Say hypothetically 1000 people were having great fun using it to cyberbully 10 people. It’s impossible to say that it’s no big deal because 99% of users loved it, unless you know exactly how much the 1% hated it.
This resonates. There are certain tasks, like dealing with any government or healthcare-related web page, that I won't even bother attempting on my phone. In my case, it's because I just know in my heart of hearts that the crummy mobile website won't be feature-complete enough for me to complete my goal.
My wife is the opposite. It doesn't occur to her that the problem may be with the janky website, not with her. She'll ask me for help with a thing out of frustration and my first troubleshooting step is to reach for my laptop. This is almost inevitably followed by "hey, wait, how come you're able to press the Submit button but I wasn't able to?" "Because the dev never tested this on a phone and it's broken." "So it's not just me being incompetent to use this website?" "Nope, never was."
No laughter here, my brother in music. This is one of the few vocal groups that I could be in the zone with, except "Fernando", because one must release their inner theater kid with that one.
I’d recommend installing a Pleroma server. It speaks ActivityPub and you can use any of the nice Mastodon apps with it. I've run a Mastodon server for the last 9 years, and wouldn't recommend Pleroma over it for a large many-user instance, but it's relatively tiny and lightweight for a personal server. You can configure it not to talk to the rest of the Fediverse so that it remains your friendly, isolated silo.
Pleroma looks to be very twitter-y. I don’t feel twitter is a great model for a small tight-knit group. For a larger less familial group, it’s probably better suited.
Like, i’m thinking photo album sharing (twitter-like makes photos ephemeral, quickly disappearing on the timeline) and conversation (twitter threading has never been strong imo).
If you were going for a social-media-y experience, I'd not recommend Pleroma (or Akkoma which is the less problematic fork) because dealing with Erlang+Elixir is a massive pain in the arse. You'd want GotoSocial[0] (single binary, reasonably straightforward), snac[1] (haven't tried it but fedimeteo runs a whole bunch of instances successfully), or one of the other small servers (Takahē, bovine, etc.)
GoToSocial looks interesting, i will probably spin one up to try it out! Still seems a little twitter-like, but worth a shot.
And as long as there is a docker container, i don’t really care what language it’s written in, tbh - tho that is sometimes useful as a signal of the code quality or other aspects
You know, after all this time Lucas Duplan doesn't seem so bad. His hubristic sin was posing for a photo burning fake hundred dollar bills. That just seems like a random Tuesday now.
I was half-joking, but if YC has a legal issue resulting from the alleged fraud (unclear currently), kicking out the company for the lesser infraction would make more sense.
Investors aren't on the hook for the bad behavior of companies they invest in. Quite the opposite: Defrauding investors (and acquirers, and creditors) is commonly the thing that lands people like Elizabeth Holmes in prison.
Ycombinator may have financially benefited from the scam operations since the company subsequently raised funds.
Considering they do due diligence before investment and are experts in IT and legal, how could they not know what is the business model when it was the unique selling point ?
Yeah, yeah... of course, of course... like telehealth companies prescribing GLP-1 Ozempic/Wegovy where there is one doctor for 10000 patients. Totally sounds legit.
Phrased a little more harshly than I would've, but I agree. SIP keeps any random process running as the device owner from running amok and paving over the system. You have to jump through just enough hoops to disable it that a rogue process can't automatically do it against you.
Until maybe 2 years ago absolutely. Huge demand for tech workers and not enough smart US workers to fill the roles… lots of low quality candidates from crappy colleges but anyone with skills or smarts (and not unreasonable salary expectations) got a job.
Since the ai data center cash suck the jobs have dried up… ai productivity gains maybe too, although we’re waiting to see meaningful results there.
From a recruitment perspective it’s still difficult to find experienced candidates with specific skill sets because any job openings get flooded with 10,000 ai generated slop applications that have to be screened by ai and the 10 excellent candidates get lost.
In my observation, job market wasn't great even before ai. AI made it worse by clogging existing channels - already not that great - with a ton of slop.
reply