It seems like that could apply to many other ways of messaging though, not all of them online. I'm not sure the "new" bit is really correct. Perhaps the potential scale or reach of a message is what matters.
Paul Graham has always been the poster child for the highly competent leader who has no idea what the people in his industry actually do or how things actually work. We all probably have experience with a boss (or perhaps your boss's boss) who runs the show with complete ignorance of what the people below them actually do.
He only ever got SSL working on his personal website in 2023. Got a valid cert installed just in time to avoid his site breaking entirely on chrome-based browsers (without the explicit http://).
And he thought that Twitter, a fairly barebones Ruby webapp at the time, was a protocol in and of itself and could be compared to HTTP and SMTP, or even TCP/IP!?
Twitter was just HTTP over TCP/IP. It was never a protocol. It was a website. Did this guy thing PHPBB was a protocol? Does he think Wikimedia is a protocol?
It's honestly astounding that the man has a Doctorate in Computer Science and co-developed the programming language that runs HN.
> He only ever got SSL working on his personal website in 2023.
I also added SSL to a site "too late" by hivemind standards. It's static HTML and contains nothing sensitive. I guess maybe a malicious ISP could theoretically inject ads or something.
I don't think he believed it was a protocol in a literal sense, but that people were using it like one. It had an open API at the time and both production and consumption of tweets was often automated. It didn't really work out that way longer-term, but it wasn't crazy to guess that it might.
(...plus came up with Bayesian spam filtering, plus wrote the book on Lisp macros, plus revolutionized startup investing).
It's only astounding because your assumptions are false. pg is nothing like a pointy-haired boss. What he is is a highly curious and lazy (in the good sense of the word) hacker who is bored by busywork. How you managed to arrive at the inverse image of that is such a feat of pathfinding that I'd be interested in the steps by which you got there.
Dang, I really respect your work here and in general you do a great job, but I think you overstepped here in auto-collapsing this comment thread and replying (edit: To clarify for anyone coming to this later, the GP comment was collapsed by moderator action at the time of writing. It is now flagged by user flags, which I think is entirely appropriate.)
PG gets a lot of flack on HN, some comments better-considered than others. Most of us are able to tell the difference and file the mindless attacks appropriately. We don't need you to rush to his defense, and in fact you doing so is likely counterproductive.
You've often said that you take a policy of moderating less, not more, when YC is involved. This interaction and moderation action breaks that pattern, which is harmful.
The problem here is the mixing of administrative powers (collapsing of subthread) and expression of private opinions.
One is ideally an unbiased, mechanical action subject to a rigid set of publicized conditions to which the comment(s) concerned are applied. So-and-so comment(s) are moderated so-and-so because they violate so-and-so in the guidelines, for example.
The other is (by your own admission) a biased, emotional, personal action subject and liable only to yourself.
The two are mutually incompatible when performed together.
I can't (and don't want to) do this job purely mechanically, and never have. I doubt it's possible, and if it is, I doubt it would make for good moderation.
Actually, though, collapsing the GP subthread was just that sort of application of the site guidelines. It's obvious (IMO) that the subthread is flamebaity and well offtopic. I reversed that decision as a courtesy to lolinder and a nod to the "moderate less when YC is involved" principle—even though it was the correct call from the unbiased/mechanical/rigid side of the ledger.
Let me put it this way then: You're mixing the professional with the personal.
Administering and moderating Hacker News is your job, that is correct. You also admitted that the rebuttal and moderation action this all stems from was driven by personal emotions (your liking Paul Graham). Your personal emotions have nothing to do with your professional job, the two are irrelevant to each other.
It's this mixing of professional and personal that is the problem. Not performing your job consistently will draw criticism, but mixing the two will cause even more fundamental criticism as was the case here.
Personally, I think the correct way of handling this would have been one of two ways: A) Engage in moderating the thread and refrain from acting personally. Or B) Engage in the thread personally and recuse yourself from the thread professionally, asking another moderator to do the work.
I don't believe the professional and the personal can be completely separated. People can't stop being human and what does "personal" mean, at bottom, but that?
It's true that we shouldn't act on each other purely out of our own emotion but that's true personally too, not just professionally.
If you try to exclude emotion from human activity, including internet moderation, it ends up running the show anyways, just more crudely and unconsciously. Better to consciously give it a place—hopefully an appropriate place.
Questions like this have come up over the years and my sense (you may disagree of course) is that the community is happier with moderators who show feeling sometimes and can be related to personally. I could be wrong about that, but if so, it should have caused large problems long before now.
I dunno, I've been personally corrected by you and I prefer that you in turn can be corrected and can show human opinion like anyone else. So I would say you're right about that, and I'm more likely to be comfortable being corrected in future as needs must.
Because it's a terrible blog post. If you applied this criticism to any other author, it would be valid.
But because it's pg it's different? No, it's still a bad post. There are a plethora of other reasons Twitter was a big deal. It being a "protocol" wasn't one of them.
I don't think the post has held up, but the "it's not a protocol, it's just HTTP on top of TCP/IP" is a lame argument. It's clearly a protocol. I've been doing protocol engineering work since the mid-1990s, and people have been saying things built on top of HTTP aren't "protocols" since HTTP went mainstream. I was one of them, in the 1990s! That was dumb of me; most new important protocols since then have been built on top of HTTP, and I expect that to continue.
The subtext of these "it's not even a protocol" arguments are that Paul Graham doesn't know what a protocol is, which is not a plausible argument. Why make it?
We get it though; gotta white knight for your meal ticket
But your prior lived experience isn’t exactly useful to the rest of us.
To the outside observer you’re Robin defending Batman, peddling anecdotes about someone you’d actually feel something for if they died. To everyone else he could have died in the ditch a decade ago and we’d never have noticed.
You know all about neuroscience but fail to spot why you’d be biased. Same old self selecting biology like everyone else.
It's true that when I'm fond of somebody, I tend to respond to false attacks on them. Not because of "meal tickets" but just human feeling.
It's true that things look different on the outside, though one might add that people who routinely jump to cynical conclusions about others don't make very good Hacker News commenters.
But what made you think I know anything about neuroscience?
> We get it though; gotta white knight for your meal ticket
This is a rather presumptive conclusion. Unless, of course, you have specific knowledge corroborated by others such that this assertion is more than a trite ad hominem.
Full disclosure: I am not associated with anyone who owns, administers, or in any way runs this site.
It's a platform - a marketplace for buying opinion.