Tessa Ondrak

i timed this on ours by wrapping four phases with timestamps: process init, module imports, first db connect, first query. on a monday the connect was 2.8s of a 3.9s request and everything else was unchanged from a weekday. thirty minutes of instrumenting saved a week of guessing at bundle size.

Local android builds are also dramatically faster once the gradle cache is warm. First one took me 11 minutes, subsequent ones under 3.

The mechanism that stops this recurring is a change note, not a better contract. Every time something new comes up, you reply the same day with one line: 'happy to add that, it's about 4 hours and pushes delivery to the 20th, confirm and I'll start'. It feels pedantic for the first two and then it becomes the normal rhythm of the project and scope creep stops existing, because everything is priced at the moment it's asked for rather than at the end.

It gets everyone. Also daytime electricity, more laundry, more coffee made at home which sounds cheap until you count the beans.

This is the single most common surprise I hear and it happened to us at almost exactly the same magnitude. The cause is that work was subsidising your life in ways that never appeared in a budget: a subsidised canteen, a phone, a laptop, someone else's heating between 9 and 5, and a commute that kept you out of the shops. Ours settled about 8% above the working-year baseline by year three, not back down to it. The good news is it's a level shift, not a trend.

Do the arithmetic on your own numbers rather than on anyone's valuation chart. 2k a month at a flat 2 percent is 480 a year, guaranteed, no expiry, no seat hunting. To beat that with points you need to consistently redeem above about 2 cents each after fees, which in economy on peak dates is a stretch: most of my economy redemptions land between 1.1 and 1.5, and the ones that made the hobby look good were long-haul business seats I booked eleven months out. Your constraint is the calendar, not the card.

Does the removable liner go in with it or separately?

Four years out. I keep roughly two years of bare-bones spending in a money market fund and another three years in short duration bonds, everything else in equities. In practice I refill the cash bucket from whatever went up that year, which some years is dividends and interest alone. In the one genuinely ugly year I didn't sell equities, I just didn't refill and let the cash run down to about 14 months, then topped it back up the following spring. Worth saying I'm not an adviser and my situation has a paid-off house in it, which changes the numbers a lot.

Yes, by about 12%, mostly by not travelling. Wasn't painful because we'd already decided in advance which line items were the flexible ones.

For next time: 50% up front, and put a late fee clause in the contract even if you never charge it. I've invoked mine twice in six years and both times the mere mention of it moved the payment forward two weeks. The clause is a lever, not a revenue stream.

One more consideration nobody raises: if you are ever in an incident with another vehicle, what you were wearing tends to come up. Not legal advice and it varies enormously by country, so ask someone who knows your jurisdiction, but it is worth being aware that the choice is not purely about your own skin.

Sell them a fixed price block of time rather than a fixed scope. Six weeks at two days a week, invoiced monthly, with a written list of what you expect to accomplish that's explicitly a plan and not a promise. They get one number for their approval process, you get paid for direction changes instead of eating them. I've run this with three clients who couldn't hold a spec still and it removed every argument we would otherwise have had.

Method book first, reference second, and use a library for the OED. That order saves you the most money and the most embarrassment.

Three years tells you almost nothing about a 26 year projection, and if you rebuild the model every quarter you'll coast, uncoast and recoast yourself into a mess. What I'd do instead is set a review date once a year and a single decision rule: if actual balance is more than X% below the glidepath at review time, resume contributions at some fixed amount until it's back on. Written down in advance, checked once, ignored the rest of the year.

I spent about eight months chasing the score and ended up sleeping worse, because I would look at the number before I was properly awake and then feel tired on the strength of it. Two friends have described exactly the same loop. I took the ring off for a month, slept the same, felt better, and now wear it a week a month at most.

The fun part being the part you give away for free is a very common freelance failure mode. Worth watching for it in proposals too.

The way I framed it was: the mortgage isn't competing with the market, it's competing with the size of the hole I have to bridge. Clearing it cuts your required annual spend for the gap years, which shrinks the bridge on both sides. I paid mine off at a similar rate and the expected value was slightly worse, but my bridge went from seven lean years to seven comfortable ones, which changed whether I could stop at all. This is what I did, not advice, and a planner would probably tell you the spreadsheet is right.

About nine months for me, and the thing that shortened it was setting up an actual monthly transfer from the cash account into my current account on the same date the salary used to arrive. Sounds silly. It completely rewired the Friday feeling, because money still showed up on a schedule and my brain stopped treating every expense as a withdrawal from a finite pile.

19 out of 40 at eight months is completely normal and honestly not bad. Six years in I average 24-26 in a good month and I've stopped trying to push it higher, because the non-billable half is what generates the billable half. The number that matters isn't utilisation, it's whether your rate times your realistic billable hours covers what you need with margin. If it doesn't, raising the rate is a much easier lever than finding six more hours in a week.