This is the stupidest nitpick, but it's not really Pavlovian conditioning (as mentioned in the last paragraph) but rather operant conditioning. Pavlovian, or classical conditioning is the triggering of a biological response after a neutral stimulus (ring a bell before feeding each time and the dog will salivate when it hears the bell even if there's no food anywhere nearby).
Operant conditioning is where the agent learns that an action produces an outcome and learns to perform (or not perform) certain actions to get the desired outcome.
Whether stupidest nitpick or not, thank you for posting this. I learned Pavlovian conditioning better from your comment. This is the kind of comment I come to HN for. Appreciate it.
it might not be anything, but the neighbor simply yielding to every little rule in the head of another neighbor.
we haven't established that the neighbor knows whether or not someone is screwing with him.
when my building had a person like that, who played these games all day for decades, some either just ignored all of it, while others simply moved out when they could.
> Ranking posts/comments by the exponential of inverse IPAddress-post-frequency
Doesn't this just incentivize posting a bunch of comments from your residential proxy IP addresses to launder them? This smells like a poor strategy that's likely to lead to more spam than not. Also, everyone has to start somewhere so your legit IP addresses are also going to seem spammy at first.
Hmm consider an established social network which sees lots of bad actor activity, these folks will likely be sharing IP's on these residential network severly damaging their reputation.
You should only see one user-id per MAC / IP. If you see multiple then its a sign of a bad actor.
Before you're established using something like a captcha prevents most spam, except for state actors, and they wont be focused on the site until its larger.
People cannot live without money. A huge swath of illegal immigrants work for money. Wouldn't it make sense to target the individuals who are _hiring_ them rather than the actual laborers themselves? This logic seems to work perfectly fine when cracking down on drug use, but seems completely ignored when it comes to immigration. (Yes, I'm aware ICE cracks down on some employers, but it's obvious this isn't their primary strategy.)
Seriously, think about it. If _you_ were tasked with cracking down on the immigration situation, what would you do in good faith? Send masked goons to check every single individual's papers and rough up people who can't show them? Or just send men in suits to every labor operation and ask for their I-9s, at 100x less cost? It's absolutely mind-boggling to me that people even assume a shred of good faith from the current administration here. This is terrorism, not law enforcement.
Like that law that says it's illegal to HIRE workers that cannot show work authorization? IIRC that carries pretty steep penalties. And if enforced, will have a huge chilling effect on the whole illegal immigration thing. But, as sibling commenters have pointed out, it's not about enforcing laws but punishing outgroups. This is only not obvious to the willfully ignorant.
IMO the whole point of KISS is to _not_ have desktops and access all of your apps via search. If I want to open my Hacker News app, I open the search, press "h" and click the first app. All apps I use frequently naturally float to the top of my history list and can be accessed with 1, or at most 2 characters in search.
Where this really shines is chat apps -- all of your chats get their own "activities" which can be searched as well, so if I want to open Discord DMs for someone, I can search their username and one of the options is the Discord DM activity for them.
Yes, this is a couple more taps than having desktops but it's 100% dynamic (never have to manage layout) and almost eliminates the chance of getting distracted when opening my home screen.
To add on to this, KISS also sorts apps/activities based on how frequently you select them, so it "learns" what you use most commonly. For example, if I type "f", the first app that comes up is F-Droid, because that's the app that matches that I use most frequently.
I have been using this launcher for longer than I can remember, perhaps 10 years now. I don't really remember using anything else (hazy memories of the KitKat AOSP launcher).
The only thing that annoys me a little bit is if I typo an app name (or Google search query) it brings up a random distant fuzzy matched contact. I wish there was a way to set a threshold for the fuzzy matching thing for certain kinds of items. Probably trivial to implement but I haven't gotten around to it.
I think it's cool that stuff like this works in C but my gut feeling is that this isn't going to play well with LSP tools. Are state-of-the-art C LSP servers aware of the preprocessor?
I don't see an advantage over C's usual error handling, because this type is lacking the support functions (like bind and fmap in functional languages) or syntax integration (like the try operator in rust) that would make it useful.
I used to think like this, but then I realized: jj-mode.el exists[0] and you can still use magic since it's still a git repo underneath. Seriously, don't let this hold you back.
I took a look at both, and I think I'd need the jj split function (comparable to staging individual lines) to be implemented in jj-mode to match my use of magit.
You can already do this if you just open up your change in magit and repeatedly commit pieces of it. I agree it's not super ergonomic since it leaves an old change behind and it requires using magit on top of jj-mode but it's 90% of the way there.
Operant conditioning is where the agent learns that an action produces an outcome and learns to perform (or not perform) certain actions to get the desired outcome.
reply