* Posts by Peter Gathercole

2953 posts • joined 15 Jun 2007

UK.gov wants to stop teenagers looking at tits online. No, really

Peter Gathercole Silver badge

I wish politicians would learn...

...that the Internet is trans-national, and as much as he would like to, he can't penalise a company outside of the UK.

All he can do is to try to get the UK ISPs to block access to offending sites, but as we've seen from TPB, that's like playing whack-a-mole.

I can sympathise with trying to keep certain content away from vulnerable people, but that doesn't mean that I can see a way of doing it without breaking the Internet!

Start learning parallel programming and make these supercomputers sing, Prez Obama orders

Peter Gathercole Silver badge

Re: Hang them by their lab. coats.

Unfortunately, there is relatively little commonality between HPC systems from different vendors, and as with most large problems, it's the interconnect between the individual system images in the clusters that is most important, and different vendors quite jealously guard their specific implementation to maximise the value of their investment.

Unlike general purpose servers, there are a lot of tricks that go into HPC servers to make them as fast as they can be. Beside the interconnect, there's different ways of packaging multiple processors in as smaller power and space footprint as possible, and once you start putting so many processors into single system images, especially if they are heterogeneous processors (think hybrid or CPU/GPU processors), the way that the memory is laid out and accessed becomes very important. All of this can affect the way that the code has to be written. This is even though there are relatively efficient abstraction layers such as MPI, OpenMP and MPIch

This means that in order to get the absolute maximum utilisation, there is a long period of tuning when porting code from one to another, For example, the installation of the Cray XC40s that are currently replacing the IBM P7 775s at UKMO is a project that is running for over a year, from purchasing decision to final switchover, and much of that is taken up with the porting and resultant checking of the models between the systems.

I suppose that normal commercial systems vs. HPC systems is a bit like the difference between a Ford Transit and a Formula 1 car. You definitely want to invest in making the F1 car as fast as possible.

Any programme that will result in a consistent, efficient programming model that abstracts the system specifics to allow increased portability of code would be very welcome by pretty much everybody in the field.

NHS England backs down over another data extraction scheme

Peter Gathercole Silver badge

"de-identified patient level data"

The lesson that they've (whoever is suggesting such things) really not learned is that "de-identified patient level data" is not de-identified enough.

They way they look at it is that once it is de-identified, that it's not usable to track individuals. Until they realise that it's still possible to tie the data to an individual by synthesis with other data sources, we will get this happening over and over again.

If I put this story together with others that have been in the news, this is someone in government trying to more accurately estimate how much missed appointments cost the NHS, in order to try to work out a policy to minimise the 'loss'. I'm not really fussed about this type of exercise, except to point out that it's a futile operation, because the results would be meaningless.

There's some naieve politicians who believe that a missed appointment is wasted time, and thus money. In reality, it's not, because of the way that pretty much all NHS appointment systems take missed appointments into account by over-booking the system. If you're unlucky enough to be in the system on a day when people don't miss their appointments, especially if you've got one late in the session, then you'll find that your appointment time is wildly optimistic, and you end up being seen hours later than your due time as the surgery or clinic overruns it's opening hours to see all of the patients that have been booked in. It's no wonder that many GPs and specialists complain about over-work.

Unfortunately, this exercise is unlikely to increase capacity, but may ultimately be used to generate revenue through 'missed appointment' fees if the bean-counters believe that the value generated from these fees would exceed the cost of administering and collecting them.

The sooner the politicians and the civil servants that advise them are given some real training in Data Protection and IT in general, the better IMHO.

Apple Watch is such a flop it's the world's top-selling wearable

Peter Gathercole Silver badge

@Frank Bough

It's all relative, and electronic watches can have all of these features without becoming complex. They're commodity items now. Standard quartz movement plus one chip, a display and a battery in a case. Not complex in this day and age. Most mechanical watches are far more complex than any with a quartz movement, and designer analogue watches, while beautiful to look at, are often so stylized to be deliberately complex, for no real benefit in function.

I wonder how many of the readers here have Breitling, Tag Heuer, Rolex or other luxury or artisan watches that make my 12 year old £40 Sekonda or my £25 Zeon tech backup look positively ordinary!

Peter Gathercole Silver badge

Re: Who actually wears a watch anymore? And why? @jake

I said relatively.

It's a dual digital/analogue watch, so the 'extras' are not so special for a digital watch. It looks like an analogue watch with a simple digital display (6 characters plus a couple of indicators) set as the face.

Complex is an Apple Watch or an Android, or something with lots of dials, big obtrusive buttons, more than two timezones or multiple timers/alarms, or pulse/blood pressure monitors.

Yours is a very simple one. Only to be more simple if it doesn't even have a sweep second hand.

Peter Gathercole Silver badge

Re: Who actually wears a watch anymore? And why? @jake

This question seems seriously out of character for you, Jake.

Normally, you come across as a very conservative, relatively backward-looking person, so as a result, it's quite a surprise to read it, unless you've either dropped out to the extent of living by the sun, or have a pocket watch!

Even though I carry my phone around with me, having a watch (a relatively simple one, time, date, chronograph, alarm and dual time zones) allows me to read the time with a relatively simple and consistent twist of the wrist, any time, day or night (I very rarely take my watch off except to make sure I don't scratch the wife when, oh - you know. This means that it's not been taken off for a while!)

If I had to rely on other devices around me, I would have to first of all remember where the closest clock is where I happen to be or get my phone out of it's belt holster and unlock it, and then hope that the home screen is showing, and has the correct time, as it comes on showing the time that it was locked, and then updates about a second later. I run with the status bars hidden on my computer screen (It's only 64 pixels of vertical space, but I want to use it for content), so I don't even see that unless I move the mouse.

And on top of that, I've been wearing a watch since I was about 10, so it's perfectly natural to me, and it would take some time to adapt to not wearing one.

Peter Gathercole Silver badge

Re: They need to fix the bugs

Surely, you mean 'cheaper' one.

Even that's not cheap.

Your poster guide: A fascinating glimpse into North Korea's 'internet'

Peter Gathercole Silver badge

Re: Internet Directory

I know about that, but it was very little about "web sites" (as stated in the original article), and more about gopher, archie and email, often down to individual's email addresses.

At the time, my major go-to was one of the sunsite ftp and gopher mirrors. I would say that what is now called the Internet started with Altavista, Excite, Infoseek et. al.

Peter Gathercole Silver badge

Re: Internet Directory


There was only a very short period of time (18 months or so) between the launch of NCSA/Mosaic (1993 on UNIX and Amiga), and thus the web proper not the Internet, and the founding of the Altavista search engine (1995). And during this time, public use of the Internet was almost non-existent, and I very much doubt other services were popular enough to merit a generally available book. And DNS operated, so you often had a starting point for looking for something.

Windows did not get Spyglass Mosaic until 1995.

So where is the gap?

If you are talking before the use of HTML, then there were lists of Archie and Gopher sites, but they were really just paper copies of the services own indexes.

The US taxman thinks Microsoft owes billions. Prove it, says Microsoft

Peter Gathercole Silver badge

Re: Corporations are not people @AC

I'm interested in which side of the pond you are, and whether you are truly self employed.

In the UK, and you really are self employed, or a 'sole trader', then you won't have filed a company tax return, as you don't run a company. You will have filed a personal tax form which includes justifiable expenses. I don't know about the US or other countries.

If you run your own limited company, although lots of financial institutions like to treat you as if you are, you're not actually self-employed according to HMRC. You are actually employed by a company that you own. Not the same thing at all, especially when it comes to liability for tax.

For the UK, the rules are outlined here.

The taxes that companies pay are taxes on profits and certain types of transaction, and profits stay in the company, they're not paid out as salary, wages or justifiable expenses, at least not in the year that they're declared. A company may have revenue measured in millions, but may end up paying no corporation tax because all those millions are paid out in wages, materials, and other business costs.

Indeed, many small companies pay pretty much no corporation tax, as their purpose for trading is to provide the owners a living, so they're arranged that all profit that comes in over non-wage costs gets paid out with the appropriate level of tax. The purpose of running it as a limited company is to provide isolation between the business and the individual, as most people do not want to lose their house if some financial problem hits their company.

What financially savvy companies try to do is to make sure that the operating profit gets extracted from the company to the owners or shareholders with as little tax due as possible. The borderline between what's allowed and not is like a battlefield, with the front lines moving all the time. And the accountants and Revenue are the generals that get to control the battle!

Here's why Whittingdale kicked a subscription BBC into the future

Peter Gathercole Silver badge

Re: But... how do you control the cards @Irongut

Sky control it for their satellite service by making all of the boxes require a card, and then charge you ~£11 per month on top of your basic subscripting per additional STB.

You also have the problem of long runs of pretty fussy co-ax cable to the satellite dish (you can't just split it because of the signal polarisation), and the requirement to have a quad- or octo- LNB.

So no. They've not controlled it. They've made it a revenue earning opportunity, like they always do.

Peter Gathercole Silver badge

DTV was all about the cost.

It all comes down to money. During the switch to Digital Terrestrial, TV consumers were being effectively told that they had to spend money in order to continue watching TV. This was never popular, but it was convenient that it coincided with adoption of flat-panel TVs that softened the blow by giving consumers larger screens and space in their homes back as part of the 'deal'.

It would have been possible to publish requirements that mandated more functional and thus more expensive devices, but if the minimum cost STB device was £70-100 rather than the £20-30 (or even cheaper for the supermarket specials) that it was, imagine what the backlash would have been.

I remember at the time I ended up buying freeview STBs for 8 TVs in the house (it's a big house with TVs in most of the bedrooms) and ended up paying a couple of hundred quid for the privilege. I would not have been happy if it had cost me £500 instead, just for the privilege of being able to use the TVs I already owned.

Of course, I don't think that any of the first set of adapters I bought are still functional (catastrophic capacitor breakdown took most of them out - with the exception of the very oldest - about 12 years old, which was still working a few months ago)

These old adapters have mainly now been replaced, along with the TVs they used to drive with TVs that receive freeview anyway. If I had to go through the whole exercise for 7 TVs (one bedroom is no longer being used as a bedroom), I would be pretty unhappy.

I'm getting pretty fed up of everybody, from the technology companies through government and down to people I know who seem to assume that everybody will be replacing tech on a 3 year cycle. It just does not fit in to many, many peoples lifestyles to replace all their tech over such a short time span! As a result, road-maps for at least the next 10 years are required to allow the public to decide whether to spend little and often or a lot, but less frequently.

The roots go deep: Kill Adobe Flash, kill it everywhere, bod says

Peter Gathercole Silver badge

Well, I can't be pwned from Word and Powerpoint....

Because Microsoft don't make them for Linux!

From other things, well, maybe.

Run Windows 10 on your existing PC you say, Microsoft? Hmmm.

Peter Gathercole Silver badge

Re: Spring cleaning time

In my experience, apart from some kernel stubs for exporting a minimal API to the adapter, most of the code to drive the adapters is in the X Server modules, and thus run in user-space, not kernel space. You're mostly isolated by KMS in kernels from about 2.8.

I was assuming that the type of problem I was addressing was one where someone ran a normal 'apply some security fixes' upgrade, which as a result broke support for the adapter they had. With most distro's, this will not introduce a new kernel, and regressing to a driver in the same version will most likely work.

But you are right. I should have said "Another possible solution, assuming that the interfaces haven't changed.....".

I was mostly thinking about distro's like Ubuntu and Fedora. All bets are off if you do a dist-upgrade. I have encountered that sort of problem with a Ubuntu 10.02 to 12.02 upgrade on a machine with an embedded Nvidia 7100 display adapter (I don't need significant 3-D capability on this system), where I got absolutely no graphical display (text mode only) at all until I worked out what had happened. And then I moved the disk into a new system with an 8800, and things got quite crazy again for a short while until I realised that I had frozen the Nvidia packages to get the 7100 working! And don't mention cheap ATI 9250 cards! I really want to forget those completely.

I'm really not looking forward to Mir and Wayland, because I mostly understand how this works with X11. More to learn and more to get wrong, and probably whole generations of older graphics cards that will not work at all, no matter what you do.

Peter Gathercole Silver badge

Re: Spring cleaning time

Ubuntu is sometimes it's own worst enemy.

It's actually very good at telling you that there are proprietary binary drivers available for your video card, and telling you what you need to do to enable the non-free repository and switching to the driver.

Unfortunately, it's not very good at telling you that the new binary driver you've just installed as part of the update has dropped support for your graphics card. The result, you put the update on, reboot the system, and hey presto, you're back in un-accelerated 640x480 16 colour world, or if you're very lucky 800x600 VESA mode, whichever is the lowest common denominator. But you should get some sign of the screen working, even if it's just text-mode.

The solution is to remove the nvidia or ATI fglx drivers, and install either the nv or nouveau driver for nvidia cards or the radeon driver for ATI cards. Nouveau and radion both provide some 3D function, although it's likely to not be as good as the binary driver (but still perfectly adequate for 2D work and even things like Google Earth).

Another solution is to work out which the last binary driver supported your card, and back-level the package to that, or even add the repository for the earlier release and back level, and then freeze those packages. But this later option can sometimes lead to strange booting effects, especially if the KMS support for the cards has changed.

BTW, a fresh install would probably just work.

Peter Gathercole Silver badge

Re: Ha

But what are you counting as an obscure peripheral?

I've had problems with a Pinnacle PCI video capture card, but to be fair, there was no post XP drivers for that anyway.

And a slide scanner. Ditto, no XP drivers, and Linux support patchy, but I can run XP in VirtualBox to access it over USB.

I did also have some problems with a broadcom wireless chipset from around 2001, but it would not work with WPA in XP, even using the available Windows drivers.

But graphic cards? The open source nvidia and ati drivers work well (at least for 2D) with most old hardware, in fact much better than the legacy Windows ones once the proprietary windows ones have dropped all support. And the open source 3D support is getting better all the time, and for reasonably current hardware the binary non-free drivers actually work very well. There was some criticism of multi-head support, but it does work, although maybe not as easily as Windows, and again, it's getting better all the time.

Similarly sound, network, wireless, USB devices. I have far more problems with Windows drivers rebuilding older machines than I do for Linux.

Generally for older hardware, if someone wrote a Linux driver for it at some time in the past, it's still there and probably still in the repositories and the module stubs are still in the kernel, unlike Windows, where the old drivers more often than not will not work at all.

One word of caution for people with older Celeron, some Atom and some Mobile Pentium processors (like the Banias Pentium M that was put in many laptops in the early noughties) that either do not support PAE (Pentium Address Extension) or report it wrongly. Modern Linux distro's often do not come with kernels that support these systems. It is sometimes possible to work around this, but generally it's not worth the effort.

HP tops HPC charts, former Big Blue biz not so super

Peter Gathercole Silver badge

IBM has given up on HPC IMHO.

IBM, having dropped the IH systems, and having sold off the iDataplex and NextScale systems, and there being no obvious successor to BlueGene/Q appear to have lost interest in HPC.

I know that there are supposed Power 9 hybrid systems in the pipeline for Oak Ridge and Lawrence Livermore in 2017, but at this time, these both look like vapour-ware, there being only limited details of either the Power9 processor or the Volta GPU. As we found out with Blue Waters, such projects don't always deliver.

Both the Top500 and the article are incorrect about IBM's market share. It's actually less than stated, as the ECMWF P7 775s at 108 and 109 are no longer there (and were actually turned off for the last time something like a year ago), and before the November list, the UKMO systems at 140 and 169 will also be decommissioned. This will drop IBM down the list still further, with no obvious real big systems in the pipeline for the next year or more to push them up the list!

For me, it will be a sad day when the UKMO P7 775 systems are turned off, because it also means that I will be looking for a new assignment.

Any offers?

I cannae dae it, cap'n! Why I had to quit the madness of frontline IT

Peter Gathercole Silver badge

Re: Oh so true! @awhit

I don't think IT is as lucrative as IT (used to be) any more. Blame commoditisation, which incidentally also feeds security issues. Obscurity is not a substitute for security, but it does help!

Peter Gathercole Silver badge

Re: @AC

Come on. How simple are you?

If Trevor were to write about this stuff, two things would happen.

1. He'd get sued for breach of contract (the NDA is a contract).

2. He's get excluded from this sort of information in the future.

In fact, he's probably on shaky ground even admitting that he's subject to an NDA, if they're worded like any of the ones I've been subject to in the past.

So if he did, he would be shooting himself in the feet, both of them.

Flash HOLED AGAIN TWICE below waterline in fresh Hacking Team reveals

Peter Gathercole Silver badge

Re: Oh Adobe... @Chris 155

"Flash was first..."

First, unless you count RealPlayer, or possibly xanim.

IETF doc proposes fix to stop descent into data centre 'address hell'

Peter Gathercole Silver badge

Nothing new?

This does not sound too dissimilar to the layer 2 heuristic bridges that I was using 20 years ago to bridge geographically separated networks over slower WAN technologies with some degree of optimisation. Of course the scope is different, but the concepts look very much the same.

Call that a mugshot? Aussie model/fugitive asks rozzers for more flattering pic

Peter Gathercole Silver badge

Re: I am more concerned... @Chemist

Yes, of course the picture is still important, but if it was just the picture and manual appraisal, they would probably be less strict about the expression, background, glasses etc, as the officers would probably prefer to have pictures that resembled you as you normally look, much as they used to do before biometric passports came alone.

It's true, I don't travel that much. Do the immigration officers ask you to take off hats, glasses and comb back your hair so that they can make an accurate appraisal of whether you resemble the picture? If not, then the picture is of limited use.

But conversely, if you travel to a country that does have the equipment, the biometric data will probably be read off the passport and recorded in a database that LEOs have access to, so that if you are picked up dead, or infringing the law, they can make a more positive identification of you. Biometric data is less than perfect, but the basic map of the face can give useful information, and it is much more accurate if the face is not obstructed and in the same orientation as the picture.

BTW. If your wife's passport is a non-biometric one (and I'm making a big assumption that it's a UK passport), it is probably close to needing to be renewed.

Peter Gathercole Silver badge

Re: I am more concerned... @Chemist

The reason for the strict instructions is not to generate a picture from which a person can identify you, it's so that the computer generated hash of the salient features of your face can be encoded into the biometric data stored on the passport (things like the ratio of the gap between the eyes and the length of the nose).

Glasses, the direction you're looking, tilting the head, obstructive hair, and even the change in shape of the muscles in the face if smiling and the background can all make a significant difference to the hash.

And again, it's not about people looking at the photo, it's about you being positively identified by machine. It's much more difficult to fool a machine (with the right data) than it is a person.

Behold the mighty Swiss SPACE JUNK NOSHER PODULE

Peter Gathercole Silver badge

How about calling SHADO

(Supreme Headquarters Alien Defence Organisation)

They used to have clean up missions (1.05 - Conflict)

And I think Thunderbird Three was recently seen cleaning up space junk!

Peter Gathercole Silver badge

Re: Space Wombles @Crisp

... But that was to Button Moon, in the Blanket Sky!

I suppose that the theme was composed by an ex. Dr Who, and Trillian from HHGTTG TV show (and also the voice of Grandma in 2015 Thunderbirds). Maybe that gives it some authenticity!

Peter Gathercole Silver badge

Re: Space Wombles

Surely the Clangers would be the people(?) to call.

Their music boat with Tiny or Small piloting would be just the ticket.

The Space Chicken might be useful too for the smaller items.

Smart Meter biz case still there, insists tragically optimistic UK govt

Peter Gathercole Silver badge

Re: I had a smart meter fitted last week

You've just outlined pretty much all of the consumer benefits (you missed out remote meter readings).

You've not seen what the downsides are, of remote control of your power system, of remote hacking, which may allow other people to determine whether the house is empty, and even turn off your power if the meter is one that does this. If it's not one of those, it will need replacing again before the program is complete.

Also, from what I've read, you've also pretty much locked yourself into a single supplier, because they use different and incompatible metering technology. If you wanted to change to a supplier that used a different meter technology, then you may have to pay an additional meter installation charge.

You could have achieved much of the same benefit for electricity (sans remote meter reading) with one of any number of clamp-on external measurement systems (OK, they're less accurate, but still can demonstrate power use in real time) without any of the downsides.

I listened to the Radio 4 Today program this morning, and Roger Witcombe, chairman of the Competition and Markets Authority (a government institution), who was a guest on the discussion about energy companies and overcharging, mentioned smart meters as an aid to choosing supplier, but in such an apologetic tone of voice that it seemed to me that he was echoing an official line while not really being supportive of the devices himself!

Cloud provider goes TITSUP? Will someone think of the data!

Peter Gathercole Silver badge

And remember

If you're also using your service provider's backup solution, unless you've got your backups stored to physical media outside of your service provider, even if they have separate tape copies, the physical media probably belongs to your service provider or upstream provider, and will be a tangible asset (and thus liable to lien).

This will be true unless you've got an ironclad contract that states the media reverts to your ownership in the case of your service provider entering insolvency, If you don't, you'll probably also lose access to your physical backups as well, and possibly any archival copies kept for regulatory compliance.

NASA's New Horizon probe rudely fires its thruster at gnome planet

Peter Gathercole Silver badge

Re: Genuine question

I'm sorry. You've done it now. I can't resist.....

"... Late, as in the late Dent, Arther Dent. It's a sort of threat, you see..."

- Slartybartfast.

Peter Gathercole Silver badge

Re: Genuine question

Remember that Pluto is in it's own orbit, and moving quite fast (4.67 Km/s), so late as in crossing Pluto's orbit after it has passed by. 20 seconds would have increased the closest distance, but probably not by much compared to the 7,750 miles distance.

But the answer is in the quoted article from NASA. "...[JHAPL] says without the adjustment, New Horizons would have arrived 20 seconds late and 114 miles (184 kilometers) off-target from the spot where it will measure the properties of Pluto’s atmosphere. Those measurements depend on radio signals being sent from Earth to New Horizons at precise times as the spacecraft flies through the shadows of Pluto and Pluto’s largest moon, Charon."

So yes, late.

Cash-strapped Chicago slaps CLOUD TAX on Netflix, Spotify etc users

Peter Gathercole Silver badge

Re: Is it legal? @arrbee


Peter Gathercole Silver badge

Re: Is it legal? @arrbee

I'm not going to quibble. I appreciate all of your points, and I sympathise with the people who are caught in this trap.

I was just objecting to the use of the term "tax".

I would actually rather prefer to have a more realistic balance between living costs and wages such that things like housing benefit and other subsidised housing (yes, I'm including council houses and housing associations) were only needed by a much smaller number of really needy households, rather than those who just can't afford to live where they do, whatever the reason why.

For what it's worth, and for reasons beyond most of these people's control, they are forced to rely on the state and it's devolved institutions for support far more than is healthy for the nation's finances. I wonder if enforcing the living wage and reducing benefit paid, and then carefully change the tax and/or NI on companies by an appropriate amount to shift the money away from benefits and on to a more income and tax basis would be a reasonable first step? Possibly Tim Worstal cold crunch the numbers and comment?

I appreciate what you are saying about buy-to-let, but the right-to-buy, which is what caused the public housing to be sold off was intended to benefit the original purchaser. Some of them will have bought and then sold, making significant profit for the people taking advantage of the right, but when they sold, it would be at the market rate. The b-t-l landlords would have paid market rates, and the only benefit thy has was that the houses were available at all. But the house transferred to private when the original purchasers bought, not when they were sold on.

What has helped the b-t-l purchasers most is the mortgage guarantee and incentive packages that were introduced to try to support the first time buyer market, and thus the whole of the housing ladder. These were not sufficiently guarded to prevent b-t-l landlords from using the schemes. Other than that, it's the relentless rise of house prices that allows a landlord to borrow against their existing portfolio to fund purchases, and then rely on the price rises to provide them a capital gain so they can either borrow more, or sell in the future and make a profit far higher than commercial interest rates.

There's lots wrong with the economy, much more than the loss of public housing.

Peter Gathercole Silver badge

Re: Is it legal? @Mage

Gawd. I'm going senile. Read Queue as Cue!

Can we have a bit more than 10 minutes as an editing period, please?

Peter Gathercole Silver badge

Re: Is it legal? @Mage

UK "Bedroom Tax" is not a tax. The media and opposition parties deliberately misrepresent it to increase the emotive impact of the issue.

It is actually a rebate on the housing benefit (a welfare payment) given to someone who lives in a house with an unused bedroom. (Rebate is used a little strangely here, because you don't normally expect it to be used to benefit state institutions).

If you do not get housing benefit, you are not affected at all by this.

Queue downvotes, but I don't think that the state should pay welfare benefits to people so that they can live in houses larger than they need, although I do think there needs to be significant exceptions for people with intermittent requirement for an extra room, for example if they need to house significant amounts of medical equipment, an occasional live-in carer, or children and members of the armed forces returning home for holidays and leave.

If it were a true tax, everybody would have to pay it, even people not receiving benefit and/or people owning their own house outright.

On the story, I thing Chicago are ill-advised to introduce a tax that is going to be difficult to enforce.

What's black, sticky, and has just 8GB of storage?

Peter Gathercole Silver badge

Re: Thin clients on wifi

I still run a lot of stuff over X both at home and at work (obviously through SSH X forwarding).

My primary go-to system is a Thinkpad T43, 2.0GHz Dothan Mobile Pentium with 2GB memory running Ubuntu Trusty Tahr and Gnome-flashback. It runs very well as an X server connected to other systems, both more and less powerful.

What kills X is the appalling way that some applications, particularly Java ones, are implemented. Too many client applications render the screen locally, doing thing like all of the font handling locally, and then sending the rendered screen to the server.

Now I know that this is the only way that a client application can guarantee that certain fonts are available, and rendered as they expect, but it's seriously ugly in use, and it breaks the ethos of X11, where very efficient network primitive operations are sent across the network rather than the screen bitmaps.

Certain things, like video, are clearly not suited to X, but properly written X applications can be exceptionally snappy.

I think back over 20 years to running IBM X-Station 130s in a live operation (actually the IBM UK AIX Support Centre) from RS/6000 320H servers, about 10 X-Stations per server, over 16Mb/s Token Ring, and it was not X that slowed things down, it was the processing power and memory on the servers running the clients (isn't the X computing model confusing sometimes!). At the time, this was a very realistic proposition, and I'm sure that the increase in processing power and network speed could make thin-clients technically feasible again, but there is no cost benefit any more.

With the rising popularity of delivering apps. through HTML, I can see future thin clients being <£100 android tablets with keyboards (and maybe mice), possibly built into desks rather than sitting on them!

A stick like this could be useful in a scenario like this, but once you take into account the cost of whatever you are displaying it on, and the keyboard and mouse, the tablet option I described above looks much more attractive. The real benefit of these systems would be in a household that does not want to have a desktop PC, but may occasionally want more than can be done on a tablet.

Peter Gathercole Silver badge

Re: Plenty of things I can do with 1G ram + 8G flash

I have a first generation Acer Aspire One (1.6GHz Atom N270, 1GB RAM and 8GB flash) which sounds like it's similar spec, and I have Trusty Tahr on it, and I still use it for basic web browsing, and playing some media.

I use gnome-flashback, partly because I prefer the interface, and partly because Unity is painful with slow graphics and only 600 vertical pixels. The other drawback is the SSD is seriously slow, possibly because it does not support TRIM.

The biggest problem is keeping enough of the filesystem free during updates. The update tool leaves a trail of downloaded deb files after they've been installed, and never cleans up old kernels.

If I did not make serious effort to clean up after every update, it would have run out of space ages ago. I hope that the version installed on this stick has been tweaked to do some of this automatically.

fwiw, ubuntu-tweak and computer-janitor are seriously useful for keeping this cruft down.

Samsung/Cheil merger goes on: Soz Elliot, not this time, says court

Peter Gathercole Silver badge

Re: knobs

I suspect that it's not a case of them thinking they're going to lose money. Sam*eil (or Ch*ung maybe) will almost certainly make them more money than most corporate investments.

It's more a case that they think that Samsung on it's own will make more money, and will be easier to influence than the enlarged combined company, in which the individual investors will have a diluted share holding.

Microsoft in Blighty reveals its 78 THOUSAND POUND Surface 3 slabloid

Peter Gathercole Silver badge

Re: Or just a dummy value...2 x mistakes = correct @Bob

I don't see a joke icon.

If this is real, let us all know whether they arrive!

Beyond the Grave: US Navy pays peanuts for Windows XP support

Peter Gathercole Silver badge

Re: Windows for Warships

Ah, OK.

But even so, Microsoft may have to support XP for the DoD until at least 2018, so one wonders whether the same "Custom Support" plan rules operate for DoD as for other people.

Peter Gathercole Silver badge

Re: Windows for Warships

I'm sure that I read somewhere that in part of the contract to supply the DoD with weapon, control or maintenance systems, there was a clause requiring a 10 year withdrawal of support notification.

This means that the supplier has to warn the DoD of the date that the kit would not be maintainable 10 years before the support was withdrawn.

That makes the 3 year notice rather abrupt, don't you think.

Why OH WHY did Blighty privatise EVERYTHING?

Peter Gathercole Silver badge

Re: But public money... @Alan

It's interesting that you can see some of what you say in the deregulation of the bus network from the Passenger Transport Executive (PTE), run by Tyne and Wear around Newcastle and the immediate area (and probably the other PTEs from around the country).

The PTE ran a fully integrated transport system that included the bus services and the Metro rail system.

When it was set up, it was arranged around interchange hubs, and your journey would often be bus to a hub, Metro between hubs and bus to your destination. This was paid for as a single ticket via 'zone' pricing, so you could buy your ticket on the bus by asking for a Ticket to the zone your destination was in (that you could find easily on a map), and that would cover all steps. Public transport across the Tyne was mostly restricted to the Metro, and all of the zones were well served by a comprehensive bus network.

The result was that almost everybody used it. It may have been a bit slower than travelling by car, but if all you needed to do was get yourself across the region, it was a no-brainer. You just used it, like you use the underground in London.

Roll forward to 1986. The bus regulation the PTE ran was removed, and the PTE themselves had to divest the bus operations (although they continued to run the Metro). Suddenly multiple operators started competing for the lucrative routes and ignoring the less profitable ones. Buses started crossing the Tyne again. Road traffic across the bridges slowed to a crawl. People did not know how to price a journey, and had to buy several tickets from different operators. Use of the Metro started to decline, because it was no longer the common trunk to link the hubs together.

Eventually, all of the local bus companies ended up being bought by the national bus operators, which essentially stifled the competition, resulting in effective monopolies.

They are finally trying to re-introduce a common fare system AFAIK. I no longer live there, so I'm a bit out of the picture.

Who wants a classic ThinkPad with whizzy new hardware? Lenovo would just love to know

Peter Gathercole Silver badge

Re: It is already there @naive

The T4X series, it was not the RAM slots, it was the ATI graphics chip that would come partly un-soldered. I've worked on about a dozen T40-T43 systems, and I've never come across one with the memory slots not working.

Peter Gathercole Silver badge

Re: Yes, I would purchase one (maybe 2 or 3) if... @picturethis

I've done the disk clone thingy with dd several times. I'm not concerned by it, although you can get a bit screwed by the UUID in /etc/fstab of some partitions under some circumstances.

The Windows side is more tricky, because moving the image to a new system will trip the Windows Genuine disAdvantage 'it's no longer the same system' in later Windows versions, and you have to change the license key to match the new system (believe me, I've done it several times). But I only keep a real windows partition going for those very rare circumstances when I can't do what I need to under Linux or in VirtualBox. It gets booted about twice a year!

The Windows drivers are not really an issue. It'll always come up with VGA graphics, and as long as the usb 1.1 drivers are installed or the optical media works, you can install the correct drivers (My Thinkpad history: 365X ->365XD ->380D ->T20 ->T23 ->T30 (multiple mobo's) ->T43, and I had an N33sx and an L40sx before the Thinkpad brand). During this time, I've gone through many disks, although for the last 10 years or so, I've only changed the disk as I've needed more space, not when I've changed machine.

Peter Gathercole Silver badge

Re: Yes, I would purchase one (maybe 2 or 3) if... @picturethis

I think you're very lucky. The T30 was a bit of a bogey machine, the worst of any of the Thinkpads that I've ever had. Don't get me wrong, I kept my T30 running for as long as I could, but...

The T30 was the first Thinkpad that was completely made in the Far East AFAIK, and they got the mechanical design a bit wrong. They are notorious for the RAM sockets to break solder joints. Finding a T30 with both memory slots still working without some paper wedged in to put pressure on the slots is a very, very rare thing. I think that the designers recognised this, because there was no T31 or later T3X systems, and the T40 was launched not that long afterwards. I think the motherboard was put under some strain, because several of them I've had have had different types of foam pads to act as strain relief, but they never really worked.

I know Thinkpads are repairable. I had 4 different motherboards in my T30, and eventually resorted to re-soldering the RAM sockets myself, but I don't have a re-soldering station, and using a normal dry soldering iron to melt the solder already there eventually burns the surface mount pads off the motherboard.

I kept it running until I could no longer find any mobos on eBay, and the ones I had could no longer be re-soldered. I eventually decided to replace it with a T43 (this machine) when I could find one with a Dothan processor (and the cost dropped to lower than a T30 mobo, even one not guaranteed to work. But the hard disk (swapped out of the T30 to keep the 'machine' the same even though it's different hardware, is flagging SMART errors, and large IDE 2.5" disks (100GB+) are also getting rare. Core 2 duo T60s (with SATA hard disks) are beginning to look cheap on eBay at the moment, so I may switch again, but this swap will require copying between disks, not just a disk swap.

T30's also has a definitely silly bit of design. If you tried to remove the disk with the lid shut, you were guaranteed to break the top left corner of the bezel. It is such a common problem, that if you see a T30, it's almost certainly broken there.

If there was a modern Thinkpad, in approximately the same form-factor as the T20-T60 ranges, available at a reasonable price, I may just skip to a new one rather than a used one. But, unfortunately, I think Lenovo will take the interest as a sign that people would pay a high price, and they will introduce them at ultra-book prices. If they do this, they've not really looked at what people want.

Peter Gathercole Silver badge

Re: Definitely as long as it's as good as it's predecessors

T20 running as a firewall. 700 MHz Pertium 3, 256MB ram and PC-CARD second Ethernet card.

Too slow even for a firewall really, but been on 24x7 for more years than I can remember.

I bought it second hand in about 2003.

Typed on a T43, my main workhorse, running Ubuntu LTS 14.04. Still fast enough for most day-to-day purposes.

Killer ChAraCter HOSES almost all versions of Reader, Windows

Peter Gathercole Silver badge

Re: C @Dan 55

If you were using ISO standard Pascal (ISO 7185), it may have made for safer code, but in order to achieve that safety, you had to put up with a language that was so limiting that doing something like writing an OS would have been a virtual impossibility.

I mean, really. ISO 7185 Pascal does not have anything remotely like a pointer. It also does not have any way of addressing data objects that have not been declared. It's also pretty difficult to handle variable length records in data files, because of the inflexible nature of the I/O system.

So what happened is that you got things like Turbo Pascal and Delphi, which were not Pascal, and introduced enough of the methods needed to write systems, which also added some of the same vulnerabilities that C had.

You should have used something like Ada as a counter to C, not Pascal. Although Ada was a strongly typed language, it's very reason for being was to write correctly coded systems for such things as military application, so it had the necessary constructs to interface with hardware.

Peter Gathercole Silver badge

Re: C @James Cane

Unless you're prepared to have massive inefficiencies in your code (like surrounding all data structures with hardware write protected regions), it's always going to be a matter of trust.

If you use a language that does strong boundary and type checking for all data objects in software, you're trusting that the compiler and/or run-time is correct, and does not contain any flaws You're also always going to find that your code runs slower with these checks.

I'm not going to suggest that there are not sound reasons to stop using C, but using a language with better protection does not guarantee total safety.

C is still excellent for what it was written for, generating code that closely fits with the underlying computer architecture. But it's not a perfect language, by any means.

When C was first developed, it was necessary to have a language that would map very closely to the system ISA (and it did map to the PDP11 instruction set very nicely), because even minor inefficiencies in code size may have pushed the kernel beyond the 56KB address space on a non-separate I&D PDP11 (I used to run a PDP11/34A, and had different kernels to drive all of the terminals without a tape drive, or to have only some of the terminals working with the tape driver compiled in). IIRC, Sun 3 680x0 systems had to have a kernel less than 1MB.

That time is fortunately past, but that does not mean that in these days of a word-processor needing 100's of megabytes of memory just to load, that aspiring to produce efficient code should not be something to strive for.

Peter Gathercole Silver badge

Re: Compared to this... @STB

The counter to this is that although you (and I, I will admit) may not have the skill to fix problems like this, we do have the ability to aid someone who does, with a formal or informal contribution of either money or equipment, and it does not even have to be the developer with the Open Source software model.

I suppose that you could give Microsoft and Adobe money and ask them to do the same, but I suspect that it would disappear into the general coffers, and not significantly affect the quality of the code.

Why is it that women are consistently paid less than men?

Peter Gathercole Silver badge

Re: I think the argument about shorter life expectancy for men @Tom 13

That's interesting, but I don't actually think that table is relevant to my post.

Firstly, all of the life expectancy figures in that table are actuarial, meaning that they are guesses based on historical figures rather than accurate predictions. And my point is that the assumptions that those numbers are built upon are changing because of really dangerous professions either disappearing, or becoming much safer. The only column that is relevant is the "number of lives".

Secondly, as I was commenting on differences due to work done by men and women, quoting a figure for children aged one is largely irrelevant, unless the US has found ways of infants working.

These figures for boys and girls below working age show something different, although I don't have a clue what it is. It can't even be misadventure, because (again) infants probably don't really behave very differently.

Thirdly, within the timespan of the table (1891-2010), there are two very major and several significant minor events that skew the figures in the US. These are the first and second World Wars, together with the Korean police action, the Vietnam war, the first and second Gulf Wars and the following actions, the US war in Afghanistan, and the actions in Balkans. All of these would significantly increase the proportion of male vs. female mortality, as wars are generally still fought by men, and even those who return may have had life altering injuries that persist in affecting the figures long after the events.

It's also interesting that the figures include populations living outside of the US, which probably include people living and working in much less safe environments than the US mainland. I would also contend that Europe probably has more job related health and safety regulation than the US.

We will not be able to see the full figures for the period I was talking about for may years to come.

I certainly wouldn't argue that the figures do show a shorter life expectancy for men, but my point is that recently, and on into the future, the differences due to the type of work people do (my post) will become less as time goes by.

RBS sticks it to customers once again as IT woes continue

Peter Gathercole Silver badge

Re: Consequence

Many years ago, Girobank once processed a cheque for £1000.00 (one thousand pounds) paid into one of my accounts as £10.00 (ten pounds), even though the cheque was correctly filled in, both words and numbers.

It took about two weeks for them to fix it after I spotted it, because they had to retrieve the original cheque from the document archive. In the meantime, £990 of my money was in limbo, having been taken from the source account but not appearing in the destination account.

They did refund all of the failed transaction charges, and fortunately, the mortgage company accepted that this was not my fault, and did not post a black-mark. Another fortunate thing was that this was the only direct debit from that account.

Biting the hand that feeds IT © 1998–2019