Dee

If you're going to keep taking them, at least take one at a time for a month each so you can tell what does anything. I ran mine as a six-item stack for a year and learned exactly nothing.

Get the return conditions agreed before you go back, in writing, with someone above your manager or HR in the room. Reduced hours for a defined period, no reassignment after 4pm, and an actual answer about the two vacancies. Not because all of it will be honoured, but because when it isn't you have a document to point at instead of a feeling. I skipped this and spent four months arguing about what had supposedly been said.

This is the real limiter for most people. The applications are never the bottleneck, the spend is.

Numb arm is not always the pillow. Mine stopped when I put a second pillow between my knees, because my top leg was rolling forward and torquing my shoulder into the mattress. Cost me nothing to test and it changed the problem completely. Worth ruling out before you spend seventy pounds.

Terse: hardwire with a voltage cutoff, impact-triggered parking mode, high-endurance card, and check the card monthly. The last one is what everybody skips and it is why so many people have a camera that recorded nothing.

Started doing the sheet the day it is set rather than the night before it is due purely so this trick is available. Biggest single grade change I made.

Small practical thing: check whether your tariff requires a working smart meter in a specific mode and whether yours is actually sending half hourly reads. Ours had gone dumb after a signal change and the switch sat in limbo for six weeks. Worth confirming before you apply rather than after.

Good call. We do have a policy document that I have never once opened.

The bot showing up in the participant list is precisely what I was worried about.

Also type the code, do not copy paste it, and once a section is done close the video and rebuild the same thing from memory. It is slow and it feels like a waste of a precious hour, and it is the only thing that moved retention for me.

The having somewhere the department does not reach point is underrated. When your work life and social life are the same eleven people, a bad meeting ruins everything.

Budget for review time, not just build time. The build is 15 to 25 minutes on a good day, the first TestFlight processing is another 10 to 40, and getting the first external testing group approved is a separate wait. My first ios release slipped four days entirely on things that were not compiling.

Month eight being the first good stretch you noticed. Four months in, I've been treating every bad day as proof it isn't working.

I have a reading list about burnout. I am currently looking at it.

I did it wrong first: thick hiking socks, because thicker seemed better. I woke up at 2am sweating with the duvet kicked off, twice a week, for a month before I connected the two things.

Switched to a thin merino liner sock and it's been fine for two winters. If you're going to try this, go thinner than feels correct. The point is the first thirty minutes, not insulation all night.

Mild disagreement with the framing that the social side is a problem to solve. Treat it as a job, turn up, do good work, go home, and the cohort question quietly evaporates. I did mine in my thirties surrounded by people ten years younger and the friendships I made were with two postdocs and a technician, all of them roughly my age and stage, none of them in my cohort. The cohort is an artefact of admissions, not a friendship group you owe anything to.

This is the best advice in the thread. The unspoken version of that dynamic destroys more of these than money does.

Careful with giving the range first, though - I did it for a year and anchored myself low repeatedly, because I quoted the band I was comfortable with rather than the value of the work. Where it goes wrong is projects that are worth a lot to the client and simple for you. I now ask about the outcome and what it is worth before any number of mine goes on the table, and I have had two projects come in above what I would have quoted blind.

Send the recording and stop scheduling anything for a while. Record the walkthrough once, keep it under four minutes, and reply to every cold conversation with the ninety second section that matches what that person complained about. It converts worse per viewer than a live call and enormously better per hour of yours, and at your stage your hours are the scarce thing.

Seven months of funding is a lot more time than it feels like right now. I wrote 60% of mine in the final nine weeks after five months of nothing. I don't recommend that route, but it is evidence that the state you're in is not the state you'll be in.

Whatever you choose, group the destructive work into one window. Rewiring means chasing walls and lifting floors, replastering means dust everywhere and everything moved out, and doing those in the same eight weeks is dramatically less miserable and cheaper than doing them eighteen months apart. If you can possibly stretch to wiring and plaster together and leave the windows for later, that is the combination that saves the most, because the plasterer follows the electrician and makes good behind them.

Half never coming back after day one is your entire problem and it happens before pricing is even relevant. You have an activation problem, not a conversion problem.

Define one action that means someone has actually used the thing, the moment where they see it work on their own data. Then measure what percentage reach it on day one. If that number is under about a quarter, nothing downstream matters.

Concretely I would do three things this week. Watch five session recordings end to end, painful but the fastest diagnostic there is. Email every one of the 20 who bounced with a single question, what were you hoping to do when you signed up. And put a real human onboarding call on the confirmation page for the next 20, even if it does not scale, because you need to hear where they get stuck in their own words.

The Maslach research is what most of the popular books are sitting on top of, and it's worth knowing the three dimensions: exhaustion, cynicism and a collapsed sense of your own effectiveness: because once you can name which one is loudest, the advice sorts itself out. My exhaustion was honestly fine. My cynicism was off the scale. I'd spent six months reading rest advice that was aimed at a problem I didn't have.

One addition to the taper: keep the timing constant while you reduce the dose. Changing both when you drink it and how much at once makes it hard to tell what is causing what.

Whatever you decide, put the runtimeVersion into your release checklist as a line someone ticks. Every one of these incidents I have seen came from a native change shipping under an unchanged runtime version.