Ah yes… and if the result was close but happened to be the one the Establishment wanted, do you think they would have called for a confirmatory vote just to be sure? Of course not.

> That also fits with the second one being emphatic.

The Playbook says spend more on comms, emphasise Project Fear, and call for another vote. Repeat until you get what you want.

> You're allowed to think that Lisbon warranted referenda in the member states

How very gracious of you…

> The government didn't take a blind bit of notice.

I think you’ll find that Parliamentary votes were required for the action to take place, and there were three years of deadlock during which the majority of MPs supported remain (an inverse of the popular vote) and certain MPs like Benn and Grieve led to legislation that made it very difficult to negotiate in the UK interests (no deal off the table, so a weak bargaining position).

Article 50 may have been triggered the year after the referendum, but the UK didn’t actually leave until 2020.

> Not by a mile, lol.

LOL indeed. 33.5 million people cast a vote, which was 72% of all people registered to vote. That doesn’t sound “pathetic” to me, unlike your comment in general. It reeks of someone who loses interest in democracy when it doesn’t align with what they want.

It might be worth examining the word “countries” there.

Both France and the Netherlands rejected the proposed EU Constitution by referendum in 2005. It was then regurgitated as the Lisbon Treaty (with only superficial changes) in 2007, which was ratified with no public vote.

The Irish people initially rejected both the EU-empowering treaties of Nice and Lisbon, and a followup vote was considered necessary. You get two bites of the democratic cherry if you have enough power.

A majority of the British people voted to leave in 2016, and in the three years that followed everything possible was done to reverse the decision.

You might be spotting here a difference in desires and power between the governors and the governed.

For the much harder Lorenz cipher used by German High Command from 1940, the Colossus machine was developed by Tommy Flowers at the GPO and became operational in 1944.

None of which involved the US Navy, which was my original point.

Sounds like typical US revisionist history.

They developed ASDIC? HF/DF? Hedgehog? Even the depth charge?

No, that was all the British.

I would say technological development plus the Enigma decrypts were the biggest factor.

The first is burn-in. I now move windows about when I can, and also tend to use Dark mode for everything. I also set up BetterTouchTool to give me 10%, 20%, 30% etc brightness mapped to Cmd plus the numeric keypad. I use a comfortable level of brightness, and only bump it up for a short while when I need to examine something. I also have some anti burn-in videos downloaded with yt-dlp to clear burn-in when I find a ghost of it on the screen.

The other problem is screen activation. There is no power button on the LG displays. I like to keep the Macbook Pro in sleep mode when not in use, so I can resume with my app windows in the same state from the previous session. My whole desk setup is behind a smart switch. Often, when resuming from sleep, one of the displays would not wake and the only workaround was to unplug and replug the thunderbolt cable at the Macbook side. I work around this by having the Macbook continuously powered by my Orico dock when in use or sleeping, and this seems to address the problem. There is also an occasional issue of a single display switching off when I am using it, and this just means a thunderbolt unplug and replug at the Macbook side. But fortunately this has been fairly rare.

So, visually great hardware that I’m thankful for, but Apple’s “it just works” doesn’t really apply here. Just mentioning these in case someone is experiencing similar issues.

This was a temporary rate limit and it told me to try again at a specified time which wasn’t far away. That’s different.

However, it could have been because of a temporary capacity issue. I hope so.

I subscribe to the AI Pro plan. I knew of a published limit of 100 Pro prompts per day, but before this month it seemed they were relaxed about it. I have now started to be rate limited on Pro when nowhere near that quota, due to too many prompts within a short time window (probably due to short prompts and not aggregating my questions). So now I use the Thinking (basically Flash) model and bump up to Pro for certain queries only.

There will always be a minority who spoil it for the majority.

I use a walking pad at home. Each night after dinner, I’ll go into that room and watch a movie (or half) on my iPad when I’m walking on it. I did over 3 miles last night (at a moderate pace), and that’s pretty typical. Not because I’m trying to work out, just because I’m into the movie and don’t want to stop. If I had friction of needing fresh gym clothes, having to drive somewhere, or thinking about if the weather is nice enough for an outside walk, I would probably be too lazy to walk most evenings. If I didn’t have the entertainment of the movie, I would probably find the exercise boring and not feel incentivised. I do it because I’m looking forward to the movie and I know I’ll feel great afterwards.

As they say, you need to design your habits for the laziest, most unmotivated version of yourself.

https://www.google.com/appsstatus/dashboard/incidents/nK23Zs...

I am nearly nine months into NSAID gastritis from only a week of daily 75mg soluble Aspirin.

I have had more painful acute illnesses in the past, but the grind of a long-term illness is new to me and it has been absolutely terrible.

What dynamically typed scripting is involved?

If it’s JavaScript, you can gradually migrate to TypeScript and have a Git pre-commit hook to compile (with incremental compilation). And standardise on VSCode or a derivative that makes programmatic typos obvious. Many IDEs will also spell check your strings.

Before systemd presented a generalised interface, there were significant differences in the init and service management systems between the popular Red Hat and Debian families of distros.

It's incredibly cheap and works reliably for me.

I have got it to paste my voice transcriptions into Chrome (Gemini, Claude, ChatGPT) as well as Cursor.

https://github.com/EpicenterHQ/epicenter

In my experience to date, ChatGPT really is a sycophant. Claude can be moderately stubborn. Gemini is usually very stubborn, practically unmovable, unless you present new facts or make a rock solid counter-argument.

And they mostly just come to a conclusion, rather than actionable advice I was able to pass on, like take a bag and avoid eating because they might want to operate soon.

My flareups and their accompanying setbacks have been greatly reduced because I keep a megathread chat going with Gemini. I have pasted in a symptom diary, all my medications, and I check any alterations to my food or drink with it before they go anywhere near my mouth. I have thus avoided foods that are high FODMAP, slow digesting, or surprisingly high in fat or acidity.

This has really helped. I am trying to maintain my calories, so advice like “don’t risk X, increase Y instead” is immediate and actionable.

The presumption that asking a LLM is never a good choice assumes a health service where you can always get a doctor or dietician on the other end of the phone. In the UK, consultations with either for something non-urgent can take weeks, which is why people are usually pushed towards either asking a Pharmacist or going to the local Emergency department (which is often not so local these days).

So the _real_ choice is between the LLM and my best guess. And I haven’t ingested the open web, plus countless medical studies and journals.

Guess what happened.