There still plenty of "this version of pytorch doesn't run reliably with Python 3.12, please use 3.10", though. It's not all sunshine and roses.
Jesus_666
Oh, I am aware that this dynamic doesn't quite work. Both sides of the equation get it almost, but not entirely, right. Because getting it right is challenging.
The older generations do get lazy as soon as their own needs are met, especially those of them who wield power. And that doesn't just mean politicians but also CEOs, large investors, etc. Working for the common good is difficult and most people follow the path of least resistance, leading to the dynamics young people complain about.
But young people fall prey to the very same dynamic. It's easier to get disillusioned and complain on the internet than to consistently go out and exert pressure. Sure, maybe they go on a protest march or two but few have the energy to consistently go on marches, be active in politics, stay on top of which companies are toxic, openly defy the law, and do a myriad other things that don't directly benefit them in order to fight for a better world. You can do that if you're rich and isolated from consequence but in that case your life is cushy already so you'd be fighting for abstract principles.
There are few people of the caliber of Bernie Sanders or Greta Thunberg because it's really hard and carries a real risk of ruining their own life. Most people who do try end up like the people who glue themselves to highways: They get ridiculed and fined and effect no change whatsoever.
So for most people all that remains is resignation. Gen X said "whatever" when they weren't listened to because they weren't willing to sacrifice their personal future to escalate things until change is inevitable. Millennials say "OK Boomer" as they find themselves in the same situation for the same reason.
(I'll gloss over the "young people are lazy" part but it boils down to young people actually being lazy in terms of failing to apply themselves in ways the older people did, with the caveat that some of those ways no longer apply and young people are applying themselves in new ways the older people don't realize are necessary now. This perception dynamic is at least as old as recorded history.)
It's exact opposite of a trend. The younger generations criticizing the older ones for not living up to their responsibilities has been just as much a constant throughout human history as the older generations criticizing the younger ones for being lazy by their parents' definition.
The world sucks, as usual, and by now Gen X are old enough to be in charge. It's sensible to call them out, just like in a decade or two it'll be sensible to call out the Millennials for the same thing.
systemctl disable activationcheck.service
GetRight was the bomb.
Then I got fast internet (we're talking 25 MBit DSL here) and servers stopped being so stingy and download managers slowly became a thing of the past.
And then, a year or so ago, I had to work on a company VM that would randomly reset the connection to GitHub, from which I needed a rather large file. So I wrote a super cheap download manager and named it GetWrong in honor of the hero of my ISDN days.
"They're going to send us where?"
Other lessons from the sign:
- You may send radio broadcasts to origami birds.
- We encourage you to dip your hands into any glasses of sparkling water.
- Do not swing progress bars around.
- Violence against butterfly skewers is forbidden.
I dunno. The intense dead-eyed stare works nicely to signal that the guidance counselor has wildly overstepped the boundaries of what Camila will put up with and that the time for pretend niceness is over. Give her a little psychotic touch, too, which IMO is more menacing than a "standard" angry face.
See, I don't have to worry about such details. I work in corporate software dev, which means that everything is an MSSQL database where most of the tables contain only an ID of a table-specific format and a JSON blob. Why use an ORM when you can badly reimplement NoSQL in a relational database instead?
In a similar vein, Germany has a neat labeling system¹ for the conditions under which animals (for meat, dairy, etc.) are kept. There are five levels, each of which has specific minimum criteria per type of animal. Basically, 1 and 2 are shit-tier, 3 is semi-decent, 4 is vaguely free-range, and 5 is "organic" (as vaguely defined as that term is).
That makes it easier to avoid buying from animal torture dungeons, plus it stands to reason that products from animals kept on better conditions have a better chance of being of good quality.
The labels are voluntary. However, you can find them on a good number of products, especially since a label with one of the higher levels has marketing value. I know I definitely prefer products that are at least level 4.
Notably, there are efforts to pressure supermarkets into abandoning level 1 and 2 products altogether, with Aldi having promised to do so for most products by 2030 and other chains giving weasely but vaguely affirmative statements.
¹ Yes, the website doesn't seem to be fully translated. But at least the level definitions are in English.
And the overuse of sugar is because the sugar can mask cheaper ingredients and lower amounts of spices.
Why sell an instant curry full of expensive spices if you can cut half of them out and just replace them with sugar and salt? Why use decent meat if you can just use cheap shitty meat and add sugar to hide the fact that it's flavorless? Why use real cream in the sauce if you can add some skim milk powder, palm oil, a thickening agent, and yet more sugar at half the price?
Or food is getting enshittified and it's having a real impact on our health. But since public health doesn't factor into food companies' bottom line that's not just tolerated but desired.
Except if they then have to run it on their machine and the setup instructions start with setting up a venv. I find that a lot of Python software in the ML realm makes no effort to isolate the end user from the complexities of the platform. At best you get a setup script that may or may not create a working venv without manual intervention, usually the latter. It might be more of a Torch issue than a Python one but it still means spending a lot of time messing with the Python environment to get things running.
This may color my perception but the parts of the Python ecosystem I get exposed to as an end user these days feel very hacky. (Not all of it is, though; I remember from my Gentoo days that Portage was rock solid.)