CapriciousCptl

Broadly speaking, two types of death are recognized with organ donation. Brain death, where there's no brain function at all and cardiac death, where there's no cardiopulmonary function.

Brain death in isolation generally provides the most viable organs, since it minimizes so-called "warm ischemia" time, which is the time that the organs are warm and without viable blood flow. However, sometimes patients don't meet brain death criteria strictly, or they do but the donor next of kin decides to donate after cardiac death (actually, the next of kin/family determines what the donor would have wanted and there is a strict legal/ethical precedence on what opinions count). In those cases, cardiac death is initiated by stopping life support and waiting 2-10 minutes after "death" to initiate organ harvest.

So, it's not that the heart caused death, it's just that it stopped beating, generally, due to lack of oxygen. Which means some of its cells certainly died, but by minimizing the warm ischemia time (by hooking up to a kind of life-support machine), that damage can be limited.

Patients do not have to matched strictly by age. Anatomically-speaking, body-size is the most important criterion. Obviously, matching needs to happen on antibodies/rejection and several other factors as well.

The device used is extraordinary. The US has been using it for a year and a half in adult trials. Transmedics also has devices for the lungs and liver. Livers and to a lesser extent lungs already get transplanted after cardiac death when they’re judged to be viable— the key being that this device could quality of those organs and therefore more could be transplanted! This could benefit a lot of people— not just hearts!

https://www.transmedics.com/ocs-hcp-heart/

It’s not a viable offline training machine but this thing flies with data munging. I have the 16gb version and pandas data frames that are 40gb or so load from csvs in two minutes with type inference for example. So long as you can use Google Collab it’s an excellent deep leaning laptop where it counts.

Sounds like what Shadertoy was built for-- https://www.youtube.com/watch?v=xDxAnguEOn8

Oh! That's a good idea. My goto has always been pipelining along a series of functions, but never thought of just using Jupyter for some reason.

Basically, the title of this post and its premise is just very wrong. Microsoft includes Office 365, LinkedIn and probably GitHub in its "cloud" revenue. AWS "cloud" revenue is not remotely comparable.

> MSFT: "Our commercial cloud revenue, which includes Office 365 Commercial, Azure, the commercial portion of LinkedIn, Dynamics 365, and other commercial cloud properties, was $16.7 billion and $31.9 billion for the three and six months ended December 31, 2020, respectively, and $12.5 billion and $24.1 billion for the three and six months ended December 31, 2019, respectively. These amounts are primarily included in Office products and cloud services, Server products and cloud services, and LinkedIn in the table above." https://www.sec.gov/ix?doc=/Archives/edgar/data/789019/00015...

Since BBC didn't post it, here's the full decision. https://www.bailii.org/ew/cases/EWHC/QB/2021/85.html.

Below, I pasted the review, followed by a bit from the judge. Written word has to meet a number of criteria to be libelous, but the core of the decision, I think, was that calling something a "scam" is a statement of fact and not opinion. https://www.legislation.gov.uk/ukpga/2013/26/section/3/enact...

""""*A total waste of money another scam solicitor* [Person X] left the company half way through my assessment and the replacement was useless. I paid upfront for a legal assessment of my case, but what I got was just the information I sent them, reworded and sent back to me. No new information or how to proceed or what the law says or indeed the implications of what was done. I Just got their false assumptions, full of errors showing a lack of understanding for the situation and the law. Once they have your money they are totally apathetic towards you. You will learn more from forums, you tube and the Citizens advice website about your case, for free"""

And a small bit from the judge's decision-- """"An allegation of dishonesty, fraud or attempted fraud will usually fall fairly and squarely on the side of fact rather than opinion. The same is true also, as I have already mentioned, where the allegation is of "reasonable grounds to suspect". Accordingly, I cannot allow a pleading to go forward in the form of paragraph 19A. It must be struck out.""""

Oh! Skepticism is good. And this type of tech is very good for goal-directed management of fluid resuscitation! Severe peripheral vasoconstriction, not so much, as you mentioned. Really, in my first-hand experience most clinicians use the tech to follow trends in the ICU or surgery.

I wouldn't call the claims Theranos-level, but at the same time I don't see how this is any more useful for the general population than the standard automated blood pressure cuffs. Maybe that will change as things evolve!

This article has a nice review of the history-- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4312947/.

You're right. If it was just one drug, I'd think they were manufacturing it themselves. Apparently, they plan to have a single factory in 2022. Most likely, significantly all of their formulary will be copacked.

Compare/contrast to my former email provider. They gave 1 month notice to some people before almost doubling prices, and others apparently didn’t get any notice at all. Left them for a more expensive service— large and sudden price increases send a message of instability and bait and switch dynamics whether or not it’s actually the case. Gitlab, on the other hand is using a phase in approach. It gives customers time to adjust and prepare. Whereas they could have just hiked prices and exploited customers who didn’t have time to react. Of course you get pricing wrong sometimes (I guess even by 2-5x). But really you have to be careful with how you raise prices. Personally I’d rather have happy customers, although captive ones pay the same.

tldr; 1. Attacker found a stackoverflow dev environment requiring a login/password and access key to get in.

2. Attacker was able to login to the dev environment with their credentials from prod (stackoverflow.com) by a replay attack based on logging in to prod.

3. The dev environments allows viewing outgoing emails, including password reset magic links. The attacker triggered a reset password on a dev account, and changed the credentials. This gives them access to "site settings."

4. Settings listed TeamCity credentials. The attacker logged into TeamCity.

5. Attacker spends a day or so getting up to speed with TeamCity, in part by reading StackOverflow questions.

6. Attacker browses the build server file system, which includes a plaintext SSH key for GitHub.

7. Attacker clones all the repos 8. Attacker alters build system to execute an SQL migration that escalates him to a super-moderator on production (Saturday May 11th).

9. Community members make security report on Sunday May 12th, stackoverflow response found the TeamCity account was compromised and moved it offline.

10. Stackoverflow determines the full extent of the attack over the next few days.

Where is the study? I couldn't find the link. "95-percent success rate" is meaningless without context. Specificity/sensitivity, prevalence and the gold standard used all matter. You, too, can detect coronavirus with 100% "success" (and much higher than 100% that depending how you calculate it).

bool carriesCoronavirus(human *subject) {return true;}

I could be wrong but the price seems too high. If you find the same, I would consider doing a bit of market research estimating how much people use the library at areas particularly far from the library, and estimating a price you would need to make it work with batched deliveries (kind of like garbage day, your library day is...). Then set your price based on that and see if it takes off.

I may be wrong. I'm reading into this that was on HN the other day--

"We collaborate with cloud service providers, including Microsoft, Google, Alibaba, Tencent, Clever Cloud, and others. We have shown we can find a way to do it. We even work with other parts of Amazon. We are always open to doing that; it just needs to be OK." https://www.elastic.co/blog/why-license-change-AWS.

AFAIK, that tweet from Vogel is the only time someone representing AWS claimed to have a partnership with Elastic. He was wrong, but I don't think it's fair calling it "dishonest."

Steve Jobs said something like Dropbox was a feature, not a product. I think Bezos feels the same about literally everything. AFAIK Azure/Google have actual partnerships with the Elastic stack, partnerships that assumedly benefit both sides and have staying power.

Part of me wonders if AWS always had planned to do this, and they were just waiting until it made business sense to fork (ie they had features and a new direction in mind but neglected to implement them because Elasticsearch was good enough as is). The alternative part is just 2 big corporations not finding a way to get along. Which means without clear direction and careful stewardship I'd expect the forks to just be cleanroom reimplementations or something like that.

This is very good! It's the most comprehensive guide of how to organize a large FastAPI project I've seen. Which is important, because FastAPI is unopinonated, so there's onboarding cost to getting a team on the same page-- strong opinions help alleviate that!

It's more Rails-like then Django/Flask-like. Meaning, top-level organization by "M"/"V"/"C" as opposed to by "Blueprint." I mean, "models/blog.py" as opposed to "blog/models.py".

Netflix Dispatch is an example project that organizes in the more Flask-Blueprint-like way (https://github.com/Netflix/dispatch/tree/master/src/dispatch).

I'm going to pour over this a bit longer, but that will take some time! So I wanted to leave some initial thoughts to people discovering this.

I'll give some context here, since I put magnets on people's pacemakers fairly often. Magnets for at least 10 years now is a standard, controlled way to trigger a secondary mode in a pacemaker/defibrillator. Almost always, if the device has both pacemaking and defibrillating capability, the magnet will turn off the ICD (although it is possible to program other behaviors). This allows clinicians to easily stop a patient from being defibrillated inappropriately. That sort of thing could happen during malfunction, electrical interference from cautery tools during surgery near the device, or refractory shockable rhythms.

If there is no defibrillating capability, magnets usually (although not always) put the device into an asynchronous mode. This presents the (extremely unlikely) risk of something serious, like R-on-T, where an electric pulse could be sent to the heart at exactly the wrong time, leading to a life-threatening arrhythmia. More commonly, it means the heart is less effective at pumping blood but probably still within allowable, if not optimal for long periods of time, output.

In both cases, unless the device is programmed strangely, the device resumes normal function when removing the magnet. However, there's the theoretical risk of some sort of malfunction-- so when switching modes with a magnet my practice is to refer patients to getting their devices checked out afterwards. Obviously, if something triggered the "magnet mode" surreptitiously, it may not be noticed until some time later, particularly in patients who are not being actively followed by medical care.

I didn't mention it's actually kind of difficult to getting a magent over these things. I generally tape it over the area that the pacemaker is in, but you have to watch for it moving because, well, it's hard to stick something on somebody's skin. Pacemakers are generally put in the left upper chest, sort of between the nipple and the clavicle. The idea you should have is-- well-- it's not easy to trigger this even with a purpose-built magnet in a slightly-higher-then normal breast pocket. But, still possible of course, and the results could be disastrous, particularly in the event of a malfunction.

The practice of using magnets is kind of universally and internationally understood by medical practitioners and device manufacturers. Which means, transitioning away could cause some harm just by medical errors. I don't think a solution to this is particularly easy, although many devices now get regular interrogations which would show if iPhones actually are triggering the mode switch. Which sort of means, iPhone "interference" could happen but probably isn't really happening as far as anyone knows.

Another way to look at wealth is as total spending power between now and eternity. Taken that way, it doesn't matter as much if your wealth is tied up in unsaleable stock-- stock afterall represents your claim on the discounted future earnings of a company, so those earnings will be available given enough time.

Estimating the true future earnings of businesses, and the amount available to current shareholders of course, isn't easy. But taking the stock market's approximation of that intrinsic value is quick and dirty way to get it. Of course, the caveat is that stocks aren't always efficiently priced.

In 2021, everything's a bubble.

  Bonds... 10-year treasury after tax amounts to less than inflation.
  Holding cash... The USG keeps giving out money for nothing, meaning more money chasing the same amount of goods, services and investments.
  Index funds... Earnings yields are at all time lows, about 3%.

WOW. The colors on the menu pictures are somewhat preserved. It's not hard to imagine things much more vibrant like you have to do for most paint from the period. It looks like an antiquity-era version of the food-trucks you see at carnivals even down to the yellow-ish background color.

> Found the issue - $TERM was incorrect, and LANG=c helped as well. Thanks! For the curious, it is a real vt220, and I have it connected to Arch Linux (fully up to date, well, except for the hardware). It's connected with a DB29->DB9 null modem cable I made from a breadboard, a bunch of jumper cables, and a DB9 header, which is in turn connected to a simple RS232 USB serial cable.

I don't follow. Can you edit the ka band image and circle what you're seeing?

Does anyone have a link to more context about the strike? I found this on the union website— http://www.seiu121rn.org/2020/11/29/press-release-nurses-hea...

Given the scope of the attack I think it’s reasonable to assume all high value targets were compromised in someway. It’s more a matter of what specifically was compromised, what data was exfiltrated, how much of the attack was offline, are Thunder going to start breaking a as la stuxnet etc.

The full title is, “The terms simp, virgin, and incel will all be banned on Twitch when used as insults,” which adds some nuance. But it’s interesting because now apparently the intent behind the usage matters. Maybe it’s not so easy always figuring out the intent but I suspect censoring those words even when not used as insults won’t be missed too much. Every generation adds its own curse words and prevents them from future use, the English language will go on.

In cash accounting there’s no A/R at all, right? In the context of their #7 description I think they’re talking accrual based.

Probably most people who’ve done a bit of PDF work know there’s no guarantee of the same output from different (or even the same) editor. So I don’t think it’s Preview’s fault per since the problem is endemic to PDF. But I don’t think you can blame the user either. Really, PDFs are just these enormously useful complex things that are always breaking in unexpected ways and some people haven’t been bitten by its problems enough yet to cope properly.

Heh, Burger King did that to offer $0.01 whopper coupons but I guess the difference is the app had to be in use and known to the user to recognize the location--

https://www.usatoday.com/story/money/2018/12/04/burger-king-...

I found a16z's blogpost of basic metrics great for starters--

https://a16z.com/2015/08/21/16-metrics/ https://a16z.com/2015/09/23/16-more-metrics/

Note on 16 metrics #7-- from what I understand bookings actually don't increase A/R until the services are rendered, so know what you're looking at if you don't get cash up front.