rdnt:~$ _

@GossiTheDog@cyberplace.social can we send them to a wild-goose chase of relinking the account from other instances and blaming it on the next instance? That'd be fun.

@richiekhoo@hachyderm.io I stopped making contributions because, well, life happens, but if I resume it will be on Codeberg. If a project is on GH and I absolutely *need* to contribute to it, I will unhappily do so while ignoring any and all LLM reviews and hope for the input of a human bean. Also: federation support groundwork for Codeberg (Forgejo) seems super interesting.

@pheonix regarding longevity: air is still passing through, you may be getting a little higher temperatures, and it will take a little more time to cool after a spike. However, you can configure the fan curves to anyway go to 100% after a configured threshold. For me, my room can at times be dead silent and anything above 50% fan speed can very easily become irritating, especially if it spikes to 100 then down to 20 and repeats. What I do suggest for longevity is to change thermal paste every 6 months or so (try to find good heat conductivity thermal paste, but avoid electrically conductive ones), and be aware of the hotspot temperature. If it spikes instantly to the temp limit (e.g. can be 85°C for GPU, 95°C for CPU, depends on model), either undervolt if you know how, or change thermal paste (to the GPU, too, again if you know how and are aware of the tradeoffs/warranty becoming void etc). High fan speeds won't be helping much if the component is actively cooking itself while not being able to exhaust that heat efficiently. Now, for the configuration, I can embed my config for fan2go (Linux) if you'd like, but you usually have to tinker around with some trial and error and arrive at something that works for you. As long as temps remain reasonable you are fine. A utility called Prime95 can help stress test the CPU, for the GPU i've been using FurMark but it may be dangerous for the components in your machine, YMMV. Have in mind that IF the temperature even by a little exceeds a hardcoded limit set by the manufacturer, the component shuts down for its own safety.

@pheonix Congrats on the build! :blobfoxcofecute: Some tips! Fan speeds can usually be lowered and be configured to be more stable and less annoying. There are tools to help you achieve this, SmartFan5 if you are using Windows, and fan2go on Linux (I've had good results with both). You can lower the avg temperature (aggregated across multiple components) by using some NVME temps multiple times, so as to make transient temperature spikes on CPU/GPU less annoying. Then with this 'slow start' avg temp, you can configure the fan speeds. The performance penalty, if doing spikey work (coding) and not intense renders or builds, is negligible.

@Gina morning :blobfoxcofecute:

@nadia@raru.re hdmi to the desktop and game time