Nobody's going to run a /8 on a single LAN segment.

If you want maximum efficiency, your hosts all have RFC1918 IPs and you route the public IP addresses to them. zero wastage, everything usable.

When I went to college in the mid 90's, the entire campus was on a /16, flat network, every machine had a 255.255.0.0 netmask. There was no subnetting, everything was switched (they said "bridged" at the time.)

With today's switches, you could do much crazier things. Definitely not an /8 though.

I have unfortunately seen that many a time.

That's the joke! Imagine not getting caught these day running a giant routable stolen class A network on a single LAN segment. It would require Trump-level audacity!

If you want to be fully pedantic, it's Élő, not Elo - the way in which you pronounce the former is dramatically different from the latter.

To be correctly pedantic, it's named after Élő but it is called the Elo rating system

if you want to be fully fully pedantic its Glicko-2 rating system on Lichess, while FIDE uses Élő.

Rather, players have a FIDE rating within FIDE, and a unique rating in each system, because no organization follows the exact, original Elo system.

For older architectures, you really want to use the BX instruction unless you can guarantee you're not switching execution mode.

as a bit of pointless trivia, MOV PC, PC does not cause an infinite loop - it skips the instruction immediately following.

Correct, PC as a source register means PC + 8.

`[[no_unique_address]]` in C++20 should also resolve this and permit a zero-size object.

Actually this is effectively a broadcast/multicast MAC address.

The first byte of your MAC address should always be an even number. The value of the second bit is supposed to indicate if the MAC is "burned-in" and basically doesn't matter. The LSB of the first byte, however, should always be zero.

You say "actually" but you seem to be agreeing with me.

The first byte of my bad MAC is '01'. This not even because the LSB is set, making it a multicast address which is bad.

An iterator is a class. It can contain any private data you desire and mutate in any wonderful and/or stupid ways you desire.

How it decides to compare equal to the .end() instance is also your call. Your example is not a compliant iterator (as in, wouldn't work in a range-based for loop) - yet the same behaviour would be achievable through your own operator++() and operator bool(),

I agree, but I feel it is a little sad that such a powerful programming concept as an iterator is hard to implement in C++ if you want to make full use of it. Other program languages have made better choices with respect to this and allow you to write cleaner, shorter code.

Why not just use those two operators instead of more and next?

"The best extra virgin olive oil I've ever tasted"

Oh shit, why was I banned?

The policy is against harassment, not a blanket ban of these words.

I always used argep.hu for pricing computer bits. Dunno who's better

And it just so happens to have created the Blume logo from Watch Dogs - https://dosycorp.gitlab.io/dosylogo/?v923418754891239875624v...

Edit: OK, the link doesn't work it seems. It's #66 which, at least to me, is identical to #134

0, 2, 8, 20 (on the first line alone) 56, 68, 118, 124, 136 (on the first page alone) 210, 222, 420, 432... and probably more... It's almost like I can find this shape on every random place I go.

And #196, and #430... I wouldn't be surprised if others.

Well, there is.

Say I'm an ISP with X million subscribers.

Given a peering with Y GBit of throughput to someone like say, Twitch or Facebook; with multicast, I now only end up with a single load of data for each livestream coming over the pipe.

However, this does require direct peerings. Multicast over the Internet is just not a thing. This could technically be built out with IPv6, but nobody's realistically going to arrange for globally-unique multicast space allocations and convince every AS in the world to start peering multicast over BGP for general public use.