Looks to be in one big pile in your inbox.

It took 72 hours to download all my mail via IMAP that was collected in my 15 year old gmail account. Mind you, in 2024 I was close to limits, so I used the web interface behind https://mail.google.com/ to do some severe cleaning. And it also showed up in that interface as everything being reduced back to around 7 GByte.

For years I use my gmail account to forward incoming mail to another IMAP mail server ad that worked fine, About 10 to 20 new messages showed up that way, all great. The company that receives the forwarded mail has issues, so I thought, Let's configure gmail then in the Thunderbird client on my Windows system. About 72 hours later Thunderbird is seeing 170.000+ mail messages, all in the 'Inbox'.

I have not configured any filter in the gmail web-interface. Because I find it both dreadful and aggravating to work with. Whoever thought up that interface, deserves to be put in front of a firing squad. And I would volunteer to be part of the firing squad. And load up my own real bullet, if they would have given me one of the blanks. Just to be sure.

So I have spent time making my own mail folder structure and making rules in the Thunderbird client, but damn, this is so ff-ing slow. Also, I have the Thunderbird client set to not read incoming messages automatically or anything. Yet all mail that is coming in via gmail is already read.

These last few days did everything right to rekindle my dislike for gmail. It does not help that the misery of their gmail interface is mirrored into my Thunderbird client. Tomorrow I will finish with all the mail filters, there are still some 35.000 messages left in the inbox that still need filters.

Youtube channel Climate Town assured me that if the US would stop importing oil and rely completely on its own production, the US could only manage to supply 30% of domestic consumption. This channel usually does their due diligence regarding numbers. Hence I tend to trust them.

So, I'm less sure about the US having enough oil. And even if there is, is there enough capacity to refine or distribute?

Purchase is still cheaper, yes.

Lets wait 3 months more and see if that statement still holds true. Business analysts expect RAM prices to increase throughout 2026, stabilize at 2026 price level throughout 2027 and increase in price yet again in 2028.

"You will own nothing and be happy about it", that is future planned for you and the rest of us slobs in the trenches.

Seriously, where is a good Robespierre's when society needs one...

In an effort to improve commercial planes, an U.S. cloud provider could be made to share their "backups" with a certain U.S. plane manufacturer (which has fallen on harder times times lately, regarding production quality and public perception), by the current U.S. administration.

At least, that is what AirBus is probably thinking, giving the tone in the quotes from the summary.

Can confirm, Ford Transit and Mercedes Sprint are very, very popular models of vans in Western Europe. Companies love both brands/models as they are reliable and versatile.

Had the same Microsoft mouse for 15 years. Very basic model, nothing special at all. Then I go for a week away to travel to a different nation as this South-American country where I live has no Dutch embassy. How the mouse managed it I don't know, but it seemed to have found a way to drop from my desk and break into pieces and yet no-one touched it.

Must have been a miracle of sorts... \s

Been using a wireless Microsoft keyboard and mouse combo since. Again, no complaints, both do what they are supposed to do, without fuzz. That was some 2 years ago. Marking on 2 keys from the keyboard have noticeably faded already though.

I do agree about Logitech having the better mouse. But these tend to suddenly stop working after 5 to 6 years.

Ask Norway. If I remember correctly, over 90% of new cars sold are EVs in that country in 2024. They don't have to care one iota about Trump's latest delusion.

To my understanding, Norway has so much oil in reserves that these make the Middle East jealous. And yet, they rather use their hydro power plants to power their fleet of cars.

Similar to the: "The beatings will continue, until morale improves!" style of leadership?

Google has the hardware to run AI with their TPUs. Google has datacenters and networking facilities already in place and it appears ther AI software isn't half bad either. OpenA| has no or not enough hardware, they don't have hardware facilities, but do have decent AI software and a huge non-paying customer base.

If I were to bet, I would bet on Google coming out of the AI bubble still reasonably well. OpenAI, even with their software product and customer base, I don't think it would be able to salvage itself from a popping bubble.

How Oracle financed their promised AI hardware roll-out, that can't deal with a bursting bubble either. And I'm quite sure that they would have to sell a lot of their holdings to have a chance on survival. And anyone who dealt once with their sales department will tell you that they hope on Oracle not surviving that.

Microsoft has hardware facilities, but is a quite a bit lacking in compute power. And if I'm honest, their own LLMs (both open source and in CoPilot) suck...really badly. Not for nothing they also offer the AI products from OpenAI and Anthropic via CoPilot to give it at least the impression of capability. Microsoft is diverse enough that they would also survive a deflating AI bubble. A popping bubble will hurt them more severely.

There are many very tiny text editors for Windows available. Most don't even require installing, they'll run just fine the 'Portable" way.
Examples you say?
EdXor - Hasn't been updated in 10 years, but is so small (77 KByte) there are hardly any room for errors in it. And yet, surprisingly capable. The 77 KByte is the size of the archive and includes a manual. Still larger than the original NotePad.exe, but not by much.
Notepad2 - Hasn't been updated in 14 years, 300 KByte, also small, yet 4 to 5 times the size of EdXor. Just as independent too. Newer versions are available, but balloon up in size spectacularly.

If ballooning sizes aren't a problem and installing software isn't an issue at your place of employment, try Zed (now also available on Windows natively, for those that care).

You can start this editor as a very basic text editor and then it starts really fast. Or you can start it with every feature enabled. Still starts fast, but is now capable of so much more. AI is optional, they support all the major players, have their own AI service and aren't fussy if you want to run your own LLM(s) on your own hardware. You can hook local LLMs up via the server functionality from both Ollama and LM Studio.

The last part is what I do at home and that works well enough for my intents and purposes. A tad slow, but then again, my GPU is a 16 GB VRAM card from AMD (R580 if I remember correctly), which is only accessible via Vulkan. About 20 tks with OpenAI's gpt-oss-20b LLM, configured with 100+ kb context window on a AMD 5600XT Ryzen CPU with 32 GByte of RAM and 3th gen PCI-E NVMe drive, running W11 24H2. Pretty meh for today's standards, but I'm still happy with the machine.

Altman delivers results? Guess he didn't flew into a rant when asked by business journalist how he spouts large numbers for future endeavors while making so little revenue with OpenAI. Oops, he did just that.

Anthropic however doesn't care that much about the common user and has a business model that caters for the enterprise. their numbers are a lot lower than OpenAI's numbers, but Anthropic is on much more solid footing, financially speaking. So if the AI bubble pops sooner than later, I see OpenAI fail so miserably that they will go bankrupt.

Understand me correctly, I don't think that the OpenAI ChatGPT models are bad, just that there are far too many users in their free plan, creating a lot of overhead and very little returns. At some point, investors want, nay, need to see returns and I don't expect Altman to be around for very long as he isn't the best person to lead OpenAI.

The Norse national investment fund, owning 11% of Tesla stock, said no to Musk. Whether that is because they don't think he earned that trillion USD or that he can make Tesla enough to let Tesla give Musk that generous package, who cares. And they are not the only big Tesla investor, who say no to Musk.

Tesla has even trouble giving their cybertrucks to police forces (for tax purposes). Las Vegas police has a few, apparently.

First generation code is more often than written with theoretical use-cases in mind. Second generation code usually written with hindsight of how the 1st generation code was actually used in practice. You seem to call that "better", I rather use "more insightful".

You didn't think hard enough. I live in a country where they just implemented "no DST". For life and light, it was indeed not too big of an impact.

However, the impact was felt in almost every digital way, but banking and telecom were hit hard. As in: "we can't oblige, it affects our ability to do (international) transactions and communications". As in: "our international licenses and regulations describe standards that we can't comply with. And these are difficult and very costly to attain, hence we simply ignore your decision of being DST-less".

So, no. going DST-less can only be decided on a nation level (if it would be Australia) and else on a continental level. Yes, it is that impactful, because else you will find very quickly which underlying (digital) systems fail. At least do a study to find out which of your underlying systems fail, preferably including both DST moments and solve those first. As one will be very unpleasantly surprised otherwise.

The smoothest transition is doing it on a global scale. Next best option is continental scale, etc.

White gold perhaps?