So? dns is a common part of modern network infrastructure. It's also possible you want to have devices use a dns according to which router they use (potentially, because the dns server may be on the other side of the router). Again, if they think dns in RA is bad design, it's still a common standard, it's not up to them to say "screw you" to those users.

A requirement that you just made up? A network can function properly with different hosts using different dns servers, there's no requirement to use the same.
Sure, RA doesn't have support for "per-machine config settings", but this is only something corporate environments need (and they can use dhcpv6). That just justifies the usage of dhcpv6 in those scenarios. It doesn't justify discarting RA completely for every user.

dns search domain are part of the network layout (just higher layer that IP/gateway). It makes perfect sense to include it into a single protocol that advertises network layout.

Again, none of these arguments justify not supporting RA. And saying that windows "fully supports IPv6" is mistaken, it only supports a certain network configuration. Not the most popular one by the way.

"full automatic configuration of v6 network details". The word "full" is quite relative here. It only works on a single specific (and very uncommon) scenario where you have dhcpv6

We were not discussing Linux nor debian. That's completely out-of-topic and irrelevant.

That's they opinion. The standard differs. Both the written standard, and common real life scenarios.

The dive can run any code, the signing key makes it look "officially approved" by Nokia.

That's just what I was thinking: if they're firms based overseas, then they can't force them to disclose SSL keys (making the communication secure-er). They shouldn't be able to force them to do anything, legally.

But doesn't facebook have a UK subsidiary, and offices in the UK (and mirrors, since I'm pretty sure they use CDNs)?

Why can't you mirror the repository into github, and update it just after every release?

If they do it right, the might be able to cheat unnoticed for a pretty long while.

Oh yeah, and you can leave inline comments in text files. Like the old setting when trying something new, or a note-to-self.

Like I said, I doesn't support retrieving DNS over RA; only retrieving IP address/prefix (it's quite silly to support getting an address but not DNS servers, by the way. Why would I want a public IP address and no DNS on a desktop OS?).

RA is the more common (and simple) option for advertising DNS. DHCPv6 requires a second (redundant) service just for DNS - seems a bit of an overkill. Also, I've yet to come across a network that uses stateless DHCPv6.

I really don't get what your post is about. Only very specific situations require you to actually use IP (4 or 6) numbers. We've had DNS for decades and mDNS for a long time as well. You only need to work with dns when initially configuring network devices (eg: routers) or DNS servers themselves. No other scenario should require that you use IP numbers.

The sad part is, they can't really transition to IPv6 because their own OS doesn't really support it.
Sure, some groundwore is present, but there's something critical missing: Windows can't retrieve a DNS server over RA. That means, it can get an IP, but not DNS servers.

While I agree, I still would prefer that nvidia open sourced it's drivers once and for all, so we can have hardware that works well and stability to go with it.

Actually, were it not for propietary blobs, there would be abolutely no necesity for them. Linux is designed to have drivers in-kernel, so no user intervention should be required to have devices working, hence, a friedly UI for users to configure devices is sort of wierd.

Seeing as how propietary drives need to be properly integrated for non-power-users to install them, the package manager usually sounds like the right place.

Text files have their huge advantage. They're easy to back up and don't require anything aside from a text-editor to restore a broken system. I can easily copy them over, and diff them. Sample configuration files are quick to compare.

None of this is true for the windows registry.

Text files may be less newbie friendy, but then again, programs do have a settings/preferences option generally for stuff newbies want to touch. Messing the config files OR a registry by these sort of users tends to end badly anyway.

Actually, that's quite wrong.
There's are standards for configuration locations, and only legacy applications and notable exceptions keep them elsewhere.
Generally, /etc is for system-wide configuration, and $XDG_CONFIG_HOME (~/.config, be default) for user-level configuration. The former is only user when configuring the OS itself, generally, and the latter for desktop applications. Most users will only care about ~/.config.

See the XDG Basedir Spec for more details.

No it's not. None of them have any obscuriy. Encryption can openly defined, it's just they keys/passphrases that are secret. Mac filtering has no obscurity at all.

I'm not repeating something someone else said. Mac filtering is like having a doorman to whom you scream your name before getting into the building. Anyone in range can hear it, and pretend to be you by screaming your name.

Actual encryption is like having a key. Nobody can get in until you give them a copy of it (eg: tell them the password).