Seems unlikely. I've seen teams at GDS + teams based on the GDS way of working in other UK gov departments solve some really knarly problems.

Counter to that I’ve seen a £37m contract for a form on gov uk with absolutely no change in process, just going from a letter received to a online form

GDS went from an internal distruptor, breaking the consultancy oligopoly to being another entry gate.

The companies who capitalised on this (Kainos/Equal Experts/ScrumConnect) are now their own oligopoly.

It's just big guys charging arms/legs for average work again.

They most definitely won't spend time on military bases though, whereas Palantir devs will essentially live there.

I'm not sure I get the link between being against screenshots and working from home during COVID ?

Remote work can involve a lot of screen sharing or screen capture.

Their past stance was a complete miss, but they’ve relented.

Now you can do it, but it’s harder and can have problems with some apps

You could take screenshots and do screen sharing with Wayland long before 2020.

You need to screen share a lot more during remote work

And of course the random choice of things displayed on the infinite scroll - deliberately emulating a slot machine effect.

You could do this, GIMP is scriptable which is half the battle - you may not even need to train a model.

This is good and we need more of this sort of thing.

I don't want things to be silo'd just because I run them in the GUI Vs the terminal.

Are Google one of the last holdouts ?

I've been on cheap Android phones and just moved from Samsung to Motorola and both have headphones sockets.

No, the Pixel dropped the headphone jack ages ago. It's why I don't have a Pixel any more.

For my recent middle class Android phone, I got a USB-C dongle that I keep in my wallet since phones with headphone jacks have moved to the lowest end of the market and almost disappeared entirely. Doesn't seem worth it anymore.

The dongle is tiny and seems to be of good mechanical quality.

You can use it with Ollama rubbing a local model.

Does anyone else find this kind of LLM written stuff tiring to read.

Admittedly I only read the titles.

Author here. I do use AI tools in my writing process. I build AI agent tooling, so that's how I work now.

But the ideas, analysis and seven years of firsthand experience are mine. And I write "with" AI very closely, not handing anything over blindly.

If you read past the title you'll find lots of personal details I lived through. It would be hard to prompt-engineer that from the outside.

'read the title' is not the same as reading an article.

Im also not sure why this would be LLM written and i have not seen a lot of / relevant amount of LLM written articles especially on hn

I have a feeling that people who shout this (I gotta be careful, I call out a bunch of articles as LLM-written) do not always know what to look out for. Em dashes aren't inherently LLM, neither are some of the phrases it uses.

One pattern I look for is; how is the end part written? Does it tie into the whole story? Is it extremely generic? Did the author put something real / raw / personal in it? If not, its a grey area, any other hints will lean my conclusion to real or fake.

The author even copied ChatGPT's website design!

This whole thing of private equity + companies getting massively inflated - only ends one way, it might not be this this buyer but one down the line, but there is something deeply wrong with the whole model, the one that starts with startups such as those funded by ycombinator.

The point being someone is left holding the bag?

That's potentially true, but not necessarily. I haven't looked into this particular case, however it's entirely possible that a lot of the EU have started divesting from Windows and into suse, which has caused a big spike in revenue here.

Or its PE doing PE things and it's all a farce.

maybe they are just IPOing it and selling it to the public?

> Nobody uses SuseLinux any more.

What gives you that impression? They had $700MM in revenue in 2022 and many HPC clusters run on Cray OS[1] (which is SLES).

> If SUSE gets 6 billion dollars

Not how sales work.

[1]: https://top500.org/statistics/list/

By "nobody" I presume you mean you and your friends? From the article;

>> "More than 60% of the Fortune 500 rely on SUSE to power some of their workloads, according to the company."

This is an Enterprise version of Linux, and unless you are in the enterprise space you're unlikely to come across it.

Also from the article; >> "The company generates about $800 million in revenue "

So again, this suggests that people are indeed using it.

Rancher/k3s is used a lot in many places as well.

There’s also harvester on top of rancher. It’s one of the very few open source competitors to RedHats OpenShift that I’m aware of.

I mostly like their use of an immutable OS as base layer for the virtualization - despite the limitations it sometimes has.

Harvester is just Kubevirt with some UI atop it, the same as Redhat Virt. Works fine if you’re hosting datacenters or whatever, haven’t seen it be suitable in smaller manufacturing environment

It fundamentally is rke2+rancher+kubevirt, but there’s a lot of packaging around it to make that work.

True, I am underselling it.

I liked it when I used it, but it wasn’t really a fit for our environment from what i’ve seen.

Over 60% are SUSE?! Sorry, but I’m with everyone else…

I remember since the start that SUSE was more popular in Europe, but no way would that be the case in the US. If anything, I’d be willing to put my money on > 60% of Linux installs being RHEL/Centos rather than SUSE

You could get the number wrong. The quote stated that 60% of the companies use Suse to power some of the workloads. So if most of these companies would use Suse to host SAP, some have a few teams using Rancher and some (more so in Europe ) are using Sles you still get to these numbers even if most of them use RedHat for most of their workloads.

Why would they lie? Hacker News simply has this bizarre blind spot about what Fortune 500 companies do and what computers are that run Linux. One of their biggest customers is Chick-fil-a using k3s for the their point-of-sale network. I'm sure there are approximately zero employees interacting with the system that realize that, but it's still there.

Also, from my own experience, SUSE used to have nearly all of the US geointelligence processing because of the HPC connection mentioned elsewhere with CrayOS, but that went away when DNI forced everyone onto the CIA's private AWS service, which only had RHEL AMIs available. The national labs and more niche intelligence processing that can't run in the kinds of machines AWS provides still make heavy use of it.

Interesting. It's the only commercial distro I could ever stomach, in fact I really like it but don't use it, (because there's a non-commercial distro that I like much more). (Edit: my point was that it would feel like a real loss if it were to deteriorate)

SuSE is about the 2nd most used distro in the enterprise, and I can understand why.

Maybe for your personal workstation this might be the experience you have. But from my experience for enterprise there is RHEL, Suse and maybe Ubuntu Pro. If you are an AWS Enterprise customer you might justify Amazon Linux

Also Oracle Linux and CIQ's version of Rocky, albeit in rather different niches.

I think Ubuntu Pro is more common in service providers that sell to enterprises rather than in enterprises themselves. It enables them to say "yes, we comply with all of these box-ticking standards that you require your vendors to have!" without bringing in much of the rest of the enterprise baggage.

SuSE is used more heavily than any of them - as others have said, they're used more or less everywhere where SAP is to be found, and they're very strong in the HPC space too.

It's still quite popular with SAP shops here in Europe at least. And I could imagine that the strong anti-American sentiment in Europe plays in its favor.

Yep. The majority of the worlds SAP-installations use SUSE somewhere in the stack. As for the desktop, opensuse is rock solid. I've used it for years without any problems. I've had colleagues who use Ubuntu and they always have glitches and hiccups.

I've been using OpenSUSE on my home PC for the past 3 years - it is a really solid Linux distribution and I rarely had any problems.

They own Rancher and Harvester. My brother, this is a good enough reason for someone to pick it up. There’s no better way to kill any serious sovereign cloud attempt than that.

If I can use this with a local LLM it could be useful.

Yeah. This seems like an area where a “tiny” (2-4GB) local model would be more than sufficient to generate very high quality queries and schema answers to the vast majority of questions. To the point that it feels outright wasteful to pay a frontier model for it.

In ollama is included default add the endpoint URL yourself