Sometimes, when I get very drunk, I'll watch Warhammer distopian videos about how it all goes wrong. Is it wrong that I think of Musk as The Emperor, come to save us?
The canonical example of harvesting the heat differential is the biolite stove[1] it uses a thermoelectric generator[2] to provide up to 4 watts (peak) of power.
I'm on beer fuel atm and have only read it briefly, but isn't that pure, free energy? With a solid state device. Why do we not have some of these things around our fridge etc.
Because if it generates energy from your fridge it means you deeply failed at insulating your fridge, and you have to spend several times as much energy putting the cold back inside.
In most places you can use significantly more efficient devices to turn heat differences into electricity.
I think he is more talking about the heat that the fridge outputs to cool the inside, which is likely warmer than the ambient temperature in the room. Just a guess.
I was; the fridge outputs quite a bit of heat, good for raising seedlings on top of. But I would imagine the expenses involved in this stuff wouldn't make it sensible for a fridge (I don't think they draw that much power anyway).
I just had never heard you could harness this, it's cool.
Edit: Bit late to the party, 1821. Bit of catching up to do.
Okay, then that means you're ventilating the output heat poorly. If there is a significant difference in temperature between the coils and the room, you're spending more power than you need to when you pump heat out of the fridge. Adding fins or fans would save you more power than you could ever recover.
Somewhere on the fridge is a bunch of metal that heats up when it's working. If that gets particularly hot, then you have the potential to save power. You could add more surface area or have low-power fans blow air across it. That would make the actual fridge unit be more efficient. The amount of electricity you would save is much more than the amount you could harvest. Any device that harvests power would also impede the efficiency more and make it work even harder.
But either one would be a pretty small amount.
It's fun as a toy, but not much more.
If you want to save power, buy a more efficient fridge. Maybe also get one where the hot part can go outside.
I wonder if using the validation of the DNS record and all the surrounding nonsense is really a useful function though.
I mean, if we created a new distributed system, and grandfathered the old one, how can the current incumbents stop it? The control over a name pointing to an address could be done with a blockchain, I'm hearing.
So we move to a new naming convention. An arbitrary prefix on the end of a name shouldn't mean $$$$.
I'm saying we (techies anyway, not money-grubbing VC's), can make the decision for them, en masse. When they say "But, we need to make money from this!", we reply, "No, we need a naming service for these internet addresses. Go get stuffed."
The finer details of how it all works can be left till later.
> I mean, if we created a new distributed system, and grandfathered the old one, how can the current incumbents stop it?
That's a weird way of putting it. Anyone can create DNS roots based on any hierarchy they want, no-one will stop you from doing that. The meaningful question is how do you get anyone to use your fancy new roots?
Yes I think any new system would have to see the old one grandfathered (give the current incumbents time to find something useful to do instead of jetting off to conferences all year).
Browsers/OS would be reasonably easy in most circumstances, but there are embedded devices, load-balancing configurations, other esoteric uses for DNS (text records, mail etc) that'd have to be considered as well.
An attack surface is ISPs and Governments strongarming them to use their own DNS roots which can then be plugged into the "new" root servers conforming with the theoretical browser vendor, os vendor alliance.
I guess it's two distinct issues to me; the tech stack that is hobbling along, and the people that control access to it (ICANN, gTLD registries et al I guess).
The latter is the really corrupt, broken thing here, and we should be able to swap that out.
This seems like it has gotten out of control - it is a basic key/value lookup, and we've created a rent-seeking, global series of registries to further suck good money from people.
I worked briefly at a registry - it is money for jam.
I don't know if the DNS protocol and all the related ones are fundamentally broken or need fixing, but it seems to me the problem is ICANN and the web that spreads out from under it.
Why can't we just create an alternate ... distributed naming system, except without the rent-seeking/government ownership attached? I mean, this focus on how important the last bit is is ridiculous, the namespace doesn't have to be artificially restricted, beyond what the protocol might require.
I guess I have a simplistic view and probably lack a bit of history, and there must be something preventing it. Do the top-level servers collude, and help to create something like ICANN, or was it taken out of people's hands etc?
> except without the rent-seeking/government ownership attached?
Because people like money. This TLD business literally invented a new economic activity, and one with great margins. It doesn't really hurt anyone, has little to no cost, and makes a pretty penny. People With Money love stuff like this; if tomorrow they could invent a marketable TLD system for physical addresses, phone numbers or anything else, they'd jump at the chance.
This is just how the capitalist world works: stuff exists to make money, everything else is a bonus.
Agreed, I guess the way that works in general is just disappointing in that it doesn't produce anything meaningful or advance anything (to my mind). It's spinning wheels and burning energy for no appreciable reason, to make some people rich.
Legal/Govermental/Regulatory issues, someone has to deal with it. No single organization is going to deal with every country and their various legal issues. Registries and Registrars help abstract a lot of that (ICANN definitely doesn't want to deal with it). I don't disagree with the ease of storing a few million records or more. That is not the real overhead and is in reality only a fraction of whats involved.
Yes the added verification etc on top of the key/value lookup is the meat of it I guess. A complicated problem, I guess it'd be nice to somehow remove the money factor from a critical function of the net.
(Firmly tongue in cheek.)