If it happens live, on the server where a problem was observed, then it dodges the whole bureaucratic hell problem, and puts a hacky patch up immediately.

It's not a real fix, of course, it's just a dodge for an observed, live vulnerability. At least maybe your computer won't be a botnet node in the meantime, if you're alert enough to notice the takeover, nor will your website be down.

Maybe you can even get it to generate some input based on your vulnerability, to more easily reproduce your problem, and a hint that links to the symbol file so you can see where it went wrong. If that's possible, you'd give half the problems a potentially quicker trip through the verification queue/trouble ticket queue/triage system, plus you'd cut out most arguments about how it's not reproducible.

Sounds fairly promising to me, though of course it'll probably have its own problems. But hey, so does GDB, and yet it still helps make certain problems quicker and easier to nail down.

Just because you're much closer to 102.5's radio tower doesn't mean you can't listen to 93.3.

They're not adding the Gaussian white noise that Shannon's Law refers to, they're going to pump at some specific frequency, so you presumably get to filter it out for your communication channel.

What's the difference between that and GreaseMonkey?

I guess that doesn't address the licensing issue, but if someone distributes code (such as the javascript on the webpage they serve) isn't it their responsibility to notify you appropriately with comments if there's licensing restrictions?

"It's cheaper" is an excellent example of a business case.

Pity that still means "NAT your IPv4". At least for another few months.

I'm a senior software engineer at a network equipment company.

We experimented a bit ago by hiring a bright fpga guy (one who could tell a good design from a bad one) who was bored with hardware for a software position, even though he didn't have much experience with software development or networking.

That's been very successful, and next time we're off our hiring freeze, I'm going to try to make it happen more often.

I can't count how many interviews I've been through with some guy with a master's degree and a resume that reads like they've been training for our open position for 8 years, only to find out they can't troubleshoot anything tricky, can't catch reasoning flaws, and can't tell the difference between an elegant solution and a spaghetti nightmare.

I'd so much rather have somebody who needs some basic knowledge hand-holding for a few months, but often has good insights in a high level discussion, and just keeps on getting better at the coding part.