strstr

I’d argue that’s continued failure of the extradition treaty, and part of why it’s hard to craft them.

You can’t make it too easy to extradite, but you also don’t want it to be too hard.

I’m not sure what the right carve out would be to avoid extradition of journalists for comparatively trivial charges. A full blown journalistic exception seems too far, but I don’t know what the right line is instead.

Extradition treaties are voluntarily agreed to by both countries. If you are British or American and don’t like the consequences of this treaty, I encourage you to contact relevant politicians and tell them to renegotiate the treaty.

I believe the right treaty is: https://en.m.wikipedia.org/wiki/UK%E2%80%93US_extradition_tr...

A more minor detail of the piece: “The US secretary of state, Mike Pompeo, says that non-Americans like Assange do not enjoy First Amendment rights to free expression.”

This is used to imply that the US does not respect press freedom. While that might be true (personally don’t think it particularly is), non-US persons don’t enjoy any of the rights of the US constitution. There are different laws in different places. Assange should enjoy the British Rights to freedom of the press/expression, whatever those are (or the rights relevant to wherever he was).

It’s cool to see other uses of the TPM in the wild! My coworkers and I maintain go-tpm-tools[0] for trying to make using the tpm straightforward.

[0] https://github.com/google/go-tpm-tools.

This number is shocking, but it is not weighted by usage. Just because a medical treatment is a treatment doesn’t mean it’s in common use.

On top of that, I find it 100% believable that doctors use understudied treatments. The reverse makes a lot more sense. Treatments get studied after people have tried them, and believe they warrant further study and more frequent usage.

Consider coronavirus treatments, at first we had very little idea what to do, beyond inferring based on past virus with a similar symptom profile and reasoning from first principles. Later on, we slowly but surely identified effective treatments (corticosteroids/prone position/...) after people tried them. Even later, we had descriptive studies of past effectiveness. I assume most treatments go through a similar progression (covid treatments go through this progression on fast forward).

Medical studies are slow and retrospective. Fundamentally that gets you this 1/10 figure.

Patient questionnaires with the right questions seem like the way to go.

When I was considering a Bankhart repair for my shoulder the number I looked at was return to sport. Sure, people can be biased about this, but this seems about as objective as you can get: “Are you able to participate in the activities you were able to before your injury?”

If you are getting a surgery without measurable outcomes, why are you getting this surgery at all?

Market manipulation is also separately illegal. Though classic pump and dump doesn’t seem possible with amzn.

Benefitting developers is a consequence. They are the middle man for housing. You can't get housing without aiding them (at least not easily).

If your goal is to avoid benefitting developers, you will have a hard time encouraging housing.

This reads like someone in his department targeted him with the conclusion in mind: termination. Someone wanted him fired. They tried every option available, and they eventually succeeded.

Oh man these are collector’s items. I’d love to have one of the hats. Not $500 want, but I’d probably have spent double the price of the hat on them. Maybe $50-60 or something.

Godbolt is great for comparing compiler versions (and compilers).

For example, you can see gcc's progression of efficiency for c atomics with https://godbolt.org/z/brsoEr. If you increment the gcc version number, you will see the (very slow) mfence disappear, and xchg show up.

Then there is Clang at O3: If an int falls in the forest, and there is no one around, was it ever incremented? No. The function turns into a bare ret.

This is caused by induced demand. This comes up a lot for car traffic [0]. If you build wider roads, you will almost always just see an increase in traffic, up to the point where the roads are full again. The metaphor is not perfect, but I think it is fairly apt.

Expanding infrastructure increases what people can do, and so people do more things. In some cases, it just decreases the cost of engineering (you can use more abstractions to implement things more quickly, but at the cost of slower loading sites). But in the end, you should not expect wider pipes to improve speeds.

[0] https://www.bloomberg.com/news/articles/2018-09-06/traffic-j...

This is typically used for anti-spam/anti-malware. FB has a lot of automation to try to stop obviously malicious content. Just, sometimes, it hits things that aren't malicious. I'm most familiar with the malicious links that, when you click them, cause you to spam your friends with the same link.

People don't notice this automation exists, except when it doesn't work. Then they either complain "Why didn't XYZ get caught! This is obviously spam! I could write better automation." or "OMG they banned dreamwidth!".

It's possible they extended this (or something similar) to filter out sites with hate speech/"community guideline violations". And in that case some sub-domain of dreamwidth posted something horrible. And then the automation slammed the hammer on the whole domain.

This is a blanket terms of use for their data store. They want you to cite them. And they don't want it to be possible to launder the data by having a non-profit/non-commercial user re-upload, so that a for-profit user can then download.

This feels like a heavy hammer for that purpose, but it seems pretty straightforward to read between the lines on these terms.

This was previously not the case. Cat is out of the bag though.

Extraordinary claims require extraordinary evidence. Doubly so when there is motivated reasoning behind the argument.

If a pharma company came out with a paper saying “we’re almost done developing a cure for cancer!”, people would be rightly skeptical. Lethal manmade viruses are less ridiculous than that, but I’d still classify them as very surprising. Motivated reasoning for those with political ends should be unsurprising when it comes to discussing whether covid is manmade.

Checking background/credentials is due diligence to ensure you aren’t being duped. I wish the world exclusive operated on good faith argumentation. Unfortunately, it doesn’t.

SGX has a stronger security model, but targets enclaves instead of VMs.

If you want something to run in SGX, you will likely need to rewrite the software (you can’t call syscalls directly in SGX since you cannot trust their results anyway).

SEV’s security model is weaker (no integrity), but lets you use essentially normal VM images.

Disclaimer: I work at Google in this space.

These questions underly a particular argument that I don’t buy. We institute fairly hefty national standards on schools. I can go and look up the average scores for standardized tests for students for the high school I attended, and, afaik, essentially all schools in the US. Why can’t I do the same thing for police departments?

Somehow, police unions are able to avoid having similar accountability.

This definitely doesn’t make sense to me. Time zones are going to be a huge issue. If a coworker told me they were going to work from an asian/european timezone, we could try to make it work, but it’s definitely a hassle. The story is no different for universities and their courses.

I’m not certain it matters. The hydroxychloroquine paper was retracted: that’s not a great sign for it’s effectiveness as a treatment.

I see this is a welcome change, particularly as (opt-in) kernel lockdown support starts to tighten down to the point where it is meaningful. MSR writes from userspace should definitely be disabled by default for locked down kernels.

For flexibility, users should be able to opt out of this (as with the rest of kernel lockdown). Sometimes I just want a holster full of foot-guns. Blanket R/W access to MSRs is that.

There are a lot of fairly powerful MSRs. Some of them might already be restricted for all I know. Here are some examples:

  IA32_SYSENTER_EIP -> the instruction pointer after calling SYSENTER. Similar MSRs exist for the syscall instruction.
  IA32_DS_AREA/PEBS MSRs -> DS_AREA is a pointer that points to a buffer of pointers that are written to by the perf subsystem. Manipulating DS_AREA to point to a user page should let you write ~arbitrary data to the kernel.
  IA32_SPEC_CTRL -> disable some of the speculative execution mitigations.

There are less obviously powerful msrs as well, which probably let you get away with some sketchy-ness:

  APIC_BASE -> you could try to get it mapped into userspace then control the APIC directly. Or you could hide particular kernel writes by moving the apic over another page (there was an attack against SMM based on this a while back).
  X2APIC MSRs -> no need to move the apic base when you could just write to the apic through msrs more directly.
  IA32_EFER -> Unexpected switch from long mode to protected. This can't do anything good.
  MTRRs -> Changing memory types underneath the kernel (potentially restricting access to kernel memory) can't do anything good to the kernel. You could also change the caching behavior, which sounds messy.

Edit: Formatting. This is probably still unreadable on mobile.

It’s a classic story about models like these (ones with many parameters).

Cynically, I think he is just trying to save his job. If he is a nationally known blogger, he risks losing it (for reasons stated elsewhere).

If he can go to his employer and say “look I deleted the blog, there is nothing else I can do”, they might be sympathetic, even if the outcome is potentially short term negative for his psychiatry patients.

There’s a pretty big gradient between “in a national news article with his name” and “only posts on his blog using tor, from a coffee shop, on a burner device, under a pseudonym, with no personal details revealed” (or whatever constitutes a “firewall”).

At this point, casual observers do not know Scott’s identity. With a few hours of time most people on HN could probably figure it out, but in that same amount of time they’d also probably calm down if they were interested in sending death threats.

Likely, he started out blogging when he was comparatively unknown. His audience grew, but he could not flip a switch a fix his prior lack of opsec without nuking his community.

It seems he is now taking the nuclear route.

Really depends on the on the mouse implementation: g pro is only 80g. It’s hard to get a wired mouse much lighter without a honeycomb shell (mice in that style can push into the 50s).

The wireless mouse has been the exception to this rule for me so far: the G Pro wireless (and some other mice) don’t have the latency issues, and have egregious battery life (and the charging cable fails over so it’s basically just a wired mouse).

The lack of a cable saves a bunch of space and avoids the floppy cable. Probably irrelevant if you don’t game, but definitely quite convenient for FPSes.

I believe it’s considered to be somewhat hard to integrate with, but worth it if you can. It was/is Fuchsia’s default allocator.

My current job at Google working on virtualization?

You can still acknowledge interrupts in between, so hardware watchdogs will prevent this.