2504 posts • joined 18 Jun 2009
Farewell then, Carbon
Per the repeated story that QuickTime for Windows branched off a sanitised version of the Classic OS APIs that was backported to create Carbon for the OS X transition and was the central thing that caused QuickTime X to be a compatibility-breaking rewrite (which quite possibly still doesn't have a codec plug-in API?), I guess this is goodbye. And Win32 becomes the new Father of the House?
Re: Chuckle ...
I guess you reap what you sow?
I can't decide whether QuickTime for Windows is worse than Flash for OS X was, both being fairly awful.
Re: yet to see a woman wearing one.
I wear a health band, this specific one's main contribution over everything my phone does already being continuous heart rate monitoring. Oh, and having the time on my wrist, obviously. Which is relevant to the lifestyle I aspire to; I'm sure a more psychologically robust person wouldn't be interested. But, regardless, the basic test is: what do I gain from having one that I would not otherwise have? Avoiding having to reach all the way into my pocket for my phone is not sufficient.
(EDITed addition: the Apple Watch isn't especially close to continuously monitoring, it samples every ten minutes; judged according to the things it does that my phone does not already, battery life is far too poor and it's far too expensive)
Re: IE must die
An Edge fan?
Re: Shortage of new product
Or it's because the cost of matching the production curve to the demand curve would be greater than the cost of a shortage. Even more so if you think that a shortage generates sales rather than restricting them, but possible regardless.
Ramping production up and down is expensive. Apple products tend to launch big but then remain available for at least a year as the only thing in that market segment. Both of which are atypical.
Other devices don't tend to do the same blockbuster opening weeks because they have more similar competition and, usually, get supplanted more quickly — other companies, including those that sell many more devices overall, offer more choice and launch products more frequently.
Re: Where's debs?
It's on GitHub. For Swift Apple seems to be using the public repository for the mainline development branch. It's not a case of developing somewhere else and then publishing.
No, they used let to mean val; it's not a modifier and it means only one thing. const is a modifier and means any of (at least, without being an expert): immutable value, don't copy this onto the stack or this method always returns the same thing. Though it's at least more coherent than the manifold meanings of static.
Re: The same memory as the BBC Micro Model A of 15 years ago...
Peter Ford: the lower half is Mode 5. Mode 6 is Mode 4 with the two blank lines between every character column; Mode 5 was 2bpp, 160 pixels across. Mode 6 is, like 4, 1bpp and 320 pixels across but only 25 8-line character rows high, spread over 250 display lines.
With no VIA and the mid-frame real-time clock interrupt not being conveniently placed, the Electron version of Elite is just Mode 4 for the entire display.
After Elite, I'll have Starship Command, please.
Re: Who wrote this rubbish?
Obviously the author buys into the idea that the 6502's zero page was a RISC-style load/store architecture large register bank before anybody had thought to name it. You know, if you sort of forget about the other half-dozen addressing modes...
My understanding is that, discovering that the book publishers were unhappy with being unable to control Amazon's pricing, Apple spoke to all at once — effectively mediating a group discussion — so as to propose the agency model (i.e. publishers set the price, Apple just takes its 30%), and that all the publishers then simultaneously demanded the agency model from Amazon. Since they wanted to switch to the agency model in order to increase prices, Apple had facilitated a group with a monopoly over the market to try to manipulate pricing in their favour.
All of which Apple appears to accept, except they say that they did it for the benefit of the market — that grabbing some power back from Amazon reduces the main monopoly seller, rather than artificially empowering the monopoly producers.
As a consumer, I'd rate that as hubris indeed. Supposedly competing parties acted together. So as to increase prices. How askew would the world need to be for that not to be a breach of anti-trust law?
It's a nice enough language if you don't mind the memory management
Less of a syntactic load than C++'s explicit unique/shared/weak pointers but no smarter when it comes to potential cycles (i.e. a garbage collecting language like Java is still safer), quite neatly handles optional reflection, as of 2.0 finally in the modern world on exceptions, and at least aiming towards a modern take on closures.
A lot of the standard libraries are still a hassle though, doing nothing to hide the Objective-C bridge. All the Swift-native collections and atomics are value types; anything returned by a really-Objective-C library that doesn't bridge to one of those collections will necessarily be a reference type. Even if it would make a lot more sense as a value type. So the semantics aren't always lovely. Also it bridges directly to C and to Objective-C but not to C++ so interfacing with C++ is a hassle of manual Objective-C++ work.
It sounds like IBM is mainly just thinking about helping the iOS/Mac developers who were likely to use Swift anyway though, so I don't suppose these considerations are relevant.
That plus the precedent it would set — the government doesn't need a master key if it can take each individual phone to its OS vendor and compel them to create and install an appropriately version-matched copy, modified as desired. It's one now; next year it'll be a handful; the year after that a couple of dozen...
Re: Apple has access to the information, they are just refusing access..
Radiation's a bad example because it definitely will become a public health risk if generated with abandon, the tools to monitor it are widely commercially available and any laws that require radiation monitoring are debated and agreed by elected representatives and either phased in for existing operators or known in advance to newcomers.
In this case we're talking about the work phone of a person that physically destroyed their personal phone and computer, that the FBI already has the iCloud backups of, and a request that Apple be compelled to engineer a new product.
So a lot of us are arguing that there's a negligible probability that there will be something on the work phone that the criminal decided not to destroy, that's helpful, but which isn't considered a sufficiently important category of data for the OS automatically to back up; and that in any case a court should not be able to compel the creation of a new product using the current legislation.
Microsoft itself says "When you interact with your Windows device by speaking, writing (handwriting), or typing, Microsoft collects speech, inking, and typing information—including information about your Calendar and People (also known as contacts) [...] We also collect your typed and handwritten words [...] Some of this data is stored on your device and some is sent to Microsoft [...] You can turn the Send Microsoft info about how I write setting on or off in Settings.".
So that's fairly unambiguous. Microsoft does collect things including "your typed ... words" and "Some of this data ... is sent to Microsoft". But not all of it and you can opt out. I'd be uneasy, I think a lot of people wouldn't care.
Re: Anyone remember the Atari Lynx?
The 6502 in an Atari Lynx is a 65SC02 and, being clocked at 4Mhz, is probably roughly twice as fast the Z80 in the Spectrum. But it doesn't need to be because the unlike pre-+2A Spectrums the Lynx has a double buffer, and unlike all of its contemporaries it has a scaling hardware blitter with some basic vector drawing capabilities and a maths coprocessor for multiply and divide. It has to surrender the bus for the former but can run in parallel with the latter.
These are some of the benefits a machine can accrue by being launched seven years later.
But the Spectrum games are better.
I think the opposite is more likely; Apple created and gave OpenCL to Khronos back in 2009 so probably still has a good relationship with them, and has its high-level SceneKit and SpriteKit as proprietary APIs up where most developers now hang out.
I think the only market advantage Apple derives from Metal is that cross-platform fare sitting atop Unreal or Unity or whatever runs better on its hardware than on a hypothetical clock-for-clock equivalent alternative.
If Vulkan offers the same advantages of Metal then I would expect Apple to adopt the two as equals (hedging against Vulkan falling behind, naturally).
Re: Old portable kit ...
All the benefits in productivity afforded by kilo-or-so computers with large, high-quality displays, plentiful storage and trivial networking have been eaten by email, Facebook and Wikipedia.
The Virgin 747s running from San Francisco to Heathrow still have 15 or 20 economy seats upstairs; it's well-worth getting one if you can. Quite apart from being quieter, window seats get a little shelf and, of course, you walk straight down to the exit, allowing you to get into the border control queue a couple of hundred people earlier than you otherwise might have done.
Re: Prosecution required.
Negligence = duty + beach + damage. Someone can sue as soon as they suffer damage. Which, even with the probable quality-of-lawyer differential, is still better protection than '90s-era HTTPS.
Re: Hmm 1997!
That's £2,180.34 in today's money to save everybody else the effort.
Re: Who is presenting? @AC
The second rule of Robot Club is... no smoking?
Re: Just give me a frikking keyboard!
I'm more interested in the return of the trackball (or equivalent), which I last saw on a Nexus One. I find that typing on a screen is a much smaller problem than trying to position the cursor. It's almost always just easier to delete and retype than to try to edit when performing basic phone tasks like messaging. I seem to be better on iOS than Android but that's probably only subjective practice.
The current rules prohibit non-doms from filing electronically
... because obviously if you live overseas then, ummm, the Internet doesn't work the same. Or something. I'll bet thousands just ignore the rule; if they could eliminate it entirely then that'd be progress. Therefore I'm selfishly all for a digital-only HMRC, as soon as possible.
@boltar Re: Change the record
I think it's more the case that: (i) Flash massively overreached on features, going a long way beyond animation and video, and managed only relatively poor hardware acceleration, especially away from Windows; and (ii) such exploits in HTML5 delivery code as will be found will likely be limited to specific combinations of browser and OS. So they'll be smaller in scope and therefore easier to avoid.
Re: What a corker!
There was a 'Next Week...' trailer; after there was also a very minor post-credits sequence that was just about tying up that episode. I don't think you're missing any important plot points — it was highly skippable — but if you want an excuse to skip back after next week then there it is.
Re: Oh Lord I am old.....
The original 1985 LaserWriter had a 12Mhz 68000; every Mac available at its launch and for the next two years had an 8Mhz 68000. Having the separation of Postscript — usually mapped to QuickDraw on the then-two-colour Mac — is more or less what made the first generation of desktop publishing feasible. Your slower computer with the approximately real-time display could stick with bitmapped fonts, and the resolutions and colour depth were low enough versus available RAM that you could use binary masking for complicated shape things like path clipping. Your printer which was allowed to take a few seconds to produce the image of a page could do so with a much more rigorous approach.
... plus it was just a helpful way to 'multitask' back when none of the high volume consumer OSes were very good at it.
My take is that Apple's Classic OS probably peaked in 1991 with the introduction of System 7, not coincidentally approximately when Apple's many failed internal attempts to write a complete replacement began. They really missed a trick in moving to the PowerPC in not then writing something more modern and forcing 68000 apps into a sandbox, but I guess there's the list of things early-90s Apple could achieve with X developers in N years and the list of things it couldn't. Microsoft then leapt ahead until Apple was so far behind that it had to buy in a solution.
... but we can still thank the Apple of that period for Truetype fonts — not the idea so much, but the specific implementation and the help in market positioning. That Adobe Type 1 monopoly helped almost nobody.
Re: What about Apple?
On the contrary, they're alleging that their businesses suffered due to illegal competition from Google. The 'illegal' part being up for grabs.
In Apple's case, the business suffering part would be much harder to establish: in Europe Apple has only a 6% share in Spain, a 10% share in Italy and around 15% in France and Germany. Its 38% share in the UK hardly redresses the balance. Android's share is as high as 90% in Spain, through to more like 75% in France and Germany.
Putting aside that nobody is being sued, it's therefore a lot more sensible to look at whether Google is performing both of the things necessary to establish anti-competive behaviour — both distorting the market, and using an unfair advantage to do so — than it is to look at Apple.
Has anybody told Theresa May?
She still seems to be running on the old memo.
There is a genuine belief amongst advertisers that people love adverts that are sufficiently well-made; and that if the subject of an advert is something the person actually wants then they'll be grateful for having seen it. I'm not sure those claims are universally true.
Re: Terrible idea @Dan 55
It's a bit of a hassle that Android's C++ support puts you into the realm of the NDK though, especially as that's probably the most used consumer OS by volume. 64-bit ARM is evidence, if it were needed, that there's not the desire to retain binary backwards compatibility that characterised the desktop. Most (all?) of those CPUs can also run 32-bit code as a transitional move but Android is supposed to work independently of the instruction set so if it becomes possible to save some pennies by not doing that, it probably won't last for long.
Libraries like Qt make a lot of sense to us technical people but rarely produce especially native-looking applications. They're therefore a source of friction for users.
Qt specifically also produces developer hassle via its C++ language extensions, which are problematic for many of the more popular IDEs — they'll inhibit highlighting, indexing, etc.
So versus the alternative of writing it once per platform, Qt is often a better solution but not always.
Re: Rose tinted
Think yourself lucky, all we had was the "Totally Awesome!" Codemasters budget equivalent, Mig 29. Which my memory tells me was fun, but I daren't check.
Re: I'll give you my steering wheel...
In true NRA fashion I think that the best way to limit drunk driving is to increase the level of drunkenness in total, so that if anyone sees a drunk driver they somehow think they should charge in and get involved.
Re: Standard kneejerk response @AC
Rotherhithe Residents Call for Better Internet Speeds: "Never mind the ‘cyber highway’, Rotherhithe is stuck in the cyber slow-lane, after a new study confirmed that it was home to some of the slowest broadband speeds in the capital. [...] As the News reported last June, there is a better broadband connection on the moon than in Rotherhithe"
Re: Copper's last hurrah
Copper definitely won't outlive CISC processors. Then we can have the year of the Linux desktop. While enjoying a good new Sonic the Hedgehog game, if the genuine peril in Doctor Who doesn't keep us awake. Having finally resolved the EU question with a referendum so that everybody from then on is in complete harmony will free up the time necessary to lay the fibre optics.
Standard kneejerk response
Any hope of a reliable 3mbps in the Docklands? Or just more blah blah blah about the difficulties of upgrading historically non-residential wiring?
Apple explicitly allows apps to provide root certificates. That's the mechanism this app was taking advantage of. The OS segues to an OS-provided set of dialogues requiring the user explicitly to confirm the installation. There's nothing surreptitious, no silent or drive-by install. Which puts Apple on the other side of the debate about what informed users should be allowed to do than it usually sits but no doubt is required by some big corporate user somewhere.
For this app it sounds like they offer a VPN that fishes through everything you request in order to remove advertising. The certificate is then necessary explicitly so that they can be the man-in-the-middle for HTTPS traffic like Google and Facebook.
I think it's not something I'd want on my phone but it sounds like a third party is being punished for Apple's attempts to support business while providing its own brand of consumer protection?
Nice if true, but most people won't pay for privacy
... at least that seems to be the thesis of today's El Reg BlackBerry article, and sounds compelling to me. I therefore don't think this is much of a sales point, regardless of its motivation. Being very important to you and me doesn't make it the basis on which one sells hundreds of millions of handsets.
QNX underpins all of the current-generation Ford consoles; it replaced the relevant Windows CE offshoot last year so was sufficiently good to persuade at least one manufacturer to throw their existing stack away.
I think QNX also already has good reference implementations of Android Auto and Apple's CarPlay, so will probably continue to gain licensees even if the market trends towards allowing the mobile companies to define the product.
I'm typing this on an iPhone, in my opinion the iPhone is amongst the top tier of devices but: what's the difference between this year and last? Who has been won over that wasn't before?
My pet theories mostly revolve around Apple getting something from Samsung's regression to the mean, the economy being a little better and the slow death in the US of the 24-month contract making some of these not newly acquired customers but merely customers of increased frequency. For all of which I have no direct empirical evidence whatsoever, obviously.
The problem with App Store submissions is that you cannot "install it [yourself] through the normal user route". You could resign it for local install — which is what you'd do to hand it off to QA — but it's always possible they found some way to detect the difference.
Otherwise I would assume it was a timed thing: probably something very basic like maliciously inserted code checks build date of binary, doesn't do anything if it's within 30 days, possibly there's a REST call "I'm this app ID, this build number, should I do anything?" with a remote database being manually updated only after someone has confirmed the build is in the App Store but I don't know how worried the authors would have been about somebody running through a proxy as part of the QA process (e.g. to make sure that your ordinary calls are occurring sensibly).
Re: Not the only app having Bluetooth problems
Same thing with my LifeTrak fitness watch; there's clearly been some sort of regression in the Bluetooth stack somewhere between the OS and a large body of apps — though whether it's because Apple fixed something or broke something I don't know; hopefully the latter because then we can expect a fix from the centre. However you can never rule out the former. Who here has never had a colleague who, rather than reading documentation, just tried to figure things out empirically? And who therefore ended up writing code entirely reliant on non-guaranteed behaviour, which later changed?
There's no save in Super Mario 3
Some pedantic ass had to say it. Warp whistles, etc: reward the player's dedication by having them eventually learn how to skip ahead.
Re: Let me know when it works on computers. @Arnaut the less
Serif, like the team behind Pixelmator, is based in the UK. I feel like I used to pass an office somewhere in Hampshire; the website says Nottingham now though. So if market research was of the US only then that'll have been as a result of other research.
Re: not setting the bar very high
Hulu just added some great FX content — I finished Fargo (the series) yesterday. What's potentially interesting is the Showtime deal: you can get that content, including new episodes the day after transmission, if you pay an extra subscription. So it's possible they're trying to shift to being a lot more like a traditional cable or satellite service, though presumably without the arbitrary bundling.
Re: Don't Buy, Sell
There was a report within the last few days that the supplier of Apple Watch internals had a low estimate of expected shipments but is surprised not even to have reached the break-even point. Which I think flows into the suspicion that maybe Apple doesn't know what it can do to sustain growth.
Is it accurate to describe it as a zero-day exploit?
I almost feel we should be shaming these companies a little more; by now this is at least a minus-fourteen-day exploit.
Re: There's plenty of good holocaust jokes
On the contrary, the tasteless school of humour is nowhere near the majority of jokes — there are puns, there's observation humour, there's absurdism and non sequiturs, satire, wit, sarcasm, and probably about a million more.
It's also not something that anywhere near everyone finds funny, as reassuring as that fiction might be to some. If you enjoy it then good for you, but keep some objectivity.