Re: has no access to the Youtube ad API

No, it's not a device sniffing tactic, it's that there is more than one API for YouTube, and not all APIs provide the same quality results. Here's my guess at what's going on, based on having written a small mobile YT player before.

Basically, you can fetch videos as Flash movies, or raw H.264 streams. The H.264 videos are playable on mobile devices that don't have Flash (these days, that's all of them). However, when requesting the raw videos, only clients that possess a special API key will be sent the links to high-definition video streams in response to their requests. Without the key, you are limited to 360p.

There was an older API that allowed better quality videos for anyone, but Google removed it. Now, to get the HD videos, you need to use a special API that only Google and Apple use. That API requires an access key.

Microsoft requested a similar access key from Google, and Google refused, saying that no private API was required, and they should use the website. But the website compares poorly to a dedicated app. Apple's iOS player app, for instance, has access the HD videos, so it's not even a Google exclusive.

MS then used a well-known loophole in the Flash-player API in their dedicated app to find the HD versions of the videos. (This is the same mechanism that all those "YouTube downloader" services use).

Google sent them a cease-and-desist, because using this source without Flash means that any ads before the content, and the ads normally overlaid on the content by the Flash plater, aren't played; only the requested content is shown.

There's was no sniffing involved, just Google not providing an access key for the API that they and Apple already use to fetch videos. MS wanted to use this API, but weren't allowed to by Google.

Google's behaviour here is irrational. As a company that makes all of its revenues from advertising, their interests would be best served by providing the API key to Microsoft that would allow Windows Phone devices to show YouTube videos with ads. MS are even paying for the development cost of this client, so there's no cost to Google. Google would gain a small increase in ad revenue, and at a cost of... I don't know, pride? Well, I don't read their financials in great detail, but I'm pretty sure that "pride" has a zero monetary value. (Protecting Android market share is a weak argument, when you compare the shares in question: anyone who doesn't have an Android phone by now just doesn't like the platform, and will never buy one)

So in the current situation, users still get to see the videos they want to, and Google loses advertising revenue. A cease-and-desist against MS won't stop other apps using this loophole, and closing the loophole will cost Google a lot of re-engineering, and possibly cut off a large number of desktop viewers in the process.

As an analogy, imagine if ITV (or CBS for non-UK readers) decided that people who use LG TV sets wouldn't be allowed to see their broadcasts anymore... just because, and that there would be nothing that LG could offer that would change their mind. That's the closest analogy to the situation here. It's an advertising platform - restricting its audience is counter-productive.

Re: Hmmm

And how long have you been a Google employee?

Because if you're not working for them, you really are a barrel of fun. Depriving other people of something because they don't follow your personal beliefs.

Sounds like the behaviour of a zealot to me. And an idiot for doing someone else's job for them for nothing in return.

Re: bloat...

Benchmarks won't tell you anything meaningful.

Benchmarks will only show you peak performance on batch tasks, they can't accurately model user interaction. A benchmark can tell you what your IO is like, or what your peak CPU performance is, but it can't tell you what your average event latency is, or how quickly it can respond to user input.

Pick the device up, use it. See how quickly the user interface moves in response to your inputs. See how often the UI drops a frame or two. Leave it be for about 10 or 20 seconds, then see how quickly it becomes responsive again. Load a webpage, scroll it up and down. On equal hardware (WP7 is single core 1Ghz Snapdragon) they should be similar in terms of responsiveness. They're not. (I'm no fan of WP7, btw, but it gets a lot of unwarranted abuse, normally from people who like Android)

" What I am driving at is that at least 128, 256 mb used to be quite common on cheap Android tablets, phones and even small netbooks. You still can get 256 mb tablet on newegg. "

And these low-memory devices are what got Android its reputation for having nasty user experience on cheap hardware. Once you installed a few apps with background services, the OS would spend so much time swapping and/or doing GC cleanup to free some RAM that it couldn't keep up with even basic user interactions. Games, which often pre-cache lots of graphics in RAM, suffer especially badly on this kind of hardware.

Actually, true multitasking like Android's (or before it, Symbian's) is actually bad for gaming, as other tasks can steal resources from the game at inopportune moments. In this respect, the dumb tasking of iOS or WP is better... but, frankly, only in that very specific use-case.

Re: bloat...

You're telling me that native code runs slower than an interpreted bytecode...?

Right.

I'm sure that if the situation was the other way round, and it was MS or Apple bunging their whole user libraries through a VM, you'd be right on their case for a making a stupid design decision. It remains a stupid design decision, but it's one that can be explained by a need to get to market quickly.

It's not just the recompilation penalty, though, it's the garbage collection. GC is the worst thing to mix into a small OS. The non-deterministic nature of GC runtimes makes it extremely hard to provide low event latency. This is why Android stutters, even on the best hardware available on the market.

The VM was built by smart people, and has been optimised repeatedly; the APIs were lumped in quickly, and because they're public, they've had to stick to them

To the guy who said Android was originally specced to run in 64 Mb, have a look at what Windows Vista's minimum requirements were. That's enough to run the OS. Apps? well, nobody said anything about apps...

bloat...

There's evidence that came up in the Oracle trial that Java is that bloat.

I find it very hard to believe that a competent engineering team, given a blank canvas and targetting a resource-constrained platform, would ever willingly implement their entire user-space application framework in an interpreted bytecode runtime. The only rational reason to have done this was to get a quick time to market.

There are a lot of good reasons to have applications written in bytecode (multi-architecture portability not least of them), but not the code libraries that these apps use.

Windows Phone 7 took the approach that I believe Android's developers wanted to: optimised, native-code libraries serving byte-coded apps. Comparing the user performance of a WP7 device with that of similarly-specified device running Android shows that it was the right thing to do from a technical standpoint.

As for the problems with manufacturer apps, a lot of the blame is Google's. The terms of the "powered by Google" licence agreement say that an OEM cannot remove key system applications. So even if the OEM apps are better, so you end up with two email clients, two music stores, two camera apps, and so on. The penalty for breaking this rule is to be denied access to all Google services. Most important of these are the Play store and Google Calendar, which are the most compelling parts of Android as a platform.

Perhaps if they worked a little closer with the key OEMs, a deal could be worked out, but as Google earn nothing from sales of Android, and the OEMs are putting up with the current situation, there's no benefit to Google in doing this.

Re: What is a standard and what is a capability?

The mobile telephony Generations were primarily about signalling:

1890s: 0G - Radiotelephone, ad-hoc call setup

1970s: 1G - Analogue voice, automated circuit-switching

1990s: 2G - Digital voice,circuit-switching

1990s: 2.5G - (e.g., GPRS, EDGE) transition to 3G by adding packet data services

2000s: 3G - Digital voice (circuit switched), packet-switched data

2013: 3G-LTE - ("4G" if you work in marketing), transitioning from 3G to 4G systems

2010s: 4G - Fully packet-switched signalling for voice and data.

Basically, a term that used to have a reasonably well-defined meaning has become a marketing buzzword with little releation to reality (cf "Megahertz", "processor cores"), so Samsung have as much right to call their experimental air-interface "5G" as AT&T had to call their 3G-LTE service "4G".

Re: valid assumption?

The software is very good, but this model may prevent people upgrading - especially in the freelance sector.

As for whether there's competition? Well, yes and no..

Photoshop is the only one that has credible competition, and you could make a good case that for 90% of what the people using pirated PS copies are doing, they could do just as well with Gimp, which is free.

Same kind of applies to DreamWeaver. As a developer doing occasional web stuff, I never saw the huge advantage of DreamWeaver over, say, SublimeText (plug: you owe it to yourself to try this editor!) and FireBug, but it has its devotees.

On the other hand, from personal experience, there is nothing as good as Illustrator for design and illustration work. Not saying it's perfect, mind you, but just that the competition, both free and paid, is a whole lot less perfect.

(I don't use InDesign, but I know people who swear it's the best part of Creative Suite.)

Re: Nah

Apple's EU sales are run out of Cork in Ireland, so registration would be with the Irish Data Protection Commissioner. Link: http://www.dataprotection.ie/ViewDoc.asp?fn=%2Fdocuments%2Frights%2FRightsHome%2Ehtm&CatID=2&m=r

Sounds like a misapplied process. The transaction value in this story is far too low for this kind of verification to be required, unless the card company thought it was a "suspicious" transaction.

Actually, I just tried to buy a laptop (not from Apple) this morning and had the same thing happen to me... phone call to VISA, and was told there was no reason except that it was unusual activity -- surely if I was buying laptops regularly it would be more unusual...

Re: Asha is not that bad

Yep, previous Ashas are pretty good, but this one takes it to a much higher level.

Looking at the in-use videos, this phone is very smooth and slick - the transitions and scolling are full frame rate, with no skips or judders: only one feature showed a delayed response (pulling down the SIM switcher drawer widget), and the whole thing is really well thought out. I'd like to see this same UI on a full-feature phone, actually, but for a sub-$100 phone, the perceived quality in the UI is astonishing, and well ahead of anything at this price.

The reason? The underlying software isn't Series40 anymore: it's Smarterphone - the Norwegian Linux-based OS that Nokia bought last year. Still the same Java ME API, though for third party apps, but Nokia mention that there will be new, device-exclusive APIs coming for these.

If they'd just add background agents for certain tasks (downloading in my case), I could even get over the non-multitasking aspect of the software. And even Java.

While you're at it...

How about these, similar, offensive generalisations...

People who eat liquorice for pleasure are paedophiles.

Nobody who drives an Alfa Romeo should be trusted with sharp objects.

People who prefer Star Wars to Star Trek torture small animals in their spare time.

Pepsi drinkers are intellectually sub-normal.

Anyone who likes Strictly Come Dancing is a sociopathic pyromaniac.

Re: Nokia cannot hide from their failure

" ..REFUSED to Sell the N9 Meego in the UK "

Let's get the blame to the right place: you got no cheap N9 because no UK operator wanted to offer N9 to their customers. The markets where it was sold were markets where a local operator decided to range it. No conspiracy, no evil plot, just simple commerce. Nokia could not sell a phone over the heads of the local operators - they just don't have that infrastructure (in the UK). The likes of Carphone Warehouse are just agents for the operators - their stock comes from the operators, and is locked and subsidised by them too. There's no free market; you don't have free choice.

You could *always* buy an N9, unlocked, in the UK, but then you'd have to pay its full retail price. Maybe you really wanted the discount from the operator, but giving that to you was their choice, not Nokia's. So bitch to them. They certainly deserve it.

Personally I want all these subsidies to go - all they do is make my bills more expensive, and restrict my choice of handset. There's also no reason for it anymore: does Sky sell you a TV when you take out a subscription? No, because you've already got one.

Re: What is vs what should be

Honestly, some people just WANT to see conspiracies. Competing against "Android and Samsung" is a perfect summary of Nokia's challenges.

Nokia is a full-range manufacturer, as is Samsung. Both companies make everything from simple voice-and-text phones all the way up to high-end smartphones.

In smartphones, Nokia competes with Android as a platform. Below that, their only significant competitor is Samsung's line of non-Android featurephones (against Nokia's Asha series) and Samsung's plain voice-and-text phones (against Nokias 1xx and 2xx series).

You know that both Samsung and Nokia make more money on every simple phone they sell than Sony Ericsson have with any of their Android efforts to date... dumbphones and featurephones might not be in the tech-nerd's line of sight, but they're popular devices, that people want to use, and that produce actual profit for their makers. If you want a future, you need a profit.

Re: LTE

The AT&T version of the 920 supports AT&T's bands (4 and 17). It's the Global, non-carrier-locked, model that uses 1,3,7,8 and 20.

(Source: http://www.nokia.com/us-en/phones/phone/lumia920/specifications/ -- note the "us-en" in that URL)

Re: Bloody hell..

" The cost of the MS WP licences. But more importantly, the 'cost' of being tied to MS dictating requirements. They can only build what MS allows them to in terms of actual SoC used, UI, layout of the screen, and many other."

WP licences are cheaper than Symbian development. Yes SoC is limited, but as a developer I think this is a strength of the platform, not a weakness. Ditto UI layout. I think allowing manufacturers to skin Android is a major flaw of the platform.

"So your claim is that they are not developing these beyond what they were two years ago and merely ported them ?"

No, I'm saying that whether they're developed for Windows Phone or Symbian, the cost of ongoing development is going to be the same, because the new development is a replacement for, rather than an addition to, what was being done before. Don't take my word for it: look at the company's financial reports. R&D costs have fallen far more than licensing costs increased. A net saving.

"I doubt that Android was the only choice. Symbian collapsed not because of the costs but because it was announced as being dead when sales were actually increasing. "

This is a myth. The decline in sales started in late 2010. Advance orders for Symbian devices collapsed at this point, as the N8's introduction turned sour (Bluntly, the hardware was excellent, but the software was not finished, and return rates went through the roof). Sales drops in 2011 were down to networks not renewing orders for Symbian devices back in late 2010. Hardware industries have long lead times: causes go back long before the visible effects. The collapse of Symbian orders, combined with the inability to finish the MeeGo software prompted the famous "Burning Platform" memo; it was not the other way round.

Memo or not, Symbian was already on it's last hurrah, and was never going to continue past Symbian^4, scheduled for 2011 but abandoned in 2010 (but many of its features, running on S^3, came into the Belle release in late 2011). The engineering efforts to get Symbian3 onto the ARM11 SoCs used in N8 and subsequent devices was grossly underestimated, and were the reason why the N8 itself, scheduled to release in early 2010 barely scraped onto the market in November 2010. Nokia's version of Symbian had poor separation of functional layers and was just too difficult to bring into today's more powerful hardware. (open-sourcing the code also sucked time and energy out of the company: perfectly good, working code was re-written to allow the entire OS to be Open-Sourced, and in the end nobody used it.)

"I still have and use an N800. It is not a phone. "

I know that, but telephony is an application; it's not intrinsic to the OS. As an application platform, which was more mature? N800 or the Android release of the time? Actually, if you insist on pedantry, compare Maemo5 on the N900 with Android Cupcake (1.6), both from late 2009.

" as [N950] was killed off so as not to compete with MS products"

No it was not. N9 was put on sale because it was too late to stop, and there were customer orders in place that needed to be fulfilled, but there was no future in the OS. MeeGo was an absolute disaster. It failed to achieve any of its goals, and Intel had failed to deliver any of the hardware that was to propel its progress in tablets - even now, where are those low-power Atom tablets? I love the N9, I think it's a beatiful combination of software and hardware, but it was two years late. The N950 is actually the hardware that N9 was supposed to launch with, but wasn't shipped because there was no software ready. You need to stop seeing this failure as a Microsoft conspiracy, it was entirely of Nokia's own making. It's unpalatable, but true.

Nokia's problems were all in full force years before they signed up as a windows licensee. I waited for N8. I waited for N9. Neither of them were on time, neither as good as they should have been (N8 disgracefully so). My main phone is still an N8. With the last software release on it, it's a pretty good phone, but, again, that was nearly two years later. Had N8 launched on schedule, with Symbian Belle on it (hell, even with Anna), there would have been no problem. It didn't.

Nokia went with Microsoft for solid business reasons. Not liking this outcome doesn't make the whole deal a conspiracy. Google was the other option, and they were approached (I believe WebOS was also considered) and it simply did not make business sense.

Here's a major reason: Google would not allow Nokia to preserve their mapping and navigation apps if they became an Android licensee. They would have been asked to ditch their entire investment in buying Navteq, and all of the work they'd done to make Symbian's maps and navigation class-leading, for no revenue return. Going with a "non-Google Android" would have let them keep Maps, but wasn't a viable choice - the only truly good things about Android are Google's online services: they'd have been locked out of Calendar, the app store, and all of the things that Android does very well, and been castigated by the press for it.

Those services are another reason why Android was a no-go. The post above says that the Android phonemakers have been reduced to box-makers on narrow margins, with no recurring revenue, which is a pretty good assessment of the situation. By choosing Microsoft, Nokia were able to bring their existing services (music and navigation) to Windows Phone that give them recurring revenue streams; those services are made available on all WP handsets, not just Nokia's. That wasn't an option on Android.

Finally, you have to consider the loss of first-mover advantage: Windows Phone 7 was less than six months on the market when Nokia signed up; Android had had nearly two years, and Samsung's position as top dog was becoming very clear.

"And Android is not trying to do that either. Google makes it available and it is up to manufacturers to do as they wish."

That is not how it works. Google churns it in secret, with no input from its customers, spits it out, and leaves it there: then it's up to the non-favoured manufacturers to try to get it working. Linux doesn't work like this; Qt doesn't work like this; WebKit doesn't work like this. Almost every open-source project out there lets you see the sources at all times, so that you know what's coming down the track. Android springs it on you, becaue it's the minimum that Google have to do to be acceptable to the "Open Source" evangelists.

We obviously disagree about this, and there's not much I can say that would change your mind, but you cannot blame Nokia's current woes on Microsoft, or on not choosing Android. They, and they alone, are the authors of their own past misfortunes. That said, their future looks better than some who chose Google's "free OS".

Re: Bloody hell..

" Nokia has _added_ software costs and their business is destroyed. "

Untrue on both counts. What costs have Nokia added? The special apps they do for Windows Phone are ports of the products they provided on Symbian (and Maemo), so no increase in cost. They have, however, relieved themselves of the platform costs of maintaining Symbian (and qualifying the hundreds of carrier and device specific releases of Symbian), of developing Maemo, of juggling the perfectly good Maemo to fit with MeeGo. They also have dropped the majority of their Ovi web/cloud product. Overall, engineering costs have decreased dramatically, even with a large increase in Series40 development. You can see that in figures if you look up their financials.

" The only conclusion is that software licence cost and business destruction, or success, are completely unrelated. "

Yes. That was my point, but I wouldn't use the word "completely". Some companies have failed miserably despite having their costs "reduced" by choosing Android, others have prospered despite doing everything in-house. Some are in between, and as you brought them up, Nokia is one of those.

You appear to be implying that Nokia's business was "destroyed" by choosing Windows Phone - which is a bit like claiming that firefighters cause the roofs of burning buildings to collapse. Causes are often much further ahead of effects than that.

A very small bit of research into that company will show that by the time WP was chosen as their primary platform, the major damage had already been done. Nokia bought marketshare from 2009 to 2011 by dumping loss-making Symbian devices into the market, because it couldn't get its software efforts completed on time. The last revision of Symbian (Belle FP2) is actually pretty competent, but it came out in early 2012: two years later than it should have. Then they wasted a whole year on merging their perfectly serviceable Maemo mobile platform with Intel's "hey-lets-work-on-everything" Moblin to make MeeGo - an effort that went nowhere: the Nokia N9, the only phone to "run on MeeGo", doesn't even run MeeGo; it's the pre-merger version of Maemo, hastily ported to the hardware. A whole year. Gone. That kind of mismanagement is why Nokia's business has collapsed, not any choice of Windows Phone over Android.

If you wanted Nokia to adopt Android, the time would have been 2009, not 2011. But in 2009, they had a realisable roadmap for an open-sourced phone OS: Maemo, and it was more mature than Android was at that time. Compare a Nokia N800 with Android 1.6, and tell me which is better? It was only Nokia's inept management after this time that makes Android look like a good choice in retrospect.

I really have no agenda: I own shares in none of these companies, and I use all of their products daily (my phone is an old Nokia, running Symbian, my two laptops are MacOS and Windows8, my desktop runs Ubuntu, OS X or Windows7 depending on what I need, and I use Google's web services). I like to see diversity in the market, because it helps new and better products to appear. Because what's out there can always be improved.

Maybe you want to see a Google hegemony, but count me out: we've only just managed to get free of one grasping monopolistic corporation that stifled progress for twenty years - why are you so eager to jump into the arms of another? Open source? yeah, right, and I've got some beads I'll swap you for this here island... Google don't give a rat's ass about Open Source, they care only about Google, and if more people realised this, we'd be better off in the long run.

Re: And the spin-offs ..

" I have a terminal session stuck in a "Twisty maze of passages all alike" here on OSX, hints anyone ? "

Sounds like you're trying to debug a medium-sized Java application that has a delegate chain longer and more convoluted than the European Commission.

Control-C and re-write it in Intercal - it'll be quicker in the long run.

Re: Bloody hell..

Android phones have always had better CPUs and more RAM than phones running any other OS. Yes, they're under-specced, but because Android is the heaviest of all current mobile platforms, not that the phones are weak in themselves.

If you're a carmaker, and you can't make a fast model without putting a 500 hp V8 under the hood, you look at cutting weight, not developing an even bigger engine, because there's only so much extra power you can throw at the problem before it becomes uneconomical. Weight reduction is a much better strategy, but Android's distribution model has deprived phonemakers of the leverage to get that weight cut out of the OS.

Maybe a real open-source Linux OS will arrive and give a better, lighter weight option, but I doubt it as long as Linux Foundation are running the mobile OS projects...

Re: No Chip and Pin in America?

No conspiracy theory needed: the answer is more mundane. Unlike elsewhere, US banks pass the full cost of the terminal equipment onto the merchant. It's a once-off fee, and the terminals then belong to the merchants. As the merchants would have to pay for new Chip-n-PIN terminals, they see little benefit in doing so, especially as the benefits aren't going to be obvious to them in terms of lower transaction costs. Similarly, the acquiring companies can't just cut off the magstrip services either, because as long as the merchants keep paying their service fees and refuse to pay for new equipment, forcing a change would result in loss of revenue.

In Europe, on the other hand, the terminal is the bank's property, and is rented to the merchant as part of their monthly service fee. So, if the bank wants to upgrade their security, they tell the merchant that they'll be sending a new terminal, and that's that.

(The US situation is a pain in the hole for security, as it's the only reason why the rest of the world is stuck with magnetic strip readers -- the easiest method of stealing card numbers)

Re: NFC

Put your mass-transit ticket card, or your building's door-access card, in front of it - or another NFC-enabled credit or debit card, for that matter.. The remote NFC snooper that everyone's so afraid of (despite it never having been demonstrated to work in a real-life setting) will trigger both cards and be unable to read from either.

Just because an attack is "high-tech" doesn't mean it's worthwhile for an attacker to pursue. There are easier and cheaper ways of acquiring card numbers than trying to radio them out of peoples' wallets. A few quid in the hand of a dishonest waiter in a busy tourist bar is much better return on investment.

Re: This case proves...

There's one thing right now that I'd happily have fewer of: your moronic comments.

Re: The council text you when they come to empty your bins?

Not the OP, but my bin collection service does the same thing, and it's very useful.

Your situation is most likely different, so the usefulness of this isn't obvious, but here's how it works:

1. There are different bins, for different things (rubbish, glass, recyclable materials, electrical or hazardous, etc.)

2. Every week, the collection service takes a subset of these bins.

3. The subset is different for each week in a four, five or six week cycle, depending on the operator.

You can either:

a. memorise the pattern and hope they don't change it.

b. every evening, find and consult the chart that the collection service has sent you

c. wait for the text that they send the night before, and then put out the necessary bins.

Not difficult.

Re: NFC

Many US merchants don't ask for CVC. Some will accept payment from the 16-digit PAN and expiry alone. These are usually fly-by-night porn outfits, and they pay for the "privilege" with high merchant fees, but the business is commercially worthwhile to them, it's the banks that pay for card fraud, not the card clearing companies, so who cares...

In any case, there's nothing new here. I have a device on my phone which uses electromagnetic radiation to acquire a copy of the details printed on any credit card. It's called a camera.

Re: Landfill?

PAYG phones are also subsidised. Neither of the models you cite are "sub-£150" if you try to buy them SIM-free. Experia is £170, Huawei £160 if you buy them not locked to a network. Close to the Nokia, but still more expensive.

I think the landfill comment is a fair shorthand for Android's ever-rising minimum requirements and the platform's poor update management. It doesn't help that many of these performance-sapping updates are applied by stealth. When new, a cheap Android is fine, especially if manufacturer has trimmed some of the heavier features from the stock software suite, but give it a year, and it will have been rendered nearly unusable by a procession of ever more demanding system updates that seem to enable the kitchen sink by default.

After a year, the Windows Phone will have kept the same performance it had when new. Battery life is the problem for Windows Phone, though (and Android, but Android phones seem to come with bigger batteries). Lumia 800's battery life was as bad as any Android. This series seems to be better, but still not good enough.

Re: first step, transparency

Other way around. To take the most extreme example: Apple in the US advertise their phones at a subsidised price, and then bully the network operators into paying them much more than that for them.

Sure, the network operators are crooks, but that money they're paying to Apple is being taken away from providing better infrastructure. The infrastructure still needs to be put in place, so the money for that comes from... customers. As usual.

In Europe, handset pricing is a little more transparent, but it varies from market to market. Finns, for example, have always been on the model that T-Mobile are now introducing (you pay for service, but the operator will lend you money for a handset at zero percent), but the UK is closer to the US model than any other country, with very low upfront prices, at the cost of higher monthly tariffs.

Re: 30th anniversary of every man and his dog releasing a Spectrum-basher

"What went wrong?"

I think you've got to look back to the 1960s to find the most probably cause. Asset Stripping. Profitable, viable manufacturing companies were bought up and shut down simply because their premises were in an area where property values were high. By the time the 1980s came around, Britain no longer had a "middle-strand" of engineering and manufacturing companies: it was a choice of two-man businesses, or nationwide behemoths crippled by inflexible union agreements.

So, into this environment, you've got a small, IP-led startup. They know how to build computers, but can't make them themselves. Had this happened in Germany (the most similar economy to the UK, historically), the answer would be simple: go to your small, local radio/TV factory, and ask them to knock out 10,000 of them for you. In the UK, those businesses had been destroyed by the likes of "Lord" Hanson, or forcibly merged into quasi-nationalised conglomerates by equally idelogical Labour governments; and investment money was moving into financial serivices and media, as these were seen to be the "future".

In hindsight, the whole process was like losing weight by cutting off your left arm.

The crash of 1983 put the last nail in the coffin of these small manufacturers. Many that were left in the hole by the now-bankrupt micro makers were bought by asset strippers, split up, sold off and shut down.

Amstrad's success in the market was down to having the computers made in Taiwan, at lower costs than any UK maker could manage, and bundling them with a monitor. The offshoring was a revolutionary idea back then, but it was the only way that a UK-based company could make things: the domestic capacity had gone. I had one - it was my first computer.

(Disclaimer: this is just my theory - it's not researched, and I'm not an economist)

Re: Poor results?

Sorry..

Re: Poor results?

I don't know if you follow football (I don't particularly), but here's an analogy...

Queens Park Rangers are currently second from bottom of the table, and will be relegated. If they had finished 17th, not 19th, it would be a good result for them. A great result, in fact.

Had Mancester United (now league champions) achieved that exact same hypothetical 17th position, it would be an appalling result, and heads would roll.

But it's still the same number of points scored, and the same position in the table. Expectations are the key.

Apple have taken so much money over the last five years that yes, $10bn IS a bad result. Their stock was priced in expectation of their already phenomenal profits rising in future; a fall, even a slight fall, gives lie to this valuation. In truth, I think there was a good old fashioned bubble in Apple stock last year, one that is only now correcting itself....

Re: Could anybody explain...

Human hearing has approx 130 dB of dynamic range. Microphones with good audio quality at this kind of range are expensive, and too large to mount in a mobile phone.

A compromise (used by just about everyone else) is to use a less sensitive microphone, with 100 dB of range, but this means that loud sounds will cause it to distort. Alternatively, it can be mechanically damped so that it will not distort at high volumes, but then you lose the ability to detect softer sounds because the microphone diaphragm is now too resistant to movement.

Nokia's idea was to use two mics, one damped one for high level sounds, and another undamped one for lower level, and use some clever signal processing to mix these dynamically. A different version of the technology uses one diaphragm (with, I suspect, a non-linear amplitude response), and better signal processing to re-form the sound signal afterwards.

It is clever, and it does work. That's why they got the patent. But this isn't about patents... HTC didn't use the method to make their own part; it seems that their supplier, STElectronics, supplied HTC with the *exact* part that they were making on Nokia's behalf. That's not patent infringement, it's breach of contract and possibly theft (STE were making these for Nokia, and Nokia alone, so the components were not STE's to give to HTC).

Re: Copyright?

Oh come on...

Tools are just that: tools. The fact that a device can be used to record a concert doesn't mean that it's infringing copyright. The person using the tool bears the responsibility.

There are many loud social occasions where you would want non-distorted audio capture (110 dBA is easy to reach if someone is shouting and cheering very near to your mic), few would involve copyright infringement. Sporting events (amateur or otherwise), parties, motorsport, funfairs are a few that come to mind straight away.

The best reading of the situation is that STE have made a huge mistake; the worst is that HTC have stolen another company's work without compensating them. The part is Nokia's design, even the packaging is the same. STE, as manufacturer, had no right to sell the parts to a third party, any more than Foxconn could start selling iPads to a white-label rebrander. The contract they made with Nokia would have made this very clear.

It's bad news, because the HTC One is a nice phone - HTC should not have taken this risk - and this is why I suspect it's STE's failure, not HTC's. Nokia are within their rights to pursue HTC for compensation, or have sales blocked until they change the part (which they are going to do).

Re: As an app developer...

... if those are your only choices, then yes, ObjC. But it is rather like being asked to chose a favourite venereal disease.

A long time ago, when I did Objective-C, I found that ObjC++ was the only way to make it tolerable. [[but [even [[then [that [[stupid [method call] syntax ]]]] got on my ]] tits]].

Re: Better then an N8?

As another N8-owner, I know where you're coming from... Only Nokia and Apple make phones that are as well made, and neither of them can match the camera quality of the N8. I really do not like iOS or Apple's always-on over-saturated colour processing, so that leaves Nokia. I just wish they'd get their camera stuff fully ported to WP...

None of the Androids have a good camera by N8 standards. The HTC One is closest, but it's not as good as N8. If you can take the drop in image quality, it's a nicely made phone, with enough features to keep you up to date for a couple of years. But, being Android, you will have to completely wipe and restore it at some point to recover lost performance. (It's a quick job, though). The Galaxy S 3 and 4 are mediocre as cameras, and the 3 flimsily built. Not a good combination for gigs.

Right now, the (hard to get) 808 is the only thing that is better as a camera. It would also be faster, as it has a CPU that's nearly twice the speed of the N8's, and a better GPU. I considered it, but as you say, it's a lot to pay for a dead-end device.

The 920 (and upcoming 928, most likely) is good in the dark, but it can't beat the N8 in well-lit scenes, although the optical stabilisation for video is exceptional (and the audio quality is also top-notch, if not quite as good as the 808's stunning audio capture).

However, there is a new camera-centric WindowsPhone device rumoured to be coming this summer from Nokia, but no details about it have emerged yet. I'm waiting to see what that one is like: the 920 was close, but a little to big for me, and the 720 produces some stunning results for its price-range. In any case, the N8 will continue for another year...

swapping batteries

" replacing a battery is not a weekly operation - "

No, it would be a twice-daily operation for field workers. If current models allowed removable batteries, it still would be.

Before Jobs made them sealed units, Mac laptops used to have a very useful "warm-swap" feature: put the computer to sleep, and you could pop the battery out for up to 10-15 seconds without affecting the RAM. Makes it very easy to swap batteries and continue working.

That, and their low power consumption meant that packing one or two spare batteries could get you a full day's use of a laptop far away from any power source (very useful for long-haul business travel with long stopovers).

Design isn't just about what a product looks like. Apple used to make well-designed products; these days their stuff only *looks* well designed...

Re: bet the company?

Ampex and Sony collaborated on professional VTR equipment, but it was Sony that launched the first successful *consumer* recorder.

It's Sony's consumer electronics business that provides the closest parallel to Apple's current business model.

Re: It's going to get a lot worse for Apple

It's starting already. T-Mobile US is finally ranging the iPhone, but at the same time has dropped ALL handset subsidies - you want an iOS phone, you pay Apple's inflated retail price, or pay a higher monthly tariff (rental + credit repayment for the handset). This is novel for the US, where typically you pay the same monthly regardless of phone, and all phones are subsidised down to the same low price.

Exposing the US customer to the real cost of their phone exposes Apple's biggest weakness in the market: value for money. Yes, an iPhone is better value to a customer than, say, a Galaxy S3 if both are $150 (and the monthly tariff is the same), but when you remove the unequal subsidies, and the Samsung is $350 to the Apple's $650, you really need to be a fan of the brand to buy the iPhone.

Under this model, The operators will still get the high-spending iPhone customers, but those customers won't change their iPhone for the next model every time their contract expires; now that they have to pay the full price, there's no incentive for them to do so. That's bad news for Apple: they need that operator-subsidised handset churn.

Verizon and AT&T are holding back for now, but they do want to move to the same model as TMobile, and the markets know it.

I think this is the major reason why Apple's shares have tanked. Nothing to do with Maps, or a mediocre iPhone 5. Just simple business: their biggest money tap (the US operators) is in serious danger of being turned off.

Re: PD is not the the next logical step on from FLOSS

You're absolutely right. I'm amazed by how many FOSS advocates don't understand this. Without copyright, there's no enforcement of open source licences.

GPL doesn't "reverse the direction of benefit"; it places the copyright with the project maintainers, or the FSF in some cases. The function of copyright in this transaction is no different to selling your software copyright to a commercial publisher: in both cases, it is IP law that allows the right to control copying and distribution to pass from author to distributor.

The only difference is that afterwards, a commercial publisher charges for the copies, whereas a FOSS distributor allows copies on condition of adhering to GPL or similar.

You can't have the GPL without copyright. You can't have copyright if you relinquish it by putting work into the public domain.

Re: where it's going

The standard is called just "Mirrorlink". It's not Google's invention, but rather a cross-industry standard, based on Nokia's original "Terminal Mode" connection system. First implementation was on Symbian, then Android.

iPhone can also connect if you write an app that embeds the libraries, but there's no OS support. Windows Phone, I don't know, but I guess it's coming, as Nokia were such a major player in the standard.

Clipping...

It may be fashionable (and justifiable) to think Apple are nothing but a hollow marketing operation, but that doesn't mean that the company never invented, or never pushed technology forward.

The QuickDraw region-clipping algorithm *was* new - other graphics systems allowed you to specify a single polygon if you were lucky, but usually only a rectangle. For irregular overlaps, this meant dividing your L-shaped exposed area into two rectangles, and then repeating your draw operation for each one. With an expensive draw operation, that meant wasted computation, or more complex code (preCompute(); for each rect in clip list: clip(); draw(); )

The QuickDraw system, unlike these, allowed arbitrary-shaped clipping regions. You could open a "region" handle, draw into that region with any of the QD primitves to define its shape, and then finalise it. Once you defined your clipping area using that Region, the clipping was done at the bit level, not geometrically (it's actually bit masking, but with some optimisations to a. pack the mask bit structure efficiently, and b. never execute fully-obscured draw commands).

The famous Evans & Sutherland patent is for geometric clipping on vector displays. The same techniques can be adapted for bitmaps, but the bitmapped nature of the display allowed much more sophisticated clipping to be performed. This is what Apple, or rather Bill Atkinson, did, and it's why he was awarded the patent.

(more history on this: http://www.folklore.org/StoryView.py?project=Macintosh&story=I_Still_Remember_Regions.txt )

Re: Understand the Lisa for what it really is!

Yes, because only Motorola could discontinue their processor line, but the reason for them killing the line was that the customers for the high-powered desktop chips were moving away from Motorola. A product without buyers is of no use to anyone.

SGI had already gone RISC, Atari had moved back to videogame consoles (and never progressed past the 68030), Commodore's Amiga was hitting the end of the line (and also never shipped with anything beyond 68030 - the 040 and 060 chips came from the resurrected organisation).

Intel's phenomenal success in the late 1980s gave them so much more money to pour into deep pipelining and other tricks to make their architecture work for them, but Moto couldn't justify it. They had very few volume sales for the high-end CPUs, and at the same time, the embedded device customers were pulling them towards low-power operation, which is where the 68k ended up.

Motorola stopped developing high-performance 680x0s, killed their 88000 RISC line, and joined the AIM (Apple IBM Motorola) alliance in the early 1990s. The idea was to produce a modern, RISC-based desktop to mainframe processor architecture between them and reap the rewards of the larger economies of scale. Apple and IBM were to provide an OS (Taligent), Moto and IBM did the fabrication, and all three would make customer hardware.

It didn't work that way: Taligent died very slowly, IBM concentrated on Power servers and Motorola ended up making PowerPC chips for embedded devices, with a high-performance variant just for Apple, who were also dying on their feet. The G4 was the last Moto chip in a Mac, but while its embedded focus made Apple's laptops king of the heap for battery life, the desktop line was falling behind the competition, not just on the customer-facing marketing point of peak clock speed, but also on real performance. The IBM-sourced G5 was the last hurrah, but as a server part, it would never make it to a portable.

I learned assembly on a 68000 (Atari ST), and 68030 (Atari Falcon030), and always remember the chips fondly...

Re: Perhaps none.

You might watch the video: you'll see it work with what looks like sunlight, but could have been artificial light. In any case, it was not laser light. The video clips also suggest that that the experiments were done at room temperature.

Re: Comprehension

I worked in Germany with a (German) guy who learned to sing every Bob Dylan song word-perfect years before he learned English. He had no idea what he was singing, but that wasn't the point. Could lead to some strange interpretations, though... ("Aalen in der Wachs stau er", anyone?)

Oh,...and it's "Sings the street a serenade"

Happy New Year!

Re: What the fuck does "overly prescriptive" mean?

Overly prescriptive = telling your contractors *how* you want something done, rather than telling them what you want to do with it, and why. The *how* part should always be the contractor's job, once they know "what". "why" is useful because it allows better solutions to be proposed...

Re: fat cheque

dobrý!

While I agree that the 920 seems to be selling very strongly, SeekingAlpha is a cesspool of stock-pumpers at best, and uninformed fanboys at worst. Believe absolutely nothing you read there.

Actually, the same goes for WallStreet Cheat Sheet and the Motley Fool and all the other "crowd-sourced" stock analysis sites. One day, if they get their act together and shape up, these sites might become a wretched hive of scum an villainy.

Re: Insanity of software patents

Yes, it's "just maths", but so is GSM modulation, which you claim to be fair and patentable.

The work in H.264 was in finding a set of processes made from "just maths" that can be performed without needing huge amounts of CPU power, and which won't reduce the output to unwatchable garbage if some of the parameters are garbled in transit and which end up reducing the amount of data transferred without degrading the subjective image quality. None of those three things are a trivial task.

Have you ever tried to develop an existing principle into a working, functional product that meets people's needs? Edison may have been a fucking thief, but he was right about the "99% perspiration" part. That is what you're paying for: not having to make all the false-starts, not having to go down all the dead-ends, not having to rework everything to make it safe to use, or make it work with other components that aren't strictly compatible with what you're doing. You're not paying for the idea. You're paying for the time and effort it took to realise that idea.

"It is not right for people to hold patents on basically other peoples work that they often don't even properly understand. (When they try to modify the algorithm in crypto it is proven that they lack the necessary knowledge)."

You know that this line of argument prevents people owning cars, so I'll assume you walk everywhere. The reason it's called intellectual "property" is because the implementor is able to sell it to someone else. If that someone else already knows exactly how to implement the idea, they wouldn't need to buy the intellectual property.

So, is technology only allowed to be used by people who make it? That's a frankly stupid contention. I doubt that you know enough about the various processes and systems to be able to build, from first principles, a device capable of connecting to the internet, so I must assume that you bought it from someone who did. And that company didn't know how to make all of the components in the box either, and so on back to someone who invented a way of doing something.

"Copy the algorithm as is into Cadance test it on an FPGA and then get an ASIC made is the sort of thing that a 3rd year uni student could do."

Yes, but copy *which* algorithm?

Sure, it's easy to burn any random process into silicon. It's also easy to type random words together on a page. But despite initial impressions, typing random words isn't the same thing as writing "Ulysses".

By the way, I wasn't the original AC poster... but fire away guys...