> Somebody even created a website to facilitate conversion but unfortunately the TLS certificate has expired and Cloudflare now blocks access.
This is one of the main arguments I was using in discussions with people advocating unconditional use of HTTPS everywhere. Yes, in theory it's a good thing. Yes, in theory it should be a solved problem and you wouldn't see any broken websites anymore. In practice, we lost a small part of the Web.
Yesterday I considered writing a web scraper completely from scratch (just sockets). Without HTTPS, this is trivial. Of course, you lose out on much (most?) of the web, but I have a feeling most small / interesting sites would still be accessible.
I have found that, given a random sampling of web content, an extremely small fraction of it is interesting or useful to me (nor indeed is hardly any of it what I would consider high quality enough to use as the basis for the future governors of mankind!)
Even if you moved the entire TLS web to non-TLS, this is no longer trivial. The web requires Javascript to render, full stop. Fetching and parsing HTML alone is totally insufficient.
> The web requires Javascript to render, full stop.
A small correction: some parts of the new web require JavaScript to render.
That's why on many websites teh experience is better without JS. To be more specific, several paywalled websites can be accessed just by turning the JS off. You could even say the opposite is true in these cases: JS is being used to prevent text rendering.
A small part of the web that was either archived already, or wasn't interesting enough to be archived in the first place.
I get your sentiment but at some point you have to let go. Many websites die every day not because of obsolescence but because the author eventually stops renewing the domain or paying their hosting provider.
The worst are the money units! Instead of writing "40 millions dollars", they often omit the number, like "millions of dollars".
This means they only use three values: millions, billions, and thousands.
My best guess for why is that it's a way to not be wrong. If you print "40 millions", and it turns out to be 39, you've lied, which is considered far more bad than being vague.
In a local publication I follow they always round up or down to make the article easier to read. Sometimes they'll prefix the number with an "about" or "roughly" or "nearly" or whatever.
The actual number (if it's available as a fact) will be printed in the article somewhere, but headings, pull quotes and other call-outs will have some rounded number.
For example, recent article's first paragraph:
"Justice Minister Thembi Simelane took a loan of more than half a million rand from a company that brokered unlawful investments into VBS Mutual Bank by the Polokwane Municipality while she was mayor of the city in 2016. Pauli van Wyk explains what happened."
Further down in the article the "half a million rand" is revealed to be R575,600
> My best guess for why is that it's a way to not be wrong.
It's also often used to make things seem better or worse than they actually are. "Thousands of dollars" sounds like it's far more than for example $2,108.
Only a fabrication if it can't be sourced; otherwise, a source was wrong and you run a correction. When you don't have a number you're willing to point to even that far, that's when you leave it out entirely.
Amazing, and so is the associated article: "In a recent piece on red-giant star Mira, we rather foolishly suggested that the "comet-tailed" body was travelling across the heavens at roughly 150,000 times the speed of the average sheep."
Somebody even created a website to facilitate conversion but unfortunately the TLS certificate has expired and Cloudflare now blocks access.
Article in Danish: https://ing.dk/artikel/lynch-nu-kan-ogsaa-journalister-faa-s...