From the article, I know that the new headline is more precise. But it's much less informative, I wouldn't have clicked the article under this name.
As for whether the headline says TikTok or ByteDance, the only thing I care about is what readers will complain about least. If we make it say ByteDance, that's correct, but people will accuse us of trying to bury the lede because no one has heard of ByteDance. (Why on earth would we do that? Obvious answer: we must be controlled by the Chinese state, or some other sinister interest. How people come up with this stuff...)
The problem is that if it says TikTok, people will complain about the title being incorrect. I've flipped it the other way for now; let's see what's worse.
I don't envy your work. :)
FWIW, I didn't suspect any foul play... the ByteDance is more accurate but also totally opaque.
> But some of those are edits made by submitters, [...]
Honestly never realized that's a thing, thanks for the tip!
M5 Hosting is also known as M5 Computer Security out of San Diego.
These firms are a strange choice for a place like YCombinator to host I think.
It's a small hosting company with good support. It's a nice fit with the spirit of this community. I have no idea what alternate names they have.
Can I answer any questions you have about us?
In this case, the article title is in my judgment misleading—not hugely so, but enough to be worth changing. We changed it to language from the first paragraph, which is where articles typically say what they're really about. If that's obfuscating, then the article obfuscates itself.
HN sort of reminds me of China -- a dictatorship with zero transparency and a penchant for manipulation.
Edit: Can't forget the spineless bootlickers who are hopelessly devoted to the state! You know you've struck a nerve when the best 'reply' they can offer is a silent downvote.
We have spent years calibrating how we handle titles and it's one of the most consistent things we do. I know it can be nonintuitive to casual readers at times, but that's because applying that guideline is surprisingly complicated in practice. If it were your job, you'd soon find that as well. Also, people only notice the cases that stand out, which tend to be the edits they dislike and feel we got wrong (which maybe we did). That's a sample bias. Probably less than 5% of title edits even get noticed. Maybe even less than 1%.
We're always happy to answer questions about it, as well as to change titles in response to user feedback, which we do regularly: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que....
It's nonobjective in that the title guidelines carry subjective judgements.
Its' nonconsistent because of the previous problem.
It's unintuitive for the reasons you described above.
I think the reason it sticks in my craw so much is what I'll call the "rake to the face" phenomenon. Much like stepping on a rake in a dark shed and getting whacked in the face, it's usually surprising, usually unexpected, and usually painful (though mentally rather than physically.. "where did that article I read earlier go? No way it got flagged off.. Oh, there it is, they messed with the freaking title again. That was a waste of 5 minutes.")
This is worse when a descriptive title that actually calls out why someone should be interested in the article is reverted to something generic, which IMO, actively makes the site worse in furtherance of The Rules, which is never a good thing.
----
That aside.. If I could float a feature request that doesn't require any rule changes? Put a [*] or some other signifier next to titles that have been modified, and have it grey out over time (kinda like how comments get lighter and lighter as their score falls) and eventually disappear. Per your own search example, title changes are rare enough that there likely wouldn't be many on the front page, and it would give people like me who lose track of renamed articles a place or two to look.
Outrage culture signaling with respect to language is always so fascinating.
I hope HN readers have gotten smart enough to notice how when commenters make grandiose claims like this about manipulation, they never provide links. That's because the facts never support the grandiose claim.
Curious to see what the real reason might have been, I skimmed through the last 30 or so titles with Blizzard in them and didn't find one we'd edited in this way, or even at all. Perhaps I missed it. But whatever we did with any such title, it would have been because of the site guideline: "Please use the original title, unless it is misleading or linkbait; don't editorialize."