* Posts by Richard Plinston

2608 publicly visible posts • joined 27 Apr 2009

Microsoft makes Raspberry Pi its preferred IoT dev board

Richard Plinston

Re: No comprende

> The W10 image for RPi's is 67MB. That's fairly stripped down from the 10's of GB's that a clean install takes up on PC.

According to Thurrot: " the download arrives in the form of a 500 MB-ish ZIP file." and it requires an 8GB class 10 SD card.

Richard Plinston

Re: No comprende

> Using Windows would mean needing a 128MB (2Gbit) flash device to store OS + application, versus a 16MB one for Linux.

The 67Mb is presumably the download size. This is compressed. W10IoT on RPi2 requires a class 10 8GByte SD card. It is reported to fail on slower cards* and won't fit on a 4GB card.

* presumably there is some sort of timeout problem.

Richard Plinston

Re: No comprende

> The W10 image for RPi's is 67MB. That's fairly stripped down from the 10's of GB's that a clean install takes up on PC.

There are Linux distros that will boot and run off a floppy disk (1.44Mb)*. They require 16Mb or so of RAM because they use much of this as a RAM disk, but the system includes firewall, gateway, web server, DNS, and much else. 67Mb and requiring 1Gb RAM is not 'cut-down' except in the sense that it doesn't do much - it can only run a single app.

*eg https://en.wikipedia.org/wiki/FREESCO

Richard Plinston

Re: No comprende

> Windows IoT (and, indeed the Pi) is pitched at a level of functionality above your bog standard microcontroller where some level of intelligence is needed in processing sensor data, where relatively complex networking could be involved or where there may be a requirement for some level of security.

W10IoT on the RPi2 may find some usage as a dedicated single-use device, such as a Point of Sale terminal, or a Kiosk. With a screen and keyboard, switch it on and it comes up into the UWP app - and can't do anything else. The downside for MS and OEMs is that this would replace a PC with full Windows. W10IoT is designed to communicate via Azure, so perhaps the main point of this is to change the revenue stream from OS sales to cloud.

Richard Plinston

Re: Self-foot-shooting again, Microsoft?

> Why did they not give away the Pi 2 instead?

Mainly because the RPi2 was not available at the time when MS decided that it must have something in this market. They could not get their cut-down GUI-less OS small enough to run on a RPi1.

Richard Plinston

Re: Raspberry Pi 2B is not an IoT device

> A look at the instructions for using it for W10IoT shows that you need a PC or Laptop running full Windows 10 to "drive" it. It is like using a kitchen stool as a step ladder to reach the top cupboard; it can do the job and is probably the most convenient thing to use, but it is not a step ladder.

Not only that, but you need a fork lift to put the kitchen stool into place.

Windows 10 IoT is using a RPi2 to emulate an Arduino - an Arduino nano can be bought for $2-$3 - but it still requires an Arduino, or similar, to get analogue input.

Richard Plinston

Display

One of the 'deficiencies' of the Galileo board, compared to the RPi2, is that it has no display capability. For typical IoT, and for many embedded systems, this is not a problem. But for Win10IoT MS want to promote 'Universal' apps and with no display what is the point of these?

iPad data entry errors caused plane to strike runway during takeoff

Richard Plinston

> the effect of the aerodynamic surface on the weight for the current conditions can be calculated,

You are obviously not a sailor or a glider pilot. If you were then you would know that wind never does what you would want it to do. Changes in speed and direction are generally unpredictable, especially where there are buildings (terminals, hangers, control towers, ..) and other vehicles that have devices designed to move air around in great quantities (propellers, jet engines).

In addition, the buffeting of the wind could change the static friction in the system from acting in one direction to the other and the expected change in loading could actually show the reverse.

Richard Plinston

Re: You have not heard of 'static friction' then ?

> Static Friction is static and known, therefore you should be able to adjust for it.

No it isn't. A typical undercarriage leg has a hydraulic ram with the plane at one end and the wheel at the other. Seals, linkages and dampers provide friction - both dynamic and static. As the weight on the leg changes, due to loading, unloading or to the effects of wind or other, the static friction will resist movement of the ram in _either_ direction. Or it maybe that sometimes the ram is actually in the appropriate position for the weight and there is zero static friction at that time.

The indicated weight may be plus maximum static friction, or minus maximum static friction or anywhere in between.

Richard Plinston

Re: Using toys as tools...

> And if it fails in flight?

If all electricity fails in flight (including all backup systems) in a modern airliner, then it will not be flying much further, and not being able to recharge the iPad (which was the actual point raised) will be the least of the problems.

Richard Plinston

Re: Using toys as tools...

>> " It is not like pilots randomly walk into unknown planes with random devices and get surprised by electricity not working."

> Sure it is.

No. It is airline pilots walking into the airline's aircraft with the airline's certified devices. There will not only be the correct power available, but there will also be a backup for the case of a failure. If the 'electricity is not working' it will be fixed before the plane is allowed to take off.

Richard Plinston

Re: Using toys as tools...

> consumer stuff without real keyboards

The fault in the input did not arise from use of the keyboard. According to the report the weight calculation was done, as it has been for decades, using pencil and paper (notebook) and the fault was 'not carrying the 1' when adding up several figures. This incorrect figure was then entered into the iPad app.

Richard Plinston

Re: Oh guys automate it !

> If you can measure the pressure of the wheels, then for sure one could completely automate this..

No. 'Pressure' is force per unit of area. You would also need to measure the area that the wheels happened to cover on the ground, which is rather difficult to do. As an experiment you could change the pressure in your car's tires and then reflect on whether the car's weight has changed.

Also, aircraft have devices (called wings) that are designed to change the amount of their weight that rests on their wheels. These work, to varying degrees, even when the plane is static and there is a wind blowing.

Richard Plinston

Re: Using toys as tools...

> I can't believe that airlines are using consumer stuff without real keyboards really designed for media browsing.

The primary function of these devices is to hold all the manuals and check lists that used to weigh many Kg as physical books in the cockpit (ie media browsing).

They have a secondary use as doing calculations that previously was done manually with paper and pencil and was more error prone, such as checking manifests and calculating weights.

Richard Plinston

Re: Using toys as tools...

> iPad and Surface are equally good.

It is likely that the airline chose the iPad, developed the applications for it, and deployed them before Microsoft noticed that people were buying these and decided they needed to get into that market.

Richard Plinston

Re: Using toys as tools...

> And how do you know that even if the charger is available the electrical system/outlet/whatever will be online and able to power it?

The pilots work for an airline, the airline owns the planes (or leases them) and the iPads. It is not like pilots randomly walk into unknown planes with random devices and get surprised by electricity not working.

Richard Plinston

Re: Using toys as tools...

>> 2) They need a product with a rich and vibrant software ecosystem and numerous developers that are familiar with writing software for that device.

> I'm sure there are plenty of developers for both iOS and Windows

Microsoft touch devices have been through several iterations of development styles, mostly incompatible. With phones in recent times there was WM6.5 with the devices and apps completely dumped for WP7, which were then dumped for the incompatible WP8. Now W10M has a different development system with 'Universal' apps. Windows RT was not quite the same as WP8. Tablets have gone from 'Windows for Pen Computing' through Silverlight and several others to now the new 'Universal'.

There may be millions of 'Windows' developers, but they could all develop in many different ways, many of which are obsolete.

Richard Plinston

Re: Yes it is

> but a static measure of weight while the plane is sitting at the gate should not be affected by friction in moving parts

You have not heard of 'static friction' then ?

https://en.wikipedia.org/wiki/Friction#Static_friction

Richard Plinston

Re: Using toys as tools...

> Is there not a load sensor on each wheel support? Add them all together, and lo and behold, you have the weight.

It may be accurate enough to know whether the plane is on the ground or in the air, but adding together the 'readings' will not give you the 'weight' but only some figure that will be less accurate than a transposed input.

This is because landing gear has dampers and seals that have high friction. Their design is optimized to reduce bounce and absorb changing load conditions, the antithesis of that required to measure weights.

Richard Plinston

Re: Using toys as tools...

> Why can't they use a proper tool like a Surface 3 with a proper typecover?

Because the Surface 3 won't run the iOS apps that the airline uses.

Also the typecover is inappropriate for the particular environment. There is no desk to place the device on and Surface with keyboard attached does not work well on a lap: the weight distribution is wrong; the keyboard plus back support is too long to fit comfortably without having the keys too close; only landscape mode is available which restricts layout; the screen is too small to used as it is too far away on the end of your lap.

Richard Plinston

Re: Yes it is

> Of course it's easily measurable. Strain gauges in the appropriate landing gear members will do the trick.

These do not work as well as you assert. The landing gear requires dampers to prevent bouncing when taxying or landing. These work by friction and thus affect any reading that may attempt to be taken.

Actual weighing cells, such as those placed under wheels, can only give an accurate result by being as friction free as possible.

Richard Plinston

Re: Using toys as tools...

> most GPS units and scanners running Windows CE?

Your information is out of date. That may well have been true a few years ago.

Also: just because it has a 'CE' sticker on the back does not mean it is running Windows CE.

https://en.wikipedia.org/wiki/CE_marking

Richard Plinston

Re: BYOD?

> Seems strange to need to bring your own iPad

Why do you think that the iPad belongs to the pilot rather than being issued by the airline?

Richard Plinston

> ...are they manually entering measurable data?

The gross weight of the aircraft is not 'measurable', or is not easily measurable within normal operational procedures. It is calculated from empty weight + fuel load + ( passengers * set estimate ) + freight (which may be measured for charging purposes) + other stuff.

Microsoft working hard to unify its code base, all the way down to the IoT

Richard Plinston

Re: The right tool for the right job...

> Also, there's a question about what's going to be done with all the data spewing from these devices.

"""Denny said the longer-term value is what he's trying to develop for: that the behaviour of 30,000 compressors becomes a data set that goes a long way to forecasting what's going on."""

It seems that what MS is aiming for is to have billions of IoT devices all sending their data through Azure where they can scrape it and sell it. The 'forecasting' is of no use to the butcher, it won't tell him the date and time his compressor will fail, he only needs to know, quickly, that it has failed when it does. The 'forecasting' would be done by collecting the failures and selling this data to the manufacturer (and meanwhile charging the butcher for this service).

The '10' in Windows IoT is the increase in cost (10 times) that a Windows IoT solution will have over a rational solution.

Microsoft shelves 'suicidal' Android-on-Windows plan

Richard Plinston

Re: Biggest problem with Windows phone continues to be mismanagement shocker?

> hoping that using just one name will confuse users enough for them to buy the products in spite of some of them still being 'not quite Windows'.

In much the same way Microsoft are advertising Continuum as "It’s a PC-like experience that’s powered by your phone". It is _not_ a 'PC-like' experience. It is a 'Windows RT' like experience (possibly without the touch) because it is still ARM and thus is still 'not quite Windows' and certainly not like Windows on a PC.

Richard Plinston

Re: Biggest problem with Windows phone continues to be mismanagement shocker?

> IBM broke its' promise to support Windows 32 bit applications

OS/2 Win3.11 could run Win32s for 32 bit Windows programs. However, Microsoft 'updated' the Win32s DLL to include a completely spurious memory access that was outside the 2Gb limit of the OS/2 virtual memory. This prevented later versions of Win32s to run.

Microsoft had also delayed the release of Windows 3.11. IBM had the rights to run any released version of 3.x inside OS/2 but anti-trust ruling had required that products not be announced more than 3 months prior to release. IBM had OS/2 with 3.11 announced and ready to ship but MS delayed releasing 3.11 for 3 months and IBM had to ship with 3.1.

I can well see that MS has dropped Android simply because Google could do exactly that to Microsoft - make it so Android apps won't run on Win10M.

Richard Plinston

Re: What 'app gap'?

> The so-called 'app gap' only exists in the imagination of teenage scribblers from the fourth estate and/or AAPL fanbois.

The 'app-gap' is likely to get worse now. Developers that were working cross platform probably dumped their WinPhone ports when they heard they could run their Android ports on Win10M without the effort. Now that has been dumped they may feel betrayed and not restart WinPhone porting to the new 'Universal' apps for 1.7% global market share.

Full time WinPhone developers may have moved to Android rather than to UAP in the expectation that this would keep their current market and expand it to the 80% market. Now they too may feel dumped on, as they did when WP7 killed WM6.x and when WP8 dumped WP7.

Richard Plinston

Re: Biggest problem with Windows phone continues to be mismanagement shocker?

> If I were running the show I'd ditch the toxic Windows and Microsoft brand names and start afresh in the hope of gaining some traction in the consumer market.

They tried that with Kin (actually a Danger product) and Zune. Look how well that worked.

Richard Plinston

Re: Catch 22 ?.

> leaving market share in the 4%-10% range ?.

Windows Phone is currently in the < 2% range (some say 1.7%) and falling.

Smartphone OS Shipments Q3 of 2015

Android . . . . 84.2%

iOS . . . . . . . 13.5%

Windows . . . . 1.7%

Blackberry . . . 0.2%

Tizen . . . . . . . 0.2%

Others . . . . . . 0.1%

Total . . . . . . 354.7M

Source: TomiAhonen Consulting Analysis 30 Oct 2015, based on manufacturer and industry data

This table may be freely shared

Richard Plinston

Re: Biggest problem with Windows phone continues to be mismanagement shocker?

> People are happy if they are told something runs on Windows.

People bought Windows RT devices and found that 'runs on Windows' did not work, and were not happy.

> Having a phone/tablet that is "the same as Windows" on their computer is a good thing in their eyes, because they don't have to learn another OS to use it. Users like familiar, even if it's not exactly optimal.

You have the chronology wrong. Windows Phone brought out a completely new UI (Metro) that was not 'the same as Windows' and did users did not buy it. Consultants advised Microsoft that the reason that it was failing was because of the unfamiliar user interface. MS in its wisdom came up with a scheme to make it 'the most familiar UI in the world' by forcing it down everyone's throats with Windows 8.

Users did not like it - because it was unfamiliar, and, mostly, rejected Windows Phones because it was like the 'hated' new, unfamiliar, Windows UI of 8.

> This is why Microsoft are finally getting the idea and getting rid of the umpteen sub-versions of "not quite Windows" that confused and annoyed users.

They are _not_ getting rid of the umpteen sub-versions, they are merely getting rid of the umpteen different names (RT, Phone, IoT, PC Desktop) and hoping that using just one name will confuse users enough for them to buy the products in spite of some of them still being 'not quite Windows'.

Decoding Microsoft: Cloud, Azure and dodging the PC death spiral

Richard Plinston

Re: What PC death spiral?

> Death of the PC

Manufacturers do not care about usage, they are only interested in sales and revenue. The 'Death of PCs' only refers to sales.

Microsoft Windows Mobile 10: Uphill battle with 'work in progress'

Richard Plinston

Re: Improvement good, significant sales increase - bleak

> They cost less than £100 each

One of the 'features' of the WP years at Nokia was that the phone division made a loss in every quarter despite the $billion sent by Microsoft each year. Microsoft seems to be continuing that in an effort to gain traction somewhere. They seem to be selling to their corporate customers by undercutting the prices to avoid having other OSes in the company.

But still the total market share falls, now less than 2%.

Richard Plinston

Re: Onedrive downgrade

> 1666 photos. At 10 photos a day

One day I took over a 1000 photos - not on a phone though.

Richard Plinston

Continuum

> ‘Continuum’ feature, allowing the phone to serve as a surrogate PC,

Well, not so much a 'PC' just a big big screen phone.

With ARM CPUs this won't run 'PC' software, just 'Universal' apps and will be similar to Windows RT in what can be run.

Useful as a big screen browser or perhaps as a thin client to server based software.

Feeble Phobos flaking as it falls to Mars

Richard Plinston

Re: Call me ignorant...

> Solid earth tides also exist.

You may also note that also there are effects of centrifugal forces. The earth bulges at the equator because of the rotation of the earth but this is not the whole story. The Moon does not rotate around the Earth, they both rotate around a common CoG which is not the centre of the Earth but around 5000km towards the moon (still inside the Earth). As the Earth spins on its axis the surface (or indeed everything) experiences a changing amount of centrifugal force from the Earth-Moon rotation. This is in addition to, or subtraction from, and is often in a different direction to, the gravitational pull of the moon. The total nett centrifugal force of the Earth-Moon rotation is equal to the total gravitational pull of the moon but the distribution is quite different.

Is the world ready for a bare-metal OS/2 rebirth?

Richard Plinston

Re: But ...

> A "successor of VMS" would have to support multiuser operation and NT fell a *very* long way short of being a viable multiuser OS.

NT was originally designed by Cutler to be multiuser but Gates insisted that be removed. He wanted to sell a full Windows to each user not one copy per group.

Citrix made it multiuser again, and then MS used that to make TSE. The main problem was making the applications (esp. Office) behave in a multiuser environment.

Richard Plinston

Re: But ...

> The OS that MS morphed into NT3.1 (there was no NT1.x or NT 2.x)

The first NT was numbered 3.1 because it had the Win 3.1 GUI.

NT was a completely new development by Cutler based on the planned successor of VMS. It had no relationship to OS/2 except NT ran an OS/2 1.3 personality. Microsoft paid DEC $100million for stealing their ideas.

You appear to have the mistaken idea that OS/2 1.x and 2.x led to NT 3.x. There is no basis for that idea.

Richard Plinston

> CentOS is still a f@#$ing pig that spends minutes doing god knows what.

Maybe, but the RHEL and CentOS that I look after (including 3 around here) are maybe rebooted once or twice a year. One went for over 1000 days.

We're not killing Chrome OS ... not until 2020, anyway – says Google

Richard Plinston

Re: Nobody likes change

@FailureCase

> If your complaint is merely that the WSJ were using the term "merge" or "fold" in a source control sense,

WSJ claimed that ChromeOS would be 'folded' into Android and would be dropped. Google has stated that ChromeOS will continue. My 'complaint' was that WSU did _not_ report it as a merging of the teams and source code (which appears to actually be the case).

> then why would you describe the article as damaging (and suggest possibly deliberately so)

If you look a little more carefully at my message you might see that 'damage' was quoted. I was replying to someone else who claimed that the article was damaging.

Richard Plinston

Re: Nobody likes change

> Oh come on, it's almost certain Google are stopping the development of Chrome OS. Why on earth, if they are merging the two, do they need to continue with an independent Chrome OS?

ChromeOS is based on the Linux kernel. Android is based on the Linux kernel. It is _very_ likely that they are merging the two kernel development efforts into one effort (one team). This does not stop development of ChromeOS, nor of Chrome browser nor of Android.

This may also give 'convergence' where ChromeOS can natively run Android apps and Android can benefit from Chrome browser enhancements to run Chrome online apps.

None of this indicates a need to dump ChromeOS, but it may get more Androidy.

Richard Plinston

Re: Nobody likes change

> this report also smacks of damage control

The 'damage' seems to have been caused by misreporting* by the WSJ. What seems to be 'folded' is not 'ChromeOS into Android' but the ChromeOS development team into the Android development team.

* accidental or deliberate.

Windows 10 growth stalls during October

Richard Plinston

Re: A pedant writes...

> Stalling an aircraft is only dangerous when you're very close to the ground...

Not necessarily true. Or, more to the point, higher altitude does not make stalling 'safe'.

For example some high-tailed aircraft (such as the Gloster Javelin), when stalled, can enter a super-stall where the tailplane is blanketed and there is no ability to alter pitch leading to an inability to recover regardless of height.

Richard Plinston

Re: Who here has these sites blocked?

> might have these websites blocked

It is not just whether users have the sites blocked, it is also whether particular groups of users visit particular self-selected sites.

It may be that the majority of the sites that include statistics gathering javascript are of no, or little, interest to, say, Linux users. Conversely, sites that are of interest to that group may not have this implanted code. I see no evidence that this site has any visits recorded in the statistics.

Richard Plinston

Re: A pedant writes...

> That's not entirely true; what happens is that the wings stop producing life and start producing drag. This doesn't mean you go straight down - certain aircraft[1] flutter down quite gently in a full stall. Some bite...

s/life/lift/

No, the wings do not _stop_ producing lift. They, most often, continue producing lift (as indicated by 'flutter') but produce less than required for sustained flight.

No, they do not _start_ producing drag. They produce drag all the time they move through the air. They produce a much increased lift/drag ratio.

Richard Plinston

Re: Interested to see how this turns out...

> I beleive the stats are based on web usage,

They are based on web usage of visiting particular self-selected sites which include javascript in some or all of their pages. It also requires that the client allows javascript to run as it is the client browser that sends the information directly to the collector.

If sites that are of interest to Windows users but not to Linux users predominate the sites that include the javascript then Linux will show low statistics. If javascript blockers are more often used by Linux users then Linux will have low statistics.

I doubt that any of my Linux machines are ever recorded at all.

Next year's Windows 10 auto-upgrade is MSFT's worst idea since Vista

Richard Plinston

Re: Linux

> the learning curve is vertical

Yes, some people do find learning to be a brick wall. If you find that to be the case than do stick with what you can cope with.

> their community is mostly unhelpful prick's who look down on everyone

I can quite see why that has been your experience.

How Microsoft will cram Windows 10 even harder down your PC's throat early next year

Richard Plinston

Re: Ah the MS hatred alive and well

> a pretty successful OS release

It may well be 'pretty successful' from Microsoft's point of view - they managed to force it down 115million throats (or at least download it that many times), but it seems that it is not 'successful' from many users point of view.

Big mistake, Google. Big mistake: Chrome OS to be 'folded into Android'

Richard Plinston

Re: So what's your point?

> re-write Android to be based on Chrome.

They are both based on the Linux kernel. They may add GNU to Android and/or Android UI and ART to ChromeOS. This would give Android apps on ChromeBooks and allow phones to become ChromeBooks/desktops when connected to a monitor or TV like Ubuntu does.

Richard Plinston

Re: So what's your point?

> only specialised software that did odd stuff didn't work,

Windows 95 completely dropped support for all Windows 2 programs which ran fine under all 3.x. I had one Windows 2 program that didn't have a suitable replacement so ran a copy of Win3.11 in a virtual machine.