Posts by AdamWill

IBM says AI is insane in the mainframe as z17 sales surge

Re: Lies, damned lies, and quarterly investor calls

"So did AI consulting really blow up, or did salespeople realize they could get a kicker on their quota fulfillment for tagging every possible consulting gig as AI? Not that I'm suggesting that IBM would ever do such a thing, and certainly not that it has ever done so before."

Y'know, I started wondering the same thing. I'm just going to leave this little table I drew up right here (you might have to cut and paste it into something monospace, though):

AI book Consulting Software AI cons AI soft

Q1 2024 >1b 5.186b 5.899b ~0.75b ~0.25b

Q2 2024 >2b 5.179b 6.739b ~0.8b ~0.2b

Q3 2024 >3b 5.152b 6.524b ~0.8b ~0.2b

Q4 2024 >5b 5.175b 7.924b ~1.6b ~0.4b

Q1 2025 >6b 5.068b 6.336b ~0.8b ~0.2b

Q2 2025 >7.5b 5.314b 7.387b ~1.2b ~0.3b

Q3 2025 >9.5b 5.324b 7.209b ~1.6b ~0.4b

Q4 2025 >12.5b 5.349b 9.031b ~2.4b ~0.6b

IBM's said each quarter that the "AI book of business" is 80% consulting, 20% software. Over the period it's been talking about it (last two years), consulting revenue has been basically flat, and software revenue has been growing far *faster* than 20% of the "AI book of business" could possibly explain.

So...yeah, my conclusion is that the "AI book of business" is basically just a label slapped on ongoing consulting revenue. To be fair, in https://cloudwars.com/ai/ibm-genai-business-soars-to-7-5-billion-drives-growth-in-mainframes-consulting/ , the CEO did more or less admit this when asked on a call:

"If I look at consulting, yes, there is a big piece of the AI book of business that is coming because people are directing their dollars towards that kind of consulting as opposed to alternate forms of consulting"

AI hasn't delivered the profits it was hyped for, says Deloitte

erm, what?

"Ali Sarrafi, CEO and co-founder of Kovant, an enterprise agent platform, told The Register in an interview that the problem with the way people use AI is that they see it as a form of fancy workflow automation

...

"We're working with this big large manufacturing company," Sarrafi explained. "They have about 7,000 suppliers. And every single time they needed to restock something, they had to coordinate with so many suppliers. It's actually the most boring job ever for everyone. But then they deploy this agent worker or a team of agent workers that basically monitors the stock levels. As soon as it goes below the forecast requirements level, it sends a preliminary email to the supplier saying, 'Can you tell us if you can supply this and what price?'""

How, exactly, is that not "fancy workflow automation"? And why the hell does it need AI? AI sounds like a rather terrible way to do it, in fact, since it will probably do what it's intended to do 95% of the time, then suddenly order five zillion widgets or email the supplier a blackmail photo instead of an order. And they'll have no idea why it did that.

For an extremely predictable workflow like this, what you need is a sensible bit of determinitive code to automate it. Not an LLM.

Anthropic quietly fixed flaws in its Git MCP server that allowed for remote code execution

Re: It's not possible to do

MCPs themselves aren't LLMs, though. They're deterministic code. The flaws here were exactly what the OP said - a failure to sanitize user inputs to code with potentially-dangerous access (to your filesystem and git repos).

The MCPs were *supposed* to be designed so that no matter what the prompt contained, you could contain the execution of anything via the MCP to a specific repository. But they were very badly implemented so they didn't do that.

AI may be everywhere, but it's nowhere in recent productivity statistics

Re: Don't follow the IBM route if you care about productivity

well, yes. That's literally why there is an increasing tendency for governments to propose increases in the retirement age.

The response to those proposals, however, tends to come under the heading of "turmoil and strife". Understandably so, but...it will probably have to happen in the end. And there's a very solid case for it: current retirement ages were generally set long ago when the expected survival from that age was much, much, much shorter than it is today. Even *without* the demographics issue, most pension systems weren't designed to work with the majority of recipients living for 20 years or more.

Re: Don't follow the IBM route if you care about productivity

Sure, but it's not necessarily the case that we need a growing population. It's more the case that, without immigration, the future of all advanced economies involves the *working age* population shrinking rapidly, while the post-working-age population grows for much longer. Nobody has a good answer as to how any country deals with that without substantial hardship and turmoil.

(The bad news is that even immigration can only be a temporary and incomplete 'solution', as fertility rates are dropping like a stone absolutely everywhere; probably within 20-30 years there won't be enough immigrants to "go around" either.)

Over half of AI projects are shelved due to complex infrastructure

Throw it on the pile of "OK, so AI isn't working, but it's because people aren't using this one thing we happen to sell" so-called "reports", then?

Capita tells civil servants to wait for chatbots to fix pension portal woes

"They said the volume of scheme members attempting to set up logins and view pension information "just to check" whether their details were correct was a contributing factor. Members have been advised to delay accessing the new portal until early spring when annual statements are due, unless they have a particular issue they wish to resolve or query."

How are they supposed to know whether they have a particular issue to resolve or query if the system can't cope with them logging in to check?

Hot for its bot, McKinsey may cut thousands of jobs

the one case...

...where AI replacing humans seems entirely believable.

Hell, you could replace 90% of these companies' headcounts with a good slide deck template.

Vibe coding will deliver a wonderful proliferation of personalized software

Suggested headline alteration

"Vibe coding will deliver a problematic proliferation of personalized software"

The problem I'm seeing with this, so far, is that, yes, people who don't really know what they're doing can vibe code a thing that mostly works!

So...they do.

They *all* do.

Now instead of having three or four tools for doing the thing, each maintained by someone who vaguely knows what they're doing and in which effort to do it well can be centralized, there are 50 tools to do it, all built by people who don't really know what they're doing and have no concept of how to maintain software in the long term, each with their own bugs and (no doubt) security issues.

Can you imagine having to sort this mess out in a few years when companies start realizing there are tens of thousands of vibe coded scripts doing stuff with their data and they have no conception of exactly how many, who is using them to do what, or what's wrong with each of them?

yikes.

IBM touts progress on tech stack for AI-enabled airline with no passengers or alcohol

Proposed tagline

Riyadh Air: We're Confident Our Planes Will, On Average, Have Approximately The Correct Number Of Wings

John Henry still leading the race vs AI in customer service

So to summarize the story, the people that are actually trying to use AI now in the real world are finding it's not really much cop, but the ones who are being asked to speculate wildly about the future are happy to imagine it'll be amazing in another two or three years?

Re: curious

No. No, it's not.

https://en.wikipedia.org/wiki/John_Henry_(folklore)

If you haven't heard it, listen to Bruce Springsteen's version of the song, with the Seeger Sessions band. It's great. https://www.youtube.com/watch?v=hIh74VC7oXc , for instance.

HSBC spies $207B crater in OpenAI's expansion goals

Not to mention that ChatGPT is not available in China, so you can drop about a billion from the total.

3 is 44% of 6.82, so if we make that 5.82, they're projecting over half the adult population of the world that *can* access ChatGPT will be using it. (And that's *also* assuming sanctions on Russia are lifted by then). And 10% of those will be subscribing to a paid plan.

Let's...let's see how that goes.

Microsoft exec finds AI cynicism 'mindblowing'

Yes

Yes, Mustafa, it is the children who are wrong.

Brits believe the bots even though study finds they're often talking nonsense

Re: Of course.

It's weird, somehow every time I go to vote, the only choices are politicians.

Canonical CEO says no to IPO in current volatile market

Re: Linux on the desktop ? You 'aving a larf ?

"Wrt "open source winning the desktop", I suspect Shuttleworth is kidding himself, or trolling for sound bite headlines or similar."

It sounded a lot more like 'paying lip service to what Ubuntu started out as and what most people still think of it as, while actually focusing a lot more on the stuff that pays the bills', to me.

Not that I'm saying he's *wrong*. Everybody who's ever tried to make a buck off the Linux desktop (including Red Hat, who I work for) has eventually either done the same thing, or gone broke. A reasonable amount of people want a Linux desktop. Nobody yet has figured out a way to get them to pay enough money to fund it.

Re: Linux on the desktop ? You 'aving a larf ?

That's an odd thing to pick to complain about. Evolution was (and still is!) a blatant Outlook clone. It's quite good at calendaring. I've been using it for 20 years.

It didn't interoperate with Exchange very well at some points, but that wasn't about a lack of will, it was about Microsoft trying very hard to make sure it wouldn't work.

You have one week to opt out or become fodder for LinkedIn AI training

if you post to linkedin, you've got bigger problems

See title.

Lloyds Banking Group claims Microsoft Copilot saves staff 46 minutes a day

"Vic Weigler, chief technology officer at the finance corp, said in a statement: "We converted 11,000 lines of code across 83 files in half the expected time.""

Wait. That's it? You're a global frickin' bank company and you've got a vague time savings in a tiny code "conversion" (from what and to what not specified of course) project, and that's *it*?!

I mean, 11k LoC is *nothing*. Just nothing. I can find that much code down the back of the couch. In what other context would the CTO of an *entire global bank* be reduced to waffling on about a project of this tiny scale?

Tech industry grad hiring crashes 46% as bots do junior work

Re: Disguised recession

Exactly. Is there any *proof* that the hiring slowdown is *actually* caused by AI, or is that just what people are claiming or assuming is the cause?

Cisco: Most companies don't know what they're doing with AI

Progress

Aha, we have reached the "you're holding it wrong" stage of the hype cycle, apparently.

Don't let them baffle you with bullshit. All the people and organizations now saying "AI can maybe help you out but it's not a silver bullet and you need skills to be able to use it effectively, lower your expectations!" are *THE SAME PEOPLE AND ORGANIZATIONS* who, just a year or two ago, were blootling on about how it was a magic revolutionary technology that would change the world and put everybody out of work instantly.

IBM Cloud to end free human support, suggests customers use enhanced AI instead

"We leave it to IBM customers to judge the veracity of the second assertion"

IBM Cloud customers? You'll have to find some first.

SK Telecom walloped with $97M fine after schoolkid security blunders let attackers run riot

Re: Did anyone get fired?

Yup, I was thinking the same. To put it another way: SK Telecom's annual revenue for 2024 was 17.9406 trillion won - $12.6bn . Its operating income (profit, more or less) was 1.8234 trillion won - $1.28bn . A $97m fine is less than 10% of its profit for one year and less than 1% on its revenue for one year. That's pathetic. It's barely noticeable. For comparison, it's like a fine of $970 for a person who earns $126,000 per year. Slightly annoying? Sure. Appropriate punishment for compromising the security of half the nation? No.

AWS pricing for Kiro dev tool dubbed 'a wallet-wrecking tragedy'

Re: Balance

Claude is a lot better at this than any other model, but it's also very expensive. Gemini is awful at substantial coding. The OpenAI models are in the middle.

Ransomware crew spills Saint Paul's 43GB of secrets after city refuses to cough up cash

ten points...

...for correct use of the word 'flaunt'!

Fedora 43 won't drop 32-bit app support – or adopt Xlibre

Re: Note: Changes aren't decided by community vote

"You ask if others have simpler clearer models. It's a non sequitur. *Everyone* else has a simpler model, because they don't have a model, because nobody bloody cares."

I mean, this seems unlikely. "Most distros are just a bunch of people, putting something together and putting it out there" is a very reductive description - but it equally applies to Fedora. Governance is just the *details* of how that "bunch of people" decide what "thing" it is they're putting together, exactly, and how they're going to "put it out there".

If "nobody cares" about these kinds of details, then this is a complete non-story: from that "nobody cares" perspective, *literally nothing happened*, did it? The only things that happened were the things you say nobody cares about - the details of one specific "bunch of people" deciding not to do a thing.

So if this is a story, well, that implies governance and community processes and all of that stuff *matter*, doesn't it?

Re: Note: Changes aren't decided by community vote

Regarding documentation - the Change process is documented at https://docs.fedoraproject.org/en-US/program_management/changes_policy/#_change_process .

I don't think anyone could "write a short explainer on how Fedora governance works", because it's quite complicated. In my experience, governance of anything of significant size tends to be somewhat complex unless it is run entirely by a dictator (benevolent or otherwise). Even in that case, once it's sufficiently large, the dictator probably needs to delegate *somehow*, and the details of that also get complex.

Is there a non-dictator-led distro whose governance you'd say is simpler or clearer to understand?

Re: Note: Changes aren't decided by community vote

Well, my comment intentionally didn't ascribe anything to any person. That was intentional, because I'm fully aware that headlines and subheads are often not written by the author of the article.

Liam, I like you and think you write good articles, that's why I read them. Adding an explanatory note is not a criticism, either personal or journalistic. You'll note my comment doesn't call you an idiot or mention your name at all.

Note: Changes aren't decided by community vote

The subhead and this text may give an incorrect impression:

"The change was rejected, and for now this won't happen. The margin wasn't huge, though. At the time of writing, it was showing 51 percent "Strongly opposed," plus another 15 percent "Opposed, but could be convinced.""

Changes are not decided by community vote. The vote is purely informational. Changes are voted on by FESCo, the Fedora Engineering Steering Committee. In this case the Change did not come to a vote because the proposers withdrew it.

Some signs of AI model collapse begin to reveal themselves

mm, well, details

"What all this does is accelerate the day when AI becomes worthless. For example, when I asked ChatGPT, "What's the plot of Min Jin Lee's forthcoming novel 'Nightshade Market?'" one of the fake novels, ChatGPT confidently replied, "There is no publicly available information regarding the plot of Min Jin Lee's forthcoming novel, Nightshade Market. While the novel has been announced, details about its storyline have not been disclosed.""

I'm not sure that's model collapse, though. Did you try asking it "What's the plot of Min Jin Lee's forthcoming novel 'The Price Of Fish'?" I expect you would've got the same result. Just because you used the name from the Sun-Times supplement doesn't mean ChatGPT actually assumed the book existed because someone had set it up to use RAG with that data as input. It could just be the usual "naive AI assumes the question you asked doesn't have any lies in it" hallucination.

AI ain't B2B if OpenAI is to be believed

terminology

"but far from the kind of ex-Oracle sub-boss"

Ahem. I think when they work for Oracle, they're called "mini-bosses".

AI can't replace freelance coders yet, but that day is coming

Re: The limitations

The introduction has all the meat, really. To summarize: they saw that OpenAI had done a fairly realistic study where they took *actual* real-world jobs from upwork/freelancer, and tried to get LLMs to solve them. They saw that the LLMs didn't do very well: "the top-performing model in OpenAI's study solved only about 26% of the independent coding tasks and 45% of the management tasks".

So they decided they'd do a much worse study which would give the LLMs bigger numbers. I am not making this up.

"In this paper, we propose a new evaluation approach that draws inspiration from SWE-Lancer but emphasizes automation and repeatability. Instead of using actual freelance projects that require human evaluation, we leverage a publicly available dataset of freelance job postings to generate synthetic coding tasks with ground-truth solutions. In particular, we use the Freelancer.com dataset by Oresanya et al. (2022), which contains ~9,193 job postings in the data analysis and software domain. We filter and process these job descriptions to create well-defined problem statements (e.g., data processing tasks, scripting challenges, algorithm implementations) that an LLM can attempt to solve. Crucially, for each task we provide a set of test cases (input-output pairs or assertions) so that a solution's correctness can be validated automatically, without human intervention"

so they took real freelance jobs and massaged them into AI benchmarks, then said "look! the AIs did quite well!"

Good lord.

Dilettante dev wrote rubbish, left no logs, and had no idea why his app wasn't working

That was below the belt, El Reg

"Self-taught coders who work in HR and have a doctorate in English tend to do that"

Please skip the insulting stereotypes. A lot more professional programmers have arts degrees than you might guess, and a lot of us are perfectly good at it. And there are plenty of computer science grads who can solve the travelling salesman problem all day long but are terrible at writing real-world software.

Nextcloud cries foul over Google Play Store app rejection

Both things can be true. This can be "just the standard hoops", and Google can have been quite conscious that the "standard hoops" would have the effect of restricting competition with its cloud service. Certainly the facts that Nextcloud has been trying to get a sensible response for several years without success, and that Google is conspicuously refusing to return El Reg's calls, are not insignificant.

Zuck ghosts metaverse as Meta chases AI goldrush

Re: FFS

well, I spent like three days this week reading the last nine months of the upstream git changelog for openQA, in preparation for updating Fedora's package and deployment of it. It's important to do this so I spot any changes which need to be accounted for in the package or the deployment config.

Today I got curious, so I fed the entire changelog into Gemini (2.5 pro; 2.0 flash absolutely choked on the volume) and asked it to summarize it for me, first for a sysadmin, then for a distro packager, highlighting important changes.

It, uh, did a pretty good job, I've gotta say. It was fairly concise, organized, and found the most important changes. It didn't make anything significant up (it got the name of one script wrong, but still explained the change to it accurately), or get anything obviously significantly wrong. Did it get every little detail? No. But then after three days my brain didn't remember every detail either, it just had a vague sense of the most important areas of change. I'd say the results of both approaches were pretty comparable, but one took me three days, the other took Gemini a couple of minutes.

I hate to think how much carbon it burned to do that, but hey, I'd definitely characterize it as *useful*.

Ghost in the shell script: Boffins reckon they can catch bugs before programs run

Re: Really?

The problem with this is you can absolutely cause gigantic security issues while you're doing the herding, notably with variable expansion.

You certainly cannot *fully* check everything a shell script does by statically analyzing it, but you absolutely *should* be checking *as much as you can* of what it does with an analyzer. Like shellcheck, which another commenter pointed out has been around for years. I've seen a definite encouraging trend in the last few years towards incorporating shellcheck in CI workflows, which is nice.

The most important experimental distro you've never heard of gets new project lead

Re: everyone's doing it

No major distributions has actually used sbin for that purpose in years, so nothing of value was lost. Fedora has always gone with the 'sbin is for privileged binaries' definition, but that got so fuzzy over the years that it became more or less arbitrary.

We are not going to start statically compiling some binaries in order to provide an ad hoc rescue environment within the distro, so there's no reason to have a separate sbin. The rescue environment for Fedora is provided externally, you can boot into a rescue environment from any Fedora installer image. It has functions to automatically find and mount an installed Fedora system so you can fix it.

everyone's doing it

so-called "conventional" distributions are doing this too, just from a different direction.

if you look at a current Fedora system, /lib is a symlink to /usr/lib , /lib64 is a symlink to /usr/lib64 , /bin is a symlink to /usr/bin , and so on: this is the 'usr merge' from a while back, and other distros are doing the same, e.g. Debian in https://wiki.debian.org/UsrMerge .

In Fedora 42, /usr/sbin is a symlink to /usr/bin . So /sbin , /bin , /usr/sbin and /usr/bin are all the same directory.

On an F42 install, the only real directories at the top level are /boot , /dev , /etc , /home , /mnt , /opt , /proc , /run , /sys , /tmp , /usr and /var . Of those, dev, proc and sys are kernel interfaces, so really we just have boot , etc , home , mnt , opt , run , tmp , usr and var.

There are clear purposes there which I don't immediately see replicated in the gobo design. For instance, multiple levels of configuration. It's becoming increasingly common to have a clear separation between upstream/distro defaults (which are in /usr), system-wide site configuration (which is in /etc), and per-user configuration (which is under /home). This is a useful concept, and having the files in three clearly different locations makes it easy to know what's what, transfer around the things you need, and keep a clear distinction between system files (which can be immutable) and local customization (which cannot be).

there's definitely more that can be done here - it would be lovely to get rid of one out of /usr/lib and /usr/lib64 at some point, for instance. I'm not convinced we really need /opt any more, certainly not by default. But everyone's rowing in broadly the same direction.

GNOME 48 lands with performance boosts, new fonts, better accessibility

Re: Built-in Javascript engine?

Note it's really the *desktop shell* that is implemented in javascript. It sits on top of a library called mutter (which itself sits on mesa), which is where any really heavy graphics rendering-related work is done, and that's written in C.

The official explanation for writing the shell in javascript is: "Gnome-shell is written in JavaScript because it allows for quick prototyping and development cycles. The difference between javascript and other dynamic languages (python, ruby, php, etc...) is that javascript doesn't depend on a specific platform. Gnome-shell can than run it with its own platform (Gjs)."

AI crawlers haven't learned to play nice with websites

Re: AI is little more than cloud/hosting providers ruining the internet in new and unwanted ways

It's not the same. There are more of these AI scrapers, and they're more aggressive. (And yes, lots of them ignore robots and screw with their user agents, and change IPs frequently).

Search scrapers mostly only bothered with 'normal' web pages. The problem for a git forge is that these AI scrapers want the code, not just the web UI views of issues and stuff. They are cloning every repo the forge hosts. That's massively more traffic than a search scraper generates.

not just sourcehut

This is also an ongoing nightmare for pagure.io (a forge used mainly for Red Hat / Fedora / CentOS-adjacent projects, for weird historical reasons). It's getting absolutely pounded by AI scrapers. I think on Sunday we had to just ban Brazil. Sorry, everyone in Brazil who isn't an AI scraper.

I'd imagine every git forge is dealing with this mess.

CentOS Connect conference announces return of Firefox

Ours is older

"and Bodhi packaging system. (I'm not sure if anyone at Red Hat knows or cares that there's been a distro called Bodhi Linux for some 15 years.)"

Our Bodhi is older than their Bodhi.

First commit to https://github.com/fedora-infra/bodhi was e26dacbb680bb65ffee0b82679e3f48a982ae4a1 was dated Jan 15 2007. The second was dated the next day and says "Initial import from CVS to Mercurial", so it wasn't even new then. I don't honestly remember when it was introduced.

The Wikipedia page for Bodhi Linux says its first release was on March 26, 2011. So our Bodhi has been around way longer.

(Also note this doesn't have much relevance to your "Red Hat people are arrogant" hobby horse - RH per se doesn't use Bodhi, Bodhi is specific to Fedora. For RHEL, RH has an entirely different thing called errata tool.)

Ubuntu upgrade had our old Nvidia GPU begging for a downgrade

Re: Linux is forever?

Just don't buy NVIDIA. This has been the appropriate advice for Linux forever and remains so at least until nvk is fully ready.

nvk is now probably more or less workable if you have the right hardware ("Turing (RTX 20XX and GTX 16XX) and later"). It's still missing raytracing, which is kinda a big deal for high-end gaming, but it's much less of a pain to work with than the proprietary driver.

Yes, this is a problem if you want a really high-end gaming system, as AMD doesn't have anything that competes with NVIDIA at the really high end (and, as noted, nvk is not fully equivalent to the Windows driver yet). But for anything AMD covers, just buy a fricking AMD card, it will just work with none of the proprietary driver PITAs. (Or at the lower end maybe one of the new Intel ones, they seem to be actually good!)

NVIDIA does do the same thing on Windows: https://arstechnica.com/gadgets/2025/01/nvidia-winding-down-support-for-older-gpus-including-the-legendary-750-ti-and-1060/ and https://www.pcgamer.com/nvidia-drops-kepler-windows-7-8-8.1-support/ . I think the cutoffs are actually quite similar - basically anything before Maxwell is on a legacy driver, on Windows and Linux. On Windows, the I can't quite tell if this is done more aggressively on Linux than on Windows, it's hard to figure. It seems that for Linux the 'legacy' cutoff is currently Maxwell - Maxwell and later are supported in the current driver, anything before Maxwell is in a legacy driver.

However, it does seem like they keep the legacy driver available for newer Windows releases, to some extent, which is significantly different. Though it looks like Windows 11 cuts off support for anything pre-Fermi: https://www.reddit.com/r/nvidia/comments/o7ymr9/nvidia_gpus_that_will_no_longer_work_on_windows_11/

Tool touted as 'first AI software engineer' is bad at its job, testers claim

well...

"As an example, they cited how Devin, when asked to deploy multiple applications to the infrastructure deployment platform Railway, failed to understand this wasn't supported and spent more than a day trying approaches that didn't work and hallucinating non-existent features."

I mean, to be fair, have you met some junior engineers?

Boeing going backwards as production’s slowing and woes keep flowing

In case anyone's wondering why they're still making 767s...

...US couriers are still buying freighter variants, apparently, as there's no 787 freighter variant yet.

Intel, AMD engineers rush to save Linux 6.13 after dodgy Microsoft code change

it *was* reviewed, but...

It went through a ton of review. The version that got merged was version 7.

The complaint is that it got merged in the end without "the x86 people" - i.e. AMD's and Intel's kernel folks, mainly Boris and Peter - explicitly approving it.

(I had fun with this one, because it also caused a weird bug where Fedora installs would hit a kernel bug and hang during bootloader install. Took me two weeks to fully investigate and bisect it. Whee. https://bugzilla.kernel.org/show_bug.cgi?id=219554 )

Public developer spats put bcachefs at risk in Linux

Re: disagree!

It doesn't matter whether he's right or wrong. That's what "irrelevant" means. It's never necessary or appropriate to talk to someone like that even if they are, in fact, entirely and unambiguously wrong. It doesn't help anything. It only causes trouble.

disagree!

"Overstreet's post is over 6,000 words long, but in The Reg FOSS desk's opinion, he does in fact make his case fairly well."

Um. Really? I disagree *heavily*.

As a general principle, I'd say that anyone who can't manage to make a simple apology for writing "Get your head examined. And get the fuck out of here with this shit." when asked to, but *can* bash out a 6000-word self-exculpatory blog post full of entirely irrelevant technical detail, is probably not someone I'd want to have to deal with at work.

If you do something jerkish, own up and say sorry, then move on. If you spend weeks or months, and thousands of words, on "actually I wasn't being a jerk but also I was right and lots of other people historically have been jerks so why am I being called out?! It's unfair! I'm the victim here!", um, that's a bit of a red flag for me.

Fedora 41: A vast assortment, but there's something for everyone

anaconda is developed and tested to work at 1024x768. I do check it on 800x600 occasionally and it's mostly fine. This is some kind of specific issue on the Miracle spin.

Re: no RISC-V? Yes, that too

riscv is in a kinda bring-up phase atm and is building in a completely separate space from mainline Fedora, see https://fedoraproject.org/wiki/Architectures/RISC-V . but yup, it's there and it...kinda works? maybe?