Posts by that one in the corner

AI has made the Command Line Interface more important and powerful than ever before

My agent enabled me to route around the damage from software grown baroque.

Something about that sentence...

Yes, the consumer software is baroque, with gaudy exterior decoration intended to make it *seem* that money was lavished on it, when all you really have is the thinnest of gold leaf over papier mache (MDF in the modern day, which is even more finely mulched and slathered with cheap glue).

But at least you can see that, "enjoy" the glitz.

Meanwhile, the "agent" really is burning serious resources every time you use it, but that is subsidised so that - quite deliberately - it *sems* like a cheap thing to use. Although the construction methods for an LLM still boil down to mulching and immersing in (incredibly expensive) glue.

Re: AI can call APIs at willl...

Still waiting for your description of what computers "ought" to be like.

Although, as you think all CLIs need everything on a single line of text, thinking you need to go out and get some more varied experience in the subject.

How do you manage with as few as 6 mouse buttons????

Redragon M908 RGB Backlight LED USB Wired Gaming Mouse 18 Programmable Mouse Buttons

Of course, there is also the 3Dconnexion SpaceMouse Enterprise but I'm not sure that really counts as a "mouse".

Hmm, now wondering if I could add a proper optical mouse module onto the back of a pocket QWERTY keyboard...

Re: CLI? API!

> Microsoft just started creating APIs to Windows' in an attempt to make up for its inability to be automated as well as Unix

To be sort-of fair, Microsoft started with OLE so that Word could subsume all your other GUI programs (although I still think they nicked the idea from pre-release[1] PenPoint OS from Go). Then they had a rethink, and a proper process-separating OS, did COM and built OLE 2 on top of that. Years, years, later, after Microsoft had finally put enough COM layers into OS components that they could use them to write SysOp GUIs, PowerShell did the blindingly obvious and made accessing COM via a shell a normal thing (flipping thing back the other way) for the Windows Management Framework.

So, even for heavy-duty SysOp stuff, the APIs weren't aimed at "automating as well as Unix". And whether PowerShell even manages to do that is up for argument[2]

[1] note: pre-release, and early docs, before anyone points out that PenPoint v1 was 1991 and OLE was 1990; and both were preceded by remote function calls, but PenPoint was all about the GUI.

[2] on those bad days if "I give up, what random set of commands does StackOverflow say will do this thing", the Unix results always make more sense, and are easier to check man pages for, than the weirdly verbose PowerShell stuff; your mileage may vary.

Xen Project quietly announced five years of support for all releases

Re: Does anyone use Xen anymore?

I use XCP-ng (*not* Citrix) - but then it's for hobbyist/at-home use; also have KVM/Qemu and VirtualBox in use, with an eye on Bhyve.

Not exactly world-shattering use of Xen, but it is something...

Musk's Grok sparks outrage after chatbot makes offensive jibes about football disasters

I didn't get where I am today without knowing a classic reference when I hear one.

You do know more than one thing can happen at a time, don't you?

And you can't make any claim that this subject should be ignored because something bigger is happening elsewhere or you wouldn't have engaged with this article.

LibreOffice learns to speak Markdown in version 26.2

Ooh, look at that page telling you how open source Typst is, and how *you* can contribute.

See at the top of the page the link to their pricing, where they'll sell you back your contributions; you can even "Reach out" to them in supplication and they'll discuss how much you'll pay to run your copy on-premise.

Yes, the command line parser/formatter is free and on Github, but then it is competing with projects like pandoc, which also provides a "modern" (whatever that is supposed to mean) input.

> It's sort of LaTeX on steroids with a built-in scripting language

Huh? TeX (and therefore LaTeX) is a Turing Complete language, used to write, well, LaTeX. And if you prefer a more "normal" scripting experience (TeX, like every language, is most easily used for some domains than others) then there are well-supported variants with that ability as well: LuaTeX being the one with traction. Given all the packages for handling all sorts of document tasks that have been created over the decades, saying that *any* new project is "LaTeX on steroids" is making some pretty big claims! The Rust/Typst community is writing addons but they've got a way to go (didn't get any hits for their replacement for TikzDucks, a vital resource IMO!) and to demonstrate that they've got the longevity required to be *sensibly* used by the sciences[1].

(Probably coming down a bit hard, as new projects & ideas are worth encouraging, and beginning from a Markdown-alike probably *is* easier for people to *start* with than LaTeX, and I freely admit to using Markdown etc more than LaTeX in practice, but ... having spotted that the pronunciation guide for "Typst" makes it clear that the ending is meant to sound like, and evoke, "hipster" - bleeugh)

[1] There is a massive problem with long-term archiving of materials; any plain-text based format is an improvement but you still to know how to process that to recreate graphs and diagrams. LaTeX has its flaws *but* it also has demonstrated ability for widespread usage across platforms.

Except - original flavour Markdown allows for mixing in HTML; CommonMark does as well, although it is far more prescriptive (read: pretty decently defined, less ambiguous).

So you can still go to town with your styling, add some inline JavaScript and keep your fingers crossed about what comes out in PDF or printed form from each of the many different Markdown processors out there (and that doesn't cover what can be done by applying external styling via options to some of those processors, but at that point you've moved away from authoring Markdown).

I've done a couple of (non-WYSIWYG) pipelines to collate sections of Markdown into Company Approved Style and decided the best approach was to start them by using one of the C Markdown parser libraries, so that I could hook the normal "this is HTML, let it pass through" routine and make it generate an error message instead. That way the nice, clean and, as you say, totally stripped of choice, Markdown could be safely processed (top sheet, table of contents generated etc).

Dutch cops warn 100 alleged scammers: Turn yourselves in or we tell Grandma

Yay for vigilantes! That can never gone wrong, can it?

US state laws push age checks into the operating system

Re: Since when was it a perfect stranger's job ...

Now, jake, I understand that, having rejected it, you had a hard time attempting to refute my premise.

However, deciding to descend to personal insults when you found yourself unable to produce anything resembling a counterargument: well, it's that is unfortunate. All that can really be said about such a lapse is that, on reflection, it is best that you understand: nobody is angry that felt you had to do that, we are simply very disappointed. After all your tales of trials, tribulations and successes, that you are left with no better way to express yourself, you have let us down. Worse, you have let yourself down.

Re: Since when was it a perfect stranger's job ...

My response covered rather a lot more than the internet, as did what I was responding to: a blanket statement that was rejecting ALL societal requirements regarding parenting, simply because of one thing that jake - and you - disagree with.

Note that I DID categorically say that this law is stupid.

What I was objecting to - and still do, despite his rather sad response, something about parties, is his use of rabble-rousing words, intended to create a knee-jerk reaction "They has no right to ever interfere with my (or anybody's) parenting. That is crude politicking and blatantly untrue: there are parts of parenting that some people need to have enforced on them.

But, TO REPEAT, I DID NOT SAY that this law was one of those.

Re: Since when was it a perfect stranger's job ...

> Since when was it a perfect stranger's job to be the parent of MY teenager

Always.

Since you decided to have that child whilst you were within a wider society that makes any attempts to protect its members. As it must, otherwise it would fall apart. You would - I really, really hope - be praising that "perfect stranger" if the laws they made meant that police could intervene to stop another member of your society from doing something harmful to their children.

The range of possible harms is wide, including rape, hobbling so they can't leave the farm (and that can be psychological as well as physical), punishment with a cat of nine tails for talking out of turn, punishment by shunning for taking an extra biscuit. Grooming or conditioning them to consider such harms as normal and encouraging their participation in them, on either side. Or exposing them, not taking sufficient steps to prevent exposing them, to situations where the above may occur. All of those can be - are - "parenting" to the perpetrators, even "good parenting, pappy would be proud".

You decided to remain part of that society and therefore accepted not only what it gives you but also the limitations it places on you. And, depending upon how your society is organised and functioning, the means by which you can in return modify the rules to suit your own practices.

All that is happening here is that you have found a part of those limitations that you don't care for, "they" have stepped over the threshold.

And your response today?

You come out with a perfectly phrased politician's/rabble rouser's response. One that you know perfectly well matches that used by the very worst of the worst. Remove all limitations.

PS this law is stupid.

> How would they even find out the person is using linux.

You are assumed to be using a computer to "participate" in modern life; if you aren't using a computer then There Is Something Wrong About You and you will be investigated.

We know you are not using Mac or Windows or other Sanctioned OS because we can see what you are doing (rather, we don't see what you ought to be doing); you are being investigated.

Please answer that knock on your door.

Re: How does this work for shared devices?

Well, our library's computers already require a per-user (per card-holder) login, and the only time I used an Internet Cafe they gave me a slip with what certainly looked like a newly generated login in id. Too old for school (except at Uni we had distinct logins)

But then, They might just be watching me and letting the rest of you revel in anonymity.

Re: I see differences...

Well, if that stops the child cruelty inherent in the user inadvertently preparing a meal containing broccoli*...

* the value of cruelty derived from specific vegetables may vary according to your circumstances

Unpacking the deceptively simple science of tokenomics

Re: I've never read so much total utter bullshit before on the Reg

> WTF is "Goodput"?

The opposite of "badput", and the thing you want more of per unit cost when you aim to "optiput" your process. If you have too much, you'll need to "downput" (although the management prefer to say they are "rightputting" it makes them feel better, no need to consider they may be making "redundiputs" as they reduce the outputs; then they can get onto their put-put boats on the way to the putting green)

Chardet dispute shows how AI will kill software licensing, argues Bruce Perens

Re: .....................it's already here, as a service company

Why do I get the feeling that this Malus project was created by, if not a certain frequent-posting commentard with - shall we just say, unshakeable - views about Open Source and its authors, then somebody with whom they'd feel a lasting kinship?

Dan Blanchard, deliberately out to cause confusion and upset

The normal way to handle creating a project that is based upon an existing one is to fork the code AND CHANGE THE NAME before making massive changes. Ok, bug fixes and/or some additional functionality, which you intend to have folded back into the original code base, leaving the rest of it untouched, in those cases you leave the name alone and request a review and merge, please. But if your changes are not accepted, or you realise you've just radically altered 60% of the code - you CHANGE THE NAME and let your new project fight it out against the original: everybody finds it easy to compare the two and make their choice because you CHANGED THE NAME and (unless something stupid - and avoidable - has occured) both can coexist and be invoked separately and distinctly.

But overwriting *all* of the code of V6 and calling it V7?

Explicitly stating that it is All New Flavour - which in software terms means that it may, or may not, have copied over all the bugs and weirdities of the old code, but it is *guaranteed* to have new bugs and oddities - and to have them everywhere, unlike (hopefully) most releases where you can narrow down the verification to the small percentage of lines that changed.

If he had just created a new project and started making a case for *that* to be part of standard Python. - fine. Let all the (vitally important) questions about the licence status of this entirely new code be argued over - without impacting on the poor sods who are using the existing library and just want to get their own systems working without being forced into a potential legal & ethical quagmire.

If that wasn't done deliberately to cause confusion and upset, then Blanchard must be staggeringly naive, not just to social reactions but also in his simple observation of how projects are controlled and used, including the basic observation about allowing the simplest path for other people to evaluate and compare his work and therefore appreciate its magnificence.

Microsoft previews tech to ease creation of keyboard-accessible websites

Re: It's all for AI

> No human uses keyboard navigation

Ye gods and little fishes, it is bad enough to come across someone who clearly just piddles their days away at the computer instead of getting stuck into a task and wishing for every efficiency gain. But to justify themselves by claiming that *nobody* ever works that way...

This is clearly somebody for whom the phrase "hack mode" means nothing. Who has no idea even that ALT-TAB exists. Who never wants to copy that complicated declaration from the online docs into the editor, without breaking stride or losing focus. Who never sees a website with content that has value* and is worth quoting in your own piece, be that the amazing innovation that'll solve The Problem that has plagued Company Ltd for the last two years or just a diatribe on The Register.

Congress puts the ISS on life support until 2032, orders Moon base plan

Re: Depends

The paper Assessing the scientific and economic impacts of the experiments conducted onboard the International Space Station includes comparisons of commercial versus public sector science on the ISS (although that is not the paper's primary interest).

There is a considerable amount of commercial research done, by a range of companies: the ISS is working as a common asset and appears to provide good access).

The concern about a commercial replacement for the ISS has to be that singular: *a* commercial replacement (there may be two, but that feels unlikely - unless we get lucky and Elon starts spending his own money). It could be built and run by a consortium that provides equal access for comparative cost, including non-commercial rates to academic and other non-commercial bodies and which ensures that results are shared in the scientific press, encouraging innovative use of its facilities. It *could* be that.

On the ground, there are many private labs run by companies that jealously guard their interests. Luckily, there are also a good number (but never enough) of labs that are run by the public sector and academia as well. These provide for the bulk of the basic research, the "not directed at profits by next quarter" research, that keeps progress chugging along.

Without the ISS, will we see that public space research go by the wayside?

And that is ignoring the value of the international collaboration and encouraging youngsters to talk to the ISS...

Trump administration spoiling for a fight over global satellite regulations

Well, we are trespassing on US sovereign territory

After all, everybody[1] knows that the USA was The First in Space[3] and owns all of it. Why else would there be stars up there, eh? And Saturn's rings, what are they but stripes that have been joined at the ends to stop them fraying!

[1] within 16.4 feet[2] of Trump

[2] trying to be accommodating to that section of the US population

[3] don't fall for those rumours that say otherwise, trust what it says in your classroom's King Trump Version of the Bible; it is well-known that the Commies infiltrated everywhere and that their sleeper agents, in deep cover as tea ladies, were carefully trained and in place by October 1957 to execute the grand strategy of standing behind anyone with a radio set and making a "bleep" noise every 96 minutes.

Broadcom says AI companies can’t make their own silicon any time soon

Re: Hock Tan - Master of the Obvious

> Trump declaring war on the entire world and believing he can win...a foolproof bullet-pointed step-by-step plan

If only it was being done with that level of precision...

Once upon a time, saving your bits meant punching holes in floppies

8" floppies

> you started with a write-enabled floppy, and if you wanted to protect it, you punched a notch at just the right spot

Hmm. All the 8" floppies I bought[1] around 1980 (upgrading from punch card, yay) had the notch pre-punched, with a small sticky label already in place from new (and some spares in the box), so no need for risking yourself, just peel off the sticky. Always thought this was a good way around, as if the floppy was a bit old and the stickum dried out so that the label dropped off at least it failed by keeping your data safe from overwriting.

[1] still have on the back of a shelf, in a big cream-coloured plastic box; bit short on working drives, though. Not sure the discs will still read, which is a shame as I had some of my best angsty poetry on there.

Google stuffs Gemini into Android Studio Panda 2 to build apps from prompts

> The more precise you can be, the more likely it is too get it right.

Presumably culminating in the suggestion from TFA, to simply give it the entire answer up front:

>> More precise prompts, such as to use the well-established open source double dummy code on GitHub, albeit in C++

although the article still doesn't sound convinced that the LM would manage the job even with a complete crib:

>> would no doubt achieve better results

Users fume at Outlook.com email 'carnage'

Re: ipv4 downvoter

> Hey what is with the downvoter?

He[1] just checked his own PC's config and every one of those messages mentioned one of *his* IP numbers, then two others claimed that they owned *his* IP, so now he thinks we are all trying to "hack" him.

[1] it is going to be a he, shes are too sensible and would have RTFMed first

AI doctor's assistant is easily swayed to change prescriptions, give bad medical advice

You can't ignore a logged error just because *today* you had the safety catch on

>> as controlled substances can't be acquired through the program.

Not at the moment, because this is just a trial; very sensible. So there won't be any of those sorts of mishaps, at least, occuring.

>> Doctronic told us that it "reviewed the prompt patterns [Mindgard] reported as part of our normal review process..."

And did - what? Just note that it didn't *actually* let Mindgard have a bottle of pills, so nothing to worry about?

>> In short, neither Doctronic nor the state of Utah seem too concerned about Mindgard's findings since no one's actually getting a prescription cut for triple-strength Oxy or tricking their local auto-doctor into dispensing misinformation.

"Not too concerned" includes seeing no reason to reassure a tech site like El Reg by stating that they are carefully logging every time the current trial-mode limiting safeguards catch an attempt to prescribe controlled substances? So that when the time comes to evaluate the trial the question isn't "did it do anything bad?" but "did it try to do anything bad but was caught in the practice safety net?".

Because when - if - the trial period is over and Doctronic is declared usable, the practice safety net is rolled up, the training wheels come off, the extra-cautious limitations are removed and (as noted above) the overworked humans start to just "do what the computer says", if they haven't logged that 7% of the time it tries to prescribe opiates for a bad case of zits (hey, the common problem with zits is all bad social attitudes and resulting mental-health issues, so if the kid ain't worrying about the pustules then job done) but instead go by counting the actual prescriptions that got past the net and issued..

Just tell Doctronic that a session hasn't started

Good grief, an LLM that is more stupid than the average politician:

*Always* treat *every* microphone as live, otherwise what you say when off the clock *will* come back and bite you!

Malware-laced OpenClaw installers get Bing AI search boost

OpenClaw has tens of thousands of forks hosted on GitHub

Tens of thousands of developers are productively working on OpenClaw, it must be moving at an amazing pace!

Was going to comment on the difficulty of handling all the pull requests that are being generated - nearly 6,000 of them open at the moment, and 5,000 issues currently open - but of course they will be using their own AI to evaluate and manage all this churn, won't they!? Which they do, using at least Copilot and Claude. Gosh, wonder if any of the issues and pulls originate with AI (he asks, betting he already knows the answer)?!

And a release nearly every day.

This is all very impressive or all very scary (the words "uncontrolled churn" spring to mind, as well as the phrases "is everyone moving the project in the same direction, following the plan?" and "did you read and understand the release notes before updating?").

Google feels the need for security speed, so will ship Chrome updates every two weeks

Anyone here playing whack a web mole?

> releases new beta cuts of Chrome every two weeks, so enterprise users will have a chance for an advance look at any new features that might spell trouble for SaaS services to which they subscribe, or for other web apps

Is it easy for you to automate Chrome to run it against your SaaS and web apps?

It must be a nightmare of a job: you get a beta of Chrome, find it buggers up some third-party SaaS (not one as widely used as, ooh, Google Docs, but one your company relies on) and the clock starts ticking. Will the Chrome release have the same issue as the beta? Will the third-party update themselves to work around Chrome? Is the script to replace Chrome with Firefox ready to run across all the machines in departments X and Y (but not Z, who use a different SaaS thingie)? Will the new CEO understand you have bleep all power to change any of these systems, other than sending in bug reports and adding another pantheon to your daily rituals of appeasement? Will The Will Of Imhotep be satiated by chicken blood now that Waitrose have removed Egret from the shelves? Will *your* blood chemistry be ok if you crunch another roll of antacids?

Qualcomm, Nvidia ready for 'AI-native' 6G, if only the world knew what it was

Re: Utter bollocks

At the risk of being a kill-joy over the enjoyable "what you on about, ain't no useful 5G here, fool"

> ever since 5G came to dominate in the earlier part of the decade

is probably meant to refer to dominating the industry media machine's self-pleasuring exercises. Now that prattling on about 5G phones has moved from the Mobile World Congress to the Carphone Warehouse telly ads.

And, sorry to be disagree, but

> “software defined radio” in general is such a half-arsed concept

SDR lets (some) radio applications be performed more cheaply by allowing more generic hardware be applied, only changing the software/configuration side of things. That is all it need do to be a useful concept: let a few (suitable) things be done a bit more cheaply.

Of course, like absolutely everything else, you'll then get it hyped as far as it possibly can be, to the point of total charlatanry - although you gotta love the way that the UFO and ghost hunters have taken to running GNU Radio on their laptops: that scrolling frequency map just looks so sciencey!

US struck Iran with copies of its own drones

Re: Proof this was long planned

What about that cosy hotel that Trump is building right on the bay down in the south eastern corner of Cuba?

Re: Proof this was long planned

DJT will carefully examine each item placed before him and autograph it: "Who should I make it out to? Pete - can you spell that for me?"

Fly me to the Moon: NASA reshuffles the Artemis card deck

Re: What is the hard science objective for this mission?

How do they plan to avoid being truncheoned by the gas stoves living there?

They could put dimes into the coin slot, instead of shillings, jamming it and then waiting for the meter to run out, but doing that relies on catching the beasties: do any of the announced Artemis crew ski?

Popular prayer program becomes propaganda pusher after reported Israeli hack

There's very little app users can do beyond

Not giving in to this fetish to UPDATE NOW! Don't stop to check release notes, UPDATE NOW!! Doesn't matter if this version doesn't fix any bugs, let alone any CVEs, UPDATE NOW!!!

Too slow, we just changed the background colour, you must UPDATE AGAIN, NOW!!!!

To heck with opportunities for State Hackers to get in and change the app, if you follow that advice you are less likely to be caught out when a perfectly useful app, which you paid for and everything, is bought out by some random company who suddenly fills it with ads, removes half the features you were using and generally takes the piss.

BTW, v17.3.5 is out, UPDATE NOW!!!!!

OpenAI’s Altman says Pentagon set ‘scary precedent’ binning Anthropic

Re: Grandad, why do we live in an underground sewer?

Hey now, no needs to shout, I's ain't deaf. Martha, kin you bring us both some lemonade-colored water substitute?

Re: "Grandad why do we live in an underground sewer?"

Speak up Sonny, yer ole gran'pappy don't hear so good no more.

Windows 11 tops market share as 10 faces extended farewell

Windows 11 at 72.57 percent and Windows 10 at 26.45

Yay, W'11 is in the winner, in the prestigious category of "all the Windows boxes we saw".

Just nudge the "of all the boxes, total" under the carpet. ChromeOS? never heard of it. iOS? Now you are just making up silly names.

Open source devs consider making hogs pay for every download

Re: Easy solution

Yeah. I had it working for a few years - not a large dev department, but we had a number of distinct projects that would be active in parallel. I took the unofficial role of the librarian (weren't using SA terminology at the time), checking which licences were in use, generating the declarations for SBOMs from the build system blah blah.

But I think I can say with absolute certainty that that all went by the wayside pretty much as soon as I left (except for contractual fixes to those older projects), as the Bright Young Things were having none of it. Including the person who was formally announced to be the SA across all the dev teams! I'm pretty sure that the SBOMs, if done at all, are now back to being manual, and as one of the last things I did there was to discover an undeclared GPLed library where none should be... And, yes, there already were excess XML libraries in gleeful abundance, also DOM'ing when SAX would have made sense (and probably vice versa).

I wish them well, but glad not to have to worry about this stuff now. Although - still use the same build system at home, so if I ever do let some poor fool have a copy of any of my for-fun projects, it'll come with a SBOM :-)

Re: Easy solution

It all gets very murky :-)

You can have a VM image that is used as the base for each time the VM is started up, so each VM instance starts identically (bar the necessary differences set by config outside that image, e.g. network addresses). That sounds very sensible and controlled.

The first week this is used, a pull is a "nice normal pull", bringing down just a small number of updates. Although, yes, each VM instance is getting the same data time after time.

It is all working well, so no need to update the base image, doing that requires change orders to be signed off and all that hassle.

Months - or years - later, those VMs have been working ok, no failures being reported back to the PHB's dashboard, so let them carry on. However, the repos being pulled *have* been changing, a lot, so now instead of the pull only calling for a few recent changes, each one is now calling for 95+% of the volume that a fresh clone would do. And no, the commands within the VM base are not properly using the "depth" parameters to stop all the history being dragged down (well, the entire issue is about the abusers not bothering to optimise the downloads, even though we are all shouting "you don't do it like that!"). And each new VM instance is repeating the same downloads as all its siblings.

Re: Easy solution

Having your software architect keeping control over the proliferation of external dependencies is not about about continual inter-team meetings.

Nor is *not* controlling your externals about "team autonomy". Not unless you definition of "autonomy" fundamentally includes wasting time by pointlessly duplicating effort. Effort not just for you: you are checking for all the potential non-technical issues surrounding an external, aren't you? Legalities?

In case you don't know how it works:

You determine the need for functionality X; hey, maybe that is your entire team or maybe it is literally just you making the determination. Then one person - just needs one, but you can take a friend - goes to the SA with these requirements, and probably one or two names of possible solutions. He points you at the already in use, tried and tested, solution - which may be another project in the company or may be an external that is already in the company library, open source or proprietary. Plus a list of who is already using it and can help with any issues you may have, and any notes about legalities your team's project now has to abide by (e.g. the correct and consistent declaration for your software BOM; seat licensing; a statement that your project is for internal use only can not be released outside of the company because it is now using GPLed code and you don't - yet - have clearance to ship the company-written portion). If there isn't already something suitable - or you have a *really" convincing argument why there needs to be yet *another* XML parsing library added to the list - then you and the SA work together to ensure all those boxes that you don't care about, but Legal and Contracts get worked up about, are ticked. And for your sins you get to be added to the list of "any questions, this person has experience with the thing".

No doubt this all a pipe dream to many, but if you are wasting both your time and the company's EITHER duplicating effort OR getting tangled up in all those meetings, instead of being able to treat a common collection of libraries & utilities &c as a trustable source that can be accessed without involving your PHB, then something has gone wrong.

Or you are really just a one or two man band and all of this is overkill - and you probably also are not the source of the problem to the repos that TFA is taking about.

Trump orders purge of 'woke' Anthropic from government

Pope Urban II sticks his hand up and waves for attention; Pope Innocent III puts down his beer.

Re: Kathy Burke said it best...

> equality of opportunity, which is a right wing belief

Bwaahaaahaa ha ha ha! (wipes tear from his eye) thanks, needed a good laugh.

Re: US military reportedly used Claude in Iran strikes...

Given the short time between the ban and the actual strike it would be almost miraculous if the US military had managed to 100% - or even 1% - obey the ban in this cae.

Making *any* kind of random change to the software in use that close to an operation would be absurd. Even if it was only a mid-level commander needing to remember to log out of his Anthropic account before asking his phone where he could get his favourite sandwich, doing so would be way down on his list of priorities, no matter what any of the "Trumpings must be obeyed IMMEDIATELY! IMMEDIATELY!" commentariat might say.

Re: Grok replaces Anthropic

== Ah, I see your confusion. It is car washes, of course.

Oops, spot the bad URL. Incredibly bad URL, in fact!That was supposed to read:

Ah, I see your confusion. It is rain that you have to keep off your Cybertruck. And keep well away from car washes, of course.

Re: Money talks - don't sing or walk

herman> Anthropic cut a deal. So much for all the Woke and Liberal BS.

Are you confusing OpenAI and Anthropic after the headlines such as OpenAI strikes deal with Pentagon, hours after rival Anthropic was blacklisted by Trump?

Or do you perchance have a citation you'd like to share with the class, one that hasn't yet hit the headlines us mere mortals read?

Re: Grok replaces Anthropic

Well, Elon said the Cybertruck will float https://www.motortrend.com/news/elon-musk-tesla-cybertruck-float-water-claim-twitter and who are we to disagree with the Muskmeister?

> I thought cybertrucks were even less water tolerant than the wicked witch of the west?

Ah, I see your confusion. It is car washes, of course.

But operating as a battleship? No problem. It is bulletproof, after all (note: guarantee void if enemy throws the shot by hand). Why, you'll even get extended range as the salt water action charges up the battery for you!

Re: This presents us with a disconcerting conundrum.

> Anthropic are AI doomer nutcases

Totally unlike OpenAI - oh, wait, wasn't it Sam who told us “I think that AI will probably, most likely, sort of lead to the end of the world. But in the meantime, there will be great companies created with serious machine learning.”. That sounds like doom to me, brought on us by nutcases (and a massive dose of "but we'll be rich and won't care", dancing as they cause the world to end) - and that Forbes article is an optimistic one!

Re: Sugardaddy

I don't disagree about the annoying lack of low-end video cards. However, in response to:

> In 2020 ... I wanted to stick a £25 card in a machine ... Cheapest I could find was over £100

I have to report that in both 2021 and 2024 I bought 2GB GT-71O cards for under £40 (and those are OTT for my requirements so, yes, I'd've gone for £25 cards had they been available - but see also admission below).

Now, I don't have quick access to purchase records for 2020 (those two, above, were from the dreaded Amazon Prime - I know, I know, but needs must etc - and were v. easy to look up) but I also don't recall 2020 as being *that* bad for low-end stuff. Yes, prices in 2020 were up for webcams, for some strange reason, and gaming rigs & consoles were more popular, and costly...

> PSU, ITX m/b, processor, 8GB, SSD, DVD drive – for under £300...that's been impossible since 2020 or so

Um. October 2024. Mini-ITX with N100 built-in, £140. 32GB DDR £50. 1 TB SSD £60. Teeny Travla case, with PSU, £37. Ok, no DVD drive, but a SATA internal or USB external is yours for about £20. So just drop the DDR to your 8GB and bish bash bosh, easily under three hundred guv.

Now, that case is - utilitarian - and uses an external power brick (supplied) but it does have a nice carrying handle; even most ITX cases are way more than that. But the board runs really cool so could be left in its inner cardboard box (with a few holes punched through, don't torture it), so I could have saved a few more quid - and it would match the "external" (i.e. SATA cables out the hole where the DVD would go) SSDs that I have since added (RaidZ if you are interested).

Re: Kathy Burke said it best...

>> usage policy prohibits directly using Claude for domestic surveillance or in lethal autonomous weapons

> What line of business did Anthropic think the DoW was in?

A massive amount of paperwork, logistics, supply lines around catering, haberdashery, cobbling, vehicle maintenance etc etc.

The actual shooting of guns, autonomous or otherwise, generally takes place outside of the DoW itself, as does the bulk of surveillance (on the basis that the bulk of the surveilled isn't traipsing through the DoW corridors).

Plenty of scope for sticking to Anthropic's terms and conditions, even if readers here may still think those applications of Anthropic's services leave too many loopholes ("Why are there 4,000,000 left boots in Maine and 4,050,000 right boots in Dakota? And why are all of them pink?")