Huston: Calling Time on DNSSEC?

I don't even disagree with that FUSSP. I would have probably read it back in the day, and I'm sure completely agreed with it. I definitely would have made many of those arguments myself.

But my thinking has changed too. Will we replace email? Will we have other messaging systems that are widely used - obviously we do. And spam is a problem, and no useful, widely used system can exist without some amount of expenditure - by someone somewhere - to combat spam. Whether because we centralise the messaging to some (1 or a handful) of giant tech corporates, who can afford the staff and compute to mass analyse traffic and classify spam and stop it early on; or whether we push that expenditure of resources somewhere else; that expenditure is today a *FACT*.

So the FUSSP - to the extent it envisaged it as an invariant that email and (more generally) Internet messaging had to continue on as email had started out, as a free way to send messages - is wrong. We *do* pay to send and receive email. Corporates pay directly in subscription fee to a small circle of tech companies, end-users pay by viewing advertising.

As for your point on botnets and compromising hosts. That does indeed happen in the kinds of systems I was talking about. There are spam attacks against blockchains, e.g.. However, they are generally short-lived, because they are expensive. There are also botnets that use compromised hosts - however, these days they use them to just "mine" the token/"coin" - they don't bother sending spam, cause the token is worth more than the profit from sending spam.

And that's the thing, when you go in that direction, the economic utility of the payment system becomes such that it becomes uneconomical to spam. Why waste the more valuable token? The negative economic utility (overall) and minimal economic utility (to the spammer) of spam versus the greater general economic utility of the payment system and its token (which the would-be-spammer can use the compromised system to generate, at a low rate) mean the spam becomes uneconomic.

Also, those botnets of compromised hosts - while still robbing resources from the poor owner - are now less of a negative economic utility overall, whatever harm they're doing, they're at least helping a little to secure a distributed Internet payment system. At some point in the future, when we go from PoW to PoUsefulW systems, they may even help do useful stuff like participate in (say) genome analysis jobs submitted by cancer researchers.

So I re-iterate what I say - though, I'll change "dumb" to "naive" or "outdated": The FUSSP critique dates from a different era. It doesn't apply, or it's supposed objections are irrelevant. Society most definitely _does_ "pay" for email, and messaging generally. That is simply a fact. It's simply a matter of how you want to pay, and what is the best way to allocate and use the resources that are /required/ to combat spam.

Oh, and to relate it back to the topic of DNS, such systems can (and are) also be used to implement naming systems, again in a distributed, decentralised manner while being robust against spam attacks - and with actual cryptographically strong proof of ownership too.

What you seem to be missing is the point where the economic incentives fail, regardless of particular current or future blockchain you pick.
If you have high cost per message, you make mining a more interesting target for botnets but also you leave out a significant fraction of users who can't or won't pay for it, while also making the reward of scams greater. It's worth investing some tokens to drain a wallet.
If you have low cost per message, then the token is also likely cheap (as you don't need as many to interact in the network), so the incentive for spamming for scams is high due to low reward of mining and high reward of scamming. But you get to keep a usable solution for the general population.
This applies to the blockchain scale issue as well. The higher the overhead, the higher the costs, and the higher the costs the larger your transactions need to be to be affordable, which means fewer users have the economic capacity to participate. But the cheaper they are, the more common it is to get fake tokens in your wallet trying to get you to engage and get drained.

The technical solution that supposedly makes the cultural issues uninteresting relies on and is solely about economic incentives, which also makes it part of the uninteresting but central human, rather than technical, issue, and those economic incentives appear to be flawed.

> which can be economically balanced such that the cost for normal use is relatively trivial, but sufficiently greater than-0 as to deter large-scale spam attacks.

> (scam attacks are economically different).

And roughly what percentage of ALL email sent is scam attacks? 90%? More?

The amount of legitimate mail, as a percentage of all mail, is probably rounding error :-( Most of the rest will be scams.

Cheers,
Wol

That is a good point about energy use. I had not realised large chunks of Europe have been eking out efficiency gains for that long. Ireland a little bit less so - we're about the same as 1.5 decades ago, though, that's basically a dip after '08 financial crisis, and then back to rising annual demand in the last decade. And Ireland is tight on capacity already. The planning applications already put in for future data-centre builds will definitely exceed our current generation capacity.

This eking out of current generation capacity with efficiency gains is pretty cool - lot more than I'd realised TBH. But I'm not sure how long it will go on for.

That said, one area where we potentially can get more efficiency is by being a lot smarter about making use of waste heat. We have parts of society that use energy for work, creating unwanted heat; and we have other parts that need heat for warmth (residences and offices in the summer); and while in some places there are schemes to use the latter for the former, in many places that doesn't exist - and the socio-economics are such that it would be hard to see that ever changing. So we house data-centres and other such things in parts of Europe with cheaper land, away from offices and residences - often consuming green-field land (least, does in Ireland); and we vent the heat to atmosphere. And then houses are being heated in winter with "dumb" energy use - again, ultimately venting to the environment.

I think it'd be nice in to future to locate as much compute as possible _with_ humans, who can then benefit from the generated heat (least in winter).

To make that generally possible requires tackling a lot of economic-incentive issues... And... well... another story. ;)

One thing though: Some of that flat-lining of energy use in Europe over decades will have been due to the de-industrialisation of Europe, rather than efficiency gains.

Ireland kind of hints that, cause Ireland was never that industrial compared to Europe - Ireland has a more agricultural economy, compared to continent, until the rise of the tech sector from the 90s on. And it hasn't flat lined as much.

> And for the same amount of electricity we get vastly more actual *work* done than decades ago. Your current phone can do more than your old laptop on a fraction of the power use.

I might beg to differ ...

How much of what is done on that phone is actually *useful* work, (I won't comment on the usefulness of games etc,) and how much of that work could be done in less time, using less energy, if only somebody used their brain instead of letting AI get it all wrong ...

Cheers,
Wol