mikekchar

Alright. If you are unable to look at any other data than the one page you have. 7,961 deaths to influenza in 2015-16. Is that the number you think will show that COVID-19 is more deadly than the flu? Because, I've got to say that I don't think that number is better for you than 23,000 deaths that the data I'm pointing to. Oh, of course there are 131,858 deaths due to pneumonia from other causes. Clearly something pretty powerful that wasn't influenza swept through that year. It's pretty strange that we keep statistics on the piddly old flu when there is something killing 20 times as many people, but is completely unidentified. I'm so glad we agreed to use this data rather than literally any other page on the CDC website!

According to Wikipedia, influenza causes an estimated 41,400 deaths per year. 2017-2018 was the worst in 40 years and totaled about 61,000 (though the tally has not been finalised and some people have reported it as being as high as 80,000). The number of hospitalisations for the flue in that year was estimated at 800,000 people (clearly that isn't ICU). The number of medical visits was 21 million. I have no data for how many people were confirmed to have the flu, but were not hospitalised.

COVID-19 has killed 114,148 people to date. The number of hospitalisations are only reported up to the end of May 30 and appears to be 82 people per 100,000 people in the populations (thank you CDC for such an epically terrible statistic!) That works out to about 270,000 people. I'm going to use PCR tests as a proxy for "medical visits" for the flu, but in reality they aren't comparible. There are 22 million tests that have been carried out to date. The reason I use that as a proxy is I assume that there is reason to suspect that someone may have COVID-19 if they get a PCR test, just like there is reason to suspect that you have the flu if you have a "medical visit". The last piece of information would be the number of confirmed cases which is 2,045,549. So just under 10% of the cases that were tested were confirmed to have the disease.

Of the people who were hospitalised in the worst flu year in 40 years, about 10% died. Of the people who were hospitalised for COVID-19, about 42% died (well a little higher because I'm using June 10th data for deaths and May 30th for hospitalisations). The number of medical visits for the flu was 21 million in the worst year and the number of PCR tests for COVID-19 so far is 22 million (about the same). It may be a bad assumption,

Just to sum up (and sorry for those on mobile):

    Tested/Visits   Confirmed   Hospitalised   Died
     21,000,000    *1,900,000    800,000      61,000
     22,000,000     2,000,000    270,000     114,000

Where the asterisk means my esitmate which may be completely wrong. Edit: The first line is the flu and the second line is COVID-19

But any way you slice it, I don't think the numbers work out the way you are portraying it. Corrections to the above are very much welcomed!

I'd love to actually read the paper to see if it is reasonable. Anybody have a link? Having said that (and keep in mind that I'm in favour of wearing masks and wear one myself when I go out in public), I think it's pretty clear (as indicated in the article) that masks alone won't bring the R value below 1.

The article says, "He [Richard Stutt, who co-led the study] said the findings showed that if widespread mask use were combined with social distancing and some lockdown measures, this could be 'an acceptable way of managing the pandemic and re-opening economic activity' long before the development and public availability of an effective vaccine against COVID-19, the respiratory illness caused by the coronavirus." (emphasis mine)

Anecdotally, in Japan (where I live) there seems to be a much higher than 50% use rate and there is still need for lockdowns. However, as the article suggests, it appears that the severity and length of the lockdowns might be reduced.

The question I have, though, is if the sheer number of people with the disease in the US and UK may be an issue. In South Korea, Taiwan and Japan, the total number of active cases never really got above 10K -- so the chances of meeting someone with the disease is really quite small. Potentially (as seems to be the case in Japan) you can get away with just dealing with cluster cases and let the stragglers go -- as long as the masks are effective enough to keep transmission rates low in those situations. But if you have millions of people with the disease, many of them in large cities, I wonder if it will be as effective.

TODO lists for which I use org mode, but you could use practically anything. I like a text editor for this rather than an app, per se, just because it keeps me in the flow. All you need is a place to jot down what you are planning to do next and to be able to arrange the order.

Usually I'll start with pretty high level ideas. If I have a story I'm working on, I'll put the description of the story in my TODO list. Then I'll think for about 5 minutes about what general things need to get done. I'll order these by some priority (doesn't really matter usually, to be honest). Then I'll start working on the first one.

Normally I need to poke into the code to really see what I have to do. I'll often add a sub-task to my first one that says, "Figure out what to do" or something like that. Then I'll do some exploratory coding for a few minutes. As I discover what needs to get done, I write it down in my TODO.

It's hard at first to stop yourself from just writing code, but pulling yourself back for the 20 seconds or so it takes to write down what you are just about to do can be surprisingly valuable. Don't censor yourself either. It's fine to guess what you need to do and then delete stuff that you realise is unnecessary later. As you are coding, any time you think, "Oh, I'm going to need X", add it to the TODO (again, difficult to train yourself to do it consistently!)

Once you get good at this, in my experience you will be quite interruptible. Any time I get distracted, or unfocussed or lack motivation, I just look at the top think on the TODO and say, "I'm just going to do that top thing". It always pulls me in.

I don't always code like this, but every time I do I'm dramatically more productive. I should always code like this, but... sometimes you want a relaxed day ;-)

It is incredibly difficult to remember something that you don't understand. In fact, your memory often elides things that you don't understand to the point where you will swear that something never happened when, in fact, it did. Statistically speaking, repetition is required for learning, but understanding is at least as important. In fact, sometimes if you can find no meaning for something, it helps to make up a meaning (for example the use of mnemonics).

On the other hand, true criticisms of Javascript, the language, are boring. Yeah, there are problems. You learn to deal with those problems. They aren't really that horrible. I worked with a lot of less convenient languages than JS (especially ES6, which is actually not bad). The standard library is pretty bad, but the language itself is easy enough to use that it's pretty trivial to implement what you need yourself. There are also a few pretty well written third party libraries with no other dependencies which you can use. The build environment is horrible, but not really any more horrible than some other environments I've had to deal with.

No, the real problem is that a lot of Javascript developers choose to stick hot pokers in their eyes. They don't read the code of their dependencies. They don't care how many ridiculous dependencies of dependencies they use. They refuse (absolutely refuse, to the point of calling you an imbecile if you even suggest it) to write their own tools. They choose the build tools that are the most wonky and are built on the most insane internal code -- because they don't care to ever look at that code. They look at the "box features" and say, "Oh, everyone is using that and it has all the features we want. You are crazy if you want that stupid boring thing that barely does anything (and yet works)". They don't do any planning for configuration management. They don't think about how they want to upgrade their dependencies, and especially don't dare think about inspecting the code in the dependencies. "Latest is best! If it breaks, we'll deal with it then".

Javascript is not really that bad. It really is that the community does not have a particularly good grasp on how to minimise risk in large projects. On the other hand, it's a common refrain on other platforms. While Javascript is not really that bad, other platforms are considerably better and you can get away with really poor practices for a lot longer. Not that they won't absolutely kick your ass eventually -- it's just going to be a couple of years away when you have moved to another company at a higher pay scale.

I also have a small caveat. If you have a docker setup or a similar style of setup, if it is not obvious how to do this by hand, then you should write documentation on how to do it. Honestly, my standard is that you should have the same level of setup instructions as you would expect from a good open source project of the same complexity.

Quite a lot of Japanese people seem to recognise characters less by their shape and more by the stroke connections, in my experience. No matter what the tool used to write it, legibility really requires writing the characters in the correct order for the most part.

Actually, Chinese characters did originally have a defined linear order (or, rather several schools existed IIRC). It's been a while since I looked at this, but essentially they were ordered by radical and then pronunciation, with some other tie breaker rules. Japanese word dictionaries, though, are indeed usually ordered by pronunciation. Kanji dictionaries have a variety of different orders, but most of them are ordered by radical and stroke count.

Not the OP, and it's definitely not my style, but some people really like taking classes. They love the environment and energy that a university life gives them. I've known one or two people who have spent decades just taking classes (and accumulating degrees).

To be fair, it is very common for Forth programmers to redefine the interpreter as they go. You literally change the language in your program. That's a very different expectation for other kinds of languages.

You are my hero!!!!

I had to read up on the "means of production" because I had no idea previously that it often refers to the capital (both social and financial) used to produce goods. It's interesting to me that Engels says (in essence) that a worker lacks the means of production and is therefore forced to sell their labour.

But, it occurs to me that this has always been a seller's market. I have, in my naive youth, thought, "What are they going to do, fire me?" Because they might, but it's not like I couldn't another job -- probably even higher paying! What exactly is the "means of production"? Is it really the capital that pays the salary of the programmer, or... is it the will of the programmer to actually write the code? By and large, the programmer is very difficult to replace. If you take note of the incredible number of programmers that are an absolute PITA (I mean, I'm sorry but we are and ornery bunch!), how on earth do they get jobs if the capitalists are not completely desperate. In fact, the big companies even lobby the government to allow them to ship programmers in in large quantities!

So can you not finish and deploy it? I wonder... What could possibly happen?

Small print: Not responsible for loss of livelihood, and social shame that results thereby.

Evil mode in Emacs is almost perfect. My biggest complaint is that undo works differently than vim (if anyone knows how to fix that, I would be forever grateful!) I originally used Emacs for about 20 years and then switched to Vim for a variety of reasons. I love the editing model in Vim. But for the last few years I've been using Emacs with Evil and it's been the best of both worlds.