2026/08/15

06:48 On Hacking

I recently wrote:

It's a bit unfortunate how Hollywood hijacked the word "hacker", the way labeling Richard Stallman or Ymir Vigfússon to a non-technical person as "hackers" and they start thinking of them hitting a computer keyboard at 120fps doing cybercriminal activity.

I was exploring Wikipedia's list of hackers. It's full of computer scientists, hackers, but also cybercriminals. Richard Stallman has a great article on what's hacking: On Hacking.

#Programming

06:52 Thoughts on opening comments.

HackerNews, Lobsters, and many other online forums dislike leaving comments open after a certain period (usually a month). I believe that leaving the section open for new comments indefinitely is extremely useful. Following is an exchange about it.

On Fri, Jun 26, 2026 at 2:17 PM Saleh  wrote:
Hi,
I was browsing some older threads, I noticed that some of them had received new comments a year or more after the original post (mostly on posts that were so old I’d say). The current system no longer allow that, once a post or comment becomes old enough, the discussion is locked.
I think there is still value in allowing late replies. Someone researching a niche topic might come across an old comment asking a question that was not answered or discussed and happen to know some useful relevant information that might  enrich the discussion, discover that the problem was fixed years later, I’m sure that many of technical people can relate to this experience. Being able to add that information would make the archive more useful imo. Has this ever been discussed?
(Another use case, sometimes I want to add a response to a post or a comment, and I would like to schedule it, instead of freely schedule doing that -since it will be a lengthy reply- knowing that HN does not allow late replies, I put it as a “deadline” instead, which is not very convenient)
—
Regards,
Saleh

From: Hacker News 
Subject: Re: Commenting on older posts
Date: 27 June 2026 at 12:32:05 AM GMT+3
To: Saleh 
>  noticed that some of them had received new comments a year or more after the original post (
That shouldn't normally be possible. Do you recall which cases?
One exception is when people email and ask for permission to respond to something; we sometimes make exceptions in special cases. But the basic rule is threads are closed after 14 days.  That has served HN so well that I doubt we'd change it - there are two many ways for it to distort discussions after the fact. (For example the infamous Dropbox thread would have been swamped by anachronistic comments by now.)
In some of the cases you mention, we'd probably be willing to temporarily re-open a thread so someone could post an update and whatnot. But we'd only do it case-by-case with review.
Daniel (dang)

From: Saleh 
Subject: Re: Commenting on older posts
Date: 27 June 2026 at 1:33:06 AM GMT+3
To: Hacker News 
dang,
That shouldn't normally be possible. Do you recall which cases?
Mostly very old post as mentioned, I expect that the rule was not enforced then.
- https://news.ycombinator.com/item?id=87019
- https://news.ycombinator.com/item?id=611382
- https://news.ycombinator.com/item?id=557406
- https://news.ycombinator.com/item?id=1
That has served HN so well that I doubt we'd change it - there are two many ways for it to distort discussions after the fact. (For example the infamous Dropbox thread would have been swamped by anachronistic comments by now.)
Newer comments can be flagged as newer, possible to toggle and hide all of them or so. I think having that harm should not the reason why the community can not benefit from a continuous discussion. I can think of other ideas like “subsequent discussion” section. The historical thread would remain intact.
That has served HN so well that I doubt we'd change it - there are two many ways for it to distort discussions after the fact. (For example the infamous Dropbox thread would have been swamped by anachronistic comments by now.)
HN’s success does not establish that every rule it currently has contributed to that success, much less that none of those rules can be improved. A successful institution can and should be able to preserve unnecessary restrictions because their costs are less visible than their benefits.
I would like to elaborate more on how I imagine that, a discussion can be understood either as an event or as a body of knowledge. with an event, freezing does make sense, but if it is also a body of knowledge, then preventing correction and further discovery makes the archive less useful than it could be and anachronistic comments should not be allowed to masquerade as contemporary ones.
I would love to know if this was further discussed internally, I spent some time looking for discussions on that but could not find any, so I thought this decision might have had been affected by a stand-by effect of internal people seeing it as a rule and accepting it assuming someone thoughtfully put it.
—
Regards
Saleh

From: Hacker News 
Subject: Re: Commenting on older posts
Date: 27 June 2026 at 4:56:01 AM GMT+3
To: Saleh 
Ah yes - those posts are so old that they probably predate this restriction.
It hasn't been discussed for a long time because it's obvious that it's a good idea. Of course, you can say "not everything that seems obvious is actually true" and I would have to agree!
dang

I felt that the last response is dismissive of my argument, as it ignores it totally. But it's understandable that it is the way they like to manage their organization.

07:40 Some quotes from Abstract Heresies: Why vibe code in Lisp?

Nice quote from Abstract Heresies: Why vibe code in Lisp? about how Lisp is designed: Most modern languages force you to describe exactly how a machine should shuffle bits around. Lisp was designed as a language for expressing high-level abstractions rather than expressing tedious implementation details. When I prompt the AI, I want it generating architectural logic, not fighting with boilerplate just to manage basic state.

I would love to give some examples to this argument. Let's say you will count things in Go:

words := []string{"cat", "dog", "cat", "bird", "dog", "cat"}

counts := make(map[string]int)

for _, word := range words {
 counts[word]++
}

We create a map, we iterate through the collection and extract the elements, then we use it as a key and mutate the counter. That's "mechanical state management". In Lisp, you could write:

(loop with counts = (make-hash-table :test #'equal)
 for word in '("cat" "dog" "cat" "bird" "dog" "cat")
 do (incf (gethash word counts 0))
 finally (return counts))

Still imperative and not magically eliminating all the mechanics, but you can still build on this the abstraction that tells "group these things and count them".

There's this nice one too;

Designed for the Elite Let’s be honest: Lisp is a language designed by and for elite hackers, not for the masses. It doesn't hold your hand, and it doesn't pander to lowest-common-denominator programming bootcamp patterns. When you use it as a target language, you are operating in an environment built for maximum expressiveness.

I had a bad writing before about Lisp's superiority: Thoughts on languages. #Programming #Lisp