I use Linux utilities everyday on macOS as well as on my servers and networking equipment—I love it to death.
But when it comes to the end-user stuff, I’m deeply invested in Apple’s ecosystem. My personal laptop, work laptop, phone, Apple TV work pretty seamlessly together and I love that aspect of it.
What saddens me most is the drive away from native app development regardless of OS. Can’t stand Windows but if I’m using it I want the applications running on it to match Windows’ design language the same as I do on macOS (Linux GUI looks awful on macOS) and Linux (depending on desktop environment).
More related to Linux though, I do tend to lean more toward BSD UNIX than Linux, but due to the lack of containerization and popularity it’s hard to make it a daily OS.
This sounds like a fun problem to solve as a developer. A web application could be written that:
If SearXNG offers an API you could even ditch the proxying of results. If you got really fancy you could store stats on which are the fastest/slowest and act on that in the future.
Me too, and it’s worse because it’s not secure.
I keep saying this a lot but I don’t know why recently (the last ~5 years) everyone is jumping on SMS-based 2FA. I remember this was really big around 2010 and as a developer all the tools for SMS-based 2FA are deprecated or unmaintained (at least in my programming language). It seems like all these websites that jumped on board 10 years late have very poor security practices.
I’ve run into issues with SMS-based 2FA (yikes) on some websites because my phone number was a landline number I purchased then later transferred to my wireless carrier.
I bring this up because I’ve noticed some websites have the typical “we’ll confirm your information with your wireless carrier” verbiage, but those generally mention they do so to determine whether the number is a landline or wireless.
I’m super unsure of what’s going on in this case, but when I first saw this screenshot this is what came to mind.
Hard agree, except I do have an issue with the last paragraph in that I think it’s far dumber than you’ve described.
Simply blocking (a shit ton of) domains can really get you 99% of the way there. I’m a web developer and it’s stupid dumb how third-party stuff is hosted. It’s either exactly that (third party hosted) or a CNAME or a third party which is easily blocked.
Look, I know how complex tracking and fingerprinting can be. But from my experience, it’s really not hard to block. Of course, I’m not really speaking to first party tracking where blocking would destroy the entire experience. But for the most part, you can prevent a profile being built about you (at least for tracking and advertising) by blocking with DNS.
I use primarily DNS blocking myself, but it’s a custom solution that pulls in a ton of blocklists. I get tired of the “just use a browser extension” as the solution for everything, and any time I bring up IP/DNS-based solutions people say “but that doesn’t block everything” as if browser extensions do.
I’ve been reporting spam for years (old iCloud email account I can’t destroy was leaked everywhere back in like 2015) and using the websites to report spam seems to have zero effect.
What does seem to work well is reporting to the originating server’s owner, but it’s mostly hit or miss.
If the email happens have suspicious links, reporting to the IP address owner and the registrar the domain is hosted at is usually very successful. I’ve filed simple reports for those and have received a “we took down this host/domain” within minutes several times.
I really wanted it to work on Fly.io but I couldn’t get it to. I’d also like to get the Tailscale software Dockerized but running multiple nodes on the same host with custom DNS was a complete shitshow.
I really love Tailscale, but the daemon and CLI seem to be absolute garbage.
I’ve been blocking Google domains completely (except for OCSP) for almost a year (using DNS). I’m sure some domains use Google Cloud and slip past the DNS blocks, but usually the only things that break are captchas and some shitty old websites that pull jQuery from a Google domain (why would anyone do that?).
“It breaks all of the internet” is a little dramatic, maybe if you block their OCSP domains that’s true.
I do agree though that 80% is low, even if only counting the traditional tracking script that’s been used everywhere for ages.
I’ve never gotten this to work with iOS devices for several years now. Yes, I’ve made sure that background app refresh is permitted, location services are on, etc.