In a TSO architecture like x86 or SPARC, every "regular" memory load/store is effectively a release/acquire by default. Using release/consume or relaxed provides no extra speedup on these architectures. In weak memory models, you need to add in acquire barriers to get release/acquire architectures. But also, most weak memory models have a basic rule that a data-dependent load has an implicit ordering dependency on the values that computed it (most notably, loading *p has an implicit dependency on p).

The goal of release/consume is to be able to avoid having an acquire fence if you have only those dependencies--to promote a hardware data dependency semantic rule to a language-level semantic rule. For Alpha's ultra-weak model, you still need the acquire fence in this mode, it doesn't help Alpha one whit. Unfortunately, for various reasons, no one has been able to work out a language-level semantics for consume that compilers are willing to implement (preserving data dependencies through optimizations is a lot more difficult than it appears), so all compilers have remapped consume to acquire, making it useless.

The big legal loophole is that the government needs a particularized warrant (per the 4th Amendment) to ask for any user data, but if the government buys commercial data, well, there's no warrant needed.

I would also submit that it's possible that sending everything through a giant computer-magic-bullshit-mixer allows you to discriminate on the basis of race while claiming plausible deniability, but SCOTUS has already constructively repealed the 14th Amendment between blessing Kavanaugh stops and the Roberts Court steadily repealing the Voting Rights Act, Bivens claims, etc.

See also: Parallel Construction (i.e. evidence tampering) and most of the times a "drug-sniffing" dog is called to "test" something the police already want to search.

On a somewhat related note, it always bothers me that the discussion is about whether it’s appropriate for the government to buy this sort of data as opposed to whether it is appropriate for anyone to sell, or for that matter collect, that data.

I would prefer if neither the government nor any data brokers or advertisers had this data.

Right; but as far as I know Palantir don't sell commercial data. That's my beef with this whole Palantir conspiracy theory. I am far from pro-Palantir but it really feels like they're working as a shield for the pitchforks in this case.

1. They provide tech that is used to select targets for drone strikes and apparently also for targeting violent attacks on US civilians. I don't know too much about how the algorithm works but simply outsourcing decisions about who lives or dies to opaque algorithms is creepy. It also allows the people behind the operations to avoid personal responsibility for mistakes by blaming the mistakes on the software. It also could enable people to just not think about it and thus avoid the moral question entirely. It's an abstract concern but it is a legitimate one, IMO.

2. I don't know if this is 100% confirmed but we have heard reports that Elon Musk and DOGE collected every piece of government data that they could get their hands, across various government departments and databases. These databases were previously islands that served one specific purpose and didn't necessarily connect to all the other government databases from other departments. It's suspected that palantir software (perhaps along with Grok) is being used to link all of these databases together and cross reference data that was previously not available for law enforcement or immigration purposes. This could enable a lot of potential abuse and probably isn't being subjected to any kind of court or congressional oversight.

I certainly think that Palantir has ethical issues; as I stated in my parent comment, it wouldn't be high on my list of choices for places to work.

But, when it comes to things like (2), this is a failure of regulation and oversight and needs to be treated as such. Note that this doesn't make Palantir "right" (building a platform to do things that are probably bad is still bad), but there's no reason anyone with basic data warehousing skills couldn't have done this before or after.

Essentially, I think people give Palantir specifically too much credit and in turn ignore the fundamental issues they're worried about. Panic over "dismantle Palantir" or even the next step, "dismantle corporate data warehousing" is misguided and wouldn't address the issues at hand; worry about government data fusion needs to be directed towards government data fusion, and worry about computers making targeting decisions needs to be directed at computers making targeting decisions.

If you couldn't go backwards Palantir wouldn't have a market. So, I would consider that a loophole.

Do they? I don't think they even do this, either.

I have really strong knowledge of this from ~10 years ago and weak knowledge from more recently. I'm happy to be proven wrong but my understanding is that they don't sell any data at all, but rather just consulting services for processing data someone already has.

One of those consulting services is probably recommending vendors to supply more data, but as far as I know Palantir literally do not have a first-party data warehouse at all.

They did! At least three different versions of it!

It's vibes and drive by cheap shots, all the way down. I get that dabbling in adtech is not great, I get that they've cycled through side bets recently without committing to them (unlike Google?!), but it's an ounce of truth with every pound of nonsense. Mozilla Derangement Syndrome.

This is becoming steadily less true over time, as LLVM IR is growing somewhat more divorced from C/C++, but that's probably a good way to start thinking about it if you're comfortable with C's corner case semantics.

(In terms of frontends, I've seen "Rust needs/wants this" as much as Clang these days, and Flang and Julia are also pretty relevant for some things.)

There's currently a working group in LLVM on building better, LLVM-based semantics, and the current topic du jour of that WG is a byte type proposal.

First of all, you're right. I'm going to reply with amusing pedantry but I'm not really disagreeing

I feel like in some ways LLVM is becoming more like C-in-SSA...

> and the current topic du jour of that WG is a byte type proposal.

That's a case of becoming more like C! C has pointer provenance and the idea that byte copies can copy "more" than just the 8 bits, somehow.

(The C provenance proposal may be in a state where it's not officially part of the spec - I'm not sure exactly - but it's effectively part of the language in the sense that a lot of us already consider it to be part of the language.)

I'd have to double-check, but my recollection is that the current TS doesn't actually require that you be able to implement user-written memcpy, rather it's just something that the authors threw their hands up and said "we hope compilers support this, but we can't specify how." In that sense, byte type is going beyond what C does.

That's my understanding too

> I'd have to double-check, but my recollection is that the current TS doesn't actually require that you be able to implement user-written memcpy, rather it's just something that the authors threw their hands up and said "we hope compilers support this, but we can't specify how."

That's also my understanding

> In that sense, byte type is going beyond what C does.

I disagree, but only because I probably define "C" differently than you.

"C", to me, isn't what the spec describes. If you define "C" as what the spec describes, then almost zero C programs are "C". (Source: in the process of making Fil-C, I experimented with various points on the spectrum here and have high confidence that to compile any real C program you need to go far beyond what the spec promises.)

To me, when we say "C", we are really talking about:

- What real C programs expect to happen.

- What real C compilers (like LLVM) make happen.

In that sense, the byte type is a case of LLVM hardening the guarantee that it already makes to real C programs.

So, LLVM having a byte type is a necessary component of LLVM supporting C-as-everyone-practically-it.

Also, I would guess that we wouldn't be talking about the byte type if it wasn't for C. Type safe languages with well-defined semantics have no need for allowing the user to write a byte-copy loop that does the right thing if it copies data of arbitrary type

(Please correct me if I'm wrong, this is fun)

And outside of that, basically everything you'd consider plonking another path already exists with some service: BART runs up the east shore of the bay, as it does west of San Bruno Mountain. You have two mountain crossings covered by BART and one by ACE. The main missing things are curving BART back into San Jose and reactivating the Dumbarton Bridge.

Still if you built that the gap between Millbrae and Redwood city is 12 miles.

I just don't hold much faith in the separation of power between state and federal government, it sounds like there has been a massive erosion of this barrier since the civil war and the abuse of the commerce clause and all that.

They will certainly attempt to dismiss any charges, but they are far from guaranteed that they will be successful.

So is it part of their official duties to walk in front of a car of someone who is trying to leave the scene, alter their path when the car turns out of the way to ensure they remain in their way, and then shoot the driver? Or is that merely the kind of excessive force that's in contravention of their training and not part of their job role?

1. Prosecution. US wont because of trump. State might not because of some threat from trump.

2. Proof beyond reasonable doubt. Video can make case for self defence so bar is high.

3. No presidential pardon.

2. We'll let the jury decide but ICE is specifically trained not to shoot at moving vehicles and not to approach them from the front.

3. The president can't pardon someone convicted of murder in a state court, he can only pardon people convicted of federal crimes.

"Minnesota officials say FBI blocked their access to ICE shooting probe"

https://www.startribune.com/hennepin-county-prosecutor-attor...

anyone who thinks otherwise hasn't been paying attention to the hundreds of analogous situations that have been happening to blacks.

literally just look it up, the exact scenario has already played out before. cop gets off. and I mean literally exact scenario - with two cops, one in front, one to the side, driver (black) tries to drive off, front cop shoots and kills the black guy, is convicted, and acquitted. seriously, look it up.

George Zimmerman stalked Treyvon Martin, shot and killed him and was still acquitted. you think this guy is going to be convicted really?

I wouldn't put the chance of a conviction here at 100%, but it's certainly well above 0%.

if you have an example of a similar situation where the cop is convicted I am very very interested.

The things the POTUS says, are intended to further the US Government's goals. The actual statements made may be true or false.

If POTUS says "We did X because Y", that's no guarantee that Y is the reason that X was done, or even that X was done at all. That just means that POTUS would like people to think that Y was the reason X was done.

That Trump is also a serial liar is not actually relevant here, this is true for every President. They make statements in service of their agenda, not in service of the truth.

Secondly, even if argument could be that 'some other, more credible president would lie' - this actually does not hold up, because nobody could operate in those terms.

The presidents statements in an official context are official, that's it. Except in rare cases.

"He tells people what to do on a whim" and "has longstanding personal beefs and gripes" - that's it.

We don't know what he's going to wake up and tweet tomorrow so all we have are his statements.

Also, I think we give way to much credit to this notion of '4d chess' - he lies in the moment because he can get away with it, not out of some well plotted deception. He's not servicing some complicated scheme - just his gut.

He'll say something else the next day, but for that moment, what he says is policy.

* "Statements", not "statement". Past statements can be used to assess the credibility of more recent ones.

* Actions speak louder than words. Pardoning the king of cocaine trafficking demonstrates just how seriously the administration is trying to counter drug trafficking.

I'm sorry, this is nonsense. "He makes things up, therefore we have to take the things he says as credible"?

The President is not an oracle of truth, nor are his words the most accurate representation we have on the intentions of US government actions.

Let's say he had said directly, "The January 3rd operation in Venezuela was a best-effort attempt by the US to take control over Venezuela".

Now let's say he had instead said "The January 3rd operation in Venezuela was a best-effort attempt by the US to take control over Madagascar".

You would genuinely, truly believe that in that moment, the capture of Maduro from Venezuela was the most effective thing the US government could do to take control of Madagascar?

The presidents statements are the legitimate statements of the State of the United States of America, it has nothing to do with what you or I think about 'Madagascar'.

He is POTUS, his words are nominally and pragmatically state policy.

If he makes a declaration of 'use of force' against another it should be taken at face value.

This would be true if were only a nominal figurehead, leaving policy to others, but he's not, he has material power and wields it.

Given the construction of the balance of power - 'He is America' at least for the time being.

The statements POTUS makes to the public are simply statements by a person and should be taken as such.

The instructions the POTUS gives, privately or publicly, to the various apparatuses of the US government, are what is nominally and pragmatically state policy. When these contradict public statements POTUS has made, it is these instructions that are what actually matter.

First this: "The statements POTUS makes to the public are simply statements by a person and should be taken as such." <-- this is definitely not true, even with a basic deference to the more traditional, formal view of the US president's role, or the role of any Head of State for that matter.

The US Presidents proclamations are policy, and always have been. Obviously - a statement at the 'correpondents dinner' is not the same thing as a quick media response, is not the same thing as a statement from behind the podium, is not the same thing as a prepared address or document - but anything above board is representative of the State.

Particularly given the current POTUS leverage over Congress and wide Judicial deference to his power.

Obviously, POTUS is going to have private discussions and give directions that are not consistent with public statements - that adds to the ambiguous nature of his statements, but his public statements are still facto policy and must be taken at face value.

A statement like 'force is on the table' internally may seem like a negotiating tactic or 'populist politics' or 'stuff tough business guys say' or even 'fodder for fox news', but geopolitically it's borderline a declaration of war. It should be taken seriously.

Sure, what a head of state says should be taken seriously. It has the potential to resemble what the country does.

That's not the same thing as "the most recent thing out of the mouth of the head of state Is The Best Idea We Have Of Current Policy". That's asinine.

You are welcome to believe everything that President Putin is saying about anything, including Ukraine.

That's a profoundly absurd statement. Appeal to authority is a fallacy, especially with a trackrecord of an "authority" lying.

If the President's words are the truth, what to do with the statements in which he contradicts himself? What about situations in which 2 presidents disagree?

>The presidents statements in an official context are official, that's it.

Official, perhaps most of the time. Truthful, definitely not.

What's up with the inability to separate "opinions/statements" vs. "facts/truth"?..

But his statements are the position of the US government aka the most 'truthful' representation of US policy.

I'm responding to the notion that because he lies and misrepresents, his statements don't count as representative somehow, which is not true.

If he says 'military force is on the table' for acquiring Greenland, we should assume he means to invade if wants.

2. "Truthful representation of US government policy"

They are 2 very different things. And even the second one can be easily debated against due to:

1. Discrepancy between what countries say vs. what they actually do. Threats, lies, dishonesty, hiding truth, creative paraphrasing, etc. are normal ways the politics operates.

2. Trump's twitter messaging. What he says does not necessarily represent even his own opinions and policy. Case in point, when he announced the no-fly-zone over Venezuela a few weeks ago. The problem? It was only a tweet. No actual commands/decisions were made/given to the diplomats, bureucracy, military. It was a fake news by the President himself.

There are other credible sources whose consensus could be checked.

An invasion with the intent of taking control of the country would not involve troops arriving in the capital, completing their mission perfectly with no losses on their side, and then everybody leaving, such that no enemy troops remain.

Your understanding of the relationship between the truth and the words being spoken by POTUS are the only discontinuity here. Update that expectation and everything makes sense.

The bet was, specifically:

This market will resolve to "Yes" if the United States commences a military offensive intended to establish control over any portion of Venezuela between November 3, 2025, and January 31, 2026, 11:59 PM ET. Otherwise, this market will resolve to "No".

FTFY