Do you expect your electrical bill to be fixed every month? Electricity is not changed by capacity. The resource being sold is electrical power.

"overage" is merely usage tiers: X $/kwh for power up to R1. Y $/kwh for power above R1.

For (consumer) internet, the resource being sold is bandwidth, regardless of volume.

Not so sure about the "it works". Installed Debian 9 on a machine, tried logging into an xvnc session. I am unable to login to both the console and the xvnc session - basic multi-user multi-head X-server stuff. Before session-manager/gnome/systemd/whatever-the-cause, that would just work. You could login twice (three times). Even if I were to figure out why it doesn't work, and figure out the correct forum, and find the correct set of developers, the response will be "Why would you want to login twice?". If it's some interaction between xvnc / session-manager / dbus / systemd / whatever - good luck getting somebody to fix it.

I tried running gthumb through an ssh session, but it seems to require OpenGL (or some kind of acceleration). "Of course you need accelerated graphics to run a photo application!"

BTW, xvnc does not do any graphics acceleration.

It seems that major components of "Linux" (or, should I say debian / ubuntu / redhat / suse) have changed from "cooperating targeted utilities" to "single-solution all-encompassing kitchen sinks": Gnome (gui, config, window manager, events, panels), Systemd (monolithic system controller).

For the projects that are becoming monoliths, the danger is that the system becomes a de-facto interface and implementation instead of the typical unix-way of having "commodity solutions" - perl and python have no problem co-existing.

The headline says that the rate of CO2 emission is unchanged over the past three years. However, that only means that our rate of pollution has been constant over the past three years. This does not mean that the CO2 level is flat. In order for the CO2 level to decrease, the CO2 emission rate must fall to below the amount that the environment (etc) can absorb/process.

The article seems to confuse or mislead as well, muddying the difference between CO2 level and CO2 emission rate.

You have come up with a prioritization scheme for your traffic. Do you feel that should apply to my traffic? Do you think my prioritization scheme should apply to your traffic? Who should determine the global traffic priority scheme (besides you)?

Alternately, let's suppose that user A only has VoIP data full bandwidth, while user B watches streaming video (arguably realtime) at full bandwidth. Should user A's traffic take priority over user B's traffic? Would you like to be user A or user B in this case?

Why should your streaming video take priority over my download? We both pay the same amount for our bandwidth.

The only "fair" approach is to treat user A's traffic and user B's traffic equally (unless one user voluntarily says that some traffic is globally lower-priority). Some fair approaches to throttling would be to randomly drop packets (maybe based on packet size), or start to throttle the source that is using higher bandwidth (so if A+B each have 10Mb connections, and are being pushed into a single 10Mb connection, each source would get at least 5Mb, or more if the pipe is otherwise idle). However, such equal throttling does not scale because the bottleneck point might not be local (ie: A+B have plenty of bandwidth to use within the ISP's network, but are competing for limited bandwidth somewhere else, so now every router needs to fairly shape A+B's bandwidth - but what if A has paid for twice the bandwidth as B? Does every router on the internet need to take into account every source's bandwidth?).