• 147 Posts
  • 3.51K Comments
Joined 2 years ago
cake
Cake day: August 15th, 2023

help-circle




  • Harmonic frequencies are more likely to be an issue.

    If you have an antenna transmitting at 2.4ghz, you will also see subharmonic bumps at 1.2ghz, 800mhz, etc. A receiver at 800mhz could potentially get “washed out”, or overpowered, by a 2.4ghz transmitter that is too close simply because of subharmonics.

    Transmitters aren’t perfect either. While you can get really strong transmissions at very specific frequencies that can propagate really far, electronics resonate at many frequencies and that resonance will make it to a TX antenna as noise.

    Unless the antennas are designed to work together, you shouldn’t put them that close together. (I am also speculating that in extreme cases, a weird configuration like that could detune the transmitter antenna in a such a way that it would blow out the transmission circuit. I dunno about that though.)





  • Air, water, AIO, whatever. If it cools well, use it. I just prefer AIOs and there really isn’t any maintenance, was my main point. There are always tradeoffs between AIO, air or a proper water rig, so there is that. (Fans are crazy quiet these days, but when I made the switch, it was mainly for noise. I always run an overclock, so my fans were always hauling ass which probably isn’t needed now.)

    Ultimately, I prefer AIOs for the way airflow is managed. It’s not better or worse than air in many instances, but I like working with a radiator rather than a chonky heatsink.

    I cannot disagree though: zero maintenance is better than maybe-maintenance. Like I said, it’s about tradeoffs. (I can still make my PC sound like a fucking jet engine, though. Noctua server fans kick ass.)






  • The quality of the app still depends on the quality of the developer. The developers ability to limit which parts of the project are modified by the agent is kinda important as it’s all too easy to get swept up in how confident AI agents seem when it suggests code changes.

    I use AI agents to code a ton of personal stuff, but I also understand the code it generates. If a person just turns an agent loose with implementing a dozen features while trying to fix bugs at the same time, they are going to have a bad time. AI coding agents can lose context, refuse to fix bugs, lie and even argue about code problems.

    AI is just like any other tool: Use it correctly and where applicable. The second a person starts blindly trusting vibe code, it’s game over.

    My personal opinion is that vibe coding is generally bad for public facing projects where security and stability is a real concern. It’s just too easy to get lazy and forget to review code.






  • Ok, on 332 and there is still an issue. Sorry for the screenshot post, especially with data, but I exclusively use Lemmy on my phone with Connect only. (My apologies if this turns out to be a red herring.)

    I had a feeling that there is a difference between the loading behavior when using two different ways to refresh the feed. Either [Right Side Menu -> Refresh] and [Main Menu -> Click All].

    Using the Main Menu and clicking All seems to yield slightly better feed loading behavior before it hangs.

    Obviously this is non-scientific, but I reloaded feed and speed scrolled till the feed hung and recorded the last page number. It’s not an even number of test scrolls, as I just kept doing it until each test set looked about the same length: