Posts by Kristian Walsh 1817 publicly visible posts • joined 10 Apr 2007
Since January 2015 VAT on digital downloads (and other telecommunications services) in the EU is charged at the rate applicable to the downloader, not the company selling the download.
This change in the rules on Place of Supply doesn't affect the supply of physical goods; it's only for "digital" online stuff.
(If you've been affected by the issues raised in this post, you can contact your local tax advisor who'll expensively explain it in great detail to you...)
Ireland won't see anything like 13 billion. If the appeal is lost, then the leaders of every other state that Apple did business in will line up outside the Irish Revenue offices with a bill in their hands.
Heading the queue will be an orange man with stubby little fingers, because the big loser in Apple's non-shore tax status is the USA.
So, Ireland would be lucky to see 500 million from the whole pot, and guess who'd foot the bill for administering those claims? Hint: rhymes with "Blyreland".
It's only the megaphones and placards class of politicians that's playing this as "Ireland throws away thirteen billion to please its masters" - truth is, there's no upside to the country.
In the USA, maintenance of the Federal interstate highway system is paid for by a tax on gasoline (about 18c per gallon). States add an average of 30c/gallon on top of this, with varying percentages of that going towards road infrastructure. A pittance by European standards, but still a significant portion of the cost of fuel.
Once electric cars become popular, they'll stop being given a free pass on the taxes. Probably in terms of some kind of weight-tax, because it's very hard to tax electricity used for cars differently to electricity used for domestic heat/light/cooking.
In Europe, there's plenty of ways already to tax electric cars (road tax, registration taxes, etc), but they're just disabled for now.
Or use Lithium-Iron (note the R!), aka LiFePO4: It's like the Lithium-Cobalt-ion chemistry used in mobile phones (and Tesla's products), but it has the disadvantage of being slightly less energy-dense and quite a bit more expensive (although much of that is due to manufacturing economies-of-scale in favour of LiCo). In exchange for that, you get a far longer lifetime, higher current delivery, dimensional stablity, much greater resistance to mechanical mishandling, and, somewhat importantly for use in a dwelling: a battery that can't spontaneously combust even if it's hit by something heavy.
Currently, it's targeted for the applications currently using Lead-acid packs, but I think it has a good future as a home storage system,
Your argument does seem to be that because the touchbar might detract from your work flow, those people it benefits don't do 'real work'.
No, my argument does not seem to be that at all. I'll re-state it for you here: by replacing a proven useful interface with one of dubious and as yet unproven use, Apple is losing sight of what people who use their equipment need. Maybe the touch thing is useful, but surely the amazing design skills of Dieter Rams Jonathan Ive could have found a visually harmonious way to accommodate a touch bar and the escape key -- tossing the existing functionality is laziness at best, form-over-function at worst.
"Nobody needs [feature] anyway" is the cry of the Apple apologist, but I've never been in that camp - my needs are representative of me alone; I've no right to tell other people what they should and should not need.
True, I've bought my share of Apple computers over the years, but I've never felt the need to use the receipts for them as some kind of identity card. In the twenty years I've used Macs, I've spent a lot of time avoiding tedious arguments about why I wasn't using Windows (or Linux), and yes, I occasionally suggested Macs to people I thought might find them useful too, but I'm not going to buy an Apple product that doesn't meet my needs just out of some misplaced need to belong.
I'm well aware that I'm not Apple's customer anymore - their product decisions made that clear to me. I was wondering if they'd make something that would change my mind on that, but the only two personal computing products that have made me look twice at them have both been from Microsoft: Surface Book last year, and the Studio last week (I'd love the latter, but I only do illustration as a hobby these days, and can't justify the price on that basis).
The irony of "boring old Microsoft" leading the way in personal computing hardware is not lost on this ex-Apple employee.
You're right about function keys, but nobody is complaining about losing the function keys. Macs never had these keys originally, so a lot of Mac software never used them. As far as I can remember, the Macintosh was ten years old before every model shipped with a keyboard that had function-keys on it.
What existing customers are complaining about, and apologists are ignoring, is the removal of the Escape key, something which is widely used in lots of existing software. Removing this key irritates users and hinders productivity, and forces you to upgrade software to work around the deficiency. You mention Adobe, but there's still a lot of Adobe CS users who bought their software outright, and don't want to go to the subscription Creative Cloud for financial reasons, so the argument that newer software will work around this just doesn't wash.
(Yes, the original Macintosh and Mac Plus keyboards didn't have Escape keys either, but every other Mac keyboard did, even if it was in a bloody stupid place like on the old MacII keyboards)
Yes, and Lenovo doesn't have a subset of its customer base that accepts anything Lenovo does, and rationalises Lenovo's dumb mistakes as being nothing more than their own inability to deal with "progress".
And this is really is a stupid, awful idea, and it's bad on more than one level. Superficially, it forces users to divert their visual attention from the screen down to the keyboard, because it removes the ability to find the Fn keys by touch. But that's not why it's a disaster for people who actually do work with their computers.
It's a disaster for users of all kinds because it removes the bloody Escape key. Does nobody in Apple's "design" group use Adobe Creative Suite anymore? Or Excel? Or vim? Or XCode? Have they never needed to force-quit an application? (Command+Shift+Escape, and it's the only way to restart Finder when it hangs). A surprising number of Mac software dialogs also dismiss with "Cancel" when you press Escape. It's something that you pick up over a decade of using an OS.
Escape has a specific meaning on Macs: it means "exit this mode, cancel". It's as much an essential command key as Backspace, Return ("confirm") or Tab ("go to next").
The only commonly-used Mac application I can think of that doesn't map Escape to a function is.... Safari.
So there you go. Apple thinks that its customers will only ever use the web browser. And that's probably true of the floods of vacuous tech-bloggers who'll treat us to articles proclaiming how nobody needs the escape key anyway just because they didn't...
637x438 mm will fit an A2 sheet, so we're getting there. There are plenty of A0-sized displays around, but they're in televisions and signage. You'd need one with at least a 9000 x 6000 resolution to keep the image sharp enough to substitute for a drawing-board...
Maybe in three or four years? Still, It won't be cheap.
In any case, NeXTStep's kernel was first written for Motorola 68030 CPUs, not x86. From there it went to x86, then PowerPC, then x86 again.
If my memory hasn't failed me, 68030 still had only two execution contexts that it had inherited from 68000: User and Supervisor, but moving between them was only a matter of a stack push and a status register bit flip (you couldn't do it explicitly: you had to generate an exception, usually via the TRAP instruction, to enter Supervisor mode, or use the RTE instruction to leave it)
London's area code is (020) domestically, or "+4420" internationally. The next eight digits are the subscriber number, and the first digit of this has no geographical significance, and there are three ranges in use at present: 7xxx xxxx, 8xxx xxxx and 3xxx xxxx.
The distinction only matters if you're phoning from another fixed line number within London; from a mobile, you always have to dial the destination area code, so it's fine.
The confusion is a legacy of the decade-long manoeuvring of London's phone-numbers in the UK's wholesale reinvention of its numbering plan between 1990 and 2000. From 1990, there were indeed two area codes (071) and (081) which changed to (0171) and (0181) in 1995 before finally being merged back into (020) in 2000.
On "Swap Shop", the (01) 811-8055 number was specially set up for the BBC by the Post Office a year into Swap Shop's run to reduce the number of mis-dials. They picked "811" as it was the most numerically-distant exchange code from existing London exchanges (when entered on a phone-dial). The original "call the BBC" number was (01) 288-8055 [yes, I did have to look that up]. After the 1990 London split, the number became (081) 811 8181, but that was long after my days of getting up early on a Saturday.
Minority participants showing a higher median income is a sign that there are barriers to entry. (Consider an opposite situation: why are the very few men working in the women's fashion industry disproportionally clustered at the top?)
Basically, the problem with women in tech is that the secondary school environment* is stacked against girls developing interest in computing, so only those with a strong aptitude for it will persist with it. This self-selected removal of "okay" and "average" from the pool that applies to university means that only the top tier of females with IT ability ever participate in the workforce; to use arbitrary percentiles, those salary surveys could be comparing the top quarter of all females with IT ability to the top three quarters of all males with IT ability.
* "environment" doesn't just mean teaching and subject options; it also means the perception of what a tech job is, and the kind of person who does it. After twenty years in this business in various settings, I can recall hardly any emotionally-stunted egomaniacs with dubious personal hygiene, but that is still the stereotype of what "good with computers" means.
But it's not just female engineers - we're losing well-rounded people in general. The mess we're in is a result of too many companies hiring people who can churn out code, rather than people who understand the needs of other humans - the poor bastards who will have to use that code.
Performance isn't just speed, it's also agility, roadholding, and response to driver inputs. Getting these right takes serious mechanical engineering, and that's a very different mindset from the silicon valley"if it looks like it's working, then it IS working" attitude.
Incidentally, the highest signposted speed limit in the United States is already over 130 km/h. I bet you can guess which state without looking it up first...
Also agree that 10.6 was the peak OSX release. Significantly it was also Bertrand Serlet's* last release as head of OS development. He left Apple after this release, handing over to a guy who'd only been in the company a wet weekend. The next version, 10.7 was the first ever MacOS X release I rolled back.. But it wasn't the last. My Mac is still on 10.9, and I think it'll retire on it.
(* An ex-NeXT manager who led OSX development and later took over as head of all software. The most senior person I had a proper conversation with in my time in Apple: I had a somewhat heated engineers' argument about application resource formats with him, *before* my then manager revealed exactly what his role in Apple was. Oops.)
My personal requirements for a laptop: Light weight, good screen, able to run Illustrator as well as bash and all you can spawn from it. Windows 10 laptops now qualify, and the "bash" bit of that deal is better than on MacOS ( genuine " apt-get" versus brew/port) Also good is that I can choose to purchase a device with a physical NIC... Apple wrongly thinks I don't need one anymore.
Downsides of Win10 is the regularity of updates that require a restart and twenty years of MacOS muscle memory still making me go to the writing place for things; upside is it's faster than OSX, less buggy (these days) and doesn't assume I own an iPhone.
... I do miss BBEdit, though.
There was a function called a "speakerphone-modem", which allowed you to use the modem as a dialler, for calls you'd carry out over the speaker*. For exec types (the kind of people who'd be given, or buy, personal computers in the mid-1980s) using the PC as an electronic phonebook like this was probably more useful than any other the other application software on it.
Can't remember who said this, but:
"Successful garage startups are mainly notable because they happen in garage-sized industries"
The mid-70's personal computing business, the early 1980s application software business, the mid-1990s Internet business, the mid-2000s social media business - these were all tiny industries at those times, and any one of the companies launching product could have made a go at it without having to spend millions.
The spelling CZech came from... the Czechs.
The spelling of the Czech language was reformed in the mid-19th century. Before this time, many words that begin with "č" were spelled as "cž"*. By the time "Cžesko" officially became "Česko", the English-speaking people of the world had already decided to use the Cz.. form, but stripped of its diacritics, of course.
To add to your post, the best approximation of "ch" for English speakers is the end of the Scottish (or Irish) word "loch", as pronounced by a Scottish or Irish speaker.
Also, as a lesson for programmers to stop using code-point sorting for lists that a human will read: "ch" is considered to be a letter in its own right in Czech (it sorts after 'c')
The Latin language survived long after Rome fell, and was used in places that never saw the Empire's rule.
Throughout the early medieval period, the "Czech lands" were firmly within the Holy Roman Empire ("... of the German peoples"), and while that state itself was neither holy, nor Roman, nor really an Empire*, it kept Latin as its language of state.
( * someone was going to say it, so I decided to get in first)
"Česko" is what most Czechs refer to their home country as when speaking Czech, "Česká Republika" is what's on the banknotes and passports, but is hardly never used in day-to-day life (although you do see "čR" frequently used in print as an abbreviation).
As I understand it, Vaclav Havel's objection was to the use of the name "Česko" as the international name for the country, and it's because the "-co" ending makes it sound like the name of a business, not a country.
It's surprising it took so long for "Czechia" to appear: it just substitutes the Czech "-sko" with its international equivalent, "-ia" and so preserves the meaning of "Czech-land". In the German language, "Tschechia" (same pronunciation) is commonly used as the name of the country already, so there's a precedent there in two of the country's neighbours.
... but yes - many, many Czechs really don't like the new name.
Yes, Apple spent much of this period on death's doorstep.
If I were a government minister in 1997, weighing up the income-tax receipts for 1,000 well-paid jobs versus the corporation tax receipts from a company that was losing money hand over fist (and I mean losing money in the old-fashioned way of spending more than it got from sales; no financial trickery needed), I'd have made the same decision.
Back in those days, one of those jobs would have been mine, so this isn't entirely dispassionate, but the fact was that this deal was a calculated risk taken on by the Irish government with the intention of maximising tax revenue at the expense of some future corporation-tax receipts. Of all the manufacturing exporters operating in Ireland at that time, nobody would have predicted that it would be Apple that would become the world's highest-valued corporation - even as late as 2005, the company was merely "doing okay", and at that point I'd suspect the balance was still in favour of the Irish government.
Apple was not the only company offered this arrangement, but the company's stratospheric profit growth right at the end of the period it was in effect does make it look worse than it was. I've no love for the company that Apple has become, but I think this ruling is stepping over the line of allowing member states to organise their own tax affairs.
Maybe, with twenty years of hindsight, Ireland's government shouldn't have assisted Apple to remain in Ireland in this way, but that's like saying that maybe, in hindsight, Decca should have signed the Beatles when they auditioned for them in 1962.
Rigorous sex-education in schools also correlates with societies that don't see a problem in young teenagers having access to contraception.
The problem I thought would arise with this "real dolls" programme was unfortunately what happened: as humans, we're good at learning, adapting and coping. Give a fourteen-year-old one of these and she'll be driven demented for the first two weeks, but after that it'll just become routine, another part of the shitness of being a teenager (c'mon, doesn't anyone remember what having to go to school every day was really like?). The fear of having to raise a baby on your own is a big factor in girls not engaging in unprotected sex; this programme only removed that fear.
If you don't accept that teenagers are having sex, and nothing you'll do will stop that, you can't tackle the problem of early motherhood. I think the money would have been better spent on sex education, supplying teenage girls with condoms, and teaching teenage boys to use them (because it's not like girls get pregnant by magic)
It's not a carrot to devs, more like the old Windows Phone apps were not built for resizeable windows, which is what Continuum looks like to your software.
Basically, if you've got a UWP app, and it's running in phone mode, the app will see a tall portrait (or low landscape) window of a certain size; but when the OS launches into Continuum, the app's window resizes to a new, larger*, landscape size to match the connected display. To accommodate these different screen aspects, the code uses a similar system to CSS fluid layout: you define "Visual States" that your layout goes through depending on the viewport width and/or height. (Yes, this does make the UI layouts much harder to design, but that's the price you pay for flexibility)
(* larger in terms of "virtual pixels", so the app's text and controls get smaller even though your phone display probably has more actual pixels than the desktop monitor you're attaching to)
The problem with Windows Phone 8.1 apps is that they're designed for a fixed-size window, and there's no API mechanism to inform the app of live changes in that window's size. Also, Windows Phone 8.1's control set relies much more than 10's on touch gestures (swiping particularly) that are clumsy with mouse, as anyone who's run their WP8.1 app in the Visual Studio phone emulator can attest.
So, rather than present a pretty crap user experience (a tall pillarbox view in the middle of your monitor, that you've got to continually drag over with the mouse to navigate) , MS disallowed 8.1 apps from using Continuum at all.
One corollary of how Continuum works is that there's no technical limitation preventing a multi-windowd variant of it in future (to work with Continuum, apps must already support window resizing), but this would mean that the phone would have to keep multiple apps running in the active state at once, which is asking a lot of a typical phone CPU and battery.
Passive cooling means it'll be drawing cool air in through vents at the bottom, as warmed air rises through vents at the top.
Your skin does offer a better thermal sink than air, but only where the incoming air is warmer than your skin temperature, and only if the heat-producing elements are not be placed directly in contact with your skin. In reality, what's against your head will be the plastic shell, then an air gap, then heat-source.
Agreed. I'd be shocked if I heard they paid money for any of these. It's not that I don't particularly like the work, it's that they all show a lack of understanding about what a brand logo is supposed to be.
A logo is not a "picture", it's a visual component that has to be used in a wide situations, from signage to promotional items as well as the more obvious documents, advertising and websites. Some of these would be a total pain in the ass to integrate into any kind of layout (Flik-Flak, #7, is the worst, but "Impossible M" is equally difficult, and the type doesn't fit the graphic properly), never mind the more inventive ways that branding is applied in the real world.
My vote would be to just adapt the dinosaur-head image that MDN has used for years, and keep the existing "mozilla" type - it's recognisable, distinctive and it can be used widely.
Criminals rob banks all the time - does that mean that a bank building is less secure than a greengrocer's shop?.. because, after all, you never hear of a greengrocer's getting robbed.
Security is not a technology issue. The technology is just a tool-kit; it still needs to be used by a human. Windows has all of the same security measures as Linux in place to prevent privilege escalation or out-of-process access, and all the other nasty ways a piece of software can do something that you, as the IT admin, don't want it to do.
But none of these protections really matter, because the easiest way to get malware onto any system, no matter how "secure", is to give the victim step-by-step instructions on how to install it and let them work around the system's security measures for you. Of course, you tell them that they're doing something else (like getting something for free), but the malware gets in there all the same.
And if you target Windows with your ransomware, you've got: 1. a far bigger target population, and 2. a higher likelihood of getting your malware into a company that'll pay the ransom.
I use OSX, Windows10 and Linux daily, and I don't really play favourites, but I do get fairly tired of the insistence from Linux and Mac users that their platform is somehow more secure simply because criminals have no real financial incentive to attack it.
it is actually surprising they never bought a Korean automaker like Kia to add to their collection in the past.
It went with a joint-venture instead. "Samsung-Renault Motors" produces a range of cars based on Renault designs for the the Korean domestic market. (Info in English here: https://group.renault.com/en/company-vehicles/discover/renault-samsung-motors/ )
Assuming you're not trying to reason from information that's nearly thirty years out of date, your old FIAT's ignition system was most likely made by Bosch, not Marelli.
Today, Marelli is widely regarded as the most advanced automotive lighting systems supplier, with a customer list that reads like a who's-who of the top car manufacturers. But, in terms of a Samsung Electronics purchase, their position as a producer of digital "infotainment" systems would not be as commanding as other companies such as Bosch (makers of Audi's fancy "virtual cockpit") and Harman.
Once again, Tesla PR is playing fast-and-loose with statistics. They cite the US National fatality rate for ALL ROAD CLASSES to try to minimise their involvement in an accident that occurred on a divided, limited-access freeway - the safest road class. To go on to compare to a "Global" figure is meaningless and insulting.
For a more valid comparison, the German Autobahn network, including those small sections where no upper speed limit applies, has an overall fatality rate of 1.9 deaths per billion-travel-kilometres (one fatality per 326 million miles travelled).
None of this absolves the driver. Unless the car acted deliberately against the driver's control to cause the accident, it's still the driver's failure - it was their responsibility to control their vehicle. Had the other vehicle been a passenger car, and not an 18-wheeler, the Tesla's very heavy weight would have meant the driver would have probably survived... and then stood trial for second-degree murder.
The "EU" didn't exist 70 years ago, but it originated in the ECSC, a transnational trading bloc to prevent future trade-wars over access to commodities (in this case coal and steel, it still being the 1950s). The redrawing of Germany's borders after WW2 left a lot of potential for regional disputes to flare up, and it was the Treaty Of Rome in 1950 that defused most of them before they could start. That's where the "70 years of peace" claim comes from, and it has only ever referred to Western Europe - those national empire-states that spent most of the previous millenium killing each other.
"the EU did little good in Northern Ireland"
Hardly a surprise. Not just because the EEC had no such powers at the time, but even if it had, member countries of the EU are sovereign, and thus "Europe" cannot intervene in their internal affairs. Mostly, though this is because the Northern Ireland crisis predates the UK's relationship with the EEC.
The whole Northern Ireland thing kicked off in 1968 (the only surprise was that it had taken so long). The situation peaked in late 1969, and it was only the deployment of British Army troops in the province to protect the Catholic population that stopped the province descending into civil war (I say this even with the memory of Bloody Sunday; without the troops, things would have been far, far worse). Before this, things were so bad in early 1969 that the Irish government actually considered the possibility of having to stage a military intervention in Northern Ireland to create a protected enclave for Nationalist refugees within the province. (The Irish government papers are brutally honest about the slim chances of such an operation succeeding, but it had been planned for)
By 1973 when the UK (and also the Republic of Ireland and Denmark) joined the then EEC, the crisis had been largely averted. What remained was a domestic UK law-and-order issue, and as a domestic issue, the EU could do nothing about it. However, common membership of the EEC did give the Irish and British more opportunities to cooperate on resolving the problems - the fringes of EEC meetings allowed the governments to meet and talk about this without it carrying the weight of a "summit".
What the EU did do later was fund infrastructural links between Northern Ireland and the Republic - the opening up of access between Dublin and Belfast has done more for peace in Northern Ireland than most will give credit for.
I hope the UK votes Remain. We actually like our next-door neighbours more that we'll ever let on, and we really don't want to have to live with border checkpoints again on what would become the frontier of the EU in the event of an exit vote.
If this was on the back of decades of the UK governments trying to reform the EU from within, I'd say "vote leave - you've done your best; they're never going to learn".
But it isn't; the UK most definitely hasn't; and I hope you stay. Be part of the solution, rather than walking away.
Andrew's article is probably the most reasoned, thoughtful argument for a leave that I've seen, although he makes the common mistake of imagining that all of the UK is like England. Others have pointed out what was happening in Northern Ireland while the UK was joining the EEC. I lived there at the time, although very young, and I'm glad that the Westminster government stepped in, rather than walking away and leaving NI to its own evil devices - as bad as having the British Army pointing guns at British citizens was, it was better than the alternative.
Stepping in, rather than walking away.
This piece gets to the heart of the problem with the EU -it's out of touch. A cosy cartel of middle-class "right-on" self-interests that has alienated the working classes to the point where only the far-right hatemongers can offer them anything. Very few in 1930's Germany would have put the NSDAP's stringent anti-Jewish policies at the top of their reasons for voting, but they certainly appreciated the ideas of "more jobs, more security, less vested interests, less corruption" and "making Germany great again" enough to enable the worst episode in Europe's history.
If we all let things as they are, we will sleepwalk into hatred. But that doesn't mean that the UK should walk away. Walking away won't fix the disease, because the UK has a bad case of it already...
If you're not a Labour or Conservative voter, do you really have any say in how your country is run? Worse: as both of those parties are on a knife-edge, they're afraid to take any kind of meaningful stand for fear of handing power completely to the other (Labour are particularly guilty here). The electoral system creates reactionary politics. Nobody will lead, and everyone defines themselves by who they're not.
Also, I'm no fan of UKIP, but an electoral system that gave 12.9% of the voting public just a single voice in a 650-seat parliament is hard to describe as "democratic". I don't care that theirs is a voice I strongly disagree with - all that matters is that over one in ten people who voted in the UK wanted it heard. In the long run, it's better for the more extreme viewpoints to get a good airing, because it robs them of their moral high-ground of being "persecuted by the system". Calling them "racists and bigots" is deflection: democracy is about accepting the decision of all the people, not just the people who went to the same kind of school as you did.
The European Parliament is elected on a more representative system, so the irony is that a pro-Leave party like UKIP gets its only significant say in how the UK is governed by virtue of its 9 MEPs (from 73). Its single MP (from 650) has practically no power.
So, yes, the people of the UK are living under a democratic deficit, but leaving the EU will probably magnify it.
Why leave out other similar glyphs like Greek and Cyrillic? "0391.xpm" and "0410.xpm".
And THAT'S why this kind of idea is such a bad one... :) Not just lookalikes, but have a look at how scripts like Thai are laid out for an "oh shit" moment. The files in this example are glyphs, not characters, not codepoints. The relationship between the displayed glyph and the codepoints that caused it to be displayed is not 1:1 - it's n:m.
Fonts contain a set of glyphs, and character mapping tables that the operating system uses to choose an appropriate glyph for a sequence of codepoints that represent the text. "Table" is an over simplification.. some of these fonts define finite state automata that are necessary to correctly combine and rearrange the input codes into a sequence of glyphs.
tl;dr? Don't try to write your own text shaping engine.. your OS (be it Windows, Linux or Mac) already has a very good one, and the information you need can be extracted from it.
"I never did such a thing. Did you actually ever pronounce "install" ? I only have typed it."
Really? Never? I regularly communicate with other humans through the medium of speech. Sometimes such conversations include the description of processes that involve manipulating objects stored in filesystems.
Er... no. The point was that the poster cited this as an example where they NEEDED both the uppercase and the lowercase image to be present in the same directory. On a case-insensitive filesystem, his (I assume "his") example cannot be implemented. My alternative doesn't require case sensitivity.
I assumed, in the absence of any given context, that this was an application that rendered text on screen and needed two "glyph" images: one for 'a' and another for "A" (ignoring the details of why doing something like this is such a bad, bad idea). If this was the intended application, then using the Unicode codepoint of the character, rather than the character itself, is far safer, more portable, and more flexible for future needs.
I'm not kidding, and you miss the point. The two characters I showed ARE the same, but coded differently. The lexical meaning of both codepoint sequences is identical: "a Latin-alphabet lowercase a with acute accent above". That the underlying codepoints are different is only visible when you hexdump the encoded stream. One version is "lowercase a-acute" (a standalone codepoint inherited from ISO-8859-1), the other is "a" then "combining acute accent" (the general case of applying an acute accent to a base letter). The result, however, is the same displayed glyph: "á".
Glyphs are what you see, characters the concept that a glyph represents; codepoints are how you specify the characters in a text stream; bytes are how you represent codepoint. C and its standard library has probably misled you, so a read of this will give a more accurate explanation of how text is actually handled in computer systems: http://www.unicode.org/reports/tr17/ (Unicode TR17: Character Encoding Model).
Case insensitivity is the dumbest, least-effort way to handle text. In effect, you're not interpreting it at all - it's just a bunch of bytes, with no meaning. Some bytes aren't allowed because you use them to delimit your directory paths, but after that, it's all up to you. The idea of byte=character is a fallout from the first commercialised computing systems being developed in the USA, that most monolingual of nations - hence a standardised alphabetic code that couldn't accommodate any other language except English). Linux takes this approach, and it's often a difficult conversation to convince someone that Ext4 filenames are just a bunch of bytes and don't have any encoding - it's only libraries like glib that assumes filenames are UTF-8.
But that's the machine's view of the world, not the user's: if it's so natural that case should matter, tell me how you would pronounce "INSTALL" differently to "Install" ?
Regarding the "A.xpm" and "a.xpm"; I'd have called them "0041.xpm" and "0061.xpm", because if you ever needed an image for "☞", it's a lot easier to map it to "261E.xpm" than any other alternative. This convention would also allow you to create a forward-slash glyph without the special-case in your code.
HFS+ Alternate Data Streams are just the same thing as Ext4 metadata. Different name, same function. They are, by definition, not the file contents, and thus should be stripped if a foreign system cannot understand them.
The HFS Resource-fork, which is what you're referring to as the "second stream", has been deprecated since 2001... Since then, OSX image thumbnails aren't stored with the file, but rather in the ".DS_Store" file in the same directory. Think of it like the Windows "thumbs.db" file with extra info (view settings, icon positions).
(The creation of empty ".DS_Store" files when none are required is a bug that's only been in MacOS since 2001, but who knows, maybe one day Apple will start addressing functional correctness again...)
HFS+ is case-preserving, but case-insensitive.
$ touch Moo
$ echo "said the cow" >moo
$ cat MOO
said the cow
$ ls
Moo
Case sensitivity is a user-hostile feature, but of course there's one very large software project, in the form of the Linux kernel, that actually requires case-sensitivity in order to compile properly.
One other very useful feature in HFS+, that I hope Apple will preserve, was its use of the Unicode Canonical Decomposition for accented characters. There's more than one way to represent a character like "é", so the filesystem APIs made sure that all such combinations are stored in a standardised, unambiguous form.
By contrast, Ext4 (Linux) just stores names as a variable-length sequence of bytes which by convention (and only by convention) are interpreted as being UTF-8 by the shell. If you think case-sensitivity is hard to explain to a user over the phone, just wait until you try to tell them why they've got a file named "amadán" and another named "amadán" and these really are two separate, distinct, files...
To clarify: without feed-in subsidies, for domestic residential use, PV is not there yet unless you've got a lot of space available for panels. With only a small amount of suitable space, you're entirely right to go for thermal solar instead - it's also what I'm doing.
However, if you are operating a business, the case for PV solar is far stronger. I'm writing from Ireland, where we have no feed-in subsidies on solar electricity, but where we do have a lot of small, residential farms. For this kind of operation, PV can be economical - they've got space to install a sufficiently large array, and the business activities to be able to claim a capital allowance on it. PV also acts as a hedge against future electricity price rises, which is important when the price of your product is highly volatile.
When you factor in capital allowances, the payback time for a modern system is 5-10 years to a business user (including sole traders). Other jurisdictions treat this stuff differently, so your mileage may vary, but even without feed-in subsidies, one can make an economic argument for PV.
The big problem with tidal power has always been the inability to build a machine with moving parts that can survive for more than a couple of years in any sea with the conditions that would make tidal power economically viable.
Wind does work, but there's a limited number of places where the wind is predicable enough to make it work. What the report is saying is that the relatively few good sites are now taken, and there's no point in building out anything else.
Despite the scepticism about the UK's climate, Solar is probably the best of the options mentioned. Solar cells convert light, not heat, to electricity. Heat is the enemy of photovoltaic power: a hot cell has higher resistance and produces lower current. The UK's cooler climate offsets its lower light levels, so that even with today's relatively inefficient panels, people with enough free space can achieve an economic return on solar, even without subsidy.
Subsidies are part of the problem, as they encourage people to invest in poor technology. Also, I personally don't like subsidised buy-back tariffs for renewable electricity, because at the bottom line they're just another tax-break for the wealthy - the economics of selling energy back to the grid only work if you don't have to borrow the capital needed to set up the generating system. Also, subsidies have to be paid for by someone, and that someone usually ends up being the poor git who can't afford to deck his/her house with fifty grand's worth of solar panels.
Yes, that was the revolution in signals intelligence. The mere fact that A was talking to B more often than normal, and only after talking to C was recognised as being of value to the Intelligence services. Traffic pattern analysis could reveal that "something" was going to happen in a particular place, and that was often enough information to work with - from there, operatives in the area could be put on alert to gather more concrete evidence.
Gordon Welchman was one of the key figures in this area at Bletchley. ElReg did an article on him last year (http://www.theregister.co.uk/2015/09/27/gordan_welchman_bletchley_park_remembers/) and there was also a very good BBC Two documentary on him recently http://www.bbc.co.uk/programmes/b069gxz7
Welchman later worked for the US military, and developed an operational communications system for them that was much more opaque to his own traffic analysis methods. The name of it escapes me, but from the limited info I've read, it seemed to operate as a message bus (or ring) of encrypted traffic, rather than point-to-point: the topology meant that by just watching the traffic, you couldn't know who was talking to whom - every station relayed every ciphertext, but only the recipient had the necessary key to read it.
Not to dismiss the work of Bletchley, but to imply that Enigma wasn't strong encryption in its time is a bit specious - it was far, far stronger than what the Allies had available.
It took a decade of continued investment by the Polish, then the British secret services, plus the invention of a new kind of signals intelligence, a new branch of mathematics and an entirely new technology for performing computations just to perform brute-force on Enigma. With better operating procedures, an Enigma system was still economically infeasible to crack until well after the war.
In the end, the biggest aid to the Allies in cracking Enigma was good old military discipline: there were enough stations that sent short, known-plaintext messages ("Station XYZ, 1200, nothing to report") that the cryptanalysts had a greatly-reduced search-space to work with.
Had the Germans been ordered to begin every message with two random words from the day's newspaper (i.e., salting the plaintext), things would have been much harder for the Allies.
Remember when Samsung destroyed Android user's phone perfromance because they thought that any old filesystem would do on the Galaxy S?
Remember when Samsung's smart TVs stopped using YouTube because they shipped with a player app that was using a protocol that was already deprecated, and Samsung never bothered to provide an in-field update?
Remember when Samsung told angry customers that it was only required to support devices for two years from date of product launch, not date of sale, despite it keeping devices like the GalaxyS4 on market four years after launch?
Remember when Samsung "forgot" to encrypt user's audio commands for its voice-activated TVs, and remember when someone discovered that they were always listening too?
Remember when Samsung interfered with the Windows Updates application installed on its laptops to countermand user requests for updates, and thus prevent users receiving ANY software updates, including security patches?
Remember when Samsung's cack-handed modification of SwiftKey's auto-updater left a back-door malware installation vector on 500 million devices?
Remember when simply installing Linux on a Samsung laptop turned it into a paperweight?
Yes, I can see this being all Microsoft's fault, alright...
Richard Skillman worked on both Harmattan and Fastlane in his time at Nokia. Fastlane was developed for a hardware profile that was far more restrictive than the N9's - without multitasking, there's no point in having Harmattan's very nice task-switcher. On the other hand, Fastlane's timeline event view is an improvement on what was in Harmattan (and most other UIs)
N9 died for one reason only: With a couple of small exceptions, the mobile networks didn't want to offer it to their customers. If you can't get AT&T/Verizon/Vodafone/Hutchison/Telefonica/Orange, etc to buy your handset, it is dead in the market and there's no amount of advertising or technical superiority that'll change this.
" it just wasn't as simple to use as a grid of icons. "
I cannot understand this comment. The Windows Phone homepage IS a grid of icons. You can make some bigger or smaller, depending on your usage of them (according to Fitts's Law: if you want to hit something quickly, make it bigger on screen), and as a plus, those icons can display notification information if you make them big enough to do so.
The mechansim to customise the start screen is even the same as iOS: hold your finger down on any item. I don't know how this could be "less simple to use".
I would argue with some of Microsoft's terminology on Windows Phone, and all the "Metro" shells generally: "pin to start" doesn't automatically suggest "put this on main screen" to people, but I think the "Metro" start screen is still the best approach to a mobile-phone launcher we've seen to date.
(For task-switching, I still haven't seen a more natural mechanism than Maemo-Harmattan's swipe-from edge gesture on the lamented Nokia N9, but that OS's launcher was the fairly unimaginative grid-of-icons carried over straight from Symbian).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
