Hacker News - It's not for hackers and it's hardly news.

[teriyakiburns]

I identify as the man behind the turkey baster in your ass

Don't threaten me with a good time.

 

[UERISIMILITUDO]

Static typing is ALWAYS > Dynamic typing

I'd rather argue that strong typing is better than weak typing. Ada has nominative typing, so equivalent types aren't compatible without explicit conversion, if allowed; private types don't allow conversion to some equivalent type, as an example. The limited types allow one to control even assignment and equality. Controlled types, lastly, allow code to run at initialization, assignment, and finalization. Still, Common Lisp is also nice, with its dynamic and strong typing.

Anyway, here's my latest roundup. We'll start with more instances of Walter Bright reminding everyone of what a good boy he is:
https://news.ycombinator.com/item?id=40042616#40045743 (archive)

Nestle getting people addicted to sugar is A-OK, and anyone who falls for it is just weak:
https://news.ycombinator.com/item?id=40067575#40068534 (archive)


I'm not even a Catholic, and I hate motherfuckers like this:
https://news.ycombinator.com/item?id=40102779#40112064 (archive)

I can imagine how a conversation with this fuckwit would go:
No, public school teachers don't count.
No, it's not pedophilia when a rabbi does it.
Why are you citing FBI crime statistics? Don't you know they're biased?

Uh oh, dang is getting pretty serious about policing those mean and subtle usernames, but remember people who hate white people on Hacker News are allowed:
https://news.ycombinator.com/item?id=40055120#40056982 (archive)

By the way, the linked article is a very pleasant read, and I recommend it.

I found this asshole on Lobsters trying to use shame to get a protocol adopted:
https://lobste.rs/s/je9hvj/ (archive)
https://whynoipv6.com (archive)


There's some interesting spam happening right now I managed to notice, although I only successfully archived one instance:
https://news.ycombinator.com/item?id=40122837
https://news.ycombinator.com/item?id=40122628 (archive)
https://news.ycombinator.com/item?id=40121934

 

[306h4Ge5eJUJ]

I found this asshole on Lobsters trying to use shame to get a protocol adopted:

No! You MUST support the new thing, because... you just must, okay? Otherwise you're a Bad Guy. A really bad one. IPv6 cultists are the weirdest fucking thing.

 

[AnOminous]

No! You MUST support the new thing, because... you just must, okay? Otherwise you're a Bad Guy. A really bad one. IPv6 cultists are the weirdest fucking thing.

What that is actually available is better? I mean we literally ran out of IPv4. Arguably we should have come up with something better since this happening was obviously coming, but something of the sort is kind of necessary now, even if like much of the Internet it's ad hoc bullshit came up with as a last minute kludge to stop everything just immediately falling apart.

 

[Tom Myers]

You should see how people react when I mention that I'm primarily a Java dev. Go back to Javascript you ignorant pleb, ya boi is going to make mo' money than you could even imagine.

I'd rather argue that strong typing is better than weak typing. Ada has nominative typing, so equivalent types aren't compatible without explicit conversion, if allowed; private types don't allow conversion to some equivalent type, as an example. The limited types allow one to control even assignment and equality. Controlled types, lastly, allow code to run at initialization, assignment, and finalization. Still, Common Lisp is also nice, with its dynamic and strong typing.

 

[UERISIMILITUDO]

What that is actually available is better?

I think it may be a better proposition to buy part of the IPv4 address space, like an Urbit fiefdom actually worth something. I don't fully understand everything to do with IPv6, but I do know that every problem I've ever had with it is solved by turning it off. It's most popular with smart phones, because those people don't control the network in any way, and it's most popular among subhuman populations as in India and Africa, which is another downside.

Regardless, since I write my libraries by myself, the tired solution of use someone else's library to paper over the differences doesn't work. The reality is that there are two incompatible Internets, and one matters much more than the other. As an example, storing an IPv6 address takes four times as much space, and I don't necessarily want to do that.

 

[306h4Ge5eJUJ]

What that is actually available is better? I mean we literally ran out of IPv4. Arguably we should have come up with something better since this happening was obviously coming, but something of the sort is kind of necessary now, even if like much of the Internet it's ad hoc bullshit came up with as a last minute kludge to stop everything just immediately falling apart.

Is it really necessary when ~50% of the biggest sites don't support it and do just fine? For what it's worth, I agree that IPv4 is a little undersized, but a lot of the fashionable (for lack of a better word) enthusiasm for IPv6 seem to boil down to "it's Progress", with the occasional dash of "think of the niggers", just like the linked site. A P2P Internet would be a great argument for moving away from IPv4, but that revolution sadly didn't happen - for fundamental human reasons I worry.

 

[AnOminous]

Is it really necessary when ~50% of the biggest sites don't support it and do just fine? For what it's worth, I agree that IPv4 is a little undersized, but a lot of the fashionable (for lack of a better word) enthusiasm for IPv6 seem to boil down to "it's Progress", with the occasional dash of "think of the niggers", just like the linked site. A P2P Internet would be a great argument for moving away from IPv4, but that revolution sadly didn't happen - for fundamental human reasons I worry.

I generally disable it in VPN because it makes me paranoid. I'm not sure it's actually as bad as claimed.

 

[inception_state]

Is it really necessary when ~50% of the biggest sites don't support it and do just fine? For what it's worth, I agree that IPv4 is a little undersized, but a lot of the fashionable (for lack of a better word) enthusiasm for IPv6 seem to boil down to "it's Progress", with the occasional dash of "think of the niggers", just like the linked site. A P2P Internet would be a great argument for moving away from IPv4, but that revolution sadly didn't happen - for fundamental human reasons I worry.

We are still slowly moving towards it because slapping NATs on NATs on NATs is a pain in the ass, IPv4 addresses keep getting more expensive, and the old stuff without IPv6 firmware is dying off. The fact that CGNAT even exists is hilarious, it's an industrial scale kludge that supports billions of dollars of commerce.

Also nice thread @UERISIMILITUDO. One thing I'm not sure you mentioned in the OP was how the front page algorithm is designed to kick stuff off the front page if the comment:upvote ratio is too high. Basically it's a flamewar filter. Dang might manually put his thumb on the scale once in a while, but I think they have just calibrated it so contentious topics get moved off the front page quickly.

 

[<3 norafawn]

As an example, storing an IPv6 address takes four times as much space, and I don't necessarily want to do that.

>4 times as much space
Nigger maybe you should use the library

 

[UERISIMILITUDO]

>4 times as much space
Nigger maybe you should use the library

I don't understand. An IPv4 address is thirty-two bits. An IPv6 address is one hundred and twenty-eight bits. Thirty-two times four is one hundred and twenty-eight.

 

[<3 norafawn]

I don't understand. An IPv4 address is thirty-two bits. An IPv6 address is one hundred and twenty-eight bits. Thirty-two times four is one hundred and twenty-eight.

I think you misunderstand my criticism - storing the full 128 bits is basically pointless, you really want to store 64 or 48 bits or spam will BTFO your service. Libraries provide all kind of subnetting features

 

[UERISIMILITUDO]

I think you misunderstand my criticism - storing the full 128 bits is basically pointless, you really want to store 64 or 48 bits or spam will BTFO your service. Libraries provide all kind of subnetting features

Yes, I misunderstood. For logging purposes, I see no reason not to store the full address.

At the least, a mechanism to distinguish between IPv4 and IPv6 addresses would be unnecessary, by using that IPv6 address space dedicated to representing IPv4 addresses.

I commented in this thread, also about Hacker News, with these thoughts:

Really, the idea of using fixed-length addresses for the network to end all networks is stupid. An address should be a variable-length string of digits, and it would be easy enough to look at only the first few digits for routing, after which point they could be discarded. Universally unique addresses are stupid.

Regardless, an awful lot of organizations, probably almost all of them, would be just fine with a single IPv4 address and a private network. A mere sixteen million addresses in just one private network allotment is enough for almost every organization, except for the wasteful technology companies.

 

[Rusty Metal Skull Gun]

There's like three or four threads on this site now which can be summed up as "Uerisimilitudo says something stupid, people less autistic than he is try to argue with him".

An address should be a variable-length string of digits

This is a dreadful idea.

and it would be easy enough to look at only the first few digits for routing, after which point they could be discarded

This is literally how it already works.

 

[UERISIMILITUDO]

There's like three or four threads on this site now which can be summed up as "Uerisimilitudo says something stupid, people less autistic than he is try to argue with him".

Those are rookie numbers, I know.

This is a dreadful idea.

I'd ask for an explanation, but we can just let it die on the floor.

This is literally how it already works.

I know. In the context of such a network, I mean that the address would be modified in-transit like this.

Hey, I'm today going to purchase this book, recommended to me by this article mentioned in one of the linked articles from my recent post, and I'll perhaps report back on any of my thoughts I see reflected in his words.

 

[SangeAndYasha]

Spoiler: A real world example of AI ethics

I know we joke about reddit and their Marvel obsession, but I had a good chuckle with this one.

 

[AnOminous]

I know we joke about reddit and their Marvel obsession, but I had a good chuckle with this one.

O NOOOOEZ AI IS STEALING THE JOBS OF REAL HUMAN WAR CRIMINALS!

 

[inception_state]

Some Scala community drama was just settled in court, TL;DR:

I am a Scala developer and speaker who was cancelled three years ago. Yesterday I attended the High Court in London to hear an apology from several prominent members of the Scala community for making untrue claims about me on 27 April 2021. I sued them for libel, and they admitted fault and settled, paying me costs and damages.

The HN thread is fairly split between people happy for the guy winning through legal channels, and people upset about the outcome or how he accomplished it. The case was settled in the UK, which is fairly notorious as a libel tourism jurisdiction.

 

[ditto]

There's some interesting spam happening right now I managed to notice, although I only successfully archived one instance:

There seems to be a lot more spam on Hacker News lately. Obvious YouTube comment-tier crypto scams, too.

 

[Dick Mason]

https://news.ycombinator.com/item?id=40164077

https://web.archive.org/web/20240426124659/https://news.ycombinator.com/item?id=40164077

Someone is plagiarizing blog posts at scale, using many of different throwaway HN accounts and fake Substack blogs.

They appear to be going through old Hacker News submissions, finding the source pages that were submitted, and then publishing those posts on new Substack blogs that they created.

Hacker News has been hit with automated comment spam attacks recently, causing DOS for a bit. Not bad enough, now a self-promotion blogspam operation has been tearing through.

https://news.ycombinator.com/item?id=40168052#40168606

https://web.archive.org/web/20240428130820/https://news.ycombinator.com/item?id=40168052#40168606

This Substack appears to be stolen - by re-creating a Substack hostname that the original author has deleted, then re-publishing the articles with back-dated dates.

This Substack plagiarism scammer is getting even more creative.

They appear to have found an old Substack that was deleted in 2022-2023 (https://web.archive.org/web/20220826223418/https://michaelde...), then registered the deleted Substack hostname (with their own Substack profile), gone through Internet Archive or another source, and re-published the posts with back-dated dates.

OP, what say ye? Am I wrong?

Substack, how is this possible?

The culprit:

https://mamddoh.wordpress.com/

https://web.archive.org/web/20240428025348/https://mamddoh.wordpress.com/

Some guy named Mohamed.

They're all focusing on the submissions and the links, I would also check the comments. Like some lolcows that post here, it seems likely that they'll employ comment fluffers on their promotions.