Admittedly the game probably has to have a devlog of some form, otherwise this question is synonymous with "what games are you waiting for?", which is far too generic a question.

The game I'm most interested in is Songs of the Eons (itch.io, most active), which wants to be like Dwarf Fortress but a grand strategy game (main post, for context). Obviously they've noted that such a project will have Dwarf Fortress esque development, so waiting is very much in order.

There's also Hyperbolica (trailer) which I've posted about here somewhat.

There's also Hytale, Creeper World 4 and to an extent KSP 2, which I like but don't follow that regularly, for not much reason TBH.

I'd probably also follow Deltarune but Toby doesn't seem to really do dev updates too often.

So are there any games that you're following as they're developed?

Hi! I've heard tilderinos talking about the gemini-verse on some other posts; I tried it out this evening and it honestly felt strange browsing in terminal and even stranger navigating the web without search engines. I was wondering if anyone had a gentler introduction than the official site? I feel like I've got a ship, but no map to this new verse.

So I thought I'd try a little experiment here. Here are 2 songs I like from the same band. They're very different songs, and here's why I like them:

I Got You - This is a song about infatuation. It conveys that feeling I had when I was smitten with someone as a teenager. It feels very intimate to me, like the singer's directly expressing his innermost feelings to the person he's infatuated with. Or perhaps thinking of what he would say if he had the guts. It's very much a new wave pop song, and is probably the most well-known of Split Enz songs, at least where I live in the US.

I See Red - This is very much a song of rage. To me it's about a guy who's been dumped, or maybe who was infatuated with someone, and now they're with someone else. Whereas "I Got You" was very poppy, this is more punky. (I mean it's still pop, but with a punk flavor.) Putting it together with the previous song makes a lot of sense to me, even though they have such different tone. It's like the 2 songs together tell a story. I also love the phrase "down the drain like molten toothpaste." There's just something so illustrative about "molten toothpaste."

Anyway, just thought I'd share these random thoughts.

As life slowly returns to normal in the UK, I've felt the need to look after my fitness more. I lacked discipline throughout lockdown to workout at home and keep my fitness. As a result I've got a nice COVID-gut, and my endurance and strength are shot. I swam regularly before quarantine hit, at least 4 times a week, and I'm keen to get back into that. But I'm also looking at picking up a martial art, for a more intense workout and fitness, as well as just to pick up a new skill. However I have no idea where to begin with martial arts, so I figured I'd start a thread for some inspiration, and go from there.

So are any tilderen martial artists? If so, tell me about it!
What do you practice?
How long have you done it?
How does it benefit you?
Do you attend classes or practice solo?
Would you recommend your martial art to a beginner?

As part of a weekly series, these topics are a place for users to casually discuss the things they did — or didn't do — during their weekend. Did you make any plans? Take a trip? Do nothing at all? Tell us about it!

Suppose I am trying to iteratively produce a completed image from some subset using a combination of convolutional/DNN methods. What Image norm is best?

The natural (for me) norm to ascribe to an image is to take the bitmap as a vector with L₂. If the input image is anime or something else, the uniform coloring makes this very likely to be a good fit in a low dimension - that is: no overfitting.

However: pictures of fur. Given a small square, the AI, set to extrapolate more fur from that single image, should be expected to get that stuff right next to the given subimage right, but further away, i want it to get the texture right, not the exact representation. So, if the AI shifts the fur far away from the image left by just the right amount, it could get an incredibly poor score.

If I were to use the naive L₂ norm directly, I would be guaranteed to overfit, and you can see this with some of the demo algorithms for image generation around the web. Now, the answer to this is probably to use a fourier or a wavelet transform and then take the L_N norm over the transformed space instead (correct me if I'm wrong.)

However, we get to the most complex class: images with different textures in them. In this case, I have a problem. Wavelet-type transforms don't behave well with discrete boundaries, while pixel-by-pixel methods don't do well with the textured parts of images. Is there a good method of determining image similarity for these cases?

More philosophically, what is the mathematical notion of similarity that our eye picks out? Any pointers or suggestions are appreciated. This is the last of two issues I have with a design I built for a Sparse NN.

Edit: For those interested, here is an example, notice how the predictions tend to blur details

I know the site is still in its infancy and cross-posting won't be much of an issue at the moment, but I was interested to see what other users thought about cross-posting, whether we should allow it and if so how it should be done?

Personally I am in favour of cross-posting but I think some site mechanic should exist that doesn't allow two separate threads to be created. Instead, the cross-post should link directly to the original thread so that discussion of the topic can be kept in a single location but the topic itself can reach multiple tildes. For example, say an article about music being created artificially by a robot was originally posted in ~music. Someone may want to cross-post this to ~tech, and to do so would only have to click some sort of cross-post button and select the tilde they want to cross-post to. Anyone browsing the ~tech tilde would see the post, but upon clicking it would be taken to the comments page of the post originally made in ~music. Some indication of where the post was originally made could be given as well when viewing the cross-post on another tilde.