Paralleling what dang's said here, I've been looking at 17 years of HN front page activity over the month or so, and am starting to tackle the question of topic drift and/or focus over that period.
I've been sort of live-blogging the experience on the Fediverse: https://toot.cat/@dredmorbius/tagged/HackerNewsAnalytics>, as well as in some of my HN posts.
My current tack involves looking at sites (as reported in parentheses at the end of each HN front-page post title) and classifying those. With slightly more than 30% of sites categorised, I can classify about 65% of all HN posts.
For the full dataset (17 years), that's roughly:
1 63913 35.73% UNCLASSIFIED
2 22589 12.63% blog
3 15112 8.45% general news
4 13823 7.73% tech news
5 12851 7.18% programming
6 8622 4.82% corporate comm.
7 8459 4.73% academic / science
8 7294 4.08% n/a
9 5324 2.98% business news
10 3803 2.13% general interest
11 2151 1.20% social media
12 2074 1.16% software
13 1613 0.90% technology
14 1463 0.82% video
15 1144 0.64% general info (wiki)
16 1009 0.56% government
17 724 0.40% misc documents
18 720 0.40% law
19 702 0.39% tech discussion
20 620 0.35% science news
Tons of caveats: this depends heavily on how I classify individual sites, a given site's
stories might well be technical, social, or political, etc., etc.
The breakdown-by-year analysis is in development, but if anything programming-specific content as increased in prevalence. Political discussion seems not to have (though it rose significantly ~2014). Cryptocurrency and blockchain-specific sites also peaked about that time (I suspect much of that discussion is now mainstream). General news has always been a huge portion of HN discussion, as have individual (and corporate) blogs.
Note again that this isn't about discussion and comments, or even the titles or article contents (I'm thinking of looking at those, it's ... a challenge for me).
But across nearly 200,000 front-page stories, on which nearly half of all HN discussion occurs (based on another API-based study looking at comprehensive posts), the overall trending seems at first blush to be pretty consistent and if anything improving over time.
(As with all preliminary results, I'm hoping I won't have to eat my words here. Though I'm reasonably confident in most of this.)
From the classifications above, the places you might find some that "suffering" would be in general news, genral interest, and social media categories. All but the first of those are single-digit percentages, and a lot of that general-news content is about technology, business, finance, and science, all of which would crowd out the sort of social and political issues which seem to generate strong feelings.
The "UNCLASSIFIED" sites are a wide mix, though most are probably a mix of blogs, corporate / organisational communications, and the like. The mean posts per site is 1.739951, so gains from additional site-categorisation are pretty slim. I have captured a lot of obvious patterns via regexes and string matches, so academic/science and major (or even minor) blogging and social media sites aren't a large fraction.
More recent discussion here: https://news.ycombinator.com/item?id=36524001>