* Posts by Kristian Walsh

1125 posts • joined 10 Apr 2007

Page:

Jobs in Ireland may be vulnerable at post merger Dell Technologies

Kristian Walsh
Silver badge

Re: Race to the bottom

If you really wanted to avoid tax, you can do it without hiring 5,000 staff.

5
1

Is Dublin becoming as unaffordable as San Francisco?

Kristian Walsh
Silver badge

Re: don't forget the bubble in Cork

Cork's rental increases are very little to do with Apple. Most of Apple's "new" staff are on low wage selling and customer support call-centre jobs, plus junior clerical/accounting positions, and almost none get stock options. Everyone else has been there a long time, and isn't in the market for accommodation. Certainly, those new jobs did increase demand for housing, but they didn't raise the availability of money to pay for it: it has become harder to get an apartment, but rents have not shot up. (Also, Apple's location on the very periphery of Cork city makes it possible to live in one of the city's northern or western satellite towns and commute in)

In any case, you're assuming that Apple is the only major employer in this city, which is completely off the mark. The medical and pharmaceutical companies that account for most of Cork's FDI employ more people, and pay them better wages, than Apple does - as such, their hiring activities have a far more significant impact to the local economy.

On Dublin, the mention of Google, Facebook and company is a bit of a red herring (and Microsoft doubly so, as they've had a major presence in Dublin for over 30 years with no recent mad-dash to hire people). What I think is really fuelling the rush is the second-tier of tech startup companies, who can't match the prestige of working with a name-brand, so offer higher salaries instead.

These guys are using their VC's money to out-bid each other for limited developer talent. Those workers need housing, but the difference from Cork or other cities is that in this case, they've got money to overpay for that accommodation and drive out lower-waged residents. In this, the analogy to San Francisco does hold.

3
0

Shares down?! But, but, but ... Apple just made $50bn – that's the way the Cookie grumbles

Kristian Walsh
Silver badge

Re: Idiotic economic system

Because Apple, like most "tech" stocks, doesn't pay a representative dividend: $0.57 dividend a year on a $100.00 share from a company that's still making $40 Billion profit a year is pathetic (my bank pays me a better return on deposits), and for all the time that the iPhone was taking over the world, there was zero dividend payable.

Without an income stream, the only way that shareholders can benefit from owning the company's stock is to trade it, and that requires volatility. That need to constantly "flip" shares to earn money is what drives the short-term thinking of the market.

5
0
Kristian Walsh
Silver badge

Re: Interesting

Well, this Mac user gave up "holding out" last year. When I need to change my remaining Mac laptop, I'll be buying a Surface Book. The Ubuntu subsystem on Windows 10 is what sealed the deal for me, as well over half the time I spend on my Mac is inside a BSD shell (plus the Ubuntu VM I host on it for when I need a genuine Linux target).

2
1

Woz says 'Jobs started Apple for money' – then says it must pay 50% tax like he does

Kristian Walsh
Silver badge

He could be counting the indirect taxes that every citizen pays: VAT (GST/Sales/MWSt), levies and duties, fuel taxes are all government revenue. In the UK, HM Revenue collects nearly as much from VAT as it does from direct income taxation.

Corporations are exempted from these, as their purpose is to ultimately to collect these taxes on behalf of the government.

In that vein, there's also a very logical argument that corporations should pay no tax at all on their profits, because their purpose is to distribute profit to their shareholders, and you can then tax those shareholders on that income, and tax them at a higher rate than currently. (for Corporate shareholders, you just keep applying the profit distribution until you reach a living human who can be taxed on income).

Doing this would prevents the ludicrous situation where shareholders of Apple receive a piddling, miniscule dividend, while the company sits on hundred billion dollars of cash that it can't spend anywhere. (... until the US treasury department does another one of its once-a-decade corporate tax amnesties)

Such a system can only work if every country accounts for income and tax in a compatible way, to avoid double-counting or non-counting. The OECD "BEPS" guidelines are a start in this, but some nations are slow to fully implement them (and not Ireland, Luxembourg or Netherlands... think a little more star-spangled)

19
0

HTC 10: Is this the Droid you're looking for?

Kristian Walsh
Silver badge

Re: Don't believe the (CD) hype (HiRes)

I'd be fairly confident that the decision to not use 24-bit coding in Red Book was not a marketing one. I could imagine Philips's engineering teams would have mutinied if asked to implement such a high data rate into a product at the price-point of the original CD players (not cheap, but still, barely, a consumer product)

We tend to forget that this development work was done in the late 1970s. Moving to 24-bit sampling, while holding the sample rate constant would have reduced the capacity by a third down from 74 minutes to a 50 minute disc. 50 minutes doesn't sound too bad (still longer than a vinyl LP), but remember that the quoted 74 minute runtime of a Red Book CD was a maximum based on the track pitch - in reality, the very early playback equipment had difficulty tracking at the very edge of the very early discs, so early releases were kept to about 50-60 minutes to be safe.

Of course, 24 bits per channel instead of 16 means forcing every component in the transport mechanism to operate at a 50% faster symbol-rate than Red Book. As the 44,100 x 16 x 2 + error-correction was the cutting edge of what was possible in a consumer digital-optical device in the late 1970s, asking for 50% above that would have made the players prohibitively expensive, and the discs even more error-prone than they already were.

In fact, the original Philips proposal was to use 14-bit sampling in order to keep the overall system data rate, and thus component cost, down. Sony pushed for, and got, 16-bit, but consumer DAC linearity and noise levels at the time meant that you really only had 13 bits of faithful reproduction anyway. It was only a decade later, when properly linear 16-bit DACs became available, that Sony's foresight paid off.

Logarithmic DACs are harder to make than linear ones (it's easier to get linearity than "the right amount of" non-linearity), and while I think a 14-bit log would have done away with the need for 24-bit linear, we're really complaining that 1970s engineers didn't solve their problem with the technologies that arrived twenty years later....

Many of the problems of 16-bit are in mastering. Red Book audio is played on everything from the very cheapest of cheap cd-players to audiophile equipment costing close to six figures. The thing is, discs are mastered with the cheap players in mind, so you get everything compressed into the upper half (sometimes upper third) of the available dynamic range. This seems to happen at the final "mastering" stage, rather than final mixing - there are famous cases of unlistenably compressed CDs being available in a much more ear-friendly version as MP3 purchases (or, in the case of Megadeth's famously bad "Death Magnetic" album, as rips from the Guitar Hero videogame that featured some of its tracks)

24-bit material tends to be less aggressively mastered, and I think it's because it will never be stamped on a metal disc.

2
0

Apple Fools: Times the House of Jobs went horribly awry

Kristian Walsh
Silver badge

Re: Newton fishing

Newton was first spun out into a standalone subsidiary of Apple, called Newton, Inc (Claris was similarly split out); later, it was "spun-in" to Apple again, mainly so that Jobs could shut it down.

Newton's last product line was "eMate" a small laptop form-factor aimed at the US school market (and a forerunner of the design language that would appear with PowerBookG3/iMac/iBook). It did okay, and looked like opening a new market, but maybe someone in Apple saw it as a threat to Apple's own, higher margin, computer sales into that market.

But the big benefit of Newton was the investment in ARM Holdings. There were only two things that kept the company running through 1997: income from the ARM shares and the $400M settlement/investment from Microsoft.

I briefly worked on the "Tanzania" 4400-series PowerMacs - I wouldn't class it as a failure, though. By the standards of the time, it actually sold well. More importantly, it provided a lot of experience about reducing Apple's dependence on expensive components, particularly its reliance on SCSI for mass-storage; a year later, iMac arrived, with many of the same PC-parts-bin bits inside, just in a fancier case.

The hardware "failures" had all occurred before the likes of the 4400 came out. The real damage to Apple was done by the bewildering "Performa" range, where some were decent, and others were complete dogs, and even people inside Apple had trouble remembering which was which.

2
0

Microsoft to make Xamarin tools and code free and open source

Kristian Walsh
Silver badge

Re: @Kristian Walsh I don't see the difference and 'var' != 'auto'

C# is perfectly capable of running on other platforms. Don't confuse C#, the language, with Microsoft .Net, the class library for Windows applications. There is a portable subset of .Net, the PCL, which I have used to run the same tools on MacOS, Linux and Windows. I'd rather use C# than python or Java for middleware. Mono/Xamarin has provided a very solid cross-platform runtime for C# - people only seem to remember that the first releases of Mono were slow, but frankly, I remember the first JVMs, and they weren't too hot either. Time passes, software improves. I don't imagine there will be much difference in performance between runtime performance of C# on Linux and C# on Windows in a year or so.

(If you really need speed, you can compile C# right down to native machine-code if needed - Microsoft's Windows10 app store does this for you on Windows: you submit a C# CIL app, but the customer receives an optimised, native ARM or x86 binary when they download)

If I need high performance, I'll always turn to C/C++, but there are very, very few times when anyone needs the highest-possible performance, and it's only ever in very small parts of the code. I've about 20 years experience with C++, and I like the language a lot, but I would prefer C# for anything with a UI, or anything that interacts with a web-based backend. It's about choosing the proper tool for the job: the right tools make work faster.

Incidentally, you might look at the C++11 spec again: the new use of the 'auto' keyword does not stack-allocate (although that was its original meaning in C): when used on its own in declarations it now means nothing more than "compiler, please work out the correct type based on the rvalue". In fact, the default object-allocator in C++ has been to use heap allocation for a very long time. This is one fundamental difference between how C99 and C++ work, and it's one that many C++ users are completely unaware of.

Fundamentally, though, I don't buy into the "language X is faster, therefore better" argument. It's more true to simply say that you're buying runtime performance with engineering effort - the languages that produce faster execution demand more effort from developers. Sometimes that extra effort isn't worth it, as sometimes it obscures the purpose of the code, and allows bugs to creep in more easily. After all, engineering is about making things as efficient as they need to be, not as efficient as humanly possible.

I'm a big believer in Dijkstra's maxim that "premature optimisation is the root of all evil". I write my code with clarity and maintainability as the overriding priorities. Once I've got it functionally correct, I analyse the performance, and only then do I optimise. I've been programming long enough to realise that using a good algorithm will beats any amount of runtime speed-boost on a bad one, and also that where I think the bottleneck might be when I'm writing the code, is rarely where it turns out to be when the users get to run it...

2
0
Kristian Walsh
Silver badge

Re: Pointless

"Different form-factor layouts hare not fun to write in XAML. "

I'll admit I've done no WPF C# development, but UWP XAML uses size-triggered Visual States that allow you to do this really easily. (it's pretty much the same mechanism as CSS's fluid stylesheets).

However, I'll happily concede a variation of your point: different form-factor layouts are not fun to design. Keeping a non-trivial view coherent at the three or four window sizes needed is considerably more work than implementing the thing.

4
0
Kristian Walsh
Silver badge

Re: Fantastic

To people who say "Java is just the same as C#": that might have been true ten or fifteen years ago, but while Java stagnated the language, and concentrated on bloating the standard library, C# has improved the core language, year after year.

It's not the big-name fatures like LINQ or async, but rather the small things, like 'var', that make life so much smoother for developers, especially when so many libraries depend on generics, and so many classes have descriptive, but long-winded names. Compare these two functionally-identical snippets:

Java WhizzoContainerWrapper<ComplicatedIndexType, BlobWithUnexpectedBehaviours> c = new WhizzoContainerWrapper<ComplicatedIndexType, BlobWithUnexpectedBehaviours>();

APIThing.StatusEnumeration s = api.callMethod();

C# var c = new WhizzoContainerWrapper< ComplicatedIndexType, BlobWithUnexpectedBehaviours>();

var s = api.callMethod();

This isn't dynamic typing: c and s have the same types in both examples. The difference is that the C# compiler writers realised a long time ago that the compiler usually knows, or can infer, the type of pretty much any variable in a piece of code, so there's not much point in asking the programmer to remember it. (C++11's repurposing of the auto keyword does the same thing)

Basically, C# is full of handy stuff like this; Java isn't.

11
1

Computer says: Stop using MacWrite II, human!

Kristian Walsh
Silver badge

Re: Customer support, not Computer Support

There were two Microsoft Word apps of "that era": 5.x and 6.x. 5.x Mac was a really nicely written piece of software; in the same vein as Mac Excel, which I preferred to any other spreadsheet on Mac.

6.x, on the other hand was awful. Whereas 5 had been written to the native Mac Toolbox, 6 abandoned this and instead ran on top of a horribly-implemented library of Windows-like UI functions. It needed more memory, more disk-space, and it ran slower and had a UI that was completely unlike its predecessor: In the end, customers complained so much that MS "de-discontinued" the old Word 5.2 for Mac.

Okay, what they were trying to do was a good thing (the core logic of Word was the same sourcecode, and so completely compatible), but the execution left a lot to be desired. But it was the late 1990s, and this kind of "cross-platform UI toolkit" thinking was everywhere.

Personally, all my engineering docs were in ClarisWorks Writer. It did numbered sections, and it did a TOC and index. What more do you need? (These were the days when "reviewing your document" still meant printing it out, and then going through the red-pen annotations with the author in person, rather than festooning the document file with rectangles of cryptic text in the right margin and emailing it back)

1
0

Woman scales Ben Nevis wielding selfie stick instead of ice axe

Kristian Walsh
Silver badge

A good apology would be...

... a large donation to the local Mountain Rescue.

Despite what people think, MR is staffed by volunteers (albeit with occasional central-government funding for equipment) - they're risking their lives for no monetary reward, so you could at least do them the courtesy of being in an emergency that you didn't directly cause*.

For anyone else considering such a bloody stupid idea as this:

1. Don't. You could quite likely die.

2. If you insist, phone someone BEFORE YOU START, and tell them where you're starting from, where you're heading, and when you expect to be back. That way, when your brain stops working from hypothermia, at least someone lucid will realise that you've got yourself into deep shit.

but, most importantly:

3. Don't. You could quite likely die.

(* words aren't bad enough for the people who then sue their rescuers for "inadequate care" while saving their self-important lives from the results of their own stupidity)

32
4

Bash on Windows. Repeat, Microsoft demos Bash on Windows

Kristian Walsh
Silver badge

Re: Stealing VMS

Surely that name is short for nothing more than "NT OS KERNEL", no reference to OS/2 at all? But, yes it was cross-platform. I remember seeing NT running on Motorola PowerPC hardware way back in the mid-1990s.

Re an earlier poster's comment: For the record, Dave Cutler didn't "do work on" VMS. He was one of the three people who led its design and implementation. Big difference.

NT, the kernel, is a very clever design, and it's this clever architecture that makes it possible for Microsoft to shim the Linux syscalls so efficiently. The facility to do this was designed in from the beginning. Basically, you've got a core kernel, very much like a microkernel in that it deals only with abstract resources, above that, you have more concrete services, but unlike a true microkernel, these services run directly in the kernel address space. You get most of the speed of a monolithic kernel, and most of the stability and flexibility of a microkernel.

The flexibility extends to hosting multiple userspaces interfaces with the same kernel, which is what MS has done here. Here, a "Linux Personality Layer" effectively allows the Linux binaries to interact with the NT kernel without recompilation. The syscall interfaces presented by the personality layer are translated into messages for the NT kernel and the various Windows services running on it.

This isn't a new idea: QNX did a similar thing to get Android apps working on Blackberry devices (QNX is another kernel with very clear separation of concerns), but the fact that it's Microsoft doing it is more newsworthy.

9
0

'No regrets' says chap who felled JavaScript's Jenga tower – as devs ask: Have we forgotten how to code?

Kristian Walsh
Silver badge

Re: Are these dynamic dependencies really a good idea?

The naivety of "unless there are bad developers" is both charming and worrying.

23
0

We wrap our claws around latest pre-Build Windows 10 preview

Kristian Walsh
Silver badge

Re: Long File Path Support ?

It's not broken. Microsoft promised developers back then that the largest possible buffer needed for those file APIs is 260 characters.

Think for a moment what would happen if the OS suddenly started returning longer strings to those applications.

That's not the only file API in Windows, though; it's just the oldest one. Use a newer API and Windows will let you use file paths up to 32,768 characters if you're stupid enough to have designed such a requirement into your application (or masochistic enough to use an application with such a requirement).

14
0

Ever wondered what the worst TV show in the world would be? Apple just commissioned it

Kristian Walsh
Silver badge

Re: Worst TV show in the world?

They have, and so have I. And I still think they're right.

A good iOS dev story involves the protagonist being screwed, abandoned, fired and/or ripped off by Apple at some point. Can't see them putting that on screen.

29
1

How one developer just broke Node, Babel and thousands of projects in 11 lines of JavaScript

Kristian Walsh
Silver badge

"static" volatile dependencies

So, what you're saying is that your organisation's software development process can be stopped at any time by a third-party in a different jurisdiction. I'd love to have the kind of Programme Manager who'd hear that and say "Oh, the builds are broken? Because a guy in XYistan broke a module? And he's not answering his mails? That's fine. I'll tell the client that the service won't ship until an indefinite date in the future, and you guys can all go home early.."

The purpose of any build system is to produce repeatable outputs from your source-code, and to provide an audit trail for your releases. Repeatable is hard when you effectively do Lucky Dip dependency resolution. A build-system worthy of its name can check out any previous release of software by ID, and produce a binary-identical output product to that. A build process is language independent: you might need different tools, but using a particular language for development doesn't magically absolve you from responsibility.

Live-downloading isn't a "static dependency". "static" means "not moving", and you cannot guarantee that from a remote resource. You can barely even guarantee that if it's your dynamically-fetched resource. (Versioning components doesn't help you; you're still relying on strangers to not change code without re-versioning...)

So, if you're live-downloading every time you make a build, explain to me how you guarantee that those remotely-fetched dependencies don't dramatically change between the developer writing the unit tests, and your automated build system running them? There's a good way to waste development time. Also, how do you guard against someone maliciously injecting a backdoor into that crypto class you download every time you make a build.

More to the point (and this is the real reason companies spend money on revision control and build systems): Imagine it's next year, and you're being sued for doing something nasty, and to provide evidence of your innocence, you've got to set up a server with your company's software the way it was on the day of the alleged offence. How the hell are you going to rebuild it? Wayback Machine? Well done, you've just handed their lawyer the downpayment on a yacht.

ALL dependences used by a project must be accounted for. If you're not doing that, you're just wasting time and effort - you've got a glorified compiler/packager that offers no better consistency or auditing than just deploying straight off a developer's workstation.

10
0

Met plod commissioner: Fraud victims should not be refunded by banks

Kristian Walsh
Silver badge

Doesn't have the same meaning. "Encouragement" is a broader term, where "incentive" normally implies a financial encouragement.

"Disincentive" is also an acceptable word for discouragement that's achieved financially.

I don't particularly like the verbs formed from "incentive" or "disincentive", mainly because there are older, shorter back-formations of those nouns into verbs in the shape of "incent" and "disincent".

8
3

Don't – don't – install iOS 9.3 on your iPad 2: Upgrade bricks slabs

Kristian Walsh
Silver badge

Re: Disappointed, but now less so...

Unaccustomed as I am to defending Apple...

It's possible that the older iPads used a cheaper video subsystem without the necessary RGB gamma curve lookup feature needed to implement this.

Normally, white-point correction (which is what this is) is implemented by the video chipset. Just before the R,G,B bytes are sent to the DAC (or, these days, sent digitally to the panel driver), each colour-component value is looked up in a gamma table. The idea is that if your display has non-linear response to, say, the red signal, you can apply an opposing non-linear curve to the signal you send to that display, so that the result is effectively linear. For driving higher-quality displays, video chipsets can map the 8-bit RGB values to corrected values within a 10 or 12-bit-per-channel colour space.

Virtually all PC video chipsets support this function, but iPads aren't PCs, and don't use PC components. Tablets' LCD display drivers, so the theory goes, are calibrated in factory, so there should be no need for a programmable RGB adjustment before pixels hit the panel, so it's a natural thing to cut from your video module (it requires 256 x 3 words of registers that need to run at the speed of the pixel clock; no big ask, but why add the space to the chip if you don't need to).

In the absence of this hardware support, every drawing operation would have to be mapped, in software, into the "corrected" colour-space. That's another layer of "slow" that the older iPads really don't need.

Of course, it's equally possible that the iPad does have colour calibration support, but Apple deliberately don't activate it, in order to boost replacement sales.

4
0

Tech biz bosses tell El Reg a Brexit will lead to a UK Techxit

Kristian Walsh
Silver badge

Ireland (Re: Freedom of movement.)

Ireland already has a series of bilateral trade and free-movement treaties with the United Kingdom that predate the EEC, and many newer ones (e.g. Good Friday agreement) that are not conditional on EU membership. Brexit won't affect these. The fact that neither Ireland nor the UK has a land frontier with any other EU member would make it possible for the UK and Ireland to continue to honour those treaty obligations even if one nation left the EU.

The economic effects are generally considered to be neutral to Ireland overall: negative in the short term, but potentially positive long term as the UK effectively removes itself from the EU foreign direct investment competition. Short term, an exit will increase the administrative cost of trade, but not necessarily its volume. Ireland's main exports to the UK are in sectors where sourcing a non-EU substitute would be problematical (in order of revenue, those are: Pharmaceuticals/cosmetics/ingredients, then food/dairy, then plastics/machinery/electronics - only the last of these is easily sourced from other countries). Plus, aside from the big pharmaceuticals companies, these aren't big deals, but thousands of small to medium sized companies trading with each other.

Longer term, Ireland has a small possibly of benefitting economically from a Brexit, particularly in Financial Services, although not anywhere near as much as a Brexit would benefit Frankfurt. A City of London that's outside the EU loses one of its big advantages to external investors.

Now tho the "Stay" campaign's scare-tactics: In the event of an exit, the UK cannot rescind residency to EU citizens already in the UK - doing so is against international law, it is a bureaucratic nightmare (especially in a country like the UK that does not track the residency of EU passport holders) and it would also cripple the country's economy for a generation. Many could choose to leave, but they couldn't be forced to.

Ireland has a good lesson to give here on the dangers of sending people "back where they came from": in the 1930s, many of the Anglo-Irish were "encouraged" (often violently) to leave the newly-independent Ireland, and it resulted in a country whose economy was on par with Denmark's in 1922 falling to the bottom of the European league table by the late 1950s. (Northern Ireland's status as the UK's economic basket-case is similarly founded on the misguided belief that national identity is somehow related to effectiveness as a worker)

2
1

Windows Phone devs earn double what poor Android devs pocket

Kristian Walsh
Silver badge

Re: Average? Yes Average, but Median is what matters here.

According to the report itself, 51% of Windows Phone devs take in less than $1000 a month, so that puts the median pretty much at $1,000.

The $1000 mark comes up at 40th percentile for iOS devs, and 55th for Android. iOS has the most equal distribution of income bands (if we can call any app store's payout distribution "equal"),

Also, much as I'd like to dream about taking in five figures a month, the rest of the report makes it clear that earning this kind of revenue is the purview of the medium and large app studios, not the one-person developers. (73% of sole developers earned less than $1000 a month)

I'd have loved to see a the surveyors capture a "$0-$100 per month" revenue band, because my suspicion is that this would have covered 25% of all paid-app developers.

My gut feeling for the difference in average revenues is that Android an iOS attract more glorious failures, whereas Windows Phone is a way of extending the reach of an already successful product: publishers only bring a title to Windows Phone once it has proved that it can earn high revenues.

7
1

Hands on with Xiaomi's Mi 5: Great smartphone, but when do we get it?

Kristian Walsh
Silver badge

@Tim Warren [Re: Dual SIM]

It's possible, but impractical. First off, running two modems is not going to help your battery very much - you might have resilience but only for about six hours at a time (as it is, Android tries not to run Wifi and UMTS simultaneously); the long setup time for 3G connections means you have to run them simultaneously if you want failover. By the time you've upped the power reserves to allow for two data links, you've got a device that's not really practical as a hand-held.

Remember too that for 4G modems, each modem needs two physically separate antennas, and these perform best if they're far apart. So four antennas are needed.

The second problem is that most phone or even plug-in dongle modems don't make very good use of the signal they receive. They're sensitive receivers, but they can't push out enough power to reach a distant base-station, so you're losing out on throughput. Dedicated M2M modem modules, with proper antennas and a decent power supply are the way to go if you want better coverage, but that's even more awkward to carry around.

And once you've gone to that effort, it seems a shame to restrict that resilient data connection to a single mobile client - what if you've got a laptop and a tablet you want to use?

If you can accept in-vehicle mounting, then something like this: www.netcelero.com, attached to a wifi hotspot would do pretty much exactly what you're looking for, but trying to do this in a handheld package is not really worthwhile. (I've tried)

0
0
Kristian Walsh
Silver badge

Re: Dual SIM

To my knowledge it does not, but there are other ways to skin cats... What application is driving this requirement?

1
0

Meet Barra's baby: Xiaomi arrives with a splash

Kristian Walsh
Silver badge

Re: Smooth

For a hand-held device, a highly-polished, gloss finish offers better grip than a matt finish does. Against fabrics, the opposite is generally true.

I've seen the same thing as you, but only with the toughened glass used for phone screens. I usually leave my phone face-down when I'm in company, but my current device (Lumia 930) has such a slippery screen that I've seen it slowly slide off the arm of a sofa more than once, and now it stays face-up afterwards, if I have to take it out of my pocket for any reason.

2
1

Become an Andre Previn in your time: DevOps for star conductors

Kristian Walsh
Silver badge

"I'm deploying ALL the right services...

... but not necessarily in the right order!"

11
0

Women devs – want your pull requests accepted? Just don't tell anyone you're a girl

Kristian Walsh
Silver badge

Re: Hmmmm...

And does it also apply to men who declare that they're men

Well, as an experiment, I just used the "man select" command on my terminal. I now feel I know slightly less about network programming than I did beforehand.

...I think you could be on to something, there ;)

7
1
Kristian Walsh
Silver badge

Re: Peer review

It'll be the "Bob" submissions, because I'm assuming that the project maintainer is male and of English-speaking origin. Change "Bob" to "Vlad", and Alice to "Kate" and you'll see the difference narrow slightly - it's not just gender bias at work.

This is a phenomenon that's already well researched in the area of recruitment. When it comes down to a close-call decision, men favour male candidates and women favour females. To be clear: it's also well proven that both men and women will choose an obviously superior candidate, regardless of that candidate's sex. It's only when it's a case of choosing between two average applicants that the gender bias comes into play, but it's also worth noting that most candidates for a position will be of average ability...

The generalised point is that people are more favourable to others they have more in common with (sex, race, nationality, sexual orientation, place of education, membership of sporting clubs, etc, etc...).

This is why many large organisations now insist that interview panels are gender-balanced and as diverse as possible within that. It's very, very hard for someone to be completely impartial, so having a diverse range of biases and prejudices on a panel is what keeps the system fair to everyone... or equally unfair to everyone.

7
2

You've seen things people wouldn't believe – so tell us your programming horrors

Kristian Walsh
Silver badge

Re: Apple's cut'n'paste blunder

Not that I think Apple is using Linux, but the programmer may have developed that habit in a previous job.

Odd.. My dim recollection of the old Apple C coding guidelines was that you were supposed to use braces after every conditional statement, regardless of how many statements it contained. Apple's Swift language copper-fastens this rule by requiring braces around conditional blocks, so the spirit of Codebert is still alive somewhere.

As a bonus example of "belt but no braces", I once tracked a bug to this seemingly-innocuous "line" of (not my) code:

if (state==-1) *result=2; return 0;

(stuff like this still isn't a good enough reason to attach semantics to whitespace, mind you...)

1
0
Kristian Walsh
Silver badge

An excerpt from the Rogue's gallery...

I used to teach programming so I've seen some absolute howlers, but nothing I'm going to repeat here: nobody starts out as an expert, and it's only the ones who've forgotten that that get a kick out of mocking the inexperienced.

On the other hand, all of the below are from people who had been paid to do the job of "programmer" for at least ten years prior...

I once slopped-out an application written by someone who used this ghastly PHP idiom everywhere:

if (sometest()) { ?›

html markup here that gets spat out for true ‹? } else { ‹?

html markup here for false ‹? print "$other thing to insert"; ?›

more shitty markup . care to guess when it'll be shown?

‹? } ?›

A different person, same place, gave out to me for using descriptive variable names, like "foreach (candidateDevice in deviceList)" rather than "foreach (d in dl)". Apparently, in the 21st century, he didn't like code where lines wouldn't fit in an 80x24 vi window.

I didn't know anyone still had VT102s...

Perhaps a symptom of the same tunnel-vision, that same co-worker used to declare a small set of application-global variables in every .c file he needed them in. When doing that caused linker errors, he used "static" on his "declarations" to make the errors go away. Then the code started to show bizarre inconsistencies in internal state. Messages would be received, but they'd only have effect in some parts of the code...

But enough about them. The rest of these are all from my own œuvre:

The Python app where I'd mistakenly typed "self.dchp_range_start = " in two different places, then wondered why the settings only took effect in some cases.

And, to show that static type checking isn't a panacea, here's a somewhat over-zealous gating check for IP4 addresses:

if (servers[i].find_first_not_of("0123456989.")!=npos) continue;

There's also the time I tried to save one whole byte in a money handling application by making a change counter eight bits wide. The first time someone needed to be refunded £25.60, I didn't feel so clever. (Actually, I wasn't there by then, but I certainly did hear about it...)

I also once got so sick of a linux video capture driver's inability to take a still that I resorted to successively grafitti-ing the video RAM to chase down where the camera image was being dumped, then when I needed a still, I sent a "picture hold" command to the video camera, and read out the video memory, parsed the YUV colourspace into RGB and wrote the file to disk. Okay, that was a pretty skanky process, but the code doing it wasn't.

Also, not actually a bug, but I'd feel sorry for the maintenance programmer of this little gem if they'd got it without the accompanying comment:

if ( x!=x ) { // (floating-point NaN: it never equals anything, itself included)

}

..

3
0

Little warning: Deleting the wrong files may brick your Linux PC

Kristian Walsh
Silver badge

Re: Sounds Really Clever?

It's nothing to do with systemd. Nothing. The driver isn't trapping the case where the user wipes the objects that the driver exposes through the /sys/ hierarchy. You don't even need rm -rf / for this; you can just cd into the appropriate directory and issue rm -rf *

This is a longstanding "problem" with the Unix permissions model. "Write" always implies "delete", but these are actually separate permissions, and other, non-Unix-derived OS treat them as such. Actually, Unix is the odd-one-out here, but its ubiquity makes people think that its behaviour is the norm.

For comparison, VMS had "Read, Write, Execute, Delete"; Windows NT unsurprisingly maps the VMS permissions to "Read, Modify[=Write], Execute, Full Access[ which allows Delete]" - on both these systems, being able to write means only that: you can open the object, and modify its contents, even truncate it, but you can't remove it from the filesystem. Even CP/M's limited model distinguished between deletion and writing: the two permissions settings possible on a file were write-access [RO=off] and deletion [System=off].

19
7

Can't upgrade, won't upgrade: Windows Mobile's user problem

Kristian Walsh
Silver badge

Re: Hmmm...

Being on the insider list doesn't mean that the performance was good enough to get onto the release list.

520 is supposed to be on the "second wave" of W10 Mobile updates, but that doesn't mean it'll get the full Windows 10 Mobile release. Insider user reports I read weren't positive about the low-feature devices: the big problem with updating the 520 isn't the running requirements of the OS, but rather getting enough free storage space on the device to store the OTA updater.

On performance, W10 renders 8.1 app UI faster than 8.1 does, and unlike 8.1., the Windows 10 Store service recompiles the .Net CLI byte-code submitted by developers into native code for download onto the device. Both of these make apps run faster, and there are some API features introduced in 10 (particularly in allowing the use of static binding of control properties) which dramatically improve UI load times if you use them. On the other hand, the W10 OS is doing more in the background than 8.1 was, so overall it's a small performance hit.

Memory is the killer on these devices, though - I've a 530 that I used for developing my recent app, and running on such a low-memory device really taught me to aggressively optimise my heap size.

5
0

Safari iOS crashing: Suggestions snafu KOs the Apple masses

Kristian Walsh
Silver badge

@Bad Beaver, Re: Another mark on the list

10.6 is an interesting one to pin as the "last good OS X" (I tend to agree, btw), because this was the last OS X release headed by someone with long experience at the company.

Bertrand Serlet was a platforms manager who had come to Apple from NeXT, and he and Avie Tevenian were instrumental in getting OS X out the door in the first place. Tevenian left in 2006, and 10.6 was Serlet's last OSX release, as he left in 2009. His replacement at OS X, Craig Federighi, only rejoined the company in 2009. Not much time for a handover...

2
0

Brit censors endure 10-hour Paint Drying movie epic

Kristian Walsh
Silver badge

Precedent...

There's a story that Andrei Tarkovsky's 1971 film "Solaris" contains a seemingly interminable scene of a trip through city traffic in purely to bore the Soviet censor into turning it off and signing the approval.

(The scene in question was filmed on Tokyo's urban motorways, and the more plausible reason for it is that including so much of this Japanese footage in the final film was to justify the great expense incurred by Tarkovsky and company to travel to Japan to shoot it. Thanks to Soviet efficiency, they had missed the World Fair, which was to have provided the original "future cityscape" imagery for this part of the film.)

5
0

Waving Microsoft's Windows 10 stick won't help Intel's Gen 6 core

Kristian Walsh
Silver badge

Re: Bassackwards

""It wasn't my assertion,"

Maybe you should have put quotes around it then."

No, you misunderstood what I wrote. I am not the person who wrote the post containing that assertion. Thus, "it wasn't my assertion".

You're still confused about the difference between the price of media and the software licence. You do not own OS software - the producer simply grants you a licence to use it. (Linux people stop sniggering - you also don't "own" your software; if you did, you'd be able to resell it under whatever licence terms you wanted). In this case, the producer is Apple, and the terms of the licence are on that panel that you clicked through without reading. Here, have a look at it now: http://images.apple.com/legal/sla/docs/OSX1010.pdf

The licence states several times that you are only permitted to use it on "Apple branded hardware". If you run OS X on a non-Apple product, you are in violation of the licensing terms, and have no right to use the software. It doesn't matter that you've paid Apple money, you are still in breach of the terms of the licence, and if Apple ever found a way to make such an OS X system stop working, they would be entirely within their legal rights to deploy it.

You can buy it separately, it's no different to Windows or Linux.

Okay. Here are the two fundamental differences that you're missing:

1. You can buy Windows or Linux licences and use them on any hardware of your choice - if you get it to install on a system, the licence permits you to use it on that hardware. But: You cannot buy an OSX licence and use it on any hardware of your choice. Only installation onto Apple-branded hardware is permitted. OS X is legally tied to Apple hardware only.

2. You can buy or build a bare PC system, and then decide what OS you want to run on it. But: every single hardware model that Apple permits you to install OSX onto is already shipped from the factory with that software pre-installed, and so the cost of developing OSX is included in the device price. OS X is bundled, without choice, onto every device that is allowed to run it.

2
0
Kristian Walsh
Silver badge

Re: Bassackwards

It wasn't my assertion, but do you honestly think that OS X "costs" £14.99? That's a service fee for getting the physical media - Apple's licence terms make sure that you can only use that disc if you've already given them money for a computer. OS X is bundled with their hardware, you can't buy it separately the way you can Windows or a professionally-supported Linux distro.

Also, the OS X disc you mention is OS X 10.8, a release that's no longer officially supported by Apple (let's not get into Apple's "love 'em and leave 'em" attitude to OS X lifecycle management). You need to install that, then update to 10.9 and then -- if you're a masochist --- 10.10 and 10.11.

Getting back to the article topic, though: the biggest problem for the PC business can be seen by comparing the Minimum System Requirements for Windows 10 with those for Windows 7, way back in 2009... I'll save you the search: They're identical. That's six and a half years and two-and-a-half OS releases later, and there's no need to buy new hardware at any of them. (I put that down to Microsoft becoming a Windows Hardware OEM itself, and having to directly experience the pain of getting Windows 10 to run acceptably on the kind of low-power CPUs it used in Surface)

As an OS X user, I can't say the same applies to Apple. Not by a long way. 10.6 and 10.9 might compare favourably, but you'd have had to stomach the horrendous 10.7 and 10.8 to get between them, and things have gone seriously downhill again after 10.9 - not just in performance, but in basic quality. I could put up with OS X being "sedate and unhurried" in its performance, but not when it combines that with the sort of flakiness I used to associate with Windows (and don't start me on what they did to "Save As..."). I've skipped 10.10 and 10.11, and don't hold out much hope of 10.12 being better.

2
0

Recall: Bring out yer dead and over-heating Microsoft Surface Pro power cords

Kristian Walsh
Silver badge

Re: SW company?

Remember that this is about power supplies. NO hardware manufacturer fully designs and builds their own power supplies. The manufacturer specifies a device connector and the required DC parameters, and the rest is sub-contracted.

The problem here sounds like it's in the strain-testing of the cables. There's an acceptance test for these things that flexes the cable repeatedly and then checks for breakage, but humans find all sorts of clever ways to damage cables, especially cables that get stuffed into luggage in a hurry.

And for all those claiming that this is somehow a sign that Microsoft is worse at hardware than the rest of the industry, you might do a search for Apple's power-supply recalls over the years...

1
0

It's Wikipedia mythbuster time: 8 of the best on your 15th birthday

Kristian Walsh
Silver badge

Re: There was always a near monopoly on encyclopedic knowledge

"As accurate" is an interesting way to describe findings for a study that found that while Brittannica and Wikipedia were equal on the (low) rate of major misunderstandings of the topic found, Wikipedia had significantly more factual errors per article, or to quote it:

averages out to 2.92 mistakes per article for Britannica and 3.86 for Wikipedia

So, "30% more errors" is the same thing as "as accurate". I'll remember that if I ever need a blood-test.

Thing is, Wikipedia's problem isn't in facts, it's how it weights those facts. Too many articles fail to show what's important about their subject, and what's just inane trivia (or vanity-linking to some postgrad's supervising professor's paper on the subject).

This lack of editorship pervades Wikipedia, and it's causing a slow decline in the usefulness of the site as more and more crap accretes to the articles with every passing year.

20
0

Longing to bin Photoshop? Rock-solid GIMP a major leap forward

Kristian Walsh
Silver badge

More than three colour channels?

Lack of support for four colour process or spot colour layers is what has kept gimp out of professional users' workflows, so it's remiss of this article to not mention that the imaging engine is no longer limited to RGB and RGBA, and now images can be any number of channels of integer or floating point data points.

... It's quite an impressive bit of work, and congratulations are due to all involved- this is the kind of thing open source development excels at.

Whether the UI exposes this is another matter. (UI design is, sadly, something that Open Source projects struggle with)

53
1

Periodic table enjoys elemental engorgement

Kristian Walsh
Silver badge

Re: Unobtainium...

It's precisely because nobody speaks Latin that it's used so much for naming. The idea is to have a language that everyone can agree does not favour any one country.

Look at Switzerland: it has four spoken languages, each with its own vocal lobby: even deciding what to put on the postage stamps was a contentious question. The solution was to use Latin, as it belonged to no group at all. So, the stamps say "Helvetia", and the car-stickers read "CH" ("Confœderatio Helvetica" - "[the] Swiss Confederation"), as does the ISO Country code and thus Switzerland's top-level domain name.

Latin's other advantage is that because there are no native speakers to object, its pronunciation has, over the centuries, been knocked into a form that's simple enough that speakers of pretty much any other language can master it.

5
0

What did we learn today? Microsoft has patented the slider bar

Kristian Walsh
Silver badge

Re: *Design* Patent

The equivalent in UK law to this is a "registered design": legal protection for the appearance of a product. Some people just see the words 'Patent' and 'Microsoft' and go nuts..

This is a look and feel suit, nothing to do with Software Patents at all... and further evidence that EFF has lost touch with reality.

5
11

Review: Star Wars: The Force Awakens offers a new hope for the franchise

Kristian Walsh
Silver badge

Re: I don't really care whether it's any good, personally

Odd that you think diagonal wipes are a 1970's styling cue. You're about 40 years out. Star Wars was a homage to the 1930's space serials like Flash Gordon... Lucas tried to get the chance to remake FG itself in the mid 1970s (that this was possible at all shows how little the studios valued scifi/space opera as a genre at this time).

I saw this one. It was a very entertaining two hours. The cast, director and writers made me care about the characters for the duration, and they knew when it was time to have the actors shut up and let their actions and the pictures talk instead (a rarity in modern cinema, especially in anything with a 'fandom').

If I have a complaint, it's the unmemorable score, but maybe it was more that there was a film going on as well: the 'prequel' films were such tosh, the music was all they had going for tihem (Episode 2 has some really good themes)

Yes, it's a hung on the frame of the 1977 film, but if someone thinks that that was some kind of revolutionary dramatic structure, they should pick up a book on myths and legends (any people's will do ... SW lifts from pretty much all of them).

0
0

Windows 10 won't come to old WinPhones until some time in early 2016

Kristian Walsh
Silver badge

Re: Just got a 950XL

If you bought your handset direct from Microsoft, make sure you've got the Lumia Offers app installed (go to http://aka.ms/lumiaoffersapp if it's missing). There could be a nice surprise in it for you...

My wife got the 950XL a couple of weeks ago, and it's mostly very impressive. The camera is the best imaging package ever put in a phone (I am including the 808 and 1020 here - their image quality might be objectively better and better for post-processing, but the 950 pair's whole package of sensor, stabilisation, intelligent HDR and colour-balanced flash produces the best images with the least amount of effort, so the newer phones just shade them).

Against that, the default apps shipped with Windows 10 for Phones are still unfinished in places; updates are arriving quickly enough (the "People" one lost most of its cosmetic bugs yesterday), but I can't help thinking that the development schedule is still ongoing for this release, and what's on the 950 and 950XL is a beta.

However, all the negatives are software issues, and the various fixes are coming quickly. One of the big changes with Windows 10 is that mobile operators don't have to certify "firmware" updates like they had to on 8.1, so now once a phone ships with 10, it will get updated automatically as soon as the software is released (just like Apple's iOS device updates aren't gated by operator certification).

1
0

Google chap bakes Amiga emulator into Chrome

Kristian Walsh
Silver badge

Re: Ahh, Amiga

Pre-emptive multitasking can work just fine without memory protection. Provided every other application and process doesn't vomit over your memory. Sadly, this was never true in my year-long experience of owning an Amiga. I traded it back in against an Atari 520STE: not as clever, no multitasking, and the games were nowhere near as good, but at least it didn't crash unless I made it crash (ah, the joys of 68k assembly!)

This isn't really Amiga's fault though - there was simply no hardware to deal with the inevitable instability that comes from multitasking. I seem to remember that the 68000's companion MMU was a bit of an afterthought, so nobody used it, and in any case the 68000 itself had no way to properly recover from a bus error, so even with protection, a process reading from an out-of-address-space location could still hose your system. (68010 allowed the system to restart after memory fails)

Ah, the old days...

2
1

Mobile developer report shows growing back-end challenge, weak Windows support

Kristian Walsh
Silver badge

Bear in mind that this is a survey of users of Appcelerator's toolkit. This immediately self-selects the respondents based on how well Appcelerator supports the various platforms: after all, you wouldn't choose a tool that's sub-optimal for Android if Android was your primary development target.

4
0

All eyes on the jailbroken as iOS, Mac OS X threat level ratchets up

Kristian Walsh
Silver badge

Re: This oculd be due to the popularity of windows...

Perhaps you meant "VMS and BSD", but that does undermine your argument a little. There hasn't been any DOS in Microsoft's OS products since Windows XP came out, whenever that was (I was still Mac-only in those days). The NT kernel was modelled on VMS.

I use both OSes daily. There really is no difference in privilege escalation between OSX and Windows. Processes simply cannot get above their station anymore on either OS, and must ask the user for the permissions they seek.

The vast majority of malware doesn't "crack" a system, it fools the user into handing over the keys. This is why tainted installers are so useful as a vector - users are less concerned that an application installer asks for temporary admin privilege. (MacOSX is the same as Windows here).

In my experience of cleaning up after this stuff (and a brief period working for an anti-malware company), a lot of users will accept any kind of unexpected privilege escalation if they think they're getting something for free.

Windows is still a more lucrative target than OS X, simply because Windows PCs are more likely to be in business-critical functions in small businesses. A cryptlocker in an small accountancy practice, two days before the filing deadline is way more likely to yield a payout for the malware writer.

24
1

National Crime Agency: Your kid could be a nasty interwebs hacker

Kristian Walsh
Silver badge

Re: So on the one hand....

The checklist/questionnaire reads like the nonsense it is if you substitute "photography", "dressmaking", "playing guitar", "jewellery making", "singing", "bicycle or car mechanics", "drawing", "baking", "carpentry", "pottery", or basically any other hobby that a teenager might have that could yield them some occasional extra cash.

The general rule is simpler: if your kid is coming into a lot of money, and they can't adequately explain where it's coming from, you need to consider the possibility that they're engaged in some form of illegal activity.

At a time when the UK gov. is desperately trying to get children into "Coding", it's funny to see another government agency trying to scare parents into dissuading them from doing just that.

5
0

Donald Trump wants Bill Gates to 'close the Internet', Jeff Bezos to pay tax

Kristian Walsh
Silver badge

Re: pull the wool off his supporter's eyes

That would be fun for the family returning from Haj etc.

Hajj? Nah... I'd like to see him go onto a transport plane full of U.S. Marines returning from a tour in the Middle East and try to tell them that some of the guys they've been fighting beside for the last six months aren't going to be allowed off the plane...

17
1

IETF's older white men urged to tone it down

Kristian Walsh
Silver badge

Re: Yeah, right.

"Why we have the internet today" is down to the work done in the 1970s and 1980s, which the authors of this RFC state was a time when IETF was more representative and inclusive than it is today.

Technical organisations should hire the most technically able people. There appears to be a problem in IETF that no applicant with two X chromosomes is considered to have technical ability, regardless of their accomplishments. This is failed logic.

4
1

Sketch dev pulls out of Mac App Store, cites slow reviews, tech limitations

Kristian Walsh
Silver badge

Re: "App Review continues to take at least a week"

Apple could easily rate-limit developer submissions to prevent such abuses. Spending a week to review a minor code change to an established application is ridiculous.

However, the bigger problem is upgrade pricing, I'd imagine. This is the big flaw in the "app" model from a developer's side of the deal: once a customer buys your app, they must get every new feature for free. Those features, unfortunately, do not cost nothing to implement.

As an example: if an iPhone user bought a utility app like PCalc (Hi James!) back in 2008, they still have it now, for free, on their shiny new Apple Watch and iPhone 6s. Seven years of a developer's time means the current one is far more useful than the original, but the user has had all of that for free.

That's just about manageable for an "indie" developer with low overheads, but if you're a larger company, with premises, staff, and customer support costs, you cannot operate on a model that prohibits you from recouping your product improvement investments.

8
0

Meet ARM1, grandfather of today's mobe, tablet CPUs – watch it crunch code live in a browser

Kristian Walsh
Silver badge

Re: Predicated Execution

I didn't downvote you, but tagged memory is still a Von Neumann architecture - the architectural principle is only that the program and data occupy the same addressing space.

Being able to overwrite program with data and vice-versa is not a required feature of a Von Neumann architecture; despite being a side effect of most implementations.

1
0

Page:

Forums