign for this idea, is that it gets pooh-poohed and shouted down, particularly by people who don't even hear the entire thing and just pattern match the security part. The idea that DRM can be useful and beneficial to society as a whole is precisely "What You Can't Say" for large swathes of the tech community and even more mainstream society.

We already have a subset network of computers where DRM lasted from 2006 to 2010 - the PlayStation 3.

Now, it's a console with some good games, but I wouldn't rate it as equivalent in social impact to electricity or nuclear weapons.

Right, it was a game console, and not even a particularly good one.

A real computing device (I'd accept tablets, but really, enterprise desktops and especially servers) would be entirely different.

What I really care about is servers which can be trusted to be "fair" by all parties -- server operators, software operators, and end users. There is absolutely nothing like that today, and it's impossible without trusted computing. It's unclear if trusted computing itself is feasible (it's theoretically possible).

If it works, we end up with Vernor Vinge's _True Names_

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact