There is a huge difference between this and science: Scientific results can be highly complex and hard to understand, but the standard practice is that they are reviewed by multiple people before being accepted as true.

The reason for this is obvious: scientists know very well that complex things are likely to be wrong at the first attempt :)

So a huge piece of software, which was not reviewed by anybody besides its authors, should never be trusted.

Not necessarily. Suppose that the database lookup gives erroneous results with probability (let's say) 10^-6 (for whatever reasons). Then if you are entering 1000 cars manually, you will have on average less than 1 false positive per year. If you harvest 10M license plates per day, you will be getting an average of 10 false positives *daily*. People are not going to be pleased by false accusations...

Not at all. Robot is derived from "robota", which is an old Czech word meaning just "work" (derived from "robit" meaning "to work"). In the middle ages, its meaning shifted to mean "mandatory work for the feudal lord", regardless of whether the subject was a free man or a serf (there were no slaves in Czech kingdom at that time). Contemporary meaning is just "any hard work".

Not exactly. Nobody but you can verify that the vote whose ID you present was really cast by you. When asked for a signed vote, you can just pick one randomly from the public log.

This assumes that the certificates on the devices are configurable. Alas, on many real devices they aren't - just a self signed cert generated by the device's firmware.

Fine, so we finally have good motivation for the manufacturers to stop trying to trick the customers by selling different products under the same name ;-)

I still prefer having my e-mail dropped by sites with silly administrators to having random legitimate e-mail dropped by my GMail account just because GMail uses an over-paranoid spam filter which cannot be tuned to my needs.

Wrong. Even if I don't dispute your assumptions (constant rate of new article creations, constant rate of edits per article), the resulting function is quadratic, not exponential.

Also, the majority of historic versions is probably seldom viewed, so they do not contribute to site load much.

This does not seem to be likely, though. As far as I know, there is no polynomial-time quantum algorithm for solving the Traveling Salesman Problem, or any other NP-complete problem.

Well, yes and no... Within a year, almost anybody can learn to fix a garbage collector.

I accept that some security bugs can be hard to fix. Still, it gives a clear message about the values held by the organization if copying Chrome's UI has higher priority than fixing security bugs.

The single fact that there was a high-security bug unfixed for at least 335 days (as admitted by Mozilla's FAQ) tells that there was something very seriously wrong in Mozilla's handling of security vulnerabilities. That is the reality and it should be passed around.

Unfortunately, DMARC breaks even mailing-lists which do not tamper with the contents of the messages at all. The reason is simple: SPF. Rewriting envelope senders is the proper way of forwarding mail since ages.

If you want to have proper integrity checks of e-mail messages, use PGP, not DMARC.

What they mention is not a list of solutions, but a list of silly work-arounds, which break well-established semantics of e-mail headers. Falsifying information about the author of the message (that is, the From header) for the sole sake of making the message compatible with DKIM is broken.