I tend to use "catch and re-raise with context" in Python so that unexpected errors can be wrapped with a context message for debugging and for users, then passed to higher levels to generate a stack trace with context.
For situations where an unexpected error is retried, eg, accessing some network service, unexpected errors have a compressed stack trace string included with the context error message. The compressed stack trace has the program commit id, Python source file names (not pathnames) and line numbers strung together, and a context error message, like:
[#3271 a 25 b 75 c 14] Error accessing server xyz; http status 525
Then the user gets an idea of what went wrong, doesn't get overwhelmed with a lot of irrelevant (to them) debugging info, and if the error is reported, it's easy to tell what version of the program is running and exactly where and usually why the error occurred.
One of the big reasons I haven't switched from Python to Go for HashBackup (I'm the author) is that while I'd love to have a code speed-up, I can't stomach the work involved to add 'if err return err("blah")' after most lines of existing code. It would hugely (IMO) bloat the existing codebase.
> I worry that the cost of exterminating people algorithmically could become so low that they could decide to start taking out small fries in batches.
My guess is that the cost of taking out a small fry today is already extremely low, and a desperate low-life could be hired for less than $1000 to kill a random person that doesn't have a security detail.
These costs would depend on the nature of the target, the nature of the country you live in and the requirements of the murder.
High profile, protected target? You probably couldn't find a random low-life to do it, much less successfully. And no matter what jurisdiction you want to commit the murder in, it will be more expensive than if your target was a random average joe, or jane.
Country is a place where the rule of law and legal enforcement are strongly applied and taken seriously? It will become harder and more expensive. Criminals are often stupid, but even stupid criminals in countries that take legal matters seriously are rarely freewheeling about contract murder that they actually mean to commit. The pool of willing potential killers would be smaller in such countries.
And finally, the nature of the murder: Need to kill someone in a way that looks like suicide or accident? That won't be something you hire a low-life to do on the cheap.
On the other hand, if you just need someone with modest to poor protection dead and you live in a country with weak legal mechanisms, then the situation becomes as favorable as you could want given your murderous needs. Assuming you have the right connections, a random gangbanger or would-be gangbanger on a motorbike can do the job for very cheap indeed. In the country I live in this is common and the people (often just teenagers) paid to do it will go for broke if offered as little as a couple grand or sometimes much less.
You're leaving out the cost of getting caught with risk factored in.
Also, if targeting small individuals, it's rarely one individual that's the issue, but a whole group. When Stalin or Hitler started systematically exterminating millions of people, it was essentially done algorithmically. The costs became very low for them to target whole groups of people.
I suspect that once you have the power of life or death over individuals, you automatically hold such power over large groups. Because you need a corrupt structure and once the structure is corrupt to that extent there is no clear line between 1 person and 1 million persons.
Also I suspect only one or a handful of individuals can have such power because otherwise such crimes can be used as a bait and trap by political opponents. Without absolute power, the risk of getting caught and prosecuted always exists.
The ACA lets people get healthcare insurance who are not otherwise eligible for it, for example: part-time workers, independent contractors, employees of companies with fewer than 50 employees, temporarily unemployeed (fired, laid off, switching careers or jobs), people who have lost coverage because their spouse lost coverage. That's a pretty broad swath of people!
Some people qualify for a tax subsidy that can be anywhere from $0 to the entire cost of a plan, depending on their income. A unique feature is that the subsidy is based on your expected income for the upcoming year, but if you make less than that (are laid off for example) or more (independent contract gets an unexpected contract), the subsidy is adjusted when you file your taxes.
Currently the ACA does not accept anyone who has a policy through work. IMO, every should have the option of getting ACA healthcare coverage. If their work coverage is better or cheaper, they can stick with that, but if their work coverage is worse or more expensive, employees should be allowed to get ACA coverage, with the employer paying part or all of the subsidy (what they would have paid to a private insurance company for the employee) instead of just the government.
I had a situation with HashBackup (I'm the author) where I needed a map-like data structure that could get rather large (hundreds of MB to a few GB) with big backups. In Python, storing this data in a dict was not really feasible because every unique integer takes 24 bytes, and since the data structure stores things like blockid and file offsets, there are a lot of unique integers.
Instead, I created an in-memory SQLite database for the data structure and it was something like 1/3 the size of the Python dict. It was around 10-15% slower, but for the memory savings it was an okay trade-off.
> Anyway, I'm staying with my old gas Honda until it dies which is probably never with proper maintenance and eventually restoration.
I would have stuck with my 2003 Honda Accord too, except that some woman, probably talking or texting on her cell, slammed into me while I was stopped at a light, totaling my car and damaging 3 others. I got $8K for my car after arguing with the insurance company, and paid $28K for a 3 y/o replacement.
The fine for texting while driving in Kentucky is $25.
Everything eventually ends up as paying + ads. Cable TV started out justifying the $35/mo fee 2 ways: better picture quality and no ads (only local channels had ads). Now you pay for channels, bundles, and you get ads. The Internet is and will be exactly the same.
IMO, checksums at the filesystem and higher level are only useful for maybe detecting software problems, not cosmic ray RAM bitflips. ECC RAM is the only way to reliably detect random RAM bit flips.
Without ECC, if you read a block of data from disk into a RAM page, do whatever checksums you want - filesystem page, database page, whatever - and THEN a bit flip occurs, you are now processing bad data but think it's okay because all of the checksums passed.
Every disk block has an ECC code recorded with it to detect and correct errors. Hardware busses have checksums to detect transmission errors. But without ECC RAM, there is no protection for RAM bit flips, and higher-level checksums won't help with that.
It's ridiculous that all computers don't come with ECC RAM, and it's mainly that way because Intel wants to charge a premium for data center servers.
> So people who don't conceptualize themselves as athletic, they may have insecurities if not outright skepticism, aren't going to last
I have always been thin and tried to start workouts on my own several times over many years, and never could do it, mostly because I didn't know what I was doing. Hiring a personal trainer, if you can afford it, is a great way to get over this hump. I quit after a couple of years and workout on my own now, but couldn't have done it without the trainer.
See the movie "The Corporation", which compares corporations to psychopaths; they're very similar in that corporations don't have a conscience.
Corporations, IMO, should not be involved in politics at all - only real people. Rich people would still have more influence over political decisions, but not to the extent that rich corporations do now.
For situations where an unexpected error is retried, eg, accessing some network service, unexpected errors have a compressed stack trace string included with the context error message. The compressed stack trace has the program commit id, Python source file names (not pathnames) and line numbers strung together, and a context error message, like:
[#3271 a 25 b 75 c 14] Error accessing server xyz; http status 525
Then the user gets an idea of what went wrong, doesn't get overwhelmed with a lot of irrelevant (to them) debugging info, and if the error is reported, it's easy to tell what version of the program is running and exactly where and usually why the error occurred.
One of the big reasons I haven't switched from Python to Go for HashBackup (I'm the author) is that while I'd love to have a code speed-up, I can't stomach the work involved to add 'if err return err("blah")' after most lines of existing code. It would hugely (IMO) bloat the existing codebase.