This. Part of these systems has to be that you cannot prove to another person how you voted, whle still allowing you to prove to yourself that your vote was correctly counted. There are schemes for that but they mostly require the voter to be intellectually able to trust mathematics.

I thought the whole point was to jump in head-first and just hope the thing gets bought by an aquisitions team from an established company or pull all the copper out of the walls on your way out and end up breaking even (and therefore having employed yourself for a year or three.)

All the things we worked around to get things to work through NAT. And a few thigs that you would be using if we could figure out how, but cannot.

You're welcome. It would have been much easier without NAT.

That is not likely to catch on in many enterprise environments, which is one reason for slow adoption -- first hop security had to be secured along with DHCPv6 snooping so that addresses could be held fixed. Yes, even for clients. Most of the auto-address self-configuration stuff is crap. It was crap in IPv4 zeroconf and is still crap in IPv6.

This I can readily agree with. NAT provides nothing security-wise than a firewall can do.

Unfortunately, no, NAT has been around long enough to pick up some "off-label" uses so to speak.
Once a server is set up to work correctly from behind a NAT people start thinking of clever tricks
to play with NAT and some of them have become an integral part of network functionality.

Especially it is used a lot in cloud service redundancy/bridging setups.

Most button cells are actually "lithium primary batteries", not considered "alkaline" in common parlance.

There are actually poorly engineered adgets out there that cut off well before an alkaline is tapped. They are the same ones that have trouble operating off NiMHs.

(Whereas the ones the SP mentions that drain the hell out of batteries need to be used with care with NiMH as they can decrease rechargeable shelf life by doing that.)

Ever since LSD NiMHs hit the market I have not bought a single alkaline oter than to put in gifts given to someone who can't handle rechargeables.

Just tax tires and eliminate the gas tax. Done. Next problem.

Well, brush motors and a dynamo back to back in a rotary convertor would have done it, even then.

Another special way DC is dangerous is because, while AC will cause your muscles to jerk around, DC will cause them to clench in one direction. Which makes it kinda hard to let go of the wire.

I was aware of that. It dodn't bear on the subject matter of my comment.

MITMs are different than just sniffing.

You can tell, in fact, that you were MITMd post hoc, because you can compare the cert that was used versus a copy of the cert obtained through other means. That's easiest to do if you have admin access to the server, of course, but those of us that do, know that MITM attacks are rare.

Is your google broken? This has indeed been addressed (by actual scientists) and the estimate of those impacts are of course refined as models improve.

Like here.

About the only argument for not encrypting that holds water is if you want an offboard IPS to see the attack packets. Caching and resources are of steadily diminishing importance.

No, it isn't. Because in the vast majority of cases your traffic wasn't interesting enough to MITM the first time you connected to the server, and after that, you've stored the key you found there and can be alerted if it changes. Also you can post-verify to see whether you've been MITMd if you care to know whether the horse is out of the barn, which isn't as useful as keeping the horse in the barn, but still qualifies as useful.