Discussed previously:

https://news.ycombinator.com/item?id=27764806

I only skimmed this but if I understand correctly the CPU needs to be in debug mode to access these instructions right? If so then it's hard to see why the authors present these instructions as dangerous - I imagine once the CPU is in debug mode you would have unfettered access to the system even without these instructions.

(not that it isn't still interesting!)

The undocumented instructions pdf paradox

Not sure if it's the same researcher, but I remember some talk about using fuzzing to find undocumented instructions on x86.

Is this the same idea, just with newer results?

Hum, it is the first time I see it. Was the paper posted before?

thanks for the link!

as many times as it takes

I keep trying to like this site and I just can't get past how bad the moderation of submissions is.

Stuff is endlessly re-posted even in the same day, and with the most uninformative clickbait-y titles that seem purposefully structured to make you go "...okay, and what exactly is Fruutluup, which apparently just hit version 3.0?" Due to fear of missing out on something cool/interesting, you click on it. Because the submitter knows damn well that if you knew Fruutluup was a MOD tracker written in Scheme, you wouldn't give two shits about it.

The vast majority of posts here are either reposts, fall firmly into the category of "less than a hundred people give a shit about this", "yet another implementation of ____ but in some programming language nobody uses", the ever-popular "nerd getting all philosophical", or some random subject where the post is just the means to an end, letting HN commenters engage in extensive navel-gazing and "intellectual .

A great example of the latter would be a recent submission about agriculture where the author of the news story noticed the traffic and jumped into the comments to ask "uhh, this is a computer tech site, right? How are so many of you qualified to discuss this subject?" and she got absolutely hammered by HN neckbeards going "well ACKSHUALLY, we're programmers and that means we're VERY intellectual people. And some of us are autistic, even. Thos people are EXTRA qualified to talk about shit we just googled an hour ago."

It's a shame your 'great example' has to deliberately misquote and mischaracterise the interaction.

The author asks "isn’t this a tech/VC blog? How is it that so many of you are so interested in and knowledgeable about agriculture?"

She asked why people were interested and were so knowledgable about a subject she didn't expect to appeal to this crowd. People explain the premise of HN and she then asks for other things this readership might be interested in. She didn't get hammered; she asked a question and got a number of honest, personal responses.

Not every website has to appeal to every person. The people who enjoy things that are presented here enjoy them. The people that don't are welcome to find a community that works for them.

I assume you were referencing: https://news.ycombinator.com/item?id=27809279

The GP is indeed completely mischaracterizing what happened in that thread, to the point where you can wonder if he really read that thread.

The account was created a few hours ago today and seems to be a sort of burner/puppet to share controversial opinions.

> "less than a hundred people give a shit about this"

Those are the best posts and are why I'm here

Me too. I love when people link random blog posts written by grad students and people in industry who are writing for a very specific audience, or even just for themselves.

If there's an article I don't care about, I don't click on it. There are 29 other things to choose from on the front page.

I'm feeling the opposite, I've read many interesting articles about very diverse subjects through HN, and am often impressed by the quality of the accompanying comments.

If you don't find most of the content here interesting, then go to a different website because clearly the community here disagrees.

Maybe you could try reddit, and then just subscribe to the subreddits about topics you care about? Redditors may not have a great reputation, but the the niche subreddits have nice communities in my experience.

While I agree with the contents of your message - that the titles of the submissions should be more informative - I don't believe it is the job of moderators to do that, but the submitters themselves.

Absolute numbers mean very little BTW. Recently some (but not all!) of the most interesting articles are those that are flagged or mysteriously disappear from the front page.