Val

Windows plus iPad is the awkward combination but it's survivable if you accept the iPad is a capture device and something else is the archive. Mine ends up as PDFs in a plain folder that syncs, and I search the folder rather than the app. Slightly worse search, dramatically better sleep, and no vendor holds four years of my degree.

Cross system rather than model count is a much more useful test than anything else I have read on this.

Agreed on ordering. Frequency-ordered and radical-ordered both work, they just feel different - radical order feels slower for the first two months and pays it back later.

Four months is early. The syntax that sticks is the syntax you have typed a hundred times, and there is no shortcut, only repetitions. If you want it faster, build things that reuse the same constructs deliberately rather than jumping languages.

Right. Android plus Windows is the entire problem. On all-Apple or all-Google this thread would not exist.

I do first-round screens for a small backend team. Last cycle we had 340 applicants for one junior role, phone-screened 22, and made one offer. The thing that killed nearly everyone was the same: they could describe the project on their CV but could not say why they chose the thing they chose, or what they tried that did not work. That question does not need you to be fast, and no assistant can answer it for you in a live call, which is exactly why it filters so hard right now.

Worth asking your lecturer how long they expect a sheet to take. Mine said six hours for six problems and half the room visibly relaxed.

The plain folder is a real answer and not a defeat. My notes are markdown files in a synced folder, opened with whatever editor is nearest, searched with the operating system. That does 90% of what a note app does with zero configuration surface to get lost in. What you give up is backlinks and a graph view, and I have never once needed a graph view for anything.

Compose until failover actually matters. Then decide whether you want to run a cluster or pay someone to run one.

Bike days are three of the five. I might be buying the wrong bag for the wrong days.

Also accept that you'll be wrong out loud for months. Germans understood me perfectly well with wrong cases. My fear of getting it wrong slowed me down far more than the wrongness ever did.

Course completion is a metric nobody will ever ask you about. The thing you built is the thing you talk about.

The real cost of free tools is setup time. The flashcard route means you build or vet your own decks, and shared decks are wildly uneven: I lost two weeks to a badly made one before somebody told me to make cards from sentences I'd actually read. What paid apps sell you is 'don't think about it', and that's worth something if thinking about it is the reason you'd quit.

Plain markdown files in a synced folder, six years. Gave up: pretty rendering, mobile editing that doesn't feel like a compromise, and any kind of database view. Still annoys me: mobile, badly. Editing markdown on a phone is bad in every app I've tried, and I have tried a lot of them. I keep it because I've watched four friends migrate off things in that time and I have never had to.

Then keep something with a streak and stop paying for it. The free tier of most of these still has the streak. What you're paying for is no ads and unlimited mistakes.

The chicken breast fatigue is a preparation problem more than an ingredient problem. Thighs instead of breast, cooked hard in a pan, are 24g per 100g and taste like actual food. Same with pork loin, turkey mince and whitefish. I rotate four proteins on a fortnightly cycle and the dread went away entirely.

so i never open the developer portal by hand at all?

the bundle id point is the kind of thing i would have discovered too late. changing it later is a new app entirely, right?

Cloud editors in the browser cover most of the first six months, and the free allowance on the hosted ones is usually enough for evening study if you remember to stop the machine when you close the lid. Two things to know before you commit: your session can go to sleep mid-exercise and you lose whatever was only in memory, and file uploads and downloads are fiddly enough that any tutorial involving a local CSV becomes a small chore. Neither is fatal. Both are daily friction.

This deserves more agreement than it will get. The useful sleep interventions are all things you already know, and none of them require a sensor. The tracker is useful for about a month while you connect behaviour to outcome, and then it is a subscription to being judged.

Two separate ideas are colliding here. new = old does not copy anything, it makes a second name pointing at the same list object, which is why the first version surprised you. old.copy() makes a genuinely new list, but the elements inside it are still the same objects as before, so a list of dicts gives you a new list holding the original dicts.

For your case:

import copy
new = copy.deepcopy(old)

That recursively copies the nested structures. Downside is it is slow on big data and it chokes on things like open files or database connections. If your dicts are plain data, deepcopy is the right call and you should stop thinking about it.

That is still a shallow copy, identical behaviour to .copy(), so it does not solve the nested dict problem they described.

Fourteen days with no card is also long enough for the trial to disappear from someone's memory. Most people evaluate on the first day or never. Shorter trial, or trial that starts when they hit the activation action rather than when they sign up, converts noticeably better in my experience.

Almost certainly a missing return inside getUser. If you wrote

async function getUser() {
  fetch(url).then(res => res.json())
}

then the function returns undefined immediately and the fetch chain runs off into the void. Add the return, or better, rewrite it consistently:

async function getUser() {
  const res = await fetch(url)
  return res.json()
}

The other classic cause is calling it without awaiting somewhere further up, for instance in a .map() where each callback returns a promise and you get an array of promises. Promise.all fixes that one.

Worth saying that this is what worked for me rather than a rule. If you have logged carefully for a month, including weekends, and genuinely nothing is moving, that is a reasonable point to talk to a doctor or a registered dietitian rather than cutting further on your own. There are ordinary medical reasons for it and a forum cannot check them.