Yeah I think Lemmy would actually work pretty reasonably. It reminds me of how lots of software and projects have Reddit communities. I agree that being able to share 1 account over many services, and especially not having to pay for infrastructure is something that drives discord use over forum-based platforms.

Personally, I’d prefer that projects use forums for community discussions rather than realtime chat platforms like Discord or Matrix. I think the bigger problem of projects using Discord is not that it’s closed source, but rather that it makes it difficult to search (since no indexing by search engines) and the format deprioritizes having discussion on a topic over a long period of time. Since Matrix is also intended for chat, it has these same issues (though at least you can preview a room without making an account).

Afaik the StarFive SOCs used in SBCs are a lot slower than current ARM offerings. Part of that might be because software support is worse, so maybe compilers and related tooling aren’t yet optimized for them?

Hopefully development on these continues to improve though. The biggest nail in the coffin for Pi alternatives has been software support.

Not sure what’s going on but if the printer was flashed with Klipper, it won’t show anything on the screen until it is connected to a computer running klipper’s host software.

That sounds about right. You can technically achieve full torque at 0 RPM, but the current required for that might be very high (and may not be practical for the small size, power limits, and cost of a 3D printer). I know this is a problem in Asian import mini-lathes - if you run the spindle too slow you won’t get much torque out because the driver can’t supply enough current.

I think it would be a better idea to use linear encoders if you were going to add encoders to the shaft, as that way you could directly readout the position of the tool along an axis without issues like backlash that would mess up your calculated position. This is what I’ve seen on (both manual and CNC) mills and lathes.

BLDC and AC servos maintain full torque at stop too, and have about 2-3× the torque of a stepper of similar size.

Huh, this is true about BLDCs as well? I remember seeing in a video that BLDCs tend to have very poor torque output when stopped and especially when at low speeds (due to very low efficiency requiring too high currents for drivers to supply), whereas AC motors have a pretty much flat torque curve until they get fairly fast. I’d be interested to know if this is true.

I believe this is the referenced article:

https://arxiv.org/abs/2311.03348

I’ve been using FreeTube since Piped was very inconsistent for me, but I guess that’s just the nature of these services. I’ll have to check out Invidious again, last time I tried it was several years ago and I stopped using it after the main instance shut down. Is it still under active development? I remember its development status being unclear, partially because the language it uses is not super mainstream, but it’s probably changed since then.

Fortunately, Invidious, Piped, Libretube and Newpipe all exist and work flawlessly so there’s no excuse to use proprietary trash like that.

Isn’t the very point of this post that Invidious and Piped don’t work flawlessly?

Can’t you still modify and distribute Grayjay, just not commercially? I understand that still prevents the app from being considered open source, but their reasoning is valid IMO (to prevent people from making ad-infested clones on the play store, which has happened with NewPipe before).

Not sure what other people were claiming, but normally the point being made is that it’s not possible for a network to memorize a significant portion of its training data. It can definitely memorize significant portions of individual copyrighted works (like shown here), but the whole dataset is far too large compared to the model’s weights to be memorized.

I’m not sure about chamfer (I haven’t used it in my sketches) but fillet is definitely included in the Sketcher workbench.

Entirely fair! I think FreeCAD is still fine for hobbyists like myself though. It does take quite a bit of getting used to (I came from Fusion360 and Inventor first) since it operates somewhat differently, but it’s good that we have at least one option.

Hopefully it’ll see more development and become substantially more viable in the future.

This is a result of the topological naming problem. FreeCAD currently doesn’t handle this well at all. There’s been a lot of work on this front though - you can use realthunder’s fork which should be a lot better in this regard. Alternatively, you can avoid creating features directly on top of other features, and instead make planes and reference them exclusively.

This isn’t true? Fillets and chamfers are available in the PartDesign workbench.

The big thing you get with frameworks is super simple repairability. This means service manuals, parts availability, easy access to components like the battery, RAM, ssd, etc. Customizable ports are also a nice feature. You can even upgrade the motherboard later down the line instead of buying a whole new laptop.

I haven’t read the article myself, but it’s worth noting that in CS as a whole and especially ML/CV/NLP, selective conferences are generally seen as the gold standard for publications compared to journals. The top conferences include NeurIPS, ICLR, ICML, CVPR for CV and EMNLP for NLP.

It looks like the journal in question is a physical sciences journal as well, though I haven’t looked much into it.

I’m curious what field you’re in. I’m in computer vision and ML and most conferences have clauses saying not to use ChatGPT or other LLM tools. However, most of the folks I work with see no issue with using LLMs to assist in sentence structure, wording, etc, but they generally don’t approve of using LLMs to write accuracy critical sections (such as background, or results) outside of things like rewording.

I suspect part of the reason conferences are hesitant to allow LLM usage has to do with copyright, since that’s still somewhat of a gray area in the US AFAIK.

Also one very important aspect of this is that it must be possible to backpropagate the discriminator. If you just have access to inference on a detector of some kind but not the model weights and architecture itself, you won’t be able to perform backpropagation and therefore can’t generate gradients to update your generator’s weights.

That said, yes, GANs have somewhat fallen out of favor due to their relatively poor sample diversity compared to diffusion models.