Using government guidelines to police content: state censorship?

Hey Dan,

This is a bit unusual, but I wanted to let you know about a very strange Racket issue. I'm not sure if HN is still running on Racket or not, or if the CL port has finished. But...

For many years, HN has been running just fine with ac.scm looking like this:

  (define x-set-car!
    (let ((fn (namespace-variable-value 'set-car! #t (lambda () #f))))
      (if (procedure? fn)
          fn
          (lambda (p x)
            (if (pair? p)
                (unsafe-set-mcar! p x)
                (raise-type-error 'set-car! "pair" p))))))

  (define x-set-cdr!
    (let ((fn (namespace-variable-value 'set-cdr! #t (lambda () #f))))
      (if (procedure? fn)
          fn
          (lambda (p x)
            (if (pair? p)
                (unsafe-set-mcdr! p x)
                (raise-type-error 'set-cdr! "pair" p))))))

The only reason I know that, is thanks to Scott. It was one of the first tips he shared with me. Vanilla arc3.1 looks like this:

  (xdef scar (lambda (x val) 
                (if (string? x) 
                    (string-set! x 0 val)
                    (x-set-car! x val))
                val))

  (xdef scdr (lambda (x val) 
                (if (string? x)
                    (err "Can't set cdr of a string" x)
                    (x-set-cdr! x val))
                val))

which became unreliable years ago.

Scott explained that unsafe-set-mcar! was a bit of a hack, but that it worked due to a few assumptions about the code. I upgraded to this and never gave it a second thought.

Fast forward to a week or two ago. Arc kept throwing very strange errors. As far as I could tell, nothing whatsoever had changed.

But Racket had changed. The latest shipped version, for whatever reason, causes # to be inserted to the `car` if you try to do (scar '(x y) '(a b))

(It took a long time to track that down...)

After poking around the racket source code, I noticed there was a function called unsafe-set-immutable-car!

On a whim, I tried switching unsafe-set-mcar! to unsafe-set-immutable-car! and everything worked perfectly.

I don't know why, or whether it's stable long term. But I wanted to let you know, in case someone on your team is running into strange Racket errors, that this is the fix:

  (define x-set-car!
    (let ((fn (namespace-variable-value 'set-car! #t (lambda () #f))))
      (if (procedure? fn)
          fn
          (lambda (p x)
            (if (pair? p)
                (unsafe-set-immutable-car! p x)
                (raise-type-error 'set-car! "pair" p))))))

  (define x-set-cdr!
    (let ((fn (namespace-variable-value 'set-cdr! #t (lambda () #f))))
      (if (procedure? fn)
          fn
          (lambda (p x)
            (if (pair? p)
                (unsafe-set-immutable-cdr! p x)
                (raise-type-error 'set-cdr! "pair" p))))))

I hope that saves someone a little bit of time.

Thanks for taking care of the troll. (And the HN API, and a thousand other things.) Best of luck.

> Even worse I am having a tough time putting my finger on what the shortcoming of these people is.

I've been fairly obsessed with this general area for several years now, and I have come to a fairly strong conclusion that the(!) root cause is the nature of human consciousness (and in turn, how individuals perceive & conceptualize reality...or, how the human mind ~renders reality to our consciousness).

> It’s not IQ as I’ve seen and met quite a few gullible sheep of this sort who would undoubtedly score high. I don’t see a strong correlation either way with intelligence. It doesn’t even seem to be one’s metaphysics or epistemology. I know devoutly religious people as well as mystics and occultist types who are nonetheless not gullible in the way that this crowd is gullible, and I have seen atheists and people with scientific backgrounds who are. Race, nationality, even educational attainment doesn’t seem to map.

While there is surely some variance by various dimensions like these, I have come to the same conclusion as you, this phenomenon does not seem to have a very strong causal relationship with any of these things, as counter-intuitive as that may seem. That said, I think it is possible to start to make some sense out of it if you (even speculatively, as a thought experiment) consider it from the perspective of consciousness (our "reality rendering/generation engine") being the fundamental or primary root cause.

>People you don't agree with are nazis?

You apparently need practice in reading comprehension.

But I do find a lot to agree with in reasonable people, so sure, I could defend that accusation for s&g's.