And, predictably, the Puppy Kickers get all riled up, because Wrongfen are making (mostly, and ignoring the Rabid faction) perfectly legitimate nominations for the wrong reasons. Or something.
I noticed a couple of weeks ago that Raymond Chandler had addressed this issue a while back:
Fiction in any form has always intended to be realistic. Old-fashioned novels which now seem stilted and artificial to the point of burlesque did not appear that way to the people who first read them. Writers like Fielding and Smollett could seem realistic in the modern sense because they dealt largely with uninhibited characters, many of whom were about two jumps ahead of the police, but Jane Austen’s chronicles of highly inhibited people against a background of rural gentility seem real enough psychologically. There is plenty of that kind of social and emotional hypocrisy around today. Add to it a liberal dose of intellectual pretentiousness and you get the tone of the book page in your daily paper and the earnest and fatuous atmosphere breathed by discussion groups in little clubs. These are the people who make bestsellers, which are promotional jobs based on a sort of indirect snob-appeal, carefully escorted by the trained seals of the critical fraternity, and lovingly tended and watered by certain much too powerful pressure groups whose business is selling books, although they would like you to think they are fostering culture. Just get a little behind in your payments and you will find out how idealistic they are.
Yeah, he was writing about detective yarns as opposed to Great Literature, but it applies just as well to science fiction stories as opposed to blobs of text with science-fictiony trappings and pretensions of being Socially Important Literature. Snobs gotta snob.
This, now, is interesting as far as it goes, but leaves out some things, including some that make the county hospital an even less appealing prospect.
Fer example: rumor has it (this is at third-hand, at least, and I don't have further details, so don't ask) that a patient was recently admitted with a nasty wound... and waited two days for the one-and-only wound-care specialist to come around and clean it properly. By which time septicemia had set in.
More definitely and less rumorally: the commissary is simply not configured for patients on restricted diets. If you're stuck there for a while, and have to subtract a few common things (e.g., dairy and wheat), good luck getting enough food to stay alive on (never mind heal) unless a friend or family member brings in a daily care package.*
And, patient records. Let's say, just for the sake of argument, that you're receiving chemotherapy at the specialty clinic that's part of the hospital complex, and that this is known to have clobbered your immune system. You show up at the emergency department whimpering in agony with a rapidly-developing infection and unable to think or communicate clearly. Does your medical record pop up with a great big warning at the top saying "PATIENT HAS A SUPPRESSED IMMUNE SYSTEM AND ANY INFECTION SHOULD BE TREATED AS AN EMERGENCY"? Nope. The patient is responsible for conveying that information. If you get past the screening and see a doctor, maybe the doctor will think to ask if you're on chemo. Because this is important... and it's not visible to the ER doc. Also, if you have known chronic conditions, you might hope that the prescription system would check any proposed drugs against contraindications based on your medical history... but nope. Again, the patient is responsible for knowing this stuff. (And don't count on being able to look up the drugs on Wikipedia; there's pretty much no cell coverage, and the public WiFi is pay-per-day and insecure, so I hope you have a burner credit card with you.)
(I happened to be chatting with the receptionist at the lab at the local branch when she returned from lunch and was logging back in. She had to enter a username and password in, if memory serves, eight different places. Record systems not integrated, much? And if a doctor in one department of the clinic wants to make an appointment for you with another department, it's done by phone. Just as if you were calling in yourself from outside. Real helpful, that.)
Huh. The scoring system doesn't even factor in wait times at the emergency department. Another important consideration.
Then there are all the problems you get with a government-run union shop: the workers are seriously a mixed bag. Most are good, but... there are some bad ones, and there's basically no way of getting rid of the bad ones. Keep that up long enough, and the good ones will get discouraged and head elsewhere.
Meanwhile, the nearest Kaiser facility gets a much better rating. Which is good news for me, in the event of my needing hospital time. But, really, I try to stay out of hospitals as much as possible, and generally succeed at it. Except for visiting, delivering care packages, that sort of thing.
* Also, the food itself is appalling. I had a suspicion that the county hospital shared a commissary with the county jail, but then heard (from someone who'd been an inmate of both institutions) that the jail food is even worse. If that's true, I'm sure it must violate some international convention.
So I'll be needing to power this gadget... and the design of the various subsystems is now far enough along that I can come up with a reasonably close power budget.
Looks like 5V at 1A, and 3.3V at a little over 1A.
Obvious approach is to use two stages of buck regulator: input voltage (needs to handle a range of 10V to 28V, and preferably quite a bit wider) to 5V, and then an easy-peasy 5V to 3.3V.
So! Look for a decently high-frequency buck regulator that'll handle at least 32V input and 2.5A output. Find several options. The first, as sorted by cheapest-first, has nice specs, including a very low ON resistance for the switch (ergo, high efficiency; I like high efficiency).
But! It has an enable input. The enable input must not be left open. To turn the regulator on, the enable input must be driven to a source of 1.2V to 5V. Absolute maximum rating for the EN pin is 6V.
Um. I don't have the low-voltage supply yet. That's what this regulator is for.
And so it comes to pass that I set aside that datasheet, and look at one for a part that costs a little more, and has a higher ON resistance, but whose enable input can be strapped to the input supply.
(Or... let's see, if I use a resistor divider, with 10V input giving 1.2V at EN, then for 5V at EN I'd have about 42V at the input. So I guess I could do that. It's just kinda annoying. Potential problem: behavior when EN is between 0.4V and 1.2V is unspecified, and for all I know it could be erratic and power-dissipating. Guys! Either allow strapping EN to VIN, or make it default to ON if left open.)
Hm. Since this is a prototyping operation, maybe I'll design in two or three 5V regulators, and for production use the one that runs coolest and/or has the best projected availability.
Latest third-party medical adventure is at least out of the hospital phase. I still need to catch up on my sleep, which requires cutting back on caffeine, which requires a few lazy days, which... well, there's still a bunch of stuff to catch up on.
Spent some productive time this weekend on a "simple" project: a 13-port, USB 2.0 hub (high-speed). Yeah, that sort of thing is never as simple as it ought to be, and there are impedance-controlled diff pairs that needed hand-routing (in addition to the usual power connections and such that always need hand-routing), and I had to protect a couple of fill areas against autorouter intrusions. It's nominally part of a project for a client (the hub will be part of a board with a preposterous pile of peripherals on it), but I foresee a need for a 7-port version in a particular form factor, so I'm treating this subproject as a learning exercise, not as billable work. (There's plenty of other work in the overall project, and that's all legitimately billable.) I think I have the whole thing laid out on a modest-sized 4-layer board with lots of test points (but no LEDs; I have a feeling I should add a few); critical hand-routes are done, and the autorouter can cope with what remains. This particular item is a testbed, hence the test points, but I'm not bothering with niceties like ESD protection.
I still need to review my design and the chip vendor's app notes.
Maybe by the middle of the week things will be back to normal, or what passes for normal around here. Assuming, of course, that nothing else gets Interesting.
I mentioned that week-overdue toaster project? And the Verilog mines?
There was an Issue with some of the logic. Specifically, it was failing the test case for limiting the duty cycle of the heating elements. The test involved wiggling the handle up and down in a specific pattern (lasting 9.8µs; it's a high-speed toaster) and verifying that no error was flagged for timing A (this was OK) and an error was flagged for timing B (it wasn't doing this).
So I spent such work time as I had Sunday through most of Tuesday trying to track this down. According to my calculations, timing B should indeed have tripped the duty-cycle limit. There I was, staring at half a dozen lines of simple Verilog, trying to figure out what could possibly be going wrong.
Then, late Tuesday, I decided to observe in detail what was going on. Having no ability to extract the contents of FPGA registers in real time (there being only a few probe pins available), I planned out a series of tests to determine what was actually happening (by that time, I'd determined that the logic was correctly accounting for the heating phase of the cycle, but the cooling phase was cooling much too fast).
So I constructed a spreadsheet, and poked in some test results for patterns that just barely tripped the limit. And guess what?
It came out exactly according to spec. (Within a reasonable window between the customer requirement and the damage threshold, anyway.)
Then I repeated my calculations from Monday. I got a different result: it matched the spec.
Conclusion: the logic was right all along; the test was wrong. And, somehow, two iterations of my calculations supported the test, until I gathered more data. After that, calculations based on the original information showed the test to be wrong.
This, kids, is why you should always show your work. When the result suddenly changes, you'll be able to go back and see what was wonky. Alas, I didn't write down what I was poking into the calculator on Monday, so all I have is the wrong results.
e-Filement is did. Checks still gotta be written and mailed. Four of them. One of them big.
Not sure how much of the current crud is stress, and how much might be something I picked up at the hospital. Symptoms are consistent with either. (Urgh. Slight fever? I'm leaning toward "cruddy little virus, aggravated by stress.")
Better have a nap now. Stuff I gotta do a bit later. And there's still that week-overdue toaster subproject.
“We live in a society in which it is widely believed that there is a just God, and that he rewards those who work hard and are good people, i.e. the job creators … and that he punishes those who are not good people, i.e. who are poor or slothful,” said Kivel. “The invisible hand is the hand of God.”
What. Is. This. I. Can't. Even.
Apparently the perpetrator of this remark is a victim of deinstitutionalization, or maybe his mother was smoking something really amazing. Or does he live in Looking-Glass Land?
Look, free-market economics is basically atheistic. Lack of intervention, divine or otherwise, is a fundamental premise.
Absent intervention or theft, those who create wealth end up with more wealth. You make stuff, you have the stuff. Is that really so hard to understand?
Dunno. Think our minds must be too highly trained, Majikthise.