smiletolerantly

joined 6 months ago
[โ€“] [email protected] 4 points 1 week ago* (last edited 1 week ago)

Oof.

My employer pays a buttload of money to CircleCI - for extensive checks (build, lint, formatting, full test suite, as well as custom scripts for translation converage, docs,... for the full tech stack) on every push. Reviews start only when everything passes.

I think you have given me a new-found appreciation for the reasoning behind that decision... ๐Ÿ˜„

[โ€“] [email protected] 4 points 1 week ago (2 children)

Lmao, what, that's wild. How did they justify this??

[โ€“] [email protected] 25 points 2 weeks ago (8 children)

Fuck Amazon, fuck Alexa.

But that wall clock is glorious. It's a decently look clock, but seeing how much time you have left on multiple timers with a single glance is so incredibly useful. Especially when you're cooking.

I'm currently in the process of migrating away from the shit Alexa ecosystem, but no matter what I end up with, I'll have to find an alternative for this clock

[โ€“] [email protected] 5 points 2 weeks ago

Did someone say Gemini?

[โ€“] [email protected] 2 points 2 weeks ago

I think if you want to copy a specific selection to a mouse-based, different program then it makes sense to use the mouse for precision selection.

[โ€“] [email protected] 3 points 3 weeks ago

For me personally, there is only two applications of LLMs in programming:

  • doing tasks I kinda know how to do, but don't want to properly learn (recent example: generate pgf plots from csv data in matplotlib. 90% boilerplate, I last had to do it 3 years ago and vaguely remember some pitfalls so can steer the LLM in that direction. Will probably never again have to do this, so not worth the extra couple hours to properly learn
  • things I would ordinarily write a script for, but aren't worth automating because they won't come up in the future again (example: convert this Lua table to a Nix set)

Essentially, one-off things that you know how to check for correctness.

[โ€“] [email protected] 4 points 3 weeks ago

Ahh those fuckers.

[โ€“] [email protected] 5 points 3 weeks ago

Especially if you buy access via 2 providers on different backbones. Haven't had a single failed/incomplete download since.

[โ€“] [email protected] 8 points 3 weeks ago

I'm slightly younger than that even, currently finishing up my master's but have been working as a backend dev for a couple of years.

I've learned an order of magnitude more about networking from just being in the vicinity of my girlfriend (who is a network technician) than from uni, and it's definitely already paying off.

[โ€“] [email protected] 2 points 3 weeks ago

+1 from me.

The Shield is a couple years old, but it handles everything you throw at it perfectly.

  • get SmartTubeNext to watch YouTube without ads, and it comes with SponsorBlock
  • use Flauncher for a home screen / launcher without any ads
  • Jellyfin, obviously
  • Steamlink also works perfectly
  • plus, the remote is amazing (though I would recommend to either disable or rebind the Netflix button)
 

Basically, the title. After years of inactivty, I'll be taking music (cello) lessons again, with my teacher of yesteryear, from whom I've moved half a country away.

She has suggested Zoom but is open to alternatives. I don't particularly like Zoom, plus I have a feeling better quality can be had through a custom solution - but I'm at a bit of a loss as to what exactly would be a good fit for this project.

Maybe Jitsi? Does someone here have experience with it and could tell me if it's possible to set something like a "target" audio quality?

For hardware, I basically have two options. Both are already in use, for different things, and have sufficient processing capabilities - albeit no GPU:

  • host everything at home. Plus: lowest possible latency from me to the server. Not sure how much that is worth though.
  • root server in the Hetzner cloud: much faster network speed. Again though, not sure how beneficial that is, the ultimate bottleneck will always be my upload speed (40Mbit)

OK, I realize that this post is a but of a random assortment of thoughts. I'd be really happy about suggestions and / or hearing about other's experiences with similar use-cases!

28
submitted 4 months ago* (last edited 4 months ago) by [email protected] to c/[email protected]
 

Hi,

not sure where else to post this. For a while now, I've unsuccessfully been trying to get WireGuard to work with Crunchyroll.

Setup is as follows:

  • dedicated server hosts a wg-quick instance in [neighboring country]
  • OPNSense acts as peer on a single IP
  • I have a rule for routing the entire traffic of some source device via that IP

This works just fine. Handshake successful, traffic is routed via the server. traceroute shows the server as the hop immediately after my device's local gateway. The connection is stable, and fast.

...except for Crunchyroll. The site / app itself is fine, but I can not, for the life of me, get a video to play. It just keeps loading forever.

I don't think this is an issue with CR recognizing that I'm not where I say I am - looking online, it seems pretty easy to use CR with a VPN. I've also tried from multiple other devices, all with the same symptom.

If anyone has suggestions, I'd love to hear them ๐Ÿ˜…

EDIT: ~~It was MTU. Had to manually set it to 1500 on both devices.~~

Nope, still the same issues. I was using the fallback interface there briefly.

EDIT: It WAS MTU related, I had to enable MSS clamping on the OPNSense.

view more: next โ€บ