Posts by Philip Storry

Microsoft teases ‘reimagined SharePoint experience’ with added AI

Re: Reimagining

Your comment confuses me.

It implies that SharePoint has worked.

Do we have evidence for that assertion?

Re: Valuable path length ...

I agree, and I partly covered that in another reply. But that was about NTFS and file lengths, and in replying to you I'd like to look at the other side...

Microsoft made a mistake when they tried to ape filenames in their URIs.

Take a look at Google Drive - it doesn't care what you called your file. Be it Doc, Sheet, or a PDF you uploaded, it gets a link that's just an alphanumerical string. Drive is effectively an Object Store, but the interface is good enough that nobody ever notices.

That gives Google a huge amount of flexibility.

OneDrive came out of Microsoft's Groove acquisition, and at that time Microsoft were still very much in their Monopolist Phase. So it was very natural that they'd try to attempt tight integration with the OS, and go with aping a filesystem on a webserver instead of using an object store. Personally I think that with the benefit of hindsight it was the wrong decision. It's too limiting on both ends of the solution - client and server.

They should have gone for an object store on the server, and a client that does local retrieval/sync/caching/searching of those objects. The client should also have an API so that local apps can work with it natively if they want to. But the danger there is that Microsoft would have to open up that API at some point, plus building this is a lot more engineering work for them. So they went for the fast option, despite it being complex, fragile and limiting.

Still, hindsight, eh?

Re: Valuable path length ...

Oh yeah, there's definitely potential for an issue on SharePoint's side.

Technically NTFS supports paths up to ~32Kb in size. You have to enable a registry entry on Windows to use that in applications, and it will break a lot of older apps that assume a 260 character buffer for path names.

For us sysadmins it's not unusual to see a fileserver with deep paths beyond the usual 260 limit because the share itself was created in a subdirectory on the disk. Also Apple clients or Linux clients care not for your pathetic legacy Windows 260 character limit, and will happily spaff files that go way past that limit. And NTFS will happily store them. Windows Explorer, however, will lose its tiny little mind trying to handle them unless you use various tricks.

I think that the URI limit for OneDrive is reputed to be around 400 characters, but it's the kind of thing that can vary depending on the action being requested.

Frankly, as most people have "C:\Users\John.Smith\OneDrive - Company Name\" placed before any file that they try to upload, they're much more likely to hit the 260 character limit issue than the OneDrive issue. Hence they blame the filesystem. Which is a shame, as OneDrive/SharePoint are much more deserving of ire than NTFS is...

Re: Valuable path length ...

NTFS is a decent filesystem, and I feel I should come to its defence.

Sure, it's not the new CoW hotness. But that doesn't make it a bad filesystem, surely?

NTFS has outlasted every one of its peers except for XFS and variants of FAT - which is terrible, but just won't bloody die because it's so simple that even ChatGPT can understand it.

Comparing it with XFS, it comes out favourably. Both are decent filesystems, fast, reliable, with a good feature set. Not identical, so most people are going to lean towards one or the other, but they're closer than you'd think. I'd compare NTFS to FAT, but that's like comparing a modern car to throwing yourself off a cliff - both can get you to a destination, but one of them has safety issues.

The main issues with NTFS are what's above it. Windows' IFS stack means that NTFS often has lots of cruft intercepting file operations before it's even handling them; the various Windows file protection systems, an antivirus shim, OneDrive, and more. If you create a second non-OS partition and format it NTFS, you'll find it performs very well.

Microsoft have tried to replace it with ReFS, and it hasn't really taken. I built a fileserver two years ago and gave thought to using ReFS, but then realised that I could be forced by my Corporate Overlords to switch AV or backup solutions at the drop of a hat. Did I really want to risk incompatibilities with ReFS? Especially when NTFS was dong just fine, thank you. So NTFS it was.

For over 30 years NTFS has been one of the safest, most reliable and performant filesystems in the industry. It should get more respect for that.

UK to demand social platforms take down abusive intimate images within 48 hours

You missed a bit...

X's statement is clearly unfairly truncated. I believe that the original stated:

"We remain committed to making X a safe platform for everyone and continue to have zero tolerance for any forms of child sexual exploitation, non-consensual nudity, and unwanted sexual content. Unless Elon thinks it's cool, in which case our position is obviously: **** you, we'll do whatever Elon wants us to."

It's shocking to see such a negative portrayal of the poor, honest, trustworthy, and reliable X. It's not like they've ever done anything to deserve it!

If Microsoft made a car... what would it be?

It's a fleet, not one car

It depends on which product we're talking about.

Windows has very solid underpinnings. It's the stuff on top of it that's the problem. In that regard it's like a very solid supermini that's got sporting pretensions. Remember when the original hot hatches came out, and everyone had to have one, so we got a "sporty" Austin Maestro? Which wasn't actually sporty because they'd fundamentally misunderstood the brief. Well it's like that, except with a much better set of running gear under it.

I'm guessing it'd be like having Pagani try to make a hot hatch version of a Honda Jazz, except that their budget for it was "three quid, all the coffee you can drink, and whatever you can find in your factory". Oh, and they weren't allowed to touch the engine or other running gear. Everything is trying to be wild and crazy and modern, but there's no good REASON for it to be like that. And whilst a Zonda's nuttiness is compensated for by the sheer thrill of driving the thing, this still moves like a Honda Jazz, making its insanity unforgivable.

Office, meanwhile, has almost the opposite problem. Its underpinnings are a rotten, ancient, godawful curse - but they're trying to make a functional and attractive modern vehicle on top of that. The Microsoft Office car is probably more like someone trying to make a modern Crossover on the running gear of an Morris Oxford Series III, better known as the Hindustan Ambassador. The senior management love the idea because they get to re-use all this existing technology, but it's simply not capable of doing the actual job properly. Everyone using it knows this, but it's still a ubiquitous car in some places for historical reasons.

Everyone will have their own opinions of what other products like SQL Server or SharePoint are.

Microsoft isn't one car. They're a fleet of cars, and your management is determined that all Company Cars will be selected from that fleet. Which kinda sucks, as it means that the low level engineer gets a good car, and the high level manager gets a clown car that falls apart every time it stops moving. (I think we're all agreed that's SharePoint, right?)

Edinburgh councillors pull the plug on 'green' AI datacenter

It was vibe planned anyway

Apparently they didn't even approach the local water company about how they'd get water.

Lucky that datacenters don't need much water, right? It's just rows and rows of computers, which hate water. No need for it at all.

Pardon?

They actually need lots of water? For the cooling systems?

Oh.

That's that plan screwed then...

AWS says you're on your own if media codec patent owners come knocking

Re: Excerpt from the Annual Meeting of Evil (Self-Described) Geniuses

No problems. Thanks for the question!

It took a bit of editing to reply, but was worthwhile so as to make it pertinent. There's a lot more to discuss - patents vs copyright, who benefits from patents, and market benefits. But they're topics for some other day... ;-)

Re: Excerpt from the Annual Meeting of Evil (Self-Described) Geniuses

No. I'm not completely against the idea of software patents, but existing patent terms are a bad fit for software. I don't think that they are a fair balance of protection for labour versus effect on the market.

A patent is a government granted monopoly. We allow the time-limited protection of that monopoly in order to encourage innovation. But software patents have the opposite effect, locking up techniques and advances for far too long.

The term for a patent is 20 years, which is an eternity in software. The 20 year period is OK for physical inventions where someone may need acquire funding, to set up manufacturing, distribution and so forth. In the physical world a patent provides protection against someone else, probably a large company or a rich person, gazumping you by repurposing their existing funding/manufacturing/distribution.

In the digital world the barriers to entry are much, much lower. DOOM was all around the world in the 1990s long before physical versions were even being printed, let alone heading towards the shops. And that's in the pre-internet digital landscape. You can now write software now and have it available anywhere on the planet within minutes.

The manufacturing and distribution stages are at least an order of magnitude, probably several orders of magnitude, easier for digital goods (including software).

I think in the current climate a term of 4-6 years would be OK. I might even be persuaded to go as high as 10 years. But 20 years is simply ridiculous. It holds the market back too much.

And this is before we even start talking about overly broad patents, and how the patent assessors are underfunded, overworked, and ill-trained to see what is obvious vs innovative.

Excerpt from the Annual Meeting of Evil (Self-Described) Geniuses

From the bowels of an intellectual hellscape that really shouldn't exist, we present: Software Patents!

Yes, that's right, we've managed to convince politicians and courts that mathematics and logic should be be patentable.

And boy, are we going to abuse this! It's very much a long-term play, but we think we could potentially hold back human progress by 20 years. And that's not a fixed term of 20 years, that's a sliding window we can keep moving into the future thanks to the magic of generating new patents or updating old ones!

Now, on to our next item on the agenda - charging people for air.

Whilst we've made great strides in getting people to pay for water, air remains a more difficult challenge...

Europe shrugs off tariffs, plots to end tech reliance on US

Nice work.

Now here's your coat, go away and reflect on what you did here... ;-)

I believe that if we're to create any kind of Sovereign Cloud it will have to be through a government-owned company.

Partly so that profits can be re-invested to assist in growth and service development, but also to avoid loss of sovereignty through sale.

The US Cloud Act means that any US-owned company is incapable of providing the required security. As does an act in China. And no government wants to be seen to stand in the way of the Almighty Market, lest the City take its revenge.

So we should prevent it from ever being a problem by making it a government-owned but arms-length and not-for-profit company.

Latest Vivaldi release surfs a wave of anti-AI sentiment

Re: Wait, What?

You're quite correct.

I had a brainfart. I simply thought "It's an awful idea, and the company name starts with P" and my brain autofilled the rest. Incorrectly.

Thanks for the correction!

I'm enjoying Vivaldi

I switched to Vivaldi last October, and I'm happy enough to have given them some money for it.

If anything, it has too many features. I'm occasionally stumbling over a new one and finding myself say "Well that's quite nice"... their focus is definitely more on power user web browsing. A good example being not just having workspaces, but also rules so that certain links always open in certain workspaces. I don't think I need that, but there's a chance...

My main reason for switching was just to see if I could survive without Chrome. Over the years I've used Netscape, Opera, Firefox (back when it was Pheonix and then briefly Firebird) before settling on Chrome. There was nothing wrong with Chrome as a product. Then Palantir said that they'd happily buy Chrome if Google were forced to sell it, and completely coincidentally I thought I should see how much friction there was in switching browsers...

Here I am, over three months later, and I'm still using Vivaldi. Highly recommended. Give it a try.

Old Windows quirks help punch through new admin defenses

Re: DOS? There's still DOS in windows?

There's no DOS code in Windows as far as I know.

But there are plenty of applications out there which still use old DOS virtual device files, so Windows has to present them or risk hundreds of thousands of apps breaking.

These virtual device files are things like COM1-COM9, LPT1-LPT9, CON, AUX, NUL.

That last one is probably the killer. These virtual device files exist in every folder in DOS, so a common way of testing a path exists was to look for <pathname>\NUL. If that file doesn't exist, then the folder path doesn't either. You may have seen this is many a batch file back in the day, but somehow the exact same test made it into the core utility libraries of many, many programmers. And therefore many, many companies.

Killing the NUL virtual device file would probably break a huge amount of software.

The "DOS device object directory" referred to in the article is the system that creates and manages the mappings that allow this compatibility.

I hope that clarifies everything, feel free to ask if you have further questions.

Succession: Linux kernel community gets continuity plan for post-Linus era

So negative...

Everyone's assuming that this would be because something horrible happens to Linus.

But it might be something good that makes him stop working on Linux.

For example: what if he gets an offer to go and work on something big and professional, like GNU?

(I'll get my coat.)

AWS flips switch on Euro cloud as customers fret about digital sovereignty

Re: Performative hyperscaling

Surely all of Amazon's mime artists are busy manning the phones on their customer service desk?

Performative hyperscaling

We have entered a brave new era - the era of performative hyperscaling!

This won't stop the US government from using the US CLOUD act to demand anything that it wants, for any reason. It cannot stop that.

But it can make it look like they're doing something about it.

Not anything effective, of course. The only effective change would be to move Amazon as a company to another country, so that they are free from the obligations of US law.

Still, as they say: "We must do something! This is something. Therefore we must do it."

It truly is a brave new world... (Same as the old world.)

Techie banned from client site for outage he didn’t cause

This brings back bad memories

I don't recall ever being banned from a site, but I do remember a colleague of mine getting very close.

I was working as the head of the support & implementation team at a small company that did software development and consultancy. We had a client we'd recently migrated to a new mail system. They then reported a problem.

I was working on another project so I sent my colleague John* to fix it. (This was in the early 2000s before remote support was commonplace.) After some basic checks John was very confused by what he saw, called the office to confirm that the system had the wrong settings, and then fixed it.

A day later, the exact same issue. I considered sending someone else, but I figured I'd send John again as he now had some familiarity with the site. When he arrived they were unhappy, but he cracked on and found... the exact same issue. We spoke briefly on the phone, and confirmed through date/time stamps on the configuration that this had been changed, and that only certain admin accounts had access to change it. John fixed the issue, reported all of this to the client, then left.

The next day they had the exact same issue. The client was livid. They were also demanding we send someone "more competent" than John. I happened to have finished my other work so I went onsite to give a show of "seniority", in the hopes it would calm them.

Same issue - the configuration had been changed manually. Before leaving the site I asked if they had a CD burner. They did, so I created copies of the configuration and burnt them to two CDs - one for us to keep, one for the client. This was so that we could prove a change was definitely happening.

Whilst I'd been on site poor John had been back at the office in a rather awkward meeting with one of our Directors about his performance. This was John's first job, and he'd not long been with the company, so this was not a fun experience for him. He was very worried due to all of this, fearing he'd have a career outlived by most mayflies. I reassured him and made sure to speak to the director the next day, making it clear I trusted John completely and didn't trust the client.

Two days later, the client has yet another failure. I went out to the client, as they were still unhappy with John. It was the exact same configuration change. But this time I invited their IT manager to compare the backup that they had with the current configuration, and showed the date/time stamps on the current configuration. I made it clear that someone with admin access to their environment was making this change and sabotaging their system. With no other explanation available, he accepted that this. I fixed the problem, took another backup for evidence in the future, and went home.

A few days later our director spoke to me about this. After some internal investigation the client found that one of their technicians didn't like the new system, and had been sabotaging it in the hopes that they'd roll back to the old one.

I made sure that John got an apology from the director, but was less than happy that they'd not contacted myself or John to apologise.

When the same client had some deployment work a few weeks later, I allocated the visit to John. He didn't want to go, but I told him that I expected them to apologise to him in person, and to be more professional with him from now on. If that didn't happen, I'd never send him there again and would tell our management that this client wasn't worth the hassle and we should not renew their contract when it expired. To the client's credit, they apologised to him and treated him well.

Some clients simply aren't worth the hassle that they bring you. Being banned or unwelcome at a site is one of the (many) danger signs.

Looking back on it, I'm mostly happy with how it went. I should have asked John to take a backup after his second visit, but didn't know that they had a CD burner at that point. We did change our standard operating procedures after that, making sure everyone took spare CD-Rs onsite and offered to create a backup of configs before leaving.

--------

* Name changed, of course.

Earlier Horizon rollout could widen net for quashed Post Office convictions

Re: Anecdotal evidence

ICL were my first employer. I was later very shocked to find out that other companies did not work the same way.

I have many stories which not only prominently feature their unusual management practices, but would be impossible without them.

So yes, it's quite possible that there was no beta or pilot.

But the one thing I'd like to stress is that a pilot would probably be billable. And being billable made almost anything possible within ICL.

Therefore I suspect the chances of a pilot were high.

When I reflect on my time at ICL I come to the realisation that baffling, trying circumstances were inevitable. The only difference between billable and non-billable bafflement was that Management would acknowledge, even celebrate, the billable bafflement. Non-billable bafflement was ignored or actively denied, despite its obviously plentiful portions.

Anecdotal evidence

Software doesn't just magically appear. There are betas, pilots, and so forth.

I worked at ICL from 1995 to 1998. I was at the FCY03 building in Footscray, Sidcup. It was a three floor building with many outsourced support helpdesks in it. The flagship client was Microsoft, but there were many others - Escom, Seagate, Compaq, Polaroid, Motorola, Apple, just to name a few. (Actually if I recall correctly Apple were in FCY02, but many of their systems were hosted in FCY03 as 02 was a tiny outhouse of a building by comparison.)

I recall that in 1998 a small helpdesk for the Post Office was set up. It was on the ground floor, near the front of the building, at least to start with. I think it was intended to be around three to six people, quite small by our normal standards.

I'm sure I was told that it would expand as we were going to be supporting a new system for post office branches. I don't recall it ever arriving whilst I was there, and I think that when I left ICL that helpdesk was mostly supporting the Post Office's Post Code Database software. But there was a constant rumour that this tiny helpdesk would expand "soon" as they were just about to start rolling out this new system to post offices.

Aaaaaany day now. Any day. No, really, it could be any day now. Honest.

The reason this sticks in my mind is that there wasn't much more space on the ground floor. If they were going to grow that team, they'd probably have to move it to another floor. Maybe even another building. Which would mean not just repatching network ports but possibly also running new connections between floors or buildings, and all the hassle that can entail.

Around mid 1998 this ceased to be my concern as I moved on to pastures new.

But it wouldn't surprise me if there were some earlier cases than 1999. My recollection is that they were definitely saying they'd be supporting it. Any day now... so there were wheels moving in 1998, and someone may well have been using it in a pilot somewhere...

AMD clocks in with higher CPU speeds, leaves architecture untouched

Incremental improvement is good

The same Zen 5 cores?

Good. I'm totally fine with that, because the Zen 5 cores are excellent.

"We took our excellent thing and made it slightly better" is pretty much exactly what I want from hardware these days.

(Cramming AI in there? That's just the market these days. Even if I can't forgive it, the ceaseless waves of AI bullshit leave me with too little energy left to condemn it.)

Doom hits KiCad as PCB traces become demons and doors

Maybe DOOM is an inevitable universal constant?

I have a growing suspicion that the only reason our universe exists is that some entity wanted to know just how many things DOOM could run on. So it span up this universe, just to find out.

The possibility that we only exist to explore the DOOM runtime space is both terrifying and yet strangely comforting, if only because it makes a lot more sense than many of the other reasons offered for our existence.

Who gets a Mac at work? Here's how companies decide

Another point of anecdata

My employer was "PC first, Mac as an Exception", but is now "PC Only".

The reason for the change was a restructuring in which they let go of half the IT department. The cost of that is that we simply can't afford to manage two platforms. Packaging & deploying software on one platform is a pain, doing it across two is much more so. And that's before we get to the additional expense of the tooling used for that. Servicing just one platform is always going to be cheaper and easier, and for us Windows is the platform that all departments can use - some departments have software which can't run (or run well) on Macs, so it was an easy choice.

To put it simply, going PC only made things much cheaper.

Our parent company deploys Macs, but they're much larger so can more easily handle the costs of that.

One observation I'd share is that on the costs in terms of managing the machines (not purchasing costs) is that Macs are oriented towards being managed by a user, PCs are oriented towards being managed by either the user or an organisation. GPOs aren't perfect, but you can do a lot with them. You often hit restrictions on Macs where Apple will simply say "we don't allow automating that for security reasons", and have to do things manually. And that's the right stance for individuals, but not for a fleet of hundreds or thousands of machines. I'm sure Apple are improving on that, but there's definitely a difference in mindsets when it comes to machine management.

Ultimately there's no single answer. It will be different for all companies/organisations.

Texas senators cry foul over Smithsonian's pricey Space Shuttle shuffle

Risks?

As ever the Republican Senators are fixated on costs, without seeking to understand the risks.

I assume that the private quotes they got include a damages clause for this irreplaceable national artefact? Right?

Yeah, I thought as much.

These schmucks are only concerned about a cheap success, and unwilling to fund a proper job that reduces the risks.

Farewell Discovery, it was nice knowing you whilst you were still in one piece.

To digital natives, Microsoft's IT stack makes Google's look like a model of sanity

His mistake is in thinking it was designed...

None of this was designed at a larger scale. It's a mess because the industry has been messy.

Taking SharePoint as an example, it was built to provide a web-based Groupware system.

Microsoft's first attempt at Groupware was Exchange Public Folders, which were pretty bloody awful yet somehow still exists today. Meanwhile the web was happening, everyone wanted an Intranet, and Microsoft had no product for this. Well - they had IIS as a server, they had Office putting out HTML, but unless you wanted to upload your files via FTP or maybe tinker with WebDAV (remember that?), you were out of luck in Microsoft's portfolio.

Meanwhile, Lotus Notes has gained a webserver and can publish your applications to your intranet without much work at all. Yes, they're ugly as hell, but this is 1998 and most websites are ugly as hell - what terrifies Microsoft is that you can convert Notes applications to web applications with a few simple changes to your servers.

So SharePoint was born. It was a supermarket sweep trolley dash of technologies, taken from various Microsoft teams - Office, Development Tools, Server. It was such a rush job that a version 2 came out fairly quickly afterwards that broke compatibility with the first version.

SharePoint was never designed to scale up, or to be easy to integrate. It had a reputation as being a big, heavy, awkward product that nobody loved, but that fitted with company's IT strategy so was used anyway.

The fact that we're still using it is mildly shocking. SharePoint should have died years ago. It's one of the least popular offerings from Microsoft - for all their audiences. Developers, administrators, users - they tolerate SharePoint, knowing that better solutions are available but that this is what the IT strategy says we have to use.

The real surprise is that, having bought Groove, they then decided to integrate it with SharePoint and later use the Groove sync client to create OneDrive. Again, that wasn't part of any long term plan. It just happened. Microsoft needed personal storage, and they rummaged around in their product bag to see what they could cobble together quickly.

Give it twenty years, people will be saying the same thing about Google's platform. Their big advantage currently is that they offer a lot less, so there's less to integrate. That can't last...

EU starting registration of fingerprints and faces for short-stay foreigners

Re: Another brexit benefit

And we can look forward to the Quitlings complaining about this.

Because Sovereignty is something that only the UK should have. And any time another country (or grouping of countries) exercises it in a way that annoys them, they're happy to be hypocrites.

If we had such a thing as a "self-awareness scanner" we could use their headlines & tweets to calibrate it to zero. ;-)

Mexit, not Brexit, is the new priority for the UK

The elephant in the room is Brexit

This is the kind of project that would be much better done with the EU. If we're serious about doing it, then having the resources and talent of 27 other countries is going to make it much easier to be successful.

There will of course be those who will say no to that. They will trot out the usual "the EU can't do this, it'll be a waste of money". But those same people are almost certainly already here saying the same thing about our own government. So if we're going to do this, then we have to simply smile politely and nod at them, then ignore the buggers and go and do it.

It would be much simpler if we were still in the EU. We could point at GDS and say "we already have a model for doing this kind of thing, why not reuse that?"

But now we're out we'll have to foot the entire cost ourselves. At a time when we have a hole in our budget.

The key to remember is that the goal is that much-misunderstood word "sovereignty". In that regard having a close partner from the outside is a boon for the EU, as it helps remind them ensure that country's sovereignty is a priority, rather than the EU's needs. The EU needs to avoid the US CLOUD Act, and so do we. Our needs and goals are aligned here.

On the technical side, start with cloud infrastructure by creating standards for blob storage, then block storage & hypervisor infrastructure. At the same time start work on client software, investing in LibreOffice and similar tools. Make sure that everything is publicly, openly specified. Work on getting federated systems that allow for choice and resilience. Aim to finance local bit barns by getting sensitive government workloads moved into them, as well as opening them up for other customers.

This isn't impossible. It's actually quite possible, just very expensive. So sharing the costs with our neighbours is the obvious thing to do. We have something to offer them, they have something to offer us. The real problem is Brexit, which has boiled enough brains that the mere idea of cooperating is somehow unacceptable. It shouldn't be.

Microsoft pops legacy Exchange public folders on the chopping block

A long-lived failure...

After the major SharePoint issues at the tail end of last week, I noticed that some coverage came with a brief explanation of what SharePoint is. Apparently many of the youth of today have been blessed with not having to use it. Lucky sods!

But at no point did anyone explain why SharePoint came to be. Which is how I ended up writing (and then throwing away) a lengthy reminisce about the origins of SharePoint and the death of the entire Groupware market.

Exchange Public Folders - and their almost complete failure to deliver what the market wanted - were central to the early part of that history.

I'm still amused that they live on. They're by far the longest-lived of Microsoft's failures, having debuted in 1996. That's almost 30 years of the world looking at Public Folders and collectively responding "Thanks, but no thanks"...

Sadly my insane ramblings were a bit too lengthy for use even at this august forum. Still it's quite the coincidence that I find myself thinking about them again within seven days.

Publishers cry foul over W3C crusade to rid web of third-party cookies

Come on in, the water's fine

My employer is moving to a new security platform, and one of its recommendations was that we should disable 3rd party cookies in Chrome.

I created a suitable GPO, a test OU, and have applied it to myself. It's been almost two weeks, and everything still seems to work.

I really was expecting it to cause issues with something. My biggest fear was that authentication via Entra ID or Okta would fail on one of our systems. But everything seems to be fine so far!

I'm sure we'll have to put some exceptions in at some point, but I'm really impressed at how uneventful it's been.

So, with the greatest of respect to the publishers... get stuffed.

PUTTY.ORG nothing to do with PuTTY – and now it's spouting pandemic piffle

This is good work

Thank you for this public service. And not linking to them is the cherry on top - absolutely the right choice.

FWIW I've used Bitvise's SSH Server software for SFTP purposes, and it was decent enough. If it were my choice I'd just fire up a Linux box and provide SFTP with that, but if you're a Windows shop and are afraid of Penguins then Bitvise have you covered.

(And being a Windows shop with little Linux experience seemed to be the common strand each time I've encountered it.)

My only reservation was the logging, which uses XML. Heavily attribute-laden XML. Each connection generated a stream of almost impenetrable XML, which would be logged into files rolled over at 64Mb. My personal opinion is that such a thing should never exist, and some kind of punishment should be meted out to whoever decided that was acceptable logging. But being an anti-vaxxer is probably too severe a punishment, so maybe the universe over-corrected here?

Nearly 3 out of 4 Oracle Java users say they've been audited in the past 3 years

Re: Free Libre Open Source Java

Yes.

Which I know isn't FOSS. But in my experience the process of "choosing" Oracle is the same regardless of whether the alternative is proprietary or FOSS.

The tactics used will be the same. The background of the people pushing Oracle will be the same. The overall process will be depressingly familiar for many of us.

In that regard it wouldn't have mattered whether the question was "Why not use FOSS?", "Why not use a full Microsoft stack?", or "Why not use AWS cloud services?". If the context is "instead of Oracle" then the answer I'd give wouldn't change.

Re: Free Libre Open Source Java

I once heard it referred to as "The Arsehole Pipeline".

Oracle have a lot of sales staff who, it is hypothesised, having made money selling the stuff then go into consultancy to tell people how to use it better. (For a very low given value of better.)

They know Oracle better than other options, so constantly recommend it. And companies tend to think that at the price they're charging, they must know what they're talking about.

I've seen ex-Oracle people push Oracle as the database for solutions, stating no other database will possibly work, even when the solution vendor's own documentation said otherwise. After calling my colleagues liars, they were confronted with the documentation - at which point they then tried the "alternatives will have abysmal performance" line. It was noted that the maximum users would be ~1000 people, and you'd be unlucky to have more than 100 concurrent operations in that scenario, which SQL Server would handle just fine. Yet it was still a fight to get them to acknowledge that SQL Server could even work.

Given that the organisation already had SQL Server expertise, experience, and infrastructure it should have been the natural selection. Instead the ex-Oracle consultant made it a hellish fight.

And guess what? Yes, that's right, the solution ran just fine on SQL Server, no problems at all.

But if the IT department hadn't pushed back, then they'd have suddenly acquired a very expensive Oracle system.

I've heard the same from a number of people over the years. There's just a pipeline of the buggers somewhere, spewed out into the world to put the con in consulting...

Re: Lube

Oh dear sweet summer child...

Oracle will charge you merely to produce the QUOTE for the lube. The lube itself... will not be cheap.

Trump guts digital ID rules, claims they help 'illegal aliens' commit fraud

Dumb

So now the standards for digital IDs and the security around them will be determined by other states. Europe, China, Japan. But not the USA.

Trump really is the dimmest bulb in the ugliest of gold plated yet surprisingly dim chandeliers...

As US vuln-tracking falters, EU enters with its own security bug database

Prediction

US: Now that the EU can do it, why should we pay for this? We're being taken advantage of! Let's end this!

Rest of World: *** facepalms ***

EU: OK, fine. Someone has to do it, after all.

*** EU takes over CVE handling ***

*** Six months later ***

US: Look at all these CVEs for good, honest, American companies! The EU is bullying us by advertising these security faults!

EU: *** facepalms ***

Rest of world *** facepalms ***

Apple: Since you care about yOuR pRiVaCy, we'll train our AI on made-up emails

How do you do, fellow kids?

On the one hand, kudos for caring about privacy. On the other hand, way to self-own.

Check out their suggested email: "Would you like to play tennis tomorrow at 11:30AM?"

Not only have I never sent or received such a message, but the few people I know that play tennis never have either. It's a ham-fisted, dry, asinine and tone-deaf attempt at mimicking human communication.

"Tennis tomorrow at 11:30?", or "Tennis tomorrow, usual time?" or "Still on for tennis tomorrow?" would be much more... human. WPOR in War Games had a better script than the one that Apple is ascribing to its users.

Seriously, Apple, just make training an opt-in. People that want AI give up their data, and accept that the training has potential privacy implications. But they at least get it trained on data that reflects their circumstances and behaviour.

And people that don't want AI just don't opt in. For whatever reason.

But training on Apple's "FBI Agent pretending to be a cool criminal" text corpus won't get anyone anywhere. So just don't do it.

New SSL/TLS certs to each live no longer than 47 days by 2029

Bad choice - it fails to understand the user outcome

This is a bad choice because it's simply too short to be usable.

It places a massive burden on IT teams to do the replacement. Not everything can be automated at present - I can't use LetsEncrypt on my switches, or my SSLVPN device, or my firewalls. I have application servers that use Apache Tomcat, which is a much more manual process.

So the end result will be that when this short timescale fails - because someone is off ill at the wrong time - we will end up training our colleagues to ignore the certificate warning so that they can still do their work.

That is a far, far worse outcome than any issues with highjacked certificates. Highjacked certificates may also require DNS access to use effectively, or other access to the target's systems. But this ends up training users to behave in an insecure manner.

I hope that they reverse course on this.

Trump yanks CHIPS Act cash unless tech giants pony up more of their own dough

Re: Meh

One of the main reasons for the CHIPS act is to ensure a stable and secure supply chain.

At the moment a lot of the processors that the US military and government want are made in foreign countries. Their largest supplier is a country called Taiwan, who China doesn't like being called a country because they'd be quite happy if we could all just agree that it's theirs. China frequently runs military exercises off the coast of Taiwan, and tends to throw a diplomatic hissy fit at countries that recognise Taiwan as a country.

Generally speaking, having an essential part of your supply chain in a disputed territory is thought to be... *checks notes* ... bad.

For many years, the western world has ignored problems like these. The thinking - what little of it that happened - was that the world was stable enough that the bad things could never happen, and that fixing the problem would mean the number would not go up for Wall Street. "Number go up" has been the most important thing for the ruling elites, so nothing was done about it.

The pandemic, followed by a ship being stuck in the Suez canal, focused minds on just how fragile things had become.

The CHIPS act is not supposed to be about quarter to quarter profit. Sure, it would be nice if the fabs made a profit. But it's supposed to be about re-invigorating an industry critical to national security.

MAIA?

Make America Irrelevant Again?

Microsoft's many Outlooks are confusing users – including its own employees

Point of pedantry

Pedant Alert!

There never was an Outlook '95. It first shipped in '97. Prior to that we had to use either the Microsoft Exchange Client or Microsoft Mail and Schedule+, depending on whether you were on Exchange or Mail as your backend. Neither of those supported POP3 - the distinguishing feature of Outlook was that it could support multiple backends.

I think '98 came fairly quickly, as '97 was quite buggy. It was also given away on a lot of cover CDs as Microsoft wanted to keep people in their ecosystem rather than use, say, Netscape's mail client for their POP3 connections.

Show top LLMs some code and they'll merrily add in the bugs they saw in training

No surprises here. They're not intelligent.

AI isn't intelligent. These are pattern machines.

Which we see here. Given a pattern, they strive to complete it as best they can. They don't actually think or reason, except in the completion of patterns. Nor do they care if the patterns are good nor bad - thy cannot comprehend that. The closest that they can get to comprehending the correctness of a pattern is merely to produce another pattern containing a commentary.

The big problem is one we don't want to admit to - we're also pattern machines. That's why we're so easily impressed by this technology. And why its failures are so surprising to us. They fit the pattern we've learned of an "intelligent" output, so we ascribe them intelligence.

They are, in fact, not intelligent and never can be. And to assume that they are is dangerous, because there are nasty edge-cases in their patterns that we should really be trying to avoid.

The sooner everyone realises this, the better.

The IT world moves fast, so why are admins slow to upgrade?

Re: Why are admins slow to upgrade?

I had to give this very serious thought before finally giving you a downvote.

Back in the 1998-2002 period I worked for a consultancy, and we had a couple of upgrades that had issues. Mostly due to something the client had neglected to mention or spot during the scoping & planning stages. All of them were surmountable, and we got the job done in the end.

Since then I don't think I've had any upgrade that had major issues. Any issues encountered were ones we had planned for, and either directly addressed in the plan or had a contingency plan for. Absolute worst case we had a small delay, but we always made sure our change windows were wide enough to cope with that.

I'm also excluding migrations to other technologies. To the business moving from one mail system to another may be sold as an upgrade, but it's actually a migration, and there are always unexpected issues with those. So those aren't - I'm assuming - what you're asking about.

Some may say I should have upvoted, but I think you're trying to determine if there are unexpected technical issues rather than ones that were created by the business or bad planning, hence a downvote.

Performing upgrades is only a small part of the job.

It is, I'll admit, one of the more visible parts. In many companies the other work sysadmins do is much less noticeable. Monitoring? Capacity management? Availability? Policy work? Risk management? Integration? Many of those are a lot less visible than upgrades.

Security has, thankfully, gotten a much larger presence over the last two decades - so when we're working on that it's often very visible.

I can see how folks would think we should be busy doing upgrades all the time, but that's like thinking that the only thing to running a shop is manning the tills or stacking the shelves.

And of course then there's whether or not the business wants to upgrade. All upgrades have a price, and some are more worthwhile than others. It's not always worth the effort.

I've dealt with a few systems over the years which had to be completely rebuilt rather than upgraded because they'd become so old there was no valid upgrade path anymore - we couldn't get hold of all the install media sets we'd need, and whilst an individual upgrade was quick the amount of time needed to perform about five upgrades with checks in between each one was simply prohibitive. So we built anew.

It's the sysadmin equivalent of developer's technical debt problem, and just as with the developers it's not solely the fault of IT. If the organisation won't prioritise keeping everything up to date, then the costs for moving to a current platform will only go up over time. Sadly, that's often not considered when people are looking at the costs for this year...

Microsoft signed a dodgy driver and now ransomware scum are exploiting it

Legacy code strikes again!

This is almost certainly code that was written in the late 90s or early 2000s, and has just sat there ever since.

"It works, so why change it" is fine for code in userland applications that won't hold sensitive data, but not for kernel code like this. (Or for some kinds of userland apps.)

So how do we get companies to audit their old code?

In the case of kernel-mode drivers, I suggest that Microsoft make signing the driver a contract. If the driver has a security flaw, then the cost to Microsoft of investigating and mitigating the flaw will be borne by the supplier. Plus a penalty fine per vulnerability if they're common types (buffer overflow, pointer misuse).

Only by introducing some kind of fiscal penalty will the audit of old code suddenly become fiscally viable, which is what we need.

30-year-old NHS supply chain system hit by 35 major alerts in 11 months

What is Rhesus?

It's all because of a f***ing Access database, isn't it?

It started out as a spreadsheet which got moved into Access. Then someone made a valiant attempt to put into SQL Server, which both preserved the existing bugs and introduced new ones. It's currently running on Windows Server 2008 R2, a similarly aged version of SQL Server, and can be accessed either by a website which requires Internet Explorer 6.0 or a creaky Access Runtime instance if you can get an ODBC connection. (Which you absolutely should not have, yet far too many people somehow do have.)

Of course, I could be wrong. It could be one of those ancient AS/400s that still turn up occasionally, hiding in the corner, surrounded by the faint glow of arcane magicks which ward away the evils of new technology such as non-token ring networking...

But seriously, what the hell is this thing? It sounds like it's so awful it's going to be kinda awesome in its awfulness!

Are you cooler than ex-Apple design guru Sir Jony Ive?

Re: Do I have to have an opinion of U2?

Huh. My assumption about the nickname was wrong, so I apologise for that, and have a virtual pint!

It still feels like a sixth form band thing though - so it doesn't change my opinion of them.

That being said, I'd see them live if someone else bought the tickets. Based on their reputation I reckon that they'll put a lot of effort into their live performance and it'd be great fun.

Hopefully your dad was forgiving. ;-)

Do I have to have an opinion of U2?

I never thought of U2 as boring. I just didn't think of them much at all, to be honest. Their music was, with a couple of minor exceptions, utterly forgettable for me.

My first impression of them as a kid in the 80's was that they looked like a student band that was trying too hard. Bono was sure he was the coolest cat in the room, when he actually just looked like a tit. Their guitarist called himself "The Edge", but didn't seem to be playing well enough to be worthy of the title. Both of these were good reasons to turn my attention elsewhere.

So I carried on with my mix of pop, hard rock, blues rock, metal and classical. Later I added a little hip hop. (The good stuff. To quote: "Guns, b*tches and bling were never part of the four pillars, and never will be.")

I don't like judging folks based on their music taste, but I struggle not to feel deeply suspicious of anyone who fails to be cheered by Queen's "Don't Stop Me Now", so that would probably be in any selection I made.

The rest would depend on my mood. After all, the idea of planning for a desert island is faintly ridiculous. If it happens by accident then you get what washes up on shore with you. If you're actually planning it ahead of time that rather suggests things have gone very, very badly for you... and perhaps your time could be better spent elsewhere?

Google binning SMS MFA at last and replacing it with QR codes

Re: "Apps on smartphones only came into being in 2008, with the iPhone 3G. "

Was it PalmGear that did the app store for Palm platforms? I seem to recall that there were two notional app stores, but the names are fuzzy in my memory.

I had a succession of Palm devices, but only the very last one did any kind of wifi, and it caned the battery life, so I kept installing apps on it via the HotSync cradle/cable. But I was aware that I could have downloaded an app that would function as a store and allow me to do the whole purchase on the device if I wanted to. As I was mostly using my Palm devices as an "offline mobile adjunct" to my computing needs, it just didn't quite fit my needs.

I'm curious, did the Treo come with the app store preinstalled as a promotional thing, or did you have to install it later? I only really had Palm devices (IIC, m505, V, Vx, Tungsten|C amongst others) so have no idea what came on the Treos.

I ask because one important criteria for "modern smartphone" would probably be that it has an app store pre-installed. Not that anyone seems to be able to agree on all the criteria!

Pinning down the first modern smartphone is more difficult than most people think. There's a lot of "It has to do this", followed by someone else saying "this phone did that before any iPhone did", and then the goalposts being moved. And the goalposts often move beyond what the original iPhone did at release.

I'd say the first modern smartphone is probably the iPhone 3G which ticks all the criteria "out of the box". But Apple's excellent PR muddies the water considerably, as does the fallibility of human memory.

Re: What about all the people who don't have smartphones?

If my old and frequently failing memory serves, PayPal existed long before the modern smartphone and apps. Well, long in software years, anyway(*).

Apps on smartphones only came into being in 2008, with the iPhone 3G. (Not, as many mistakenly think, the original iPhone. Everything was supposed to be web-delivered for that, until they realised that a full web experience on 2G was terrible and had to change their strategy.)

I'm pretty sure I paid for some eBay purchases via PayPal as early as 2001. It's quite possible to have a PayPal account that predates the smartphone and has never been used via an app.

--------

(*) Software years are like dog years, in that there's about 7 of them to a standard calendar year. They are unlike dog years in regards to intelligence or loyalty. The jury is still out on comparisons involving fleas and toilet habits.