No, it would be different: NSA to Microsoft: Hand over the code and release this fix so we can backdoor account X. And the next update the account can be compromised. This update will only spread to updates for server Y which holds account X.

And if they are really good at understanding risks they know that overregulation increases the chance of people not accepting the policy (and ignoring it in total).

They are different in the fact that most missiles are not configurable platforms.

Although, if you keep a max ceiling of 120 meters, I'm not sure why the area around the airfield is chosen to be so large. No aircraft will fly that low unless he is in the pattern.

Then use Flarm ;-) http://flarm.com/

If everyone has autonomous weapons, every insane person will have one.

The new 'onion'?

There is no 'best' way to spamfilter bad emails. The definition of spam is unknown and being redefined constantly by spamfilters (which is a little different from what recipients think spam is).

The problem is that the process of 'what is spam' is quite complex for humans. For instance: Suppose I gave you permission to send me bulkmail and you only start to use it two years later. Is that spam? For one person it isn't, for the other it is.

But I didn't know either...

Because we don't have proper, easy to do and efficient while running, code reuse. The coding is this world is 'pre-industrialisation' era where the guilds are companies and noone has created simple stuff like an m10 bolt. No efficient standards.

sed 's/you can identify spammers./you can identify spammers pretending to be you/'

No, DMARC authenticates you as a sender to fight phishing but does not tell you what the rest of the spamfilter concludes. You can still send spam using DMARC. The reporting feedback of DMARC is necessary to find problems in the authentication. An added bonus is that you can identify spammers.

Microsoft has SNDS which is pretty helpful in finding moments when things go bad. It does not identify your customers, I have to find that myself.

There is not much use of strong encryption on govt. stuff if the rest of the economy and infra is down or compromised.

Not when the keys get found and their own security is breached beyond repair.

As if there is ever a shortage of new exploits. Since security is not the first design criterium in current system design it will not be secure. We don't even know how. Just hope you catch the flukes that surround an actual break-in.