We need a fossil-free internet by 2030

I completely agree that is where the real issue is. Although, the Software industry has had a free lunch with Moore's Law for decades and now the bill has come due - we write a lot of (almost irresponsibly) inefficient software that certainly doesn't help the situation. We should really fix it all, but I agree the major onus should not be on the consumer, since it merely follows the producer. The producer has to set the tone.

Look, it is not hard to determine where major data centers get their electricity. Yes there are greater policy concerns, and really the economics of solar and wind make this a foregone conclusion in the long run.

"The internet doesn't intrinsically consume fossil fuels and never has." - This is almost an offensive statement in whitewashing/passing the buck.

Does the trucking industry say "we just transport trailers, the issue of whether the tractors burn diesel is a different issue". COME ON. Like trucking, where specific engineering needs to be applied to produce BEV or emission free transportation for their use cases, specific engineering and policy will be needed for the major industrial segment of power consumption: the internet.

Major datacenters site location and design are explicitly built around hydroelectric and other sources since power is such a large cost component. There are bitcoin mining centers in outer mongolia and other places that can be directly pointed to why they are sited: ultracheap coal and no environmental controls.

Related to this notion in the article is the big steaming (literally) issue of bitcoin network power.

Anwyay, AWS could absolutely be branded on the basis of zero-carbon-emission power. So could GCE and a lot of other major websites. The fact that is can be a very widespread and branded movement like gay rights and other things that corporations rapidly adopt and fight at the state, federal, and local level for is a tantalizing opportunity. Especially since fundamentally it drives down costs in the long run.

AWS and the like represent a centralization of major internet service, for better or worse, and therefore at a minimum they should be pressured / incented to build that basis on alternative or non-carbon-emitting power. The choice of IAAS or PAAS should have a marketing component of zero emission or not. We're talking a substantive slice of overall power generation being dedicated to this.

So this conclusion is utterly wrong IMO: "It's a waste of time, attention, and political capital that we can't afford"

If we don’t have it by 2030 they’ll publish a sterner article with a later date!

I think it does at least try to answer this question, their answer is that if we invest money now in renewables to power the web, then that investment will accelerate the reduction in price of renewables, and by 2030 renewables will just be cheaper for everyone everywhere and will be widely adopted.

Can't we apply the same reasoning to any well-established company that would make similar commitments?

Should we then be suspicious of all of them?

So what does it mean by "the website is hosted green"? It seems to me that this is just a list of website hosts known to be green by some criteria, regardless of whether the website will eat your CPU with tons of scripts or communicate with N different services?

It's a bit old, but should give you some ideas to start your own research:

https://www.networkworld.com/article/2229029/computer-factor...