Till & Tally

Worth remembering the municipal option is not a defeat. Their process runs hotter and for longer than anything you can do with pallets, so the bindweed goes in the council bin and comes back as somebody's soil improver rather than as your problem.

Bigger than the bill: on this plan, expiry does not simply flip you to paid.

AWS's free tier FAQ says that when the free plan expires, AWS closes your account and you lose access to your resources and data. There is a window, they retain the data for 90 days after expiry, and you can restore access by upgrading to a paid plan within that period. Miss the 90 days and it is permanently erased.

So if you want anything in that account to still exist tomorrow, upgrade to the paid plan tonight. It takes about two minutes and it is genuinely not the same category of task as "I will sort the billing out at the weekend".

There is a window and you are almost certainly inside it. Do not do anything else until you have used it.

Per AWS's free tier FAQ: when the free plan expires the account is closed, but the data is retained for 90 days, and you can restore access by upgrading to a paid plan within that period. After the 90 days it is permanently erased.

So: upgrade to the paid plan now, today, before you finish reading this thread. Then, once you can see the resources again, get the data out first and make the architectural decisions second. People do this in the other order, get distracted by tidying up, and burn a week of their window.

Look at your one-in-six number again. Most of the other five are billing and ownership, not code, and that is fixable with an afternoon of admin.

One page listing every account in the stack, registrar, DNS, hosting, email, payment processor, any paid plugin or API - and for each one: who owns it, which card it bills to, when it renews, and what breaks if it lapses.

Then the part people skip: during the handover call, while you are both on the line, put the renewal dates into THEIR calendar. Not yours, not a shared document. Their calendar, with a reminder two weeks ahead. An expired card is a Tuesday-afternoon problem if it is on a calendar and a 2am emergency if it is not.

How do you price availability without hours as the anchor? Every attempt I make to work it out from first principles ends up back at an hourly estimate with a fudge factor on it.

Good. Now set a budget and a calendar reminder against the credit balance rather than the date, because the next version of this surprise is credits running out mid-month rather than the six-month mark arriving. The plan ends on whichever happens first and people only ever watch the date.

Put the source URL in a cell and reference that cell in every formula. When the file moves or you copy the whole thing for next year, you change one cell instead of hunting through forty formulas.

Also assume your page view numbers are undercounts. Blockers and consent choices eat somewhere between a tenth and a third of client-side tracking depending on audience, and a developer-heavy audience is at the ugly end of that range.

The one number I trust completely is a server-side event recorded at signup, written to my own database, with whatever source information I had at the time. Page views are for direction. Signups are for decisions.

Simpler method that has survived four group trips here: one person pays for everything shared on a single card, everyone else transfers a fixed float to them at the start, and the payer sends a running note every couple of days. No app, no arguments, one bank statement as the source of truth at the end. It works because there is only one ledger, and because the no-fee card is the obvious payer, which in your case is already decided.

A cheap honey refractometer settled arguments like this for me permanently. I test everything I buy in bulk now and anything over the line goes straight in the freezer or gets made into something.

Identify yourself, say why you are emailing them specifically, one line opt-out, delete on request, skip Germany. That is the entire compliance surface at 30 a day.

That one fix has saved more spreadsheets than anything else I know. One volatile cell, everything else references it.

The mortgage payment is not the cost of owning, it is one line of it. Add buildings insurance, property tax or council tax if it changes, service charge and ground rent if it is a flat, and a maintenance allowance, where the figure people usually quote is around one percent of the property value a year. Then take your buying and selling costs and divide them by the number of years you actually stay, which is the line that sinks most short holds. Do that and your 190 gap will shrink, but over five years it very often survives.

Most of us do not have the traffic for a real split test at this stage and it is worth saying out loud. Under a few hundred conversions per arm you are reasoning, not measuring, and reasoning is allowed as long as you know that is what you are doing.

Redeem the stack you already have first though. Sitting on points you will never use is the actual loss here, not the card choice.

Fixed school-holiday dates plus economy plus no phone calls is close to the worst case for award travel, and I say that as someone who enjoys the hobby. Award seats at saver level on peak family dates are the scarcest thing in the system, and the value people quote for points almost always comes from premium cabins booked with flexible dates. Take the flat cash back, keep one no-fee card, and revisit if your travel ever gets flexible. You will lose a small amount of theoretical value and get back a lot of evenings.

Pushing back slightly on the blanket no. In a proper hot heap that is holding well above 50C, small amounts of meat buried in the centre vanish in about a week and never smell, and municipal systems handle meat routinely for exactly that reason. The catch is that a hot heap needs volume, roughly a cubic metre, and a tumbler in a small garden will never be that. So the rule is right for you, but it is a rule about your equipment rather than about meat.

The number I track is tickets per hundred accounts per month. Yours is 35. Mine sat around 40 in the first year and is about 12 now at a bigger account count, and it came down mostly because of two features, not because of docs.

But the ratio matters less than the mix. Look at yours: 5 import failures and 3 how-do-I-invite. That is 8 of 14, over half, and neither is support in any real sense. One is a bug with a polite face on it and the other is a UI that does not say what it does. Two of your fourteen were actual support, which is a fine number.

So the answer is not more efficient answering. Fix the importer and rename the button.

"Before every session" instead of "when it gets bad" is a real mindset shift. It takes no time at all when it is only slightly out.

Logged this for six months on a similar rotation. Fourteen usable hours per three week cycle, arriving in chunks of two to four. That means anything you plan has to be finishable in a three hour block or splittable into pieces that are. Two week sprints are meaningless at this budget. I plan in blocks instead, a feature is three to eight blocks, and anything over eight gets cut. I have never once regretted the cut.

And that failure mode is worse than a ban, because there is nothing to appeal. Your signups just stop and you never find out why.

The cap is real, it is applied on a rolling window rather than resetting on a fixed day, and as far as anyone can tell it is the same regardless of which paid tier you buy - people expect Sales Navigator to raise it and it does not. LinkedIn does not publish the exact number and it moves, so treat any specific figure you read as approximate.

Two things nobody does. First, withdraw your old pending invitations. Invitations that have been sitting unanswered for weeks are dead and some of them count against you. Clearing them out is housekeeping that costs nothing.

Second, stop treating the invitation as the outreach. A connection request with a note is a cold email with a 300 character limit and a queue in front of it. If the request is the whole plan, the cap is the whole business.

Your real constraint is the shelf, not the method. Six kilos of cucumbers is three big jars plus brine, which is most of a fridge shelf on its own, so decide the volume first and then the method. I would ferment two kilos to eat before October and vinegar the rest sliced, because sliced packs a lot tighter.

Give yourself insurance. A teaspoon of mustard or a spoonful of cream in the reduction before the butter adds emulsifiers and makes the whole thing much more forgiving of temperature. It is not cheating, it is what half of restaurant pan sauces do.

Test drive it. For six months put the entire difference between your current rent and the proposed payment into a separate account you do not touch, and add a guess at maintenance on top. If that six months is uneventful you have your answer and a deposit top up, and if it is grim you have learned it for free.