pacala

Ah. Thanks for the data. Now that you mention it, I realize that plenty of drivers run red lights regularly. And yet I very rarely registered it as an emergency situation, especially compared to cyclists. What is going on?!

* Some cyclists blatantly run a red light in the middle of the cycle, oftentimes out of nowhere. Overall, perceived as unpredictable behavior.

* Some drivers tend to run the yellow, and sometimes pass through the intersection a bit after the red has been on. It just happens that the red cycles overlap a bit, then the other drivers are still stopped, and alert for the road to clear. While annoying to lose 500ms of the start, it is fairly predictable and rarely leads to sudden brake slamming.

I have seen maybe once or twice in a decade a driver running a red light. I see about 1 in 10 cyclists running a red light. The proportions aren't even close.

City driving: some drivers have the unpleasant habit of approaching an intersection from a secondary road with no visible slowing down, slamming the brakes at the very last moment. If I happen to glimpse such a vehicle, I hit the brakes myself. Right-of-way won't pay for damage costs.

Can you please give an example of React hooks 'dynamic scoping'? See https://stackoverflow.com/a/22395580 for an introduction to dynamic scoping.

React hooks 'state' variables are scoped to a single function. They are not arbitrary variables, but represent inputs to the program. The mental model is a function that produces the same output given the same inputs. Think of a dropdown, where the associated state variable D my have values a, b or c, depending on what the user chooses. The render function simply does not care how the value of D was set, and renders the exact same jsx given a specific D value: jsxa for a, jsxb for b and jsxc for c. That is as pure as it gets. Furthermore, the dropdown state variable D never represents the intermediate computation of some other component[s], and it's never changed by other components arbitrarily based only on the variable name.

The use of "dynamic scoping" to describe React Hooks state is unnecessarily imprecise, implying that a fairly well designed system is a specific kind of mess. Please don't engage in FUD.

Tip: Never call setFoo functions from render code. Only call setFoo from event code.

Memoization is getting me too. I end up memoizing everything, which is more verbose than I'd like. Perhaps we're missing some obvious pattern?

You'd be surprised. The only deeper dive into Components is https://reactjs.org/docs/state-and-lifecycle.html of the "Main Concepts" section, which I skipped in favor of https://reactjs.org/docs/hooks-state.html of the "Hooks" section. Haven't got to "Advanced" yet, hope to stay clear of that.

Of course.

* "A Critique of React Hooks"

* "A Critique of the Change Costs Induced by React Hooks".

Objections to React Hooks are stronger if they don't invoke change costs as the primary three concerns.

Huh? You can treat state as arguments to your function. For a given function evaluation, the state is stable. The fact that state can be changed during event handling is 100% irrelevant to the evaluation. There is no dynamic scoping.

1. It's not more stuff to learn, it's different stuff to learn, arguably less. Learning Components is not a prerequisite for learning Hooks. Or perhaps I missed the memo, as I've built an app using Hooks, and still haven't need to learn what 'componentWillMount' is supposed to do.

2. Don't mix Components and Hooks.

3. Agreed, change is hard. It's also the only way to avoid stagnation. In the long term, change wins. Or else we'd be programming in JS1995.

4. Insufficient example. What is the business case for a memoized hook returning hooks? Perhaps there is a simpler design, can't comment.

5. There is no global control flow. There is only per function component control flow, which proceeds from top to bottom. Possibly preempted by a hook/hookfn execution, if my early learning curve is to be believed. Which shouldn't matter if one is thinking in terms of 'pure functions returning jsx', as preempted functions do not return, thus have no observable effect.

Tip: Only change hook state from event handlers, never from render function code.

I guess we have to wait. Do you expect the counts to shoot up 10x or 10%?

* Not sure how to read the Japanese study. There were 3700 passengers on that ship, the CT study only tracks 101 of them. I could not figure out the selection process. Furthermore, how does this compare to the annual flu? Quick Internet search: "Chest CT findings of influenza virus‐associated pneumonia in 12 adult patients" https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4941884. Sounds ominous, and yet here we are.

* Yay for a better website!

* Who is better positioned to succeed in asking countries for permission to release the data: random citizens, or the researchers that already have connections to the respective countries health departments on this very topic? Stop hiding behind bureaucratic excuses. Pull whatever strings you have to pull to get the data published.

Edit: Another presentation of the euromomo data: https://www.economist.com/graphic-detail/2020/04/16/tracking.... Looks quite a bit more balanced than the NYT piece.

Thanks for the link. If anything, these graphs are assuaging. I notice 2 things:

* The spike is steep, but it appears to have quickly receded. Perhaps an artifact of what you are calling 'not done counting deaths'. Would you elaborate, what's holding back the data, when can we count [sic] on accurate counts?

* The spike is not that tall. In deaths/week, 2017 peaks at 70k deaths and 2020 peaks at 86k over all age groups. Going to the 15-64 demographics, turns out that 2018 peaks at 8.9k, whereas 2020 peaks at 9.9k. So we've got 1k extra deaths per week for a few weeks at a population of 200M+ people, the end is nigh?!

https://www.euromomo.eu/graphs-and-maps/

PS. Speaking of sensationalist, this is a great time to start presenting data using honest charts that start at zero, and not arbitrary high values [40k/7k] to magnify small variability.

PS2. Does anyone know how to get the euromomo dataset? In a middle of a global pandemic, a bit sad that a critical piece of data is unavailable for independent verification / presentation.

... in a sensationalist manner.

If you scroll down low enough, you stumble upon per country multi-year comparisons. Notice how for Sweden, Switzerland, France or England the spike is not much higher than the seasonal 2017 flu.

Moreover the peak of the 2017 flu is cut out of the graph for France and England, as a couple weeks before/after new years are cut out from the graph for aesthetic[?!] reasons. For a publication that wants to be taken seriously, cutting out the most relevant comparison point is embarrassing.

And, of course, no breakdown by age.

The spike has all but receded. Maybe be we're just having 3 flues in one year. Reason enough to stop the world?

How about snooping the traffic through a load balancer service managed by AWS? That's exactly 'identifying trends on a competitive marketplace and taking advantage of them by launching competing product', except that instead of looking at sales data of products on your shelves, you look at URL access patterns for sites hosted on your platform.

At the very least, they own your IP traffic. From there, every single value-add service you use gives them an opportunity to eavesdrop your data. Take, for example, https://aws.amazon.com/elasticloadbalancing. All of the sudden your URL traffic is 'fair' game.

That end is incredibly sad. We still have to pay our mortgage and our dentists, but can't ask for a fair pay for our highly skilled work. I guess we deserve the ad-based monetization model that has taken over the world. Brave new world, let's shill for megacorps with all the clickbait we can possibly muster.

The more restrictive the license, the more it drives potential users and collaborators away.

That being said, do you have a concrete licensing model in mind?

How do you solve for 'work for 5 years on open source project, get it to some degree of popularity, watch AWS making a service out of your code without paying you a dime'?

[Not the person you addressed this question to].

I don't. Do you? We are 3 months into this mess and have wildly divergent facts that can't be reasonably reconciled.

Some attempts:

* The regular flu season uses 75% capacity of the system. A disease that's 3x worse will lead to 225% capacity needs.

* Maybe this disease is remarkably more contagious than the flu. In a tightly packed place like NYC, you might get all the cases that you're ever going to get flowing through the system within x weeks, whereas a less infectious flu season might trickle the same caseload within 2x weeks.

* This is out of the ballpark deadly, 10x-100x worse than the flu.

That attitude reminds me of certain type of software engineers: We are engineers, our time is too valuable to write tests, hire someone else to do so. When I hear that, I run away. Fast. The world needs results, and there are little. How about we stop the excuses and roll our sleeves instead?

FWIW, that's the exact same point I made in the GP, with s/HIV/old age/ and s/flu/covid/.

I agree, the fatality rates reported in the media smells of overhype. A lot. Over here in WA, the testing policy is along the lines of "Test patients hospitalized with severe lower respiratory illness or people working in critical sectors, and then maybe consider people with 100.4+ fevers and/or shortness of breath" [0]. That population selection has an elevated death risk, and also muddles the positive/negative ratio: symptomatic people test positive, whereas health workers test negative. We get high negative test ratios, falsely indicating we're early in the outbreak, and also high mortality ratios among positives, falsely indicating a high fatality rate. Extrapolating to the entire population is wrong and irresponsible.

The flu and vaccine point is also well taken. For the entirety of human history the flu was a health risk for the elderly and life expectancy [excluding under 5s] was 60-70 [1]. The contemporary expectation that everyone should make it into their 80s and 90s is a strong outlier.

OTOH, people counter by pointing out that NYC is already at 0.16 [14000 / 8400000] fatality rate [2], and it's unclear how many more until this wave recedes. Then anecdotes about dead healthcare workers, reinfections, mutation rates, etc.

[0] https://www.doh.wa.gov/Portals/1/Documents/1600/coronavirus/...

[1] https://www.quora.com/Excluding-child-mortality-what-is-the-...

[2] https://covid19tracker.health.ny.gov/views/NYS-COVID19-Track...

I would love to bound the death rate, both lower and upper, ideally by age group. Right now I'm torn between 'it's a nasty flu, soldier up' vs. 'it's a supernasty disease that's going to severely upend life once it becomes endemic'. With the caveat that, given how easy it spreads, we might not be able to realistically do anything to prevent it becoming endemic.

Would you please be able to expand on the 'seasonal flu' point you made?

Nitpick: The percentages sum up to 100, and represent the percent of total deaths, not deaths among people of certain age.

Not sure pointing fingers helps. Someone could also point fingers and claim "so you're willing to let people starve in the streets, how altruistic of you". It's a highly emotional and volatile situation, let's keep our cool.

Conducting experiments is expensive in both money and effort. Let's all sit in our comfy chairs and build fancier models with garbage data instead. If we could divine just the right formula, it will magically paper over the holes in the data. Sounds like every other scientific area I'm aware of. For whatever reason, the experimentalist is a dying breed.

Edit: Do you happen to have a link to a good aggregator for listening to "what's being published"?

The disease does not affect age groups equally. From what I've seen, most victims are in the 60+ age groups. From NYC fatality report [0]:

    0-9: 1 0%
    10-19: 6 0.1%
    20-29: 56 0.4%
    30-39: 227 1.6%
    40-49: 535 3.7%
    50-59: 1410 9.8%
    60-69: 2863 20.0%
    70-79: 3811 26.6%
    80-89: 1753 25.4%
    90+: 12.2%

Think of people having a "vitality score" that decreases with age, and of disease as a stressor that kills people with the "vitality score" under a certain threshold. That's how the yearly flu works, by culling people under the threshold. We attribute the death to "the flu", but we could easily also call it "death by old age".

Most working age people will be just fine. Except they are jobless. Terrible times.

[0] https://covid19tracker.health.ny.gov/views/NYS-COVID19-Track...

January. February. March. April. It's been almost 4 months since we know about Covid-19. From the rock I'm living under, it appears that the scientific community at large has produced remarkably little reliable data. Where are the epidemiological studies?

Producing reliable data is the highest priority.

The conflict of interest part is fair.

The "Note that the senior author Bhattacharya and of course Ionnadis were advancing this theory before data collection began" is not. That's how science is done: you advance a theory, then you conduct experiments to validate / invalidate the theory.

They made a hypothesis and conducted a study to validate it. Perhaps they are wrong. Would be far from the first time studies are proven wrong. At least they did something. Make a better study. Make a hundred better studies. Prove them wrong!

We are already drowning in misinformation. We have no clue how deadly this disease is, and yet we are using what appears to be [gross?] overestimations to drive public policy to the tune of trillions of dollars printed with a flick of a pen.

Produce and spread good information instead.