* Posts by Peter Gathercole

4197 publicly visible posts • joined 15 Jun 2007

Will containers kill VMs? There are no winners in this debate

Peter Gathercole Silver badge

@karlkarl

IBM LPARs are not containers, at least on Power. They are full blown OS instances, so are Virtual Machines, not containers.

On System Z, there may be slightly more debate, because of the way that linux (in particular) instances effectively can have shared OS code (similar to read only shared text segments for UNIX processes) for different instances of the same Linux VM.

The AIX feature most like BSD Jails is WPARs, especially when used with WLM.

Peter Gathercole Silver badge

@brian re. VMs

Your view of hardware VMs is very, very naieve.

Almost none of the VM systems available today, including IBM's PowerVM and their mainframe VM, and certainly not VMware and derived products are actually hardware.

There were once some real type 1 Hypervisors around (I used Amdahl's MDF about 35 years ago, and I'm not even sure about that), but now, at some layer or other, all hypervisors are type 2, with maybe the hypervisor running on the bare metal, but still very frequently a specialized derivative of a general purpose OS.

I've seen some of the internals of the PowerVM (the IBM hypervisor for IBM PowerPC systems), and inside the hypervisor, you have a full blown Linux kernel with much of the standard tool chain running as as a black-box turnkey system. I agree that this can be fairly hands-off (but must get involved with sub-CPU VM scheduling), and sets up and uses the hardware features of the platform, like CPU security rings and affinity to VM images, memory encryption, above OS level memory page table control, and many more that I'm not going to enumerate, but it's not a hardware hypervisor.

The control of all of these feature is by a software layer, and it's security is only as good as the security of that layer.

I've actually commented on this before. Putting Hypervisors in the stack just moves the stack down. The hypervisor replaces an OS and schedules multiple OS images as if they were applications, and everything else moves down a tier. Containers on VMs just adds yet another tier.

Moving to containers, and particular Docker. This has the concept of the Docker Engine. The Docker Engine in it's simplest form is a VM system or possibly you could call it an OS Binary Interface (can I coin OSBI as a term?), This is what allows you to run applications from other OSes than the hosting OS (as opposed to, say Solaris Containers, where the application still has to match the hosting OS). I admit that it is stripped down to it's barest minimum, with the required support layered on top of this minimum, and may not actually be derived from the OS environment that it provides, but it's purpose is to allow an application to run in an apparent OS environment, with some separation from other containers. This sounds so similar to a VM to me that the point is moot IMHO.

One of the primary reasons for having some form of encapsulation for an application is for resource control. But OSs have these features, in things like Linux Cgroups, BSD Jails, Solaris Containers and IBM WPARs and Workload Management (WLM). But before all of these are two things that were in AT&T (Bell Labs.) UNIX in the 1970s and 80s, being chroot and then the Fair Share Scheduler which has almost disappeared from memory, but is mentioned in the "Design of the UNIX Operating System" by Maurice J. Bach (first published in 1986), which I had contact with in the late 1980's. None of this is really new.

With these in mind, I feel that a valid deployment should still be multiple applications running on single OS instances with the correctly configured OS isolation (and some High Availability configuration), but I agree in this day and age, the resources owned by the OS are almost negligible and are easily dwarfed by most of the applications themselves, and the runtime isolation of separate OS images do actually provide some maintainability advantages for monolithic applications.

I look after the OS and hardware that a set of large Oracle databases run on. These systems have ~400GB of memory allocated to each of them, together with 20 Power8 processors, and the OS requirement is less than 8GB, or about 5% of the memory resource of each system.

Russia spoofed AIS data to fake British warship's course days before Crimea guns showdown

Peter Gathercole Silver badge

Re: To sadly turn this political

It's a matter of size and ability to resist.

If a Ukrainian ship (not that they have anything particularly sizable) had tried this, the Russians would have threatened, boarded, and impounded the vessel, subjecting the crew to some humiliating questioning and threats of 'espionage' before letting them back into Ukraine, although they would probably keep the ship.

The fact that it was a major vessel with significant offensive and defensive capability, from a navy that is well respected, which was testing an internationally agreed (with the exception of the Russians) trade route, they would be mad to do anything more than just posture.

If a Russian trawler obtains a UK fishing license, it would be free to fish in UK territorial waters. If they didn't, they could be quite legally prevented from fishing in these waters. We have fishery protection vessels to make sure this happens.

UK health secretary Matt Hancock follows delay to GP data grab with campaign called 'Data saves lives'

Peter Gathercole Silver badge

Re: All they need to do...

And the scope can be extended by secondary (i.e. not debated in Parliament) legislation. I don't think so. This is the same thing that scuppered the UK ID card system.

Let's hope the BBC and other news organizations keep flagging the concerns so the public have a chance to understand, not that I think that most of them would bother to take any note.

Peter Gathercole Silver badge

Re: Fix the existing problems before making a profit

I think that saying the hospitals are separate entities is slightly disingenuous. The hospitals are grouped together under regional trusts, which have between 1 and 9 hospitals depending on their size.

I live in Somerset, and the Somerset NHS Trust manages two general hospitals, and then a number of community hospitals that are mainly used to allow recovering patents to be moved from the large hospitals to closer to their homes (this is important in one of the larger UK regions with largely distributed populations), as well as offering some outpatient facilities. I sometimes envy people in cities and large towns who have much better access to hospitals without traveling.

As far as I am aware, this is also a clinical commissioning authority that is the financial controller of procedures.

Various stats. suggest that there are 223 NHS trusts, managing or commissioning work from about 1270 hospitals.

But I agree that these are managed separately, within rules set by the Department of Health and Social Care and other bodies such as MHRA. Over the years I remember the NHS, there has been an ebb and flow between more and less centralisation, and I doubt that will change, although the current shape is probably more amenable to privatization.

Large parts of the support, building and cleaning, as well as equipment procurement for the NHS is already outsourced, so you could say that it is already partly privatized.

As I remember it, the main problem with the original central NHS patient database has little to do with data formats, and much more to do with the way that access to users of the database were tracked and managed. One stat. that I think I remember was that they had identified over 500 job roles requiring different levels of access, and adding the required controls to the data and grading all of the people working in the different trusts with different job titles, roles and responsibilities in each trust was beyond the scope of the project was capable of.

Oh, and the fact that they were wanting to put it into just two very large data centres for the whole country was probably going to cause problems.

Windows 11: Meet the new OS, same as the old OS (or close enough)

Peter Gathercole Silver badge

Don't just blame Ubuntu. Aim at Redhat!

This will be the move away from X11, and possibly the vendor drivers for your video card which silently remove support for older cards.

I think that Wayland will break a lot of X11 functionality, and may also prevent cross network application support from remote machines from functioning correctly, because I don't actually trust the Wayland developers to really understand what it was for. They're an arrogant bunch of pricks who think they understand everything, and actually believe that VNC and RDP are superior to X remote support.

Systemd 249 release candidate includes better support for immutable OSes and provisioning images

Peter Gathercole Silver badge

Re: Thinks I like about systemd @Def

Linux supports dynamic swap files stored on a filesystem like Windows uses. It's just not the default way it is set up in most distros (and Linus actually pulled kernel support for this feature temporarily during a recent release candidate because someone introduces a bug that corrupted the filesystem containing the page file when paging occurred. It's back in now.)

The reason it is not installed on most Linux distros is that putting a page file through a filesystem is actually much slower than putting it on a native raw partition, but of course as soon as you use a partition, it is fixed in size. Similarly, the Ubuntu installer will by default only set up a small paging (swap is such an outdated term) partition. You can change it during the install, but I guess you just accepted the defaults.

I have to admit that after a little digging around, I found that Linux does not support anything like the SIGDANGER signal that AIX has had for 25+ years (AIX 3.2.3 I think), so the first time an application knows that it's regarded as a transgressor is when it receives the untrappable SIGKILL signal. I'm very surprised, although there have been many suggestions to add it. Long standing advice suggests using cgroups to trap memory conditions, which is more than I would expect a normal user to have to configure.

I don't use Windows that much nowadays, so my experience is a little dated, but I wonder whether your Linux may be similar.

Peter Gathercole Silver badge

Re: Thinks I like about systemd @Def

Hmm. Protecting the OS in a multi-user, multi-tasking OS is an important thing.

As far as I am aware, the out-of-the-box virtual memory system in most Linux distros is designed to keep the system running by sacrificing greedy applications. This is so that other users of the system will be less affected when the whole system hangs because of resource overcommit.

Using Linux as a personal system may change this core desire, as the system may just be that greedy application as far as you are concerned, so this could be undesirable behaviour. But I've seen many Windows systems brought to their knees for similar problems, and the normal remedy is to reboot the system...

It may be that in your case the application is more mindful of how to manage it's own memory in a Windows environment. I assure you that it is possible to write an application to care for it's own memory usage in Linux, the tools are all there, and allow it respond to requests from the OS to trim it's memory image when there is contention.

I have problems with exhaustion of memory on Ubuntu, but that is mainly because of the insane and resource hungry behaviour of some applications. Many applications just don't manage their memory usage, and just keep grabbing more and more memory. And with things like Firefox putting tabs in isolated sandboxes for security, making it's memory footprint much larger than it strictly needs to be (I feel that using processes with shared re-only text segments would be a better model than sandboxed threads). Once the paging space is exhausted, any OS needs to take serious actions to recover.

Peter Gathercole Silver badge

Re: What part of "temporary" do you not understand?

No, they weren't.

Often they had a housekeeping script, triggered on boot and sometimes in cron which did an ordered cleanup, often based on the type and age of the files. Others that didn't normally had half-competent sysadms who would write their own scripts.

These being scripts were easy to find, disable and/or modify for exactly what you wanted to do, and were not replaced during an update.

Why do I get the feeling that the balance of people here is skewing away from the greybeards to millennials.

Debian's Cinnamon desktop maintainer quits because he thinks KDE is better now

Peter Gathercole Silver badge

Re: But wouldn't someone who wanted Cinnamon choose Mint in the first place?

I have to admit that I am becoming disillusioned with the whole GNU/Linux distro. environment.

I've been an Ubuntu user since Dapper Drake (6.06), which is what I moved to when the old Redhat Desktop releases stopped.

When the whole Unity/Mir thing kicked off, I nearly abandoned Ubuntu, trying Mint, and mainline Debian, and LDME, but decided that I liked where Ubuntu sat in the distribution hierarchy (sufficiently functional to mostly work out of the box, and not so far down the line that it was dependent on too many other distros.), and instead worked to use a desktop other than Unity, eventually settling on Gnome Fall/Failback.

This is now long in the tooth, and does not appear to play nicely with the most recent LTS X11 and older Nvidia cards that I have installed in some of my machines. Stupid things like systems not fully powering off or suspending correctly, and not getting the screen size from X11 correctly (this is quite a bizarre problem that makes full screen apps fall of the side of the screen that I've never seen before - and it does not seem to be a virtual desktop problem).

On my most powerful desktop, I've got Cinnamon on Ubuntu 20.04, but I'm not totally happy with it. I don't know what it is, and if I did, I'm sure I could change the config, but something in the font handling just looks... wrong.

My main laptop is still running Ubuntu 16.04, so is still running Gnome Fall/Failback, but as this has just dropped out of support, I need to do something. But systemd just pisses me off with it's complexity and lack of readable user documentation, and as always, the online blogs and stuff is now becoming out-of-date compared to the latest releases of systemd, making it difficult to sort the no-longer-useful stuff from the rest,

I have an old Acer Netbook as an ultra-portable for diag. work, which has limited memory, and I run Ubuntu with LXDE on that, but that always felt functional but a bit too cut-down.

I'm forced to use RHEL desktop and Gnome 3.82 on my client provided laptop, and still don't like the UI design choices and restrictions that Gnome impose on what you're allowed to do.

I've even dabbled with Devuan and Raspian (with their default GUIs), but somehow I move away from them and never go back (the Raspberry Pi is running in a PiDP11, so I actually have the GUI turned off on that system for most of the time).

Jake keeps telling me to look at Arch or even one of the *BSDs. Maybe I should finally do it, and then decide on the desktop afterwards. Maybe I should just accept that I'm getting too old for this game and go to a mainstream consumer OS, but this would betray my last 25 years of trying to get away from them.

Royal Yacht Britannia's successor to cost about 1 North of England NHS IT consultancy framework

Peter Gathercole Silver badge

Re: Great British Engineering

Ocean was also built to civil shipbuilding rules (with some military uplift), so was always going to be a stop-gap measure with limited life.

I hope that Brazil are able to keep the ship in good repair, because my feeling is that it's going to require a lot of TLC to keep it in service. But then I guess they have a history of keeping old warships running.

Peter Gathercole Silver badge

Re: Great British Engineering

Appledore announced that they were ceasing shipbuilding after the current order book is complete about six months ago.

Peter Gathercole Silver badge

Re: Great British Engineering

Actually, at that budget, they could probably take the Type 31 frigate design, remove the weapon systems and the more military of the radar fit, and reduce the installed power, and end up with something interesting.

Might have some problems re-fitting the interior to make is more yacht-like, but the budgets are not that dissimilar.

Peter Gathercole Silver badge

Re: Great British Engineering

I think Sunseeker in Poole still build yachts. But I don't think that they are in the same league as the type of ship that I expect the Government want.

You ought to bear in mind that HMY Britannia was built with dual purposes in mind. She was technically a ship of the navy, and was equipped to allow it to be converted to a hospital ship, although that never happened during her operational life.

Mensa data spillage was due to 'unauthorised internal download'

Peter Gathercole Silver badge

Re: A bit bonkers @So negative

And there you go incorrectly associating knowledge with intelligence, like so many people.

Typical knee jerk response from someone not understanding what Mensa tests are about.

IBM pulls up the ladder behind some supercomputer customers

Peter Gathercole Silver badge

Not BlueGeneQ

But the Power 6 575 and Power 7 775 supes. also had a lift tool.

They was used to wind a drawer out of the rack, and to raise/lower it to a height where it could be worked on.

The tools for the 775s I was involved with were not powered, but allowed you to attach a a battery powered drill as a power source so you were not constantly winding handles to move the drawers.

When the clusters were removed, the de-installation team took the lift tools with the systems, but did not want the drills (we had 2 of them). Unfortunately, the IBM CE decided that he liked them too much to share one of them with me!

It is interesting about the ladders. If you go to any IBM or previously IBM site, you are bound to find these ladders knocking around. As far as I am aware, they used the same model for virtually all mainframes, RS/6000 SP/2s, BlueGene, and tall rack Power systems for the last 30 years at least. They come as part of the maintenance package of these large systems unless removed with the order, and generally are so useful that they get snaffled by the local electricians, aircon. and other engineers unless clearly marked to be left with the machines or locked away.

We don't know why it's there, we don't know what it does – all we know is that the button makes everything OK again

Peter Gathercole Silver badge

Re: My suspicion

Not sure about Calcomp plotters, but I do know that sometimes sending a poorly formatted or garbage print to a plotter will generate no output, but will require the plotter to actually receive the whole plot at whatever the serial speed is before it will accept another job, causing a delay in the output.

This did normally set some of error indicator, though.

If it was attached to a VAX running either VMS or UNIX, chances are that the plotter was just set up as a batch printer in the spooling system, so maybe someone sent a document to the wrong queue.

Peter Gathercole Silver badge

Re: Plotters

That would certainly make sense... if it was actually in-line with the serial cable.

They already said that the serial cable went straight from the patch panel to the plotter, avoiding the box.

I would have been interested in hearing exactly where the wires attached to the grey box actually did go. I presume there must have been power for the light, but did it actually go somewhere else on the plotter? It is just possible that if there was a wire running to the plotter, it might have been to a broken or normally inaccessible switch wired externally, that may have actually done something to the plotter (like resetting the serial interface on the Calcomp plotter - quite often the serial port was an add-on board that contained it's own Intel 8042 or 8049 microcontroller).

Pre-orders open for the Mini PET 40/80, the closest thing to Commodore's classic around

Peter Gathercole Silver badge

Re: OK, but why?

I remember when I was a member of the 6th form consultative council at my sixth form collage, back in 1977, and myself and a couple of other members campaigned to get them to spend their discretionary fund on a Pet 2001, but they decided to buy a second hand mini-bus instead. Probably a better choice, but we did try.

Several years later, after Uni. (where I played with KIM-1s [and UNIX]) and my first job, (where I didn't), I started working at Newcastle Poly. They had several Pet labs. with 4032 (I think) models with real keyboards and an external cassette (possibly shared with a switch box between 4 Pets), but I have to admit that I was never that impressed by them as a teaching resource, and I remember that they were forever breaking down, normally fixed my opening them up and pressing all the chips back into their sockets.

I was not directly involved in maintaining them, as I was primarily looking after the PDP-11/34E that ran RSX-11M and UNIX version 6 and 7 for the School of Maths and Computing, but ended up managing the replacement of one of the Pet labs. with a networked BBC Micro lab. which IMHO was a much better resource, but hey, they were a much younger and more versatile design.

The Poly. central Computer Unit went a different way, and installed networked RML 480Zs, and then IBM PCs, but I still think that the BBC lab. was the most versatile and interesting resource we had, but then, I was biased as I had built it.

I'm going through a nostalgia phase at the moment (CoVid lockdown to blame, I'm afraid), as I've resurrected one of my BBC model Bs, and have just bought a vintage Epson HI-80 pen plotter when it popped up on eBay (I forgot that I had set a watch up for one some years back), one of my favourite toys from the lab. Only problem now, where do you get pens for a 35 year old discontinued plotter! I'm trying with a bit of success to make some with ball point pen refills, and maybe fineliners, and it might be just the springboard I need to get into 3-D printing.

Apple settles with student after authorized repair workers leaked her naked pics to her Facebook page

Peter Gathercole Silver badge

Re: Details

The second question you need to ask is whether, if Apple or one of their Authorized Repairers refused to fix your iThing, which they seem to do on a depressingly regular basis if you can believe the independents, how much risk you would be prepared to take to actually have it fixed rather than thrown away, and possibly all the data on it lost.

If your judgement is "I would not take any risk", then great. All Hail Apple, and you will accept any gougiug gratefully. But currently, you have some choice.

But if Apple can get away with shutting out the independents for a completely bogus and false privacy reason that they cannot defend, which is one of their main arguments and which is undercut by this case, then neither you, nor the people who have less concerns will have that choice at all.

How many remote controls do you really need? Answer: about a bowl-ful

Peter Gathercole Silver badge

Re: Give me buttons ON THE TV!

I have a Sharp TV which has no obvious buttons, but has something that resembles a small mushroom on the back. I had to download the manual before I realised that it was actually a joystick to drive the menu system, with a press action for enter. Talk about reducing the controls!

Today I shall explain how dual monitors work using the medium of interpretive dance

Peter Gathercole Silver badge

Re: salesman

What makes you think that the computer salesman doesn't lie. I turned down a very lucrative technical presales jon because I would not be able to hold my tounge if I was present when an untruth was told!

Peter Gathercole Silver badge

Re: Laptop + Monitor = two computers?

What are you talking about, Unix escape characters? There ain't no such thing. Terminals or terminal emulators have have escape characters. Unix itself, not so much (unless you count "screen" codes, and there are not too many people who use "screen".

Mine's the one with the copy of /etc/termcap and the ANSI X3.64 standard in the pocket.

Dominic Cummings: Health secretary's 'stupid' targets delayed building UK test and trace system to combat COVID

Peter Gathercole Silver badge

Hindsight

It is very clear that the UK government, along with most of those in the Western World were caught flat-footed by the speed and scale of Sars-Cov-2.

But there was always going to be a balance in how to approach it, and different people will have a different idea about where that was likely to be.

From a purely 'Lets save all the lives we can', then the best approach wold have been for the government to confine the entire population to their homes, except for the people who would have been needed to distribute the necessary food and supplies, and run the essential services to keep everyone alive. But the economy would be devistated, and there would be no clear exit plan until Sars-Cov-2 had disappeared from the entire world.

For the 'We've got to keep the economy working' position, then doing nothing might have seemed the best plan, and hoping that it would burn through the population quickly, without destroying the NHS.

But neither of these would actually have been completely sensible. I'm getting quite sick of the "Oh, but if we'd followed the scientific advice...' argument, because although this seems simple, science is never simple, and even in the scientific community, there were differences in opinion. If you gave one scientist a range of different desired outcomes, they may well have come up with more than one type of advice. Science is not (unlike what many people think) absolute, and the way the scientific method is means that you go with the current theory until it is disproved.

As more information about mutation rate, transmissibility and even the source and movement vectors become known, it is very likely that the path that was chosen may prove not to have been the best one. But without this information, a huge amount of what was likely to happen was informed guesswork, and the only countries that seem to have come out slightly better are those where the people were more willing to be controlled by their government, but even their handling was not perfect (and I have a strong distrust of what those regimes actually say happened in their countries).

The only thing that we can hope that will happen is that next time (and there will be a next time), we will be more prepared.

The Home Office will need to overturn a long legacy of failure to achieve ambition of all-digital border by 2025

Peter Gathercole Silver badge

Re: Curious

The UK has a difficult relationship with identity, especially in recent years when providing identity to take part in society is becoming more important.

I have several relations who have had problems over the last decade or so.

For people who don't drive, don't travel abroad, don't pay bills (because they live in bills included accommodation or with relations), or maybe just get their bills delivered online, proving your identity is getting to be difficult.

My mother-in-law was wanting to rent a house while she was moving from one region of the country to another. Only had a paper driving license. Did not have a passport. Shredded bank account statements as soon as she had reconciled them. The letting agency wanted more documents than she could easily provide. It was very difficult to persuade them who she was, and it seemed like it was necessary because of a change in the law with regard to illegal immigrants.

My youngest son. Left school before he set up a bank account. Did not have a driving license or passport, did not pay any bills. Had no picture ID at all. Obviously did not have bank statements. We had significant difficulty setting up a bank account (eventually had to apply for a provisional driving license just for ID, and not everywhere accepts provisional licenses).

My wife. She's a beneficiary in a will. The solicitors what picture ID, but again, she does not drive, and does not have a passport. She doesn't work, so doesn't have payslips or receive any government benefits. She's also only on many of the bills as Mrs Gathercole. They won't even accept a person of importance (we know a local government Councillor) vouch for her. We're still arguing this one.

A non-compulsory ID card with biometrics, but without the large database backing it up (just enough to prove the card was issued by the government) would solve all of these. When applying for a bank account, credit, or any myriad of different accounts and services, it would negate the need for a huge list of 'identifying documents' that many organizations want.

What most people in the know objected to with the last ID card system in the UK was not the card itself, but the database and identity number that the Government was wanting to use to pull all of the government information about someone together as a super-index that could be added to existing databases relatively easily. The bill to introduce the whole system would have allowed secondary legislation (laws that don't have to be debated in parliament) to extend the database in any way that the Government saw fit, without any scrutiny (this is where it upset the House of Lords).

The other thing was that the government wanted people to pay for the privilege of having a government issued ID card, and to keep paying every 10 years or so, for something that they wanted everyone to have.

Since that time, the Government have realized that for all people of working age (and children, as they issue them while at school), the National Insurance number forms a perfectly good cross-data-source unique identifier, which is why they are asking for your NI number when interacting with vehicle and driver licensing, the NHS, and so may other things that really don't need the NI number.

Whilst it may seem obvious to people outside the UK that governments may want to marry say tax, benefits and criminal records together, we have a history of all of these things existing in disjoint databases that have been difficult to cross-match. We don't like the idea that the Government will obtain significant powers to monitor what we are doing, but at the same time, many of us are prepared to let the internet giants, banks, credit card companies, mobile phone operators, and any company issuing 'loyalty' cards or apps track us in minute detail.

Even though I am generally against ID cards, I would support a card that proves who you are without any anbiguity, although I would not want to make carrying it compulsory.

Parliament demands to know the score with Fujitsu as Post Office Horizon scandal gets inquiry with legal teeth

Peter Gathercole Silver badge

Re: STOP. GIVING. FUJITSU. GOVERNMENT. CONTRACTS.

Are Amstrad even a company any more?

It used to be that many Sky boxes used to say "Made by Amstrad", but now they say "Made by Sky", at least the ones I've seen.

Hmm. Amstrad is still listed at Companies House, but it has been declared as dormant, with a total value of £16!

The directors (3) also all seem to be directors of other Sky companies.

Are you ready to take a stand? Flexispot E7 motorised desk should handle whatever you dump on it – but it's not cheap

Peter Gathercole Silver badge

Re: Ah, think how this would have struggled with CRT monitors

I have a 42" 4K TV which I use as a monitor on occasion (it actually works pretty well, even at 4K used with a Raspberry Pi 4), and that is very light, easily light enough to be picked up with my bad arm.

I guess it depends on the monitors age, and newer is mostly lighter.

I'm reminded of the 16" IBM 5081 monitor that I had that contained a block of concrete to keep it balanced. That was so heavy it was as much as one person could do to lift it!

Peter Gathercole Silver badge

Ah, think how this would have struggled with CRT monitors

I think that many people forget how light modern flat panel monitors actually are.

I would not be surprised if some motorized standing desks are troubled with an original IBM model M keyboard!

Peter Gathercole Silver badge

Dabbsie came up with a cheaper alternative

https://www.theregister.com/2016/11/11/a_cardboard_desk_i_wont_stand_for_it_actually_i_will/

This week in AI: Man arrested after cops say he rode in backseat of Autopilot Tesla

Peter Gathercole Silver badge

Re: Darwin Candidates

The idea behind the Darwin awards is that the candidate removes their genes from the gene pool.

Unfortunately, many of the people who can afford Teslas have already procreated, so they eliminate themselves from being eligible for the awards, unless they also want to be entered by proxy by taking their children out with them.

For the marketeer that has everything – except a CPU fan

Peter Gathercole Silver badge

Re: Microsoft Error Messages at their Best

That keyboard error is about as old as PCs with POST checks.

It definitely was there in the original IBM BIOS on IBM PS/2 model 80s back in the early '90s, and might have been there in PC/ATs.

I'm not sure about the original 5150. Their BIOS was a really simple thing that did not have an RTC or battery backup for settings.

Google Docs users, you are on notice: Code rewrite may break browser extensions

Peter Gathercole Silver badge

First using the web...

Do you mean the Web, or the Internet.

The Internet existed years before the Web, using things like Archie, Gopher and let's face it, FTP.

The Web proper started with NCSA Mosaic, although there were HTTP enabled programs the predated Mosaic.

Peter Gathercole Silver badge

Re: Chrome?

Remember when Firefox was the lightweight alternative to Inernet Explorer and Netscape Navigator?

'Biggest data grab' in NHS history stuffs GP records in a central store for 'research' – and the time to opt out is now

Peter Gathercole Silver badge

Re: Why opt out ?

If you read the NHS announcements, it puts this data into the "significant national interest" exemption of GDPR, which does not require a specific opt-in.

Go to the referenced announcements if you want the exact details.

Thousands of taxpayers' personal details potentially exposed online through councils' debt-chasing texts

Peter Gathercole Silver badge

Re: Please click the link to read

What a surprise. I got another text from them the other day, effectively saying the same thing, but lookie lookie. A full URL, not a shortened link.

Amazing. Thank you for listening Wessex Water.

Sometimes it actually seems worth raising these things.

IBM teases new AIX release – the first since 2015

Peter Gathercole Silver badge

Re: I worked AIX kernel 20 years ago - It should be dead by now

AIX is a legacy OS now, and the customer base is largely people already running AIX. Many organisations still like it as an OS, because of it's proven track record, and continuing investment in the hardware that it runs on.

It did get a bit of a boost when Oracle bought Sun, many Sun customers moved to AIX, as a UNIX OS of last resort.

But it would be very strange for a new user to install AIX unless there were some very specific reasons.

I still think that AIX has some features that make it attractive over other platforms, but as time goes by, these are being whittled away.

Regardless of what some people say, Linux is not a direct replacement for everything that AIX does, and Linux is actually being driven in directions that many traditional users don't like.

But I am expecting IBM to start pushing Red Hat as the OS of choice on Power hardware sometime soon. If they enhance it to use the RAS features of Power hardware, then that will be a good path, but they may divert to the less robust OpenPower platforms rather than the current systems.

Peter Gathercole Silver badge

Cloudy AIX...

A bit late to comment here, but

AIX and Power have actually implemented many of the core concepts that define a cloud for some time (over 10 years for many of these things). It has:

1. Virtualization technologies that allow system images to be resized on the fly, and even moved between physical systems while they are running

2. Scalabillity from 1/100th of a Power CPU all the way to the largest system that IBM ships

3. Disk technologies that allow scalable access to storage small to large (things like GPFS)

4. Container technologies (WPARs)

5. System administration tools that allow vast fleets of AIX systems to be managed and deployed

6. Virtualized networking to allow discrete networks between partitions to be set up

7. Huge amounts of RAS and system resilience to component failures to keep services running

8. A proven history of long uptime.

Now, what it doesn't have is the in-vogue software layers sitting above this hardware, but that is software, and could readily be adapted to run on AIX (although I think that IBM have SoftLayer running on AIX). But Linux (and I suppose Windows) are the OS's that have caught on, because Linux is cheap and malleable, and Microsoft have to be able to have their software technologies available for them to remain credible.

Back in 2000, I was involved in setting up an AIX application farm upon which applications could be deployed, moved between shared and dedicated systems, and managed in a proto-cloud environment. This was before the term 'cloud' had even been coined. So yes, AIX is suitable for Cloud, it's just never been shouted about.

Another week, another issue: Virgin Galactic mulls test flight restart as VSS Unity fixed – but VMS Eve might be borked

Peter Gathercole Silver badge

Re: Sending the wealthy to space

Yes, but do we have to bring them back?

Microsoft embraces Linux kernel's eBPF super-tool, extends it for Windows

Peter Gathercole Silver badge

Re: This will be an affront to God and man

Christ on a bike!

Reading that referenced document makes me wonder whether future Linux systems will remain maintainable.

Sure, I can see uses for it, but if you can modify the behavior of any LSF hook by attaching eBPF code to it, the OS becomes a support nightmare, especially if the error messages in the quoted examples remain so generic.

When faced with a problem, you will need to be able to take in the entirety of these attached modifications before you can understand it. It's already difficult enough, but being able to change fundamental behavior on-the-fly is, IMHO, a disaster waiting to happen.

Peter Gathercole Silver badge

"The entirety of MS Office is bought in"

Don't think it is.

Word was definitely written in-house, although the people who wrote it came from Xerox, it appears.

I know that Excel was bought in. Although R:Base was bought, it appears that Access was internally developed in parallel as Project Cirrus. Outlook appears to be internally developed, but Powerpoint was an acquisition (Forethought).

Visio was bought.

SQL Server (not srtictly part of Office) came from a database company they bought (Sybase?).

Peter Gathercole Silver badge

Re: eBPF != BPF

OK. I've got it. I had completely missed the transition from BPF to eBPF in he Linux kernel, and uBPF to divert this to a userspace interpreter.

Kernel 3.18, apparently. December 2014.

Note to self. Must keep up.

Peter Gathercole Silver badge

Re: hope I'm missing something here

There have been a number of vulnerabilities already reported with BPF over the years.

Although it is a virtual machine, it is really pretty simple, and is implemented to do some complex pattern matching on network packets. It is normally programmed by a tool such as tcpdump or wireshark (and firewall implementations), which take the required pattern, and then create a simple program that can be run against the packets to filter the ones of interest,

On Linux and UNIX systems that implement it, it is necessary to have a privileged session/process already to inject the generated program into the kernel based BPF. Obviously, if you have privilege already, this is a less of a threat.

The kernel based VM that runs the BPF code is sandboxed inside the kernel (and may actually be running in a kernel process, I've not looked that closely that at it), although the quality of the security is in strength of the sandbox.

Copper load of this: Openreach outlines 77 new locations where it'll stop selling legacy phone and broadband products

Peter Gathercole Silver badge

Re: "how does that work then?"

As I understand it, with FTTP, the provided optical network terminator (ONT) on the outside of your house is powered down the data line from your router that will be inside your house. And the phone line, if there is one, will either be plugged into a digital converter either in the router, or in the termination unit.

What I've seen has fibre to the optical network terminator, and copper (normally gigabit Ethernet) into the house. The pole-top services are passive optical splitters and combiners that use unpowered frequency division multiplexing and multiple fibres to provide a single frequency fibre connection (or maybe two, one for download, and one for upload), so there is no conductive path to run power down to the ONT, which needs power, and thus needs to obtain it from the house.

Once we have moved to this model, phones will not work during a power outage unless there is a battery backup unit somewhere for the router and/or the ONT.

Preliminary report on Texas Tesla crash finds Autosteer was 'not available' along road where both passengers died

Peter Gathercole Silver badge

Re: experience

That's why planes have flight simulators to train the pilots. The idea is that you throw enough really bad situations at a pilot/crew in an environment where nobody gets hurt that they gain the experience for when they reallly need it.

Peter Gathercole Silver badge

Re: As a motorcyclist....

I seem to remember that there was a study in the UK about the effects of not repainting the lines on roads in cities and large towns. See https://www.theguardian.com/commentisfree/2016/feb/04/removal-road-markings-safer-fewer-accidents-drivers

They came to the conclusion that it increased safety, because all the drivers drove slower and took more care. To me, this is a crazy conclusion, especially when people who visit towns infrequently have to do so.

As I sometimes drive in unfamiliar places, not having hints painted on the road about whether the feft lane is a left-turn only, and losing stop/give way markings on roads, this whole idea fills me with horror.

I've half left the rat-race, and live in a rural town, and when I do go into more heavily populated areas, I'm horrified by the four and sometimes 5 lane roundabouts that appear to be in favor with road designers (examples being the A34/M4 interchange just north of Newbury, the M42/A556/M6(toll), and it seems the M5/A358 in Taunton and I'm sure there are many, many more), where the thought of no road markings fills with dread.

I have no idea how autonomous cars are meant to cope with no road markings.

Terminal trickery, or how to improve a novel immeasurably

Peter Gathercole Silver badge

Re: Black magic

One further point. There were problems with multi-byte function keys, but not with lost characters.

The problem is that how does the computer know whether a sequence of characters has been generated by a function key, or typed. The answer is that it would time how long it would take to receive the complete sequence of characters.

But this caused problems. If you use a terminal or terminal emulator with some networking products, like a terminal concentrator or even from one system to another, quite often typed characters were buffered until either several had arrived, or until a timeout. This meant that a sequence of 4 or 6 characters could arrive with no gaps, even though there were appreciable gaps (measued in 10's of seconds) between them being typed.

One place I worked used HP2392a terminals (very nice, but quite expensive) and some cable modems (Itil, Itec or something like that). The HP terminal used simple 2 or 3 character function key strings, and someone had written terminfo entries for all of the keys on the keyboard, including "Clear Display", which was mapped to the terminfo key "ed". In vi had the function of clearing from the cursor to the end of the file.

Unfortunately, the sequence generated was "ESC" "J". Anybody knowing vi will know that in command mode, "J" joins the current line and the next line together.

We had one person who used this quite frequently, and got into the habit of typing "ESC" to leave insert mode (as we all do to make sure we're in command mode, don't we) followed by "J". Where this was buffered by the cable modem, it arrived at the other end with no gap between the two characters, and vi promptly deleted the rest of the file in the buffer (because it saw it as the "Clear Display" function key).

This person was getting more and more annoyed as time went by, until I sat down and had a think about the problem, and finally worked out what was happening. So I generated a new terminfo entry with the mapping for this key deleted. Not many people think about this, but it is possible to have a private terminfo file over-riding the the system one, which is what I did for this person.

In a similar but opposite vein, sometimes long escape sequences can be broken into two network packets with a gap in between, causing curses/vi to think that function key presses have actually been typed. IIRC, this can be fixed using the TIMEOUT ksh environment variable (not to be confused with TMOUT), which can be used to set the time in miliseconds that is allowed to be treated as a single key press.

I believe that there is also a terminfo setting to specify this as well.

Peter Gathercole Silver badge

Re: Black magic

This was not a problem (in fact not a problem at all) with just the DEC terminals.

Unfortunately, the range of commands that you could implement using just single non-printing characters would effectively have crippled the terminal, leading them to be very 'dumb' with no cursor addressing or other advanced features.

All terminal manufacturers had multi-character key sequences, and I speak from experience having been the termcap, terminfo and curses SME at many places I worked at. Manufactures such as Wyse, Lear-Siegler, HP, Beehive, Heathkit and many others I could name all had their own multi-character function keys sets.

As I said before, look at the /etc/termcap from a BSD system to see a pretty good snapshot of pre-ANSI terminals across the ages,

I would be pretty worried if the code that you were using to read the UART could not keep up with the speed of the serial line, because things like modems have existed for ages, and would squirt characters in at the line speed.

When you look at mini-computers, generally the handling of the UARTs was offloaded to dedicated hardware, so the CPU would not be interrupted for every character.

In fact, RS-232 was designed to be asynchronous, with the character timing loop being triggered by the leading start bit transition, so that the monitoring hardware can be idle and interrupted when needed when data was arriving. It would be tremendously wasteful if it were implemented any other way,

‘Staggering’ cost of vintage Sun workstations sees OpenSolaris-fork Illumos drop SPARC support

Peter Gathercole Silver badge

VT220 terminals were text-only. The ReGIS terminals in that generation were VT240/VT241 which were monochrome and colour respectively.

Some manufacturers made VT220 compatible terminals with Tektronix 4014/4015 emulations built in. My favorite was the Falco 5220 (Falco was set up by some engineers who left Wyse to set up their own company), which allowed me to write a terminal definition for the Access 20/20 spreadsheet to switch modes to allow text mode for spreadsheets and Tektronix mode for graphs in the same session.

It also had other interesting features, including a second RS-232 port which could be used as a second session to another system, either in a separate scroll region or a hot-key swapable screen, or to attach a printer. I worked out a way to control this to allow text files to be copied from one machine in the first session to a second machine on the other session. Before the days of general networking, was often the easiest way of moving files between to systems, especially when used with uuencode/uudecode.

All a bit passe nowadays, but fun and useful at the time.

Compsci boffin publishes proof-of-concept code for 54-year-old zero-day in Universal Turing Machine

Peter Gathercole Silver badge

Re: The illusion of absolute security

Back in the early days of one of the major UNIX variants, one of my colleagues actually wrote in the "Remediation" section of a serious security problem report sent to the developers something along the lines of "Turn the system off, unplug it, put it in a secure cupboard and throw away the key".

He felt that this was the only way to prevent this particular security exposure from being exploited.

43 years and 14 billion miles later, Voyager 1 still crunching data to reveal secrets of the interstellar medium

Peter Gathercole Silver badge

A sign of (past) times

This is a testament to the days when in order to try to ensure that something would last it's expected life, it was over-engineered.

We now know so much about materials science that we can engineer to the limits of the materials. These new things will not last in the same way.

I'm not sure whether this is a good thing or not. On one hand, we get much more bang for the buck during the initial mission. On the other, we're less likely to get these amazing windfalls in the future.