I have one use case that I haven't seen addressed yet (ha, ha): reverse proxy for high availability.

Okay, let's say I have two internal servers that I want to expose externally via a single IP in order to facilitate reachability in a scenario where one server goes down. I am aware of round-robin DNS for load balancing, but that doesn't address the scenario wherein a client has a previously-established end-to-end connection and the server they are explicitly addressing goes offline. For simplicity's sake, we can assume we are discussing a stateless app server, but the concept applies the same in application clusters that share session state among the cluster.

With v4, I can have the NAT device handle failover so that subsequent requests from the client are seamlessly rerouted to the other server at the network layer—all behind the abstraction layer that NAT provides.

I've read a little about RFC 6296 (NPT), but it's not entirely clear to me that this would solve my use case. As far as I know, the simplest/best solution is to insinuate a proxy between the public IP and the private IP, because interface bonding doesn't fix the scenario where the app server has gone down but the server host is still up. In an interface-bonding configuration traffic can still be sent to the host with the downed app server (FAIL).

I have read that some people are attempting to hack NAT on top of IPv6, but I have seen many posts over the past few years claiming that "NAT is never the solution/breaks the internet/causes cancer/etc"; therefore, I am wondering what the best practice actual solution is in this case.

I definitely agree with what you are saying. However, things are simpler in Europe with your VAT.

In the US, we have a patchwork of sales taxes that are are state, locality and even "occupation" based. Furthermore, at least in my state, it's illegal for a vendor to advertise prices including tax, unless the vendor is selling goods/services from a very short, explicitly-defined list (examples include gasoline and cinema tickets). I guess the idea is so that consumers will know they paid the sales tax, because it is required to be a separate line item on the receipt.

By way of example, here's what the breakdown on the charges for prepared food in my locality: $X menu price * 1.025% "restaurant occupation tax" * (1 + (.055 state sales tax + 0.015 "local option" sales tax)). Note that as a consequence of this formula, the restaurant tax ends up being taxed in itself... yes, sometimes we have to pay taxes on our taxes.

The reason the situation evolved this way is that each state is allowed to decide how much sales tax to charge (or whether to charge any at all!). It's not a nationally-set thing. Then, each state can decide whether to allow cities to tack on additional sales tax. For example, in my state, the state law allows cities to tack on 0%, 0.5%, 1.0%, or 1.5% based on what the city populace/government decide (yes, it has to be one of those four options). So, even within the same state, sales tax varies from 5.5% to 7.0%. It's strange, because driving just a mile out of town can drop the sales tax rate by 1.5%.

The "occupation tax" for specific industries (eg. hotels, telephony service, restaurants, etc in my city) is a backdoor/loophole that my city uses to transcend the state limits on local sales tax options. An "occupation tax" is not the same as a "sales tax", even though the effect on purchase price is almost exactly the same—except occupation tax gets meta-taxed as I described above.

Fun, eh? Municipalities want to raise revenue, and if they can't use sales tax then often their only choice is to jack up property tax rates. Not every state allows cities to charge income tax. So, as insane as this all is, it's not like we are doing it for the lulz.

How do European cities raise revenue? Do they have to lobby for VAT-turnback (ie. "y% of VAT raised in your municipal area will be given to your city budget")? Or is it all property tax-based? Or are you assessed city-based income tax in some cases?

Why is it unfortunate that nature stores hydrogen as hydrocarbons?

Hydrocarbons are far superior energy storage than any battery or capacitor technology on the horizon. As a bonus, they are the very definition of stable storage (especially when compared to caps or batts).

People need to stop conflating the source of energy with the energy storage medium. Oil can be carbon neutral if it is made via reverse combustion. Link this solar reactor's water-cracking H2 production with a similar reactor cracking CO2 into CO (similar reactants involved in a solar reactor), send it through the Fischer Tropsch Process, and get carbon-neutral hydrocarbons as output. As a bonus, you can reuse all the existing petroleum distribution infrastructure.

Furthermore, if you equip a vehicle with a next-gen solid metal oxide fuel cell then you can exceed Carnot efficiency in energy extraction from the hydrocarbons.

Since most of these reactors use an inert gas inside the reaction chamber, what are the losses that you propose? Also, it probably would help to consider the ZnO more as a catalyst than as a fuel, so the "carbon costs" of producing the ZnO are a one-off.

The endothermic step is the dissociation/cracking of the ZnO to produce Zn. Elemental zinc is highly reactive, which is why it forms a passivating zinc carbonate layer when exposed to the atmosphere—which is a key reason the reaction chamber is sealed & filled with inert gas.

Essentially, the only elements present in the reaction chamber are Zn, ZnO, H2O and a few inert compounds. Most of the writeups I have read mention primary concerns about re-oxidation of the Zn that was just formed back to ZnO (cf. "elemental zinc is highly reactive"). If the primary issues are along those lines, then all that represents is an overall loss of efficiency in the thermochemical cycle. That's completely different than poisoning the zinc "catalyst" in a way that would require replacement.

Besides, even if something like zinc hydroxide were formed, all they would have to do is "burn" that in the reactor in the same manner as ZnO, to form Zn + H2O (again, this would be an efficiency issue rather than a "poisoned catalyst" issue). Don't forget that the dissociation step is being conducted at very high temperatures (~2000K).

Finally, it's not as if the zinc is being rendered radioactive. Even if impurities were to slip into the reaction chamber and result in zinc compounds that were "poisoned" for use in this cycle, I'm sure they would just enter the normal economic stream. This would offset a corresponding amount of those other zinc products' production.

Also, the concept of solar cracking is generic. For example, the cerium(IV) oxide-cerium(III) oxide cycle works in a very similar manner.

Unless you are being unnecessarily pedantic, the ZnO should be considered as a one-off, sunk cost and therefore this does indeed represent "carbon-free energy":

Zinc-Zinc Oxide Cycle

The reaction regenerates the ZnO at the end of the cycle (reminiscent of a catalyst); therefore, the net reaction is H2O -> H2 + 1/2 O2. So, while the reactor requires some quantity of ZnO to bootstrap itself, very little (or no) additional ZnO should be required to keep it operating. If this particular prototype reactor doesn't fully regenerate & reuse the ZnO, then that is a limitation of the particular implementation and not a limitation of the thermochemical cycle itself.

However, if you were intending to be pedantic in the sense that almost *nothing* can be built without generating some sort of carbon dioxide emission (eg. if you consider wind energy to be "non carbon-free energy" because CO2 is produced during the manufacture of wind turbines), then please accept my opprobrium for your pedantry.

Economists will fight over the interpretation of the empirical data all day long, and they have been arguing bitterly about it for a *very* long time now.

The fundamental issue is the underlying complexity of economies, coupled with the inability to perform any controlled experiments on an economy. As a bonus, no economic theory adequately describes even a simplistic version of the historical data. Furthermore, the predictive value of these theories is negligible—ask yourself how the downturn of 2008 caught nearly every economist by surprise. Even Greenspan and Bernanke were caught off guard; one would hope they would be paying attention to the state of the economy, right?

Effectively, all that's left is for economists to:
1) Argue about whose model is least grotesquely inadequate to describe the historical data.
2) Offer conjecture about what's actually inside the black box that is the economy. (given the situation, this is about as practically useful as philosophical musings on ontology)
3) Contrive excuses about why they didn't actually perceive (and their model didn't predict) the latest bubble in the economy that just burst and caused a major recession. Once the public opprobrium from their latest failure has blown over they can get back to 1), which is their "real job". Usually, they do this by squinting at the burst bubble and then shoehorning the empirical data into their preconceived notions. At this point, they claim that their model actually *did* predict the bubble burst. In hindsight. Every time, of course.

Here are two very amusing EconStories videos to illustrate the point:
"Fear the Boom and Bust" a Hayek vs. Keynes Rap Anthem
Fight of the Century: Keynes vs. Hayek Round Two

Yes, and gold fringe around American flags in courtrooms is a key example of this conspiracy. *cough*

Face it, these sovereign individual/allodial title/income-taxes-only-apply-to-government-employees/the-US-government-is-a-corporation/admiralty law/etc conspiracy theories are all a crock. Here's how you can tell: even if this conspiracy theory were true, the common people want it this way.

Let's imagine that the conspiracy were true and were exposed. Suddenly everyone is declared a sovereign individual. Hooray, no income taxes, etc, etc. It would take about 3.7 seconds for a new Constitution/Constitutional amendment to be passed to reinstate present income tax levels once everyone saw that all government functions were shut down due to lack of funding.

So, if there's a conspiracy, it seems to be resulting in a state that causes contentment for the overwhelming preponderance of the populace. Even written laws cannot defy a culture's will for long.

Cf. a "real" conspiracy: the backdooring of federal intrusion into all aspects of life via abuse of the Constitution's Interstate Commerce Clause. This was derived from cultural shifts in the early 20th century. The populace is happy with the result, even though the level of cognitive dissonance in these judicial decisions is farcical.

Same deal here: let's imagine that the US Supreme Court were somehow instantly packed with hardcore libertarians. All these unconstitutional federal intrusions would be struck down. Social Security & Medicare would be instantly shut down. Food & drug contents would no longer be regulated. Vehicles would no longer be required to meet safety standards. How long do you think it would take for a Constitutional amendment to be passed to reauthorize all of this presently-unconstitutional federal intrusion?

The culture wants it this way, so even written law cannot defy that mandate.

Consider this as a de facto vs de jure argument against these types of conspiracies.

"You must be new here..."

Slashdot is the 'home away from basement' for every Captain Obvious who erroneously believes they are the next Wernher von Braun.

I'll see your "contrived use case" and raise you your own comment:

I'm not the one changing the scenario around; you are. I arbitrarily limited the scenario to be "all 4 minute, incredibly simplistic song representations".

O Rly. Let's take your cited scenario:

I presume you will agree that there is more than 1 octave available in music. I chose 8 octaves, which I believe is a reasonable number, but it doesn't really matter (as you will see presently). 8 octaves * 12 notes = 96 available notes. Fine. Now, 8 notes per chord: 96 choose 8 == 132,601,016,340 possible chords. Forgive me for switching to logarithmic notation at this point, because the numbers are just too enormous to post here. Apply this to 16 chords per measure, 4 measures.

(96 choose 8)^16^4 == 10^(log(96 choose 8)*16*4) ~= 10^712

Congratulations, you have narrowed the space to 10^712 permutations. This disregards the explosion of permutations that would happen if you *also* allowed chords of smaller than 8 notes or allowed silence for one or more beats.

Feel free to insinuate this isn't what you meant by saying that you could fit static copies of this data into a few GB, but we both know the truth. Much like 10^1200, 10^712 is also a very good approximation of infinity.

Sounds useful! Just grab all 7 billion people on Earth and have them listen to your samples for ~10^703 years. FYI, the ultimate heat death of the universe (after all the black holes have evaporated) is scheduled for ~10^100 years, so you might have to develop some new physics to finish your experiment.

Well, given that your copyright would have expired before you could even begin to scratch the surface of that particular experiment...

The idea behind copyright is to promote the progress of science/art for society as a whole. I'll rebut your claim that I am appealing to unnamed authority (ie. without citing legal precedent) with a counterpoint that you are asserting without proof that having the USPTO grant you sole copyright over all copyrightable material would serve to promote the progress of art and science in society.

Because, for all the changing of your proposed attack vectors, that is your stated goal: to own all copyright and somehow foment massive changes to US intellectual property law. It is patently obvious that it doesn't serve the public interest to grant this kind of meta-copyright and so it won't be allowed.

I am not even going to engage in a futile search for case law on this, because it is so absurd. I will leave you with an adage from US jurisprudence: "The Constitution is not a suicide pact" . If even the US Constitution can be set aside in certain circumstances for the common good, what do you think your semantic games on a $35 online copyright registration will count for when weighed against the potential for destroying all intellectual property production for 75 years? Again, this doesn't cover the arguments that you know lawyers could come up against this, even based on extant law & regulations.

Your meta-copyright would be first mocked as frivolous, then retroactively rejected if you ever managed to get it granted and subsequently tried to assert it. By all means give it a shot, but don't forget I told you what was going to happen.

Don't mistake my claim that your attack is unworkable as support for the status quo in the US IP regime. In fact, the polar opposite is true: I am almost to the point that I believe that all copyright should be abolished. I just don't believe your approach is tenable in the legal system.

I presume you mean my argument that this kind of copyright claim could be trivially thrown out via constitutionality argument? (The "promote the Progess of Science and useful Arts" quote was an excerpt from the Copyright Clause, if it wasn't clear)

Courts generally refrain from resorting to making decisions regarding constitutionality if there are alternative rationales available. Regardless, you aren't going to find a copyright clerk or a judge who will support your claim of an infinitely-long "work" that encompasses all possible works within it. Your claim of holding the world hostage to your "exploit" copyright is trivial to reject because it does not promote progress to assign all copyright in the world to a single entity, especially one who merely submitted a single recurrence relation as a frivolous claim to hold all possible copyrightable works. That's all: your attack is done for that reason alone. However, I'm certain that lawyers could get your claim thrown out as frivolous without having to resort to the constitutionality angle at all.

Perhaps that's because you keep changing your proposed attack vector.

First, you proposed copyrighting static output. I presume you were talking about that, because the output of a copyrighted program isn't itself copyrighted by the software author's copyright holder (eg. Microsoft does not own copyright to the documents you make with Word):

"An example might be to create a software program that outputs every possible combination of notes permissible under the rules of standard musical notation, then file copyright on it."

This was shown to be impossible, though you were shopping the "copyright the static output" idea to other people:

"You could do it systematically, creating "choatic symphonies", each one being a permeutation of the possible space.[...]Theoretically, I could have my 8 core i7 "composing" 8 symphonies at a whack."

...it's never going to work. The permutation space is too large. Never fear, you switched gears to a different approach:

These contain a section for "sample" data, then a programatical sequence. Essentially, a program that generates the music, rather than a static representation of the music.

...at which point you have to demonstrate that it will contain all possible permutations. You can do this a variety of ways: eg. via searching a transcendental number (an idea which you rejected as pointless), or you can make it deterministic. You suggested an approach:

"The location as a timestamp would be solvable in polynomial time, and because it is rational, and not irrational like pi, it can be bounded and defined with precision."

...which I refined to be "try to copyright the natural numbers" (trivially mathematically represented as a recurrence relation), which makes claims easier for you because you wouldn't need to search at all. Not that you are going to be able to copyright a recurrence relation to expand all the naturals, but hey, this is your idea here. Why make it more complicated than that?

Finally, you presume that the humans interpreting the laws and regulations will allow your attack to bring the entire intellectual property system crashing down over some vague semantics that you attempt to exploit in a single meta-copyright claim. Yeah, this kind of approach is great for smashing stacks and exploiting code on computers. However, it's not going to play out that way in the legal system.

Feel free to try it if you don't believe me. Hell, email the EFF and post their reply in the thread. I would love to read it.

I think you are presuming that the legal system operates like a computer, via strict logical operations and firm, clearly defined semantics.

That isn't the case.

The humans involved aren't going to accept that your copyright submission is legitimate. My guess is that you will run afoul of the definition of the submission of a "complete" copy as, by definition, any such algorithm's output would require an eternity to play/output/whatever. Barring that, your copyright would likely be thrown out as unconstitutional because it could be trivially argued that it does not "promote the Progess of Science and useful Arts".

As with the meta-patent ideas, you are free to try to file this and then bring suit when your claim is rejected. I just don't believe that an infinitely-long work that copyrights all possible permutations is going to be accepted by the copyright office. Even if you "sneak it by" the copyright registration process without any human realizing what you were claiming, your copyright would just be retroactively rejected as soon as you made your first infringement claim.

Furthermore, I don't believe that a rule for the rejection of infinitely-long, impossible-to-ever-actually-express-in-fullness "works" would even require a new law to be passed. Your claim is going to be thrown out by the copyright office, or if you persist in wasting money, by an appellate court (this wouldn't even meet the criteria for a Supreme Court case).

Your copyright assertion would be inane:
"My song contains all possible permutations of notes because I am asserting copyright on all possible permutations of notes."

Just try to copyright the natural numbers instead, because that's what you are insinuating. Then there is nothing to solve at all, because the content of the work that is "infringing" is its own offset in your copyrighted work.

As a bonus, you get copyright to all copyrightable works made after the date you filed (as derivative works, which you show by performing arbitrary mappings of their content to the numberline as well). So, while this approach is not necessarily mathematically impossible, it's as vapid as all those ideas about filing meta-patents. ("I'm going to patent the idea of patenting something! Take that, stupid USPTO!")

This isn't a workable attack.

In that case, your algorithm is essentially transcendental and normal. Difficult to prove that it will include all possibilities, so you may as well just copyright Pi instead of inventing your own (it's at the same state of proof as what you propose). As a bonus, you can sue everyone who produces anything at all after that date, because you just apply a "trivial" (*cough*) mapping from the offset within your musical notes to textual or binary representation.

I am being somewhat satirical, because even if you were granted such farcical copyright protection, you would still have to prove that the other material infringes by being able to cite your time offset. You couldn't make some hand-wavy assertion that "it's somewhere in there, you dastardly infringer!"

Look at the Pi-Search system that indexes the first 200 million digits of Pi and search for a your favorite up-to-120-digits-long number. Just for fun:
"The string 1234567890 did not occur in the first 200000000 digits of pi after position 0."

That's a 10 digit long number. Didn't occur within the first 200 million digits of Pi. Most songs are going to be more than 10 notes long. The odds of finding the "infringing" material rapidly drop to essentially 0 for any nontrivial search key.

Your approach is impossible. You may as well try to copyright Middle C and then presume that you can sue anyone that "infringes" on your copyright by including that musical note in their work.

Okay, let's take a look at the math for what you propose.

Our incredibly limiting assumptions will be:
1) Ignore harmony and focus solely on melody
2) Melody is defined as single note per beat
3) Only deal with songs <= 4 minutes long
4) Since this is a geek site we should cover electronica, so we need to represent tempo up to 150 bpm
5) 8 octaves are available

Taking Western notation, we will map the 12 notes per octave (ie. lettered notes plus sharps and flats) * 8 octaves. We will also represent silence for one beat as one "note". Therefore, there are 97 notes in this representation.

150 beats/min * 4 minutes = 600 beats. This implies that there would be 97^600 possible songs. Call it order of 10^1200 possibilities. (what's multiplying by an extra few billion times between friends?) This is a reasonable approximation for "infinity" for all practical purposes, as there are presumed to be "only" around 10^82 particles in the entire observable universe.

Yes, I have made some trivializing assumptions with how to model the permutations here, but don't forget that we have also simplified the model to exclude many aspects of music (multiple notes played simultaneously, longer songs, etc), which would cause the number of permutations to explode.

As rare as it is for governments to actually do that, I propose your apples-to-apples comparison would be to a nonprofit corporation. Alternatively, a nonprofit foundation funded by a for-profit corporation might be an appropriate comparison to your hypothetical benevolent government.