Mutiny

kip/preview/tune what I'm listening to is nowhere near as powerful as I'd like. I've written library curation tools in the past, these always expected me to spend significant dedicated time in front of a screen (e.g. a similar tool like the cool looking https://github.com/kristopolous/music-explorer, I think). This has strong parallels to how older people consumed music - either totally passive curation (radio), or very deliberate, like finding music in record stores, at a friend's place. Also replay involves selecting records/CDs in your own bookshelf. Today's ephemeral digital libraries are much lower effort, are huge and curation/selection tools are not easy enough to use, so I tend to fall back onto old favourites or recommendation engines that usually don't satisfy me. A solution might be a much more configurable curation assistant that is also super easy to use (and, in my case) very accessible on a mobile device with 0-1 clicks (because I'm busy doing other things). Music discovery tools that don't allow in-situ music playing is thus also a no-go. It wouldn't be super hard to build an interactive tool, but as always, making a super intuitive and useable UX experience is the hardest part. Most streaming tools are giant swiss-army knives for listening use-cases.

m/kristopolous/music-explorer/tree/master/t... You'll see many languages in there but you won't see posix sh. It's slow, it's clunky, it's a pain in the ass. If you don't like my practice then I guess don't use it. I've been using/developing these particular tools nearly every day for over 3 years and it works well for me. I'm not going to say bash is awesome but it's pretty great for programming. I use zsh as my interactive shell though, it's miles ahead of bash in that department. Watching people who really know what they're doing in zsh is a total mindfuck.

tend to some sprawling project Example https://9ol.es/pl/

tening to them. Essentially it's designed to do a single pass. There's a REPL and it interfaces with mpv using their JSON and Lua interface. Then you rate each release and the adjacent information, such as the people working on it (on its discogs catalog page) becomes another vector for possible traversal. It's not a constant scraper, it'd go on forever. Essentially it's manually run and there's a tweakable threshold. For instance. Say you highly rate multiple releases that share a producer, artist, and label. Well, by golly, if there's a few other releases in that grouping you'll probably like that too. I mean, duh, obvious, right? A lot of this approach came from the world of italo disco where the real names often got obscured behind short lived projects and labels which required a bit of digging to find the related content. Take Aldo Martinelli for instance. Here's his fb banner which is the best catalog of his work I know (https://scontent-lax3-1.xx.fbcdn.net/v/t31.18172-8/fr/cp0/e1...). Same person doing all of that. For instance, "topo & roby" and "moon ray comanchero" are essentially the same people, but getting from one to the other or uncovering the relationships, that's the challenge this infrastructure tries to solve. This stuff gets unusually complicated with specific regards to italo disco. See the profile of radiorama for instance (https://www.discogs.com/artist/73973-Radiorama) or groups like The Flirts, Fun Fun or Den Harrow, where models were the faces, singers were the vocals, different song writers and musicians would come and go, they were all huge tangles. This method completely ignores popularity or what others are listening to and falls back on the networks of the people making the content. For me at least, it's produced great results. But it's built exclusively for me at this time. I don't even have most of the traversal code posted The code is over here https://github.com/kristopolous/music-explorer I'd be happy to make it more accessible but I'm multiple standard deviations from the mean when it comes to how I approach music so I've got no idea whether there's a large interest here If you want to see what results look like, now this is exclusively catered to my tastes, I have a pretty basic hack to listen to things on the go over here http://9ol.es/pl/ I guess the focus here is to see the label/artist diversity even if the style of music isn't your thing. --- As far as listening while you work, the tool is instrumentable. For me I can press a quasimode key (say the windows key) along with some other key and do things like pause/play, label the release, track/release navigation, get info, open the webpage that it was on, etc. These are hooked at the window manager level. This way there's as little task switching as possible and I can stay in whatever application I'm doing.

ity was the first application of WIMP. It makes the goal of intuitively navigating multidimensional problems through geospatial and tactile interfacing less abstract. The actual script is here https://github.com/kristopolous/music-explorer/blob/master/t... I was exploring everything from browser tab controls to instrumenting software debuggers to window management with this thing so it's kinda a mess. There's a lot of code scattered in my various projects on this topic The biggest UI fumble is the value repetition and tactility is great until the context of the control changes and then you get a discontinuous jump. The "obvious" workaround of using relative values has the random walk problem. You need a disengagement mode to reset it from wandering off to one of the edges. That's a flow disrupting cognitive shift (https://en.m.wikipedia.org/wiki/Cognitive_shifting) I tried using 2 cheap arduino joy controllers (example https://satkit.com/image/cache/catalog/products/arduino-joys...) and allocate one joy for mode and the other one for value (inspired from orbitouch https://c1.neweggimages.com/ProductImage/A446_13123722022688...) it works fine with logarithmic relative values. (Going to value 0 and max are common) Also both these models are about giving the thumbs more jobs then just a single key, an idea I stole from Maltron (http://www.micwil.com/images/gallery/maltron_keyboards_dual_...). Sliding a control right next to a conventional keyboard base should be doable on the cheap and help get towards the goal. The joys were a better solution for this but then I'd have to "sell" something which I'm trying to avoid. My earlier method involved a USB foot pedal which acted as a quasimode for control (example: https://www.geeky-gadgets.com/wp-content/uploads/2010/01/usb... ... it can be assigned to something like "Hyper" in Xorg parlance, see here https://github.com/kristopolous/tools/blob/master/bin/footpa...) and then my lua scriptable window manager captured the key strokes and delegated things accordingly (example: https://github.com/kristopolous/crappy-notion/blob/master/mo...). It's still just "buttons" at the end though and this whole project is about integrating sliders knobs and joys to general computing. I just hope to get everything done satisfactorily with the midi, I don't have to manufacture anything and it's a far easier technical investment for casual people and the device in the video is only $40. I don't want to "own" anything here. I want people to liberally iterate and innovate as much as possible. I'm certainly not the smartest person in the room.