I found that systemd actually simplified all the things I was doing on sysvinit. BUT, I did hold out until Debian testing stopped supporting sysvinit, and I think waiting gave me a better experience.
With X11 -> Wayland, the main thing holding me back finding a tiling compositor that will work under Plasma and is packaged for Debian and the learning at least the basics. My XMonad configuration isn't that special, but I'm really quite used to not having to re-arrange my own windows, and being able to move/resize/refocus all with the home row and modifier keys. So, I'm probably going to wait until Debian testing ships a Plasma that doesn't support X11, and have to do some learning then.
Maybe it's just the gourmand in me, but this is also my priority. It seems like we should be able to ensure no human goes hungry. Production isn't an issue, it's distribution/logistics.
First 3 replies to this post asking for a guest pass will get one from me. It's a free month of viewing anything through Nebula. (I recommend "The Prince" and "Lynchpins".)
My parents would say you just haven't been hungry enough. Their parents lived through the great depression. I wouldn't know, but I hear people are having to make food/medicine trade offs, which seems more dire than flavor/texture preference tradeoffs.
That said, I don't know a protein source that's as available and cheap as beans, but you might try insects if cheap is the priority or poultry if availability is your priority.
You can buy a large bag of frozen vegetable blend and steam it fairly simply. You can either steam single serving and keep the rest frozen OR steam the whole bag in bulk, and refrigerate for up to a week, reheating single servings as you need them.
I have always hated soaking beans, which is why the Instant Pot has been one of the single greatest cooking inventions I have ever used.
Exactly why I bought mine. Any pressure cooker will do. Beans (red, pinto, or black) 1 : 2 with water for 40 minutes, followed by natural release. I use roughly a pint of dried beans (1lb bag, then topped up out of a mixed-beans bag), to get 9 large servings.
I also do quinoa in the same cooker 1 : 5/4 with water (or sub up to half the water with stock) for 0 minutes (just bring up to high pressure), followed by natural release. I use 3 cups dry to make 9 servings.
Depending on your spice budget, you might feel like you are getting more by applying right before eating. But, if you want the spice flavors to permeate the beans, it's best to add them to the pot and warm them just a bit with the saute setting before adding the beans (or quinoa/rice/grains) and water.
If you eat meat, miscut ham is also a good addition to the beans before cooking -- they will share lipids and flavors.
I use nooch as a topping for mine, to try to keep it vegan, but what I really like is a Mexican shredded cheese blend.
Sure, eventually, I'd like a language with Haskell-ish syntax to compile to Linux x86_64 and webassembly and use the language to make better software. If my language existed today, I'd probably work on writing my own ActivityPub software, and improve/port https://github.com/NARBEHOUSE/Ben-s-Software- because my father might want it soon.
https://gitlab.com/bss03/grtt is my published code. But, I have far more intuitions that I need to write code for than finished code.
While evaluating something well-typed under a context, the heap: does not need to contain a value for a binder with modality 0, must contain a single, strict value for a binder with a modality 1, must contain a single, lazy closure for a binder with a modality of ?, must contain multiple references to a shared, strict value for a binder with a modality n, must contain at least a single reference to a strict value for a binder with a modality of +, must contain at least a single reference to a lazy closure for a binder with a modality of *. Since the typing rules propagate the modalities to subterms precisely, we should be able to identify the exact point a closure must be forced to a value (or dropped) before runtime. That's in addition to being able to compile linear functions to heap updates, eliminating at least some allocations.
There's some similarities with both the exact-use-count and relevant-or-erased semirings, but I think some things (e.g. around sums) are hard/awkward/impossible to type and the ?/+/* modalities make some make things easier while still allowing the abstract machine to know exactly when to "optimize the heap" based on a runtime flow that "activates" a particular static analysis.
Of course, it's still MLTT "compatible" -- anything that would type-check in MLTT should type-check in my variation of GRTT by "simply" using the * modality everywhere -- so you get full proofs-as-programs and a total language.
I'm probably a bit off in the weeds, but it still makes my brain buzz to think about and occasionally I'll make progress. I've been a little bit distracted with https://gitlab.com/bss03/nested which should allow me to write the abstract machine as a fold, but as proven to be place I can also put a lot of programming time into (again, with sporadic real progress).
I think co-ops are the way to go, but I can understand that someone "just" wanting to purchase the good/service might not see the difference between a co-op and corporation like Amazon.
I don't think it's a size issue really, but co-ops generally stay smaller in part due to how they are internally organized compared to a "median" corporation.
I also think that the government actually does a pretty good job at managing things; it's just their failures are public. Private boondoggles might drive many people into bankruptcy, but they aren't publicized any more than absolutely necessary.
Same here, but I can understand why someone might want to. For many people, even those that are comfortable on the command line, a GUI is a more comfortable experience. And, I have (rarely) needed to do some filesystem management as not my primary user account.
Do you have sugar diabeetus? I want to send you important information you should know!