I got a car recall email that appears to be legitimate (has the correct name and VIN, talks about an actual recall that I found in news, etc) but looks *every* bit like a scam email.

Like, it's got the "Full name: .........." "Address: ........" form at the bottom that you usually see in the Spam folder, and it's got the "IMPORTANT!" header that looks like it was drawn in MS Word, and embedded as an .emf image of all things.

And the sender company *appears* to be the official Opel representative or something, but does not have a website of any sort, and where did they get my contact information from?...

https://network.geant.org/wp-content/uploads/2024/03/GEANT-IP-Service-Description-March-2024.pdf

I wonder what exactly makes option 1 (both paths going through PoP A) more resilient than option 2 (diverse paths to PoPs A and B)?

some (many?) years ago I realized that no, I don't *really* need to log every single track to last.fm

[to the extent that I had a udev rule to automatically upload the listen counters from my iPod whenever I connected it]

now, in a similar way, I'm realizing that I don't actually need to have a public dotfiles.git on GitHub.

(I mean I still have a dotfiles.git and it's technically public for the time being, it's just no longer on GitHub.)

slowly I've also been moving other discrete projects into plain old SSH+cgit, and instead of the fancy README.md as is now the style, they'll get .html pages under website/projects/.

(although the trigger was different – with last.fm it was the Android apps that slowly broke due to restrictions on what apps can read notifications; with github it's more of a general antisocial thing.

still, in both cases, it's a little freeing, both in terms of privacy, and in terms of what software can interact with it, e.g. I can just open any .mp3 player and make a playlist without caring about the scrobble stats, etc. Same goes for my own IMAP mail server, too.)

"speed=17.4x" ffmpeg on Core Ultra 5 is fast. I'm too used to seeing 2x-3x on my ancient Sandy Bridge server...

(I usually do that kind of heavy work remotely, because downloading the result to home is fast while uploading it *from* home would take hours.)