I found the most interesting part of the NIST outage post [1] is NIST's special Time Over Fiber (TOF) program [2] that "provides high-precision time transfer by other service arrangements; some direct fiber-optic links were affected and users will be contacted separately."
I've never heard of this! Very cool service, presumably for … quant / HFT / finance firms (maybe for compliance with FINRA Rule 4590 [3])? Telecom providers synchronizing 5G clocks for time-division duplexing [4]? Google/hyperscalers as input to Spanner or other global databases?
Seriously fascinating to me -- who would be a commercial consumer of NIST TOF?
I never saw a need for this in HFT. In my experience, GPS was used instead, but there was never any critical need for microsecond accuracy in live systems. Sub-microsecond latency, yes, but when that mattered it was in order to do something as soon as possible rather than as close as possible to Wall Clock Time X.
Still useful for post-trade analysis; perhaps you can determine that a competitor now has a faster connection than you.
The regulatory requirement you linked (and other typical requirements from regulators) allows a tolerance of one second, so it doesn't call for this kind of technology.
> I never saw a need for this in HFT. In my experience, GPS was used instead, but there was never any critical need for microsecond accuracy in live systems.
> The required accuracy (Tables 1 and 2 in that document)
no, Tables 1 and 2 say divergence, not accuracy
accuracy is a mix of both granularity and divergence
regardless, your statement before:
> The regulatory requirement you linked (and other typical requirements from regulators) allows a tolerance of one second, so it doesn't call for this kind of technology.
> accuracy is a mix of both granularity and divergence
I respectfully disagree.
In context, "granularity" is nothing more than a resolution constraint on reported timestamps. Its inclusion adjacent to the specified "divergence from UTC" is a function of market manipulation surveillance objectives as discussed in preamble item (2), and really doesn't have anything to do with accuracy proper.
any time i am certain of something i never capitalize and i do not end my sentences with periods or use any punctuation because i like people to believe i am an omniscient narrator who cannot be interrupted
My guess would be scientific experiments where they need to correlate or sequence data over large regions. Things like correlating gravitational waves with radio signals and gamma ray bursts.
Would the police actually try to investigate from where came the jammer? Might the competing firm possibly even finance an investigation themselves privately? And if so, would the police then accept the evidence?
The victim firm would definitely notice, they’d tell the FCC, and their investigators will show up with a device that literally points them to wherever the jammer is. If you do this for stupid, silly reasons you will get fined[1], if you do it in commission of another crime you will probably get made an example of. It doesn’t matter how evil you are, it’s hilariously easy to get caught doing this.
> “Mr. Bojczak claimed that he installed and operated the jamming device in his company-supplied vehicle to block the GPS … system that his employer installed in the vehicle,” the FCC decision stated.
I'm not surprised that somebody would try and do this. However it is just so stupid at every level.
Sorry, you are correct. As soon as the subject of HFT came up I was thinking about London and the things they do to reduce latency to the exchanges in North America. It's too late to edit or remove my previous message.
The parent was saying HFT firms would do this to other HFT firms. They would care about doing this kind of thing - it’s not a white collar crime. And foreign adversaries would care about doing this during peacetime, especially for very unclear benefit.
But they do not need absolute time, and internal rubidium clocks can keep the required accuracy for a few days. After that, sync can be transferred with a portable plug, which is completely viable in tactical/operational level EW systems.
Google doesn't use chrony specifically, just an algorithm that is somewhat chrony-like (but very different in other ways). It's called Google TrueTime.
Ah yeah. For VPS tenants it makes sense they would default to Chrony. They have a public facing pool of NTP servers at `time.google.com`, and for tenant use they provide `metadata.google.internal`, which is probably where the Chrony config file points. IIRC TrueTime is not actually open source and is only used internally on their infrastructure.
I was optimistically hoping some of the MV3 changes would result in Chrome webstore policy enforcement being standardized, but that hasn't happened.
Sensor Tower (https://sensortower.com/) makes a lot of popular extensions, like StayFocusd https://www.stayfocusd.com/. They seem to resell ad data (in violation of [1]?) and ship likely obfuscated code [2] (in violation of [3]?), but there's no enforcement or even clear reporting mechanism.
MV3 makes it considerably harder to introduce a security vulnerability, but it doesn’t really help with outright malicious extensions. In the end this isn’t an issue which can be solved by technical means. It’s a moderation issue, and Google currently seems to be scaling back moderation despite not being great at it to start with.
Event with MV3 you still have access to `chrome.webRequest.onBeforeRequest` and content scripts, so this particular issue won't be 100% solved.
I don't think the solution is technical. The solution would be a strict policy, and nuke every extension and publisher from the store who even hints at doing this kind of BS.
Agreed! This specifically is for adjuvant osimertinib for EGFR+ Stage IB–IIIA completely resected NSCLC.
The headline here is really strong -- and the actual abstract is much more sober: "5-year OS rate was 88% with osimertinib vs 78% with placebo" [Full abstract is here: https://meetings.asco.org/abstracts-presentations/219805 ]
P.S. Hi Chris! (I think I picked up a summer student from you last week!)
Fair point, but then the article is demonstrating that people trained to do specific types of care actually might have better outcomes than generalist education.
This still seems significant to me.
Most of the comments so far seem focused on the fact this was done in a resource-limited setting, and comparing specialists to generalists. I think that's important to keep in mind, and maybe the title was misleading (I didn't post the article).
However, a study like this would be difficult to do in less resource-limited settings due to all sorts of issues, some of which are due to prudence, and some of which probably less so. It's typical of medical research in developed countries to not lower the standard of care, so this sort of study might never get done otherwise.
So, if you take it for what it is, it's suggesting that a type of procedure classically pointed to as a reason for strict medical licensing forms in fact does not necessarily work the way you think in terms of training background and outcomes. It doesn't point to getting rid of licensing, it just suggests that a particular type of educational and training background does not necessarily result in better outcomes.
This type of finding isn't uncommon in different areas of medicine, and the research is often fighting against double standards, in the sense that you're not just comparing training models, you're comparing time with training program experience as well: the alternative training tracks are often newer and involve less opportunities to have worked out problems, improve certain issues etc.
The irony is that this sort of thing is playing out in the US anyway, under the radar. For example, medical schools are often reducing coursework to a 1.5 years or even 1 year, meaning that a PA with an additional 2-3 years of training post degree often has as much experience in the clinic as a new MD. If you took them and gave them 3-4 years of additional training, they'd probably look similar in outcomes to someone coming out of a residency. I might be wrong about that, but healthcare administrators are increasingly voting with their dollars and apparently don't really see a significant difference in outcome.
The comparison makes sense in the context of Sierra Leone, where the study was done. The full article mentions that there are very few surgeons there, and that basic surgeries are usually done by general MDs or ACs, the types of training compared.
Certainly relevant to Sierra Leone and other poor countries, probably not relevant to rich countries that generally have surgeons do this type of surgery.
A minor clarification, it compares MDs to ACs. It says ACs are associate clinicians who have an amount training somewhere between a nurse and a doctor.
At least the way it's used in the US (I know hacker news is not only the US) a medical assistant is someone who has less training than a nurse and is often doing administrative work.
Oberlin has an endowment valued at approximately $1 billion — this is ~3% of that. (Understanding all those funds may not be liquid, they certainly have some assets.)
Their annual revenue is ~$180M. If any appreciable portion of their student body turns on them it's going to start to hurt very quickly. A sober assessment says that very few students would actually leave but there's a tail risk that it turns into something unsustainable.
That's a gross oversimplification. From the video the poster provided:
* V16 of the Epsilon OS by Numworks removed the ability to install custom operating systems to the device in response to pressure from the Dutch education department, who had adopted the calculator as the standard for their education system. This was due to information provided by a different Youtuber accusing the platform of being used for cheating, and then provided modified versions of the OS that enabled cheating support by working around the Exam Mode functions.
A fair and true statement would be that the NumWorks calculator was originally marketed as open hardware, and that functionality was removed (similar to Sony removing OtherOS from PS3's).
Carbon costs per-vehicle can be calculated based on your local grid power source, duration of ownership, and more: https://www.carboncounter.com/ -- be sure to click the "Customize" tab
If your annual driving distance is low (<5,000 miles) and your grid is relatively dirty (e.g. the midwest [SRMW] grid), a range of EVs have more CO2 emissions/mile than conventional internal combustion vehicles.
> If your annual driving distance is low and your grid is relatively dirty, a range of EVs have more CO2 emissions/mile than conventional internal combustion vehicles.
Right now. I'm sure you know but the beauty of this is that those cars' emissions can be reduced without changing anything about the car simply by changing the fuel used to generate electricity which can (and has to) happen in the future.
Yes this is a key point lost on many when I hear this argument. It doesn’t mean to not buy an EV, it’s still a great idea, we just need to do more work on our power sources
Within a decade from when my Volt was manufactured to when I sold it, the carbon emissions per kWh dropped from 500g/kWh to 372g/kWh. Gasoline emissions have not dropped and might even be getting worse as we have to reach further to get nonconventional oil from tar sands, etc.
Those cars' emissions can also be made worse over time without changing anything about the car simply by changing the fuel used to generate electricity....
Such as Germany currently decommissioning nuclear plants...
That didn't stop Germany from reducing its oil, coal and gas consumption though. Those reduced over time or at least stayed the same while renewables steadily increased.
Related data on CO2 intensity on grids, this shows a map based on EPA data. It shows what the equivalent CO2 emissions would be between an average EV compared to a gas car. The worst in the lower 48 is SERC Midwest (parts of IL and MO) at around 42mpg, the best is upstate NY at 255mpg.
Customize, click on West Virginia preset and graph uses 560 gCO2/kWh to EPA's 860 for WV for Electricity (charging). Select Idaho and graph uses 310 gCo2/kWh compared to EPA's 96.
Carboncounter seem to be using the eGRID subregions, but I'm not sure how much green power actually flows from southern Wisconsin all the way to western Virginia. Maybe there's a power expert here that can comment on whether state or eGRID subregion numbers are more appropriate.
That the graph origin is not 0,0 by default and includes tax credits doesn't inspire much confidence in their impartiality.
One thing to note is power isn’t consistently generated M from the same source. Sometimes more wind, sometimes more solar, sometimes more gas. There’s currently a surplus of wind power overnight in that region and There are several companies teaming up with utilities to shift EV charging to lower cost, lower CO2 producing times.
The fundamental difference obviously being that electricity production can change significantly during the lifetime of a vehicle whereas energy production within the vehicle usually can not.
Seems like the overwhelming majority of the midwest is outperforming Missouri/Illinois, but yeah, that region is performing poorly.
I’m also disappointed that the second link didn’t include the emissions in the manufacture of the vehicle itself, which will be higher for EVs than gasoline cars, but from what I’ve read[0] the lifetime emissions of an EV are far far lower (breakeven point is about 10-15kmi).
Physician scientist here: this study is a bit dated. Many of these issues have been "solved" (depending on your threat model) within the last ~7 years. Most healthcare systems have adopted Imprivata [1] for SSO, where physicians tap a badge and are connected to (usually) a VDI session of Epic.
What this study misses is the real driver of EMRs: billing. EMRs exist to facilitate billing documentation to charge for patient care. Yes, they have other benefits (like viewing lab results), but if you ever see true critical care (at the bedside, in an ER, or in an ICU) little depends on the EMR (or even labs for that matter).
A few comments here talk about patient notes: in most clinical environments, inpatient notes are useless, and a tedium required to bill. They're filled with copy-pasted jargon to meet insurance company requirements. True patient care happens in less than ~1 paragraph of text called a handoff. [2]
> A few comments here talk about patient notes: in most clinical environments, inpatient notes are useless, and a tedium required to bill. They're filled with copy-pasted jargon to meet insurance company requirements. True patient care happens in less than ~1 paragraph of text called a handoff.
That's very interesting, and explains many somewhat baffling experiences. So most of what I tell or give a doctor is effectively lost? That's the perspective from this patient, who finds it very frustrating to repeat myself, that the doctors are ill-informed and often perform that way.
But being open-minded: Is it that why I tell (or what most patients tell) the doctor isn't really useful to them? What is useful? Or is it just a consequence of time pressure?
Patients also don't usually know what is useful or not for a Doctor to diagnose them (or any medical provider usually). That has to be picked out as part of the Doctor or other medical provider questioning them and getting their vitals.
Some of it is implicit, like 'is the person able to form cogent sentences', or 'does the patient seem to be struggling to stand'. Others are more explicit, like heart rate, BP, 'what medicines are you taking', etc.
Often the ones who need the most help are the worst at getting anything useful out.
Sure, but I've told apparently critical info to doctor A, and then doctor B didn't know it (when I've told doctor B, it has changed their conclusions).
The patient's time has zero value to most physicians. It's often faster, and more clinically useful, for them to just ask you for your complaint and medical history. This allows them to guide the conversation and quickly focus in on the exact information they need. Searching through old chart notes is often a waste of time since the systems are slow and the data is frequently incomplete or irrelevant.
> Searching through old chart notes is often a waste of time since the systems are slow and the data is frequently incomplete or irrelevant.
I believe this can change if we change the amount of patients a physician must assess and treat in a given time period. Because that would be super costly (way more physicians required on staff) Nurse Practitioners and PAs would be a more cost effective way to delve into important and relevant information that takes time to uncover.
NPs and PA time isn't exactly cheap either, and they aren't sitting around waiting for stuff to do. They can be effective at treating routine cases under supervision, however it just wouldn't make sense for physicians to delegate patients histories and chart review to them. They won't necessarily know which questions to ask, and too much relevant information can be lost.
Yes that's correct. Every year, students graduate from medical schools but are unable to enter clinical practice due to a shortage of residency slots. We need more government funding to remove that bottleneck.
The UK has no link between doctors and hospitals, for a shining example. This reason alone is why so many expats I know return and receive better care from any old HMO just because the billing system glues together the case histories essential for infrequent attendants to primary care due to travel etc.
The lack of a stockpile is due to limited shelf stability and long-term bacterial growth.
It's unfortunate Reason doesn't expand on the other side of this issue: the formula industry lobbied the FDA to reduce bacterial testing frequency (and inspections overall), with an emphasis on Cronobacter risks, arguing that the FDA "overestimat[ed] the expected annual incidence of Cronobacter infection". [1]
It seems that the regulatory system is broken and I don't know how to fix it other than to remove the barriers of entry to this industry.
Four manufacturers, consisting of 89% of the nations supply, have so much influence over a regulatory agency insofar that they can write their own rules. Four manufacturers are able to shape the competitive landscape because the government has deemed their voices vital.
I've never heard of this! Very cool service, presumably for … quant / HFT / finance firms (maybe for compliance with FINRA Rule 4590 [3])? Telecom providers synchronizing 5G clocks for time-division duplexing [4]? Google/hyperscalers as input to Spanner or other global databases?
Seriously fascinating to me -- who would be a commercial consumer of NIST TOF?
[1] https://groups.google.com/a/list.nist.gov/g/internet-time-se...
[2] https://www.nist.gov/pml/time-and-frequency-division/time-se...
[3] https://www.finra.org/rules-guidance/rulebooks/finra-rules/4...
[4] https://www.ericsson.com/en/blog/2019/8/what-you-need-to-kno...