My initial reaction was that, come the European launch, if the iPhone doesn't have 3G/UMTS/HSPDA then it would be laughed out of court. However, on reflection, it sounds as if Apple's attitude is:
GPRS is good enough to check your EMAIL and gives good phone coverage. If you want a decent web-surfing experience on the train, subway or in a coffee shop, your best bet is if some bright spark has installed WiFi. So lets do a phone which makes a much better job of doing WiFi than the competition and not weigh it down and waste battery life by putting in 3G capability. We're Apple - maybe we're influential enough to put some momentum behind WiFi coverage.
(PS - am I right in thinking that EDGE/GPRS and 3G/UTMD/HSPDA are two incompatible "family trees" of protocols, and a phone that supports both needs a certain amount of duplicate "gubbins" inside?)
I have a (UK) phone that does GRPS*, UMTD & HSPDA*, bluetooth and WiFi and while 3G coverage here is ok (and HSPDA being rolled out - and very nice when you can get it) it wouldn't be much good without GRPS as a fallback. You certainly can't use the internet reliably on a train (I've tried - and did manage to send an EMAIL from a train here but it was a labour of love and certainly wasn't HSPDA!) The phone (MDA Vario II - AKA HTC TyTan) is a bit of a brick and it certainly doesn't flip seamlessly between WiFi hotspots (cough)WM5(cough).
PS - real Apple Fanbois should, of course, equip themselves with a backpack containing a laptop with a HSPDA data card and a compact WAP. Then they can whip out their iPhone and impress people anywhere with HSPDA coverage...
(* OK - PCMCIA, as they say, so I'm trying out multiple permutations for the ETLAs and DETLAs here:-)
OEM has EVERYTHING to do with it. In my business (education technology), we can't seriously consider the choice of switching to Macs and run WinOS because we'd have to buy an XP or Vista license for every Mac.
Nope - OEM is irrelevant to the anti-VM clause issue. The OEM versions of "ultimate" and "business" which allow VM use are still a fraction of the price of corresponding full retail versions. AFAIK Apple could offer those if they desired. Odds are, if you are ordering in bulk for (probably networked) systems, you didn't want the "home" edition anyway. If you were buying in quantity, you could probably get a volume licence (again, they probably don't even bother to do that for "home"). As I said, if the whole plan was to stop OEM + VM use it would have been much simpler - and almost reasonable* - to put the clause in the OEM license.
Apple is technically an OEM, and could offer copies of Vista at a discounted price. 'All of this paints a picture in which Apple could use OEM pricing to offer Windows for its Macs at greatly reduced prices and running in a VM.
First, the regular OEM price isn't a "greatly reduced price" - its the de-facto going rate for Windows (exactly what price the big boys, Dell, HP etc. pay - and what "Important considerations" they offer in return is another question).
Secondly, the current OEM license already has some major extra restrictions about how and where you can install it - with crossed fingers and a following wind you could probably argue that these already blocked VM use - after all, where would you stick the hologram?:-) They pretty clearly mean that joe user isn't supposed to buy an OEM copy and install it on his VM. If MS had changed the OEM rules to explicitly block "VM bundling" then I doubt it would have raised the sort of negative publictiy than claiming that you can't even use a full-price version on a VM has.
Thirdly, I stand to be proven wrong by posterity, but the only circumstances under which I can see Apple bundling Windows is if they throw in the towel, drop OSX and become a vendor of designer Windows PCs with an iLife-for-windows bundle (while plausible, I'm sure that ain't exactly Plan 'A'). There's a huge PR difference between the current stance of "hey - if you want to use our hardware to run Windows here are some solutions" and saying "actually, so many people can't get buy with OSX alone that we've decided to bundle windows". The latter comes under the heading of "pulling a Ratner*". I'm sure Microsoft would love to be thrown into that briar patch.
Anyway, if you've actually tried to use Parallels (or, by reputation, VMware) you'll know that while they are excellent products, they don't yet come close to the Apple "it just works" standard (on which Windows sets something of an upper limit, anyway!). Coherence is very effective, and certainly solves the screen real-estate issues with running apps inside a VM window - but to say it "makes Windows apps look and work just like OSX" is rather optimistic.
Nope - explanation "A" is that TFA was warm - this policy makes it cheaper to buy a Dell than a legit copy of Windows Vista for your Mac or Linux box. Explanation "B" is that VMs can be used to circumvent DRM and that forcing pirates to buy a full-price copy of Ultimate gives Microsoft enough plausible deniability to stop the MS board from being carted off to G'tmo under the DMCA. Of course, Explanation B only makes sense if you are sufficiently far down the DRM rabbit hole to think that evil terrorist-loving pirates are going to lose sleep over breaching their windows EULA - or will even relate the concepts of "buy" and "windows" , but then...
(*Sorry - I've googled "pull a Ratner" and it might not mean in US English what it means in English English, which is "publicly describing your own product as 'crap' and then wondering where your business went")
The emphasis, as I read it, of Dick's novel was that no matter how real something seems, it is never as good as the real thing. No matter how realistically a replicant could look or act, it would never - ever - really be human.
Strange - the take home message I got from the book was "what is the difference"? The lives, ambitions and purposes of the humans in the book were just as synthetic and imposed as the replicants fake memories. By the end of the book Deckard is pretty definitely real, but most of his life has been exposed as fake and/or a control method: the fake religion, the "mood organ" that lets you adjust your temprament to fit in, the fake DJ, the social pressure to keep an animal (even if its fake) the entire fake police station...
For me the biggest thing missing from the film was the details of the "Mercerism" religion, which make it clear that the VK test used to identify "andys" is mainly a test of religious orthodoxy, and the significance of replicant's inability to take part in the empathic "communion" that is central to the religion (and later revealed as electronic trickery).
...we can't seem to do a good job of controlling problems with climate, etc. in our own world - shouldn't we focus on that first?
Aha! Mars has an atmosphere that is nearly ALL CO2 yet it is freezing cold! So much for the pinko liberal commie "global warming" myth! (Ps - JOKE!)
Seriously, even if we never went through with it, planning and simulating the terraforming of a planet would be a pretty useful learning experience for the human race... and if we did pull it off then, TADA! backup! Why not play around with Mars? OK, mabe a bit more exploration is in order to make sure that it really is just a rock with, at most, a few bacteria, but unless it is one giant superintelligent hive-mind of bacteria living in a few mm of icy soil (...now, where did I see that one - Star Trek?) there doesn't seem to be any intelligence or biodiversity. Truly barren rocks and deserts don't contribute much (problem is, on earth, even apparently "barren" rocks tend to have thriving ecosystems).
I don't see any point in avoiding your example of "security by obscurity kludge" when the whole idea of the motherboard/chip/OSX restriction is security by obscurity in the first place.
OK, so all DRM is broken, because while cryptography is the advanced science of getting a message from Alice to Bob without Eve being able to intercept it, DRM is a futile attempt to stop Bob then handing the message to Eve in return for a good seeing-to. That's a given. I wasn't defending it.
My point was that a cryptography-based system can use an open, stable protocol which a hypervisor could "proxy" - with clean hands - and let OSX satisfy itself that it is being hosted on a Mac. Anybody found reverse-engineering the ID chip or hacking the software can then be carted off to Guantanamo under the DMCA, so that's OK. If its a kludge, the hypervisor would have to know the secret and if its just a collection of hardware characteristics that Mac hardware happens to have it could be (a) very difficult to emulate and/or (b) hard to keep secret.
I can understand that Apple will probably never allow this though, because the developer is not the target market.
Apple's main concern is, presumably, to stop people running OSX on PC hardware. Currently, any OSX running virtualized must have had the hardware limitation hacked out. That would require Apple, VMWare & Parallels to get their heads together to sort out a solution - which also assumes that they are using some sort of strong cryptography-based system with a key held on the motherboard somewhere that the hypervisor could securely "proxy" - rather than some security-by-obscurity kludge.
Microsoft gives you at least a (costly) option. Apple (correct me, if I'm wrong) doesn't.
True-ish, and Apple certainly can't chuck any bricks in that particular greenhouse. However, there are a couple of mitigating factors:
Apple do not have a 95% monopoly of the desktop market. If you don't like Apple's policy, vote with your feet (sounds like you already have). OTOH the group who are disadvantaged most by MS's policy are those who don't like Windows and are trying to switch to Mac or Linux but - because of the MS software monoculture - can only do so if they still have a way of accessing Windows.
The issue of virtualising OSX is in a chicken-and-egg state - I don't know if VMWare or Parallels support EFI (needed by OSX) or support OSX's graphics requirements (of course, no one reputable will admit to having tried it).
I don't think there's a huge market for it (once you dismiss the "I want to try OSX on my PC" brigade) - the big demands for windows-hosted virtualization come from the developer and server consolidation markets. The Apple world doesn't have the huge army of in-house developers that buy VMWare Workstation , and I'd guess that OSX Server is used predomniantly for high performance file sharing, render-farming etc. - not the sort of things you virtualize. The money in OSX-hosted virtualization is from users who need to run Windows. (Cross refernce with above point). I'm guessing they only support other non-MS guests because they were already supported by their existing windows-hosted products.
If Apple doesn't sort this out soon they're going to start hacking off developers - virtualization is so darn useful. This will come to a head when 10.5 is released and betas of 10.6 go out and developers have to juggle past, present and future major versions of the x86 based OS - but the initiative will have to come from developers, via Apple - Parallels and VMWare have no strong incentive to break a sweat over it.
P.S. Also bear in mind that the last thing Apple want is, officially or otherwise, a "try-before-you-buy" route for OSX: even if the implementation was non-flakey, the first impression of playing with a new OS is always frustration because of the differences and the fact that your instinct is to plunge into "clever stuff" rather than work through the basics. Better if you are sold on the idea by an evangelist, part with cash, and have a $2000 incentive to get over having to press the fricking pretzel key instead of "ctrl".
Evidence 2: How Microsoft explained that they changed their mind back on virtualization of Basic/Home? "The company said virtualization presents inherent security risks". Oh... My... God... They aren't even TRYING. What kind of damn security risk are we talking about?
I think we're talking security risks of the "customers might be able to circumvent DRM" using VMs which (e.g.) make it easy to change the MAC on your ethernet kind.
That people will buy cheap Windows Basic and run it on Parallels on Mac, isn't that the one. Pathetic.
Hey, that would certainly make MS feel insecure. Why do you persist in assuming that they're talking about our security?:-)
Basically, they are good at giving you a broad, objective categorization of individuals.
Unfortunately - in many fields (particularly school/college) the test is also the de-facto definition of the curriculum. It would be nice if 'twere not so, but its pragmatically inescapable. Writing a psychometrially valid test is one thing - writing a psychometrically valid test that also exemplifies how the subject should be taught is a bit of a toughie - and might mean sacrificing some of the psychometric rigour.
The real problem occurs when you have the people who understand the subject in silo (A), the people who understand teaching in silo (B), the psychometricians in silo (C), and - for computer based testing - the programmers in silo (D) all passing messages via the Administration. I really, really, really wish that I'd never seen that happen.
This is how pro tests get to be so good and so reliable.
As long as you realise that the relationship between "reliable" and "good" is one of correlation, not causation.
Ensuring that a test is measuring some factor in a stable and reliable fashion has, indeed, become a pretty exact science. The danger is that all the impressive statistics this produces will give a false impression of validity and detract from the massive inference that this some factor is the thing you are actually purporting to measure - and, of course, if the test is vital, it will Heisenberg the educational process into one that teaches some factor.
TFA mentioned bar exams. Now, IANAL but I'm pretty sure that representing someone in court does not consist of answering a battery of multi choice questions covering the entire spectrum of jurisprudence. How reliable the tests are - year on year or against other (probably similar) tests does little to support the assertion that the test measures what it claims to measure.
I've encountered math questions where, when you actually observe and question schoolchildren working on them, the actual thought processes involved have nothing to do with the what the question purported to test - however, the bright kids cope better and the psychometric analyses see no problem. Best example was a computer-based question for young kids about telling the time where, on observation, sucess or failure clearly depended more upon fathoming out the question's user interface than telling the time. Looking at the psychometrics gave the question a clean bill of health. Other questions on high stakes assessment test math but in a inappropriate way - e.g. probability questions that require you to multiply two probabilities that, in context, are clearly interdependent; statistics questions that require you to say that statistically insignificant data prove a hypothesis; plus loads and loads of little isolated techniques tested separately with no requirement to "synthesize" them into a credible application.
Psychometric analysis is a valuable tool - when combined with other measures, acknowledgement of the underlying axioms and a healthy dollop of judgement and experience - but what it offers is so seductive in today's pseudo-accountability, management-by-algortithm environment that it can obscure the truth: testing is an inexact science for reasons that go far deeper than the statistics and we should not place so much cultural emphasis on the differnce between 49% and 51%.
Computer-adaptive testing! Another sexy subject!
Basically, all tests really need to be good is a lot of data.
He should have stuck with the more useful observation that almost* any test with a very low pass rate will be unreliable.
All tests have a margin of error, although its a rather taboo subject - when did you ever get a test result that stated the 95% confidence interval? If only a small proportion pass, there is a danger that these errors will dominate.
There are:
simple "cockup" factors like mistakes in marking, or miscounting of marks;
systematic factors like the artificial stress of an exam affecting different people;
sampling errors - you can't test the whole subject, so a proportion of candidates will be "lucky" and get questions that they've drilled for. A low pass rate suggests that the domain you are sampling is sparsely populated and, therefore, that sampling is invalid. Someone who "passes" such a test has only shown competence on a few specific topics.
deeper flaws in the whole concept of an exam - such as the potty notion that you can represent ability in a complex subject by a single number, and other sweeping assumptions about independence and normal distributions made by the statistical methods used to "calibrate" tests.
The distorting effect of the exam itself - if its important, people will learn to do the exam, not learn about the subject.
Now, an issue with multi-choice is the "guessing" problem, but there are (as TFA points out) work-arounds. TFA misses out the most important way of reducing guessing - which is designing the questions carefully so that each alternative is seductive and/or represents a common error. The real problem with multi-choice is the last two bullets above - it really is the most artificial and superficial form of test possible. Done well, its a good way of quickly romping through a large domain to offset the "sampling" problem, but it should never be the totality of a test. The depressing problem is that its so easy to mark and administer - and is cheap to deliver on computer (c.f. more ambitious computer-based testing, which is expensive to develop).
*I'm sure its possible to contrive a counter-example.
If you want a quad- or octo- core xeon workstation-class computer that doesn't sound like a helicopter taking off, and you either mainly do 2D work or at all or are in the market for a $1k+ pro graphics card then Apple is good value. However, if you want a 1 or 2 core mini-tower with a gaming-quality 3D card then, not so much.
If you want a tiny but reasonably powerful and very quiet desktop "micro PC" then waffle waffle Mac Mini However, waffle waffle..
(Rinse and repeat across the whole Apple range)
Get the idea? Apple have a small-ish range c.f. Dell, and concentrate on a few not-for-Walmart niches. Whether they are good value depends on exactly what you want, how you rate Apples design and quality and how much value you place on OSX.
Oh, and they don't update their product range nearly as frequently as big box-shifters, so the picture varies monthly...
Folks! Can we not distinguish between what I wrote:
Yeah, but Apple don't really [i]do[/i] non-US keyboards, do they?
...which was an obvious hyperbole referring to the fact that Apple UK keyboards and (as far as I can tell*) several other latin-alphabet countries are basically the US layout with the bare minimum changes for currency symbols etc. and differ considerably from the de-facto standard national layouts used by the PC.
and the statement that some people appear to be reading:
I hereby state definitively that Apple Inc. do not produce any form of regionalised keyboards even for countries that use accented characters, non-QUERTY layouts or, heaven forbid, non-latin character sets.
...which is obviously complete horse droppings.
In future, I promise to make any slashdot postings completely literal and will avoid hyperbole on pain of deat... oh, blast!
(* Mainly from threads on the Parallels support forums from UK and other EU Mac users asking why they can't type a "@" in windows...)
Geez, that's like saying apple doesn't "do" US keyboards because there's no num lock key.
Nope. The dust lies thick on my num lock key. Having several commonly used (especially when coding) punctuation keys juggled - and one hidden completely - is rather more annoying. Biggest problems with this are (a) Mac Mini: bring your own keyboard - then scour the internet for a correct keyboard definition and find that some applications (Carbon?) helpfully switch back to the Apple layout and (b) switching OSs with Parallels, Boot Camp etc. which have tools for re-mapping alt/apple/command etc. but don't accommodate other differences.
"Native" is a pretty subjective term. I am always shocked when looking for freeware, how little there is for Windows in many cases.
By "native" I mean Aqua apps that live in Mac OS land rather than Unix land. Examples that spring to mind are GUI text editors and GUI file archive utilities (that let you browse inside archives). I'm not saying they don't exist - but they're harder to find (jEdit is swing-based and isn't really native, and bbEdit is very definitely non-free).
Go find a hardcore BSD or Linux or Solaris geek developing really cool server software and the chances are they're doing it from a terminal window in OS X on a MacBook.
Yes, but to a "hardcore BSD or Linux or Solaris geek" a GUI has always been just a device for running 16 copies of vim side-by-side while looking as much like NeXTStep as possible, so they're in hog heaven with MacOS - and won't cry if their save as dialog looks a bit unixy. Anybody who has ever tried to get anything else done on a *nix desktop environment will quickly deduce that *nix developers never actually use them:-)
Seriously, there's no doubt that OS X is very well served for *nix shell-based and X11 software, both bundled (emacs! gcc! grep!) and via macports/fink and that these are better integrated than (say) Cygwin on windows. That's a big deal if that's what you want, and one of the reasons I'm using a Mac.
I think that - although biassed and with some suspect statements (NeoOffice requires X11 and is impossible to install? WTF??) there is a grain of truth in what he says. OSX is not the worlds greatest platform if your main requirment is GUI-based Open Source.
I certainly share the impression that there is less native free (beer) ware than Windows - and that although most of the FOSS stalwarts have been ported they do often rely on X11. This is rather second-best, Since the Unique Selling Point of OSX over Linux/FreeBSD is its GUI, many of the advantages of which disappear under X11, this does rather defeat the object. There's quite a lot of reasonably priced shareware though, and I get the impression that things are stepping up a bit post-Intel.
So, basically, if you want a totally free ride, use Linux or FreeBSD - its no great revelation that OSX is aimed mainly at people who are either going to use iLife + (maybe) Office or shell out $$$HOW MUCH!? for professional creativity gear.
I've been using OSX for web development (targetting Linux servers) using Eclipse, PHP PostgreSQL and its largely great - proper unix filesystem (unlike Windows - what's ths point of living if you don't have symlinks?) better/more responsive GUI than Gnome/KDE, easy testing on Firefox & Safari & fire up parallels for testing on (multiple versions) of IE. However - I've had a few issues with the PostgreSQL/MySQL GUI tools not being up to snuff on the Mac.
PS - last time I looked there were multiple sources for Apple-compatible RAM - which isn't so much non-standard as not-the-cheapest (e.g. SODIMS instead of regular sticks). Crucial and Kingston will even arrange - for far less than Apple's price - for the traditional seventeen virgins to journey to the summit of Mount Fuji laden with gold and crushed lotus blossums and obtain from the ancient and venerable hermit therein the rare and valuable FB-DIMM chips coveted by the Mac Pro.
You actually need to do stuff from like early 90s Linux to make it work with non-US keyboard layout and this is pain.
Yeah, but Apple don't really do non-US keyboards, do they? They just replace the # key with the local currency symbol (presumably because USAians call it the "pound" sign). What? You still need to type a # sometimes? This is a minor pain when using Parallels/Bootcamp especially if you do something really deviant like plugging in an actual, non-Apple non-US keyboard.
When I first tried it, on a Mac Mini G4, NeoOffice was a bit of a dog.
However, on a decently powerful machine (e.g. Mac Pro) NeoOffice is eminently usable - haven't done any timings (pointless unless you reboot between each test), but I'd say it feels faster than the "release" Mac Oo running under Apple X11. Main issue is that Neo tends to be a point release or two behind Oo so you get a slightly different bugset - and the Neo people are disinclined to waste valuable porting time patching known bugs in the Oo codebase (very sensible, but still annoying if you are being bugged by one of them).
This section is stupid and ridiculous and is likely to get struck down by the first courtroom judge that looks at this thing as being too vaguely worded.
Of course, the experience of Mic^H^H^H SCO vs IBM is that it takes four years of expensive legal shennanigans before the judge is allowed to even speculate about the possibility of considering making such a judgement.
Anyway, you seem to be mistaking this for something that is intended to be legally enforcable.
The idea is to imply that lots of products are in violation, without making any concrete claims that could be tested and possibly disproved. If the open source community make the GPL less acceptable to industry in a well-meaning but doomed effort to nail this jelly to a tree then that's just a bonus.
the most dangerous litigants are companies not themselves in the software business
The difference is that certain large traditional software companies have a motive to burn some of their spare cash - or risk or having a few patents invalidated - in order to cripple the pesky open source industry. Patent trolls - sock-puppet shenanigans aside - are only in it for the direct profit.
At worst, trolls are an equal threat to the whole software industry, not just open source. At best, the open source industry should be less attractive to them - attack an open source company with a plausible patent case and there is a risk that they'll go titsup.com before you get your damages. You certainly won't see any continuing royalties. Worse, lots of big players who would just sit back and eat peanuts while you went after a commercial competitor, have a vested interest in the same bits of FOSS and might gang up on you while every geek on the internet searches for prior art. Best stick to closed-soruce companies who have a budget for patent extortion.
The real glass-half-full aspect is that these clowns are helping discredit the patent system, and upsetting the Mutually Assured Destruction status quo that keeps the big players on the pro-patent side.
a helicopter equipped with signal-jamming equipment
So much for mobile phone radio frequencies interfering with saftey-critical avionics! I guess milirtary helicopters don't have the most vulneable equipment (namely the credit card readers in seatback phones).
In other news: President stung to death by bees driven into a frenzy by mobile phone radiation... (Yes, yes I know the mobile phones affect bees thing has been debuinked).
If someone wants to write a conventional wordprocessor they can choose to give it away for free without taking on any costs or liability. If someone wants to use that wordprocessor they can test it to their satisfaction and be fairly certain that it will then continue working. Worst case scanario - it doesn't keep up with some OS update in the future - and obviously they've checked that it uses an open file format, so that they won't lose their data.
OTOH, a software-as-a-service wordprocessor is as much use as an inflatable dartboard unless someone is going to provide that service and make some minumum level of service guarantee (including data backups). That costs.
Now, it would be great if there were Free/Open Source software-as-a-service SERVERS for people or companies who wanted to run their own "personal" centralised system - but as the main source of applications for a "Free" OS it just ain't gonna fly.
I can't believe this guy's 1969 maths book literally said:
(a / b) x (c / d) = a c / b d
as opposed to writing it out properly with fractions - which kinda makes the problem go away. This is simply an artefact of not bothering to fire up Equation Editor/LaTeX/etc. Substituting "/" for a vinculum (the line in a fraction) loses information - you need extra brackets to remove the ambiguity. Film at 11. Move along.
OK, you still get an ambiguity when even the traditional division sign rubs shoulders with implied multiplication - which is probably why algebra traditionally uses fractions (and computer languages rarely implement implied multiplication 'cos it really complicates the parser).
Slashdot Mirror
My initial reaction was that, come the European launch, if the iPhone doesn't have 3G/UMTS/HSPDA then it would be laughed out of court. However, on reflection, it sounds as if Apple's attitude is:
GPRS is good enough to check your EMAIL and gives good phone coverage. If you want a decent web-surfing experience on the train, subway or in a coffee shop, your best bet is if some bright spark has installed WiFi. So lets do a phone which makes a much better job of doing WiFi than the competition and not weigh it down and waste battery life by putting in 3G capability. We're Apple - maybe we're influential enough to put some momentum behind WiFi coverage.
(PS - am I right in thinking that EDGE/GPRS and 3G/UTMD/HSPDA are two incompatible "family trees" of protocols, and a phone that supports both needs a certain amount of duplicate "gubbins" inside?)
I have a (UK) phone that does GRPS*, UMTD & HSPDA*, bluetooth and WiFi and while 3G coverage here is ok (and HSPDA being rolled out - and very nice when you can get it) it wouldn't be much good without GRPS as a fallback. You certainly can't use the internet reliably on a train (I've tried - and did manage to send an EMAIL from a train here but it was a labour of love and certainly wasn't HSPDA!) The phone (MDA Vario II - AKA HTC TyTan) is a bit of a brick and it certainly doesn't flip seamlessly between WiFi hotspots (cough)WM5(cough).
PS - real Apple Fanbois should, of course, equip themselves with a backpack containing a laptop with a HSPDA data card and a compact WAP. Then they can whip out their iPhone and impress people anywhere with HSPDA coverage... (* OK - PCMCIA, as they say, so I'm trying out multiple permutations for the ETLAs and DETLAs here:-)
Nope - OEM is irrelevant to the anti-VM clause issue. The OEM versions of "ultimate" and "business" which allow VM use are still a fraction of the price of corresponding full retail versions. AFAIK Apple could offer those if they desired. Odds are, if you are ordering in bulk for (probably networked) systems, you didn't want the "home" edition anyway. If you were buying in quantity, you could probably get a volume licence (again, they probably don't even bother to do that for "home"). As I said, if the whole plan was to stop OEM + VM use it would have been much simpler - and almost reasonable* - to put the clause in the OEM license.
(* for a given value of "reasonable")
First, the regular OEM price isn't a "greatly reduced price" - its the de-facto going rate for Windows (exactly what price the big boys, Dell, HP etc. pay - and what "Important considerations" they offer in return is another question).
Secondly, the current OEM license already has some major extra restrictions about how and where you can install it - with crossed fingers and a following wind you could probably argue that these already blocked VM use - after all, where would you stick the hologram? :-) They pretty clearly mean that joe user isn't supposed to buy an OEM copy and install it on his VM. If MS had changed the OEM rules to explicitly block "VM bundling" then I doubt it would have raised the sort of negative publictiy than claiming that you can't even use a full-price version on a VM has.
Thirdly, I stand to be proven wrong by posterity, but the only circumstances under which I can see Apple bundling Windows is if they throw in the towel, drop OSX and become a vendor of designer Windows PCs with an iLife-for-windows bundle (while plausible, I'm sure that ain't exactly Plan 'A'). There's a huge PR difference between the current stance of "hey - if you want to use our hardware to run Windows here are some solutions" and saying "actually, so many people can't get buy with OSX alone that we've decided to bundle windows". The latter comes under the heading of "pulling a Ratner*". I'm sure Microsoft would love to be thrown into that briar patch.
Anyway, if you've actually tried to use Parallels (or, by reputation, VMware) you'll know that while they are excellent products, they don't yet come close to the Apple "it just works" standard (on which Windows sets something of an upper limit, anyway!). Coherence is very effective, and certainly solves the screen real-estate issues with running apps inside a VM window - but to say it "makes Windows apps look and work just like OSX" is rather optimistic.
Nope - explanation "A" is that TFA was warm - this policy makes it cheaper to buy a Dell than a legit copy of Windows Vista for your Mac or Linux box. Explanation "B" is that VMs can be used to circumvent DRM and that forcing pirates to buy a full-price copy of Ultimate gives Microsoft enough plausible deniability to stop the MS board from being carted off to G'tmo under the DMCA. Of course, Explanation B only makes sense if you are sufficiently far down the DRM rabbit hole to think that evil terrorist-loving pirates are going to lose sleep over breaching their windows EULA - or will even relate the concepts of "buy" and "windows" , but then...
(*Sorry - I've googled "pull a Ratner" and it might not mean in US English what it means in English English, which is "publicly describing your own product as 'crap' and then wondering where your business went")
Strange - the take home message I got from the book was "what is the difference"? The lives, ambitions and purposes of the humans in the book were just as synthetic and imposed as the replicants fake memories. By the end of the book Deckard is pretty definitely real, but most of his life has been exposed as fake and/or a control method: the fake religion, the "mood organ" that lets you adjust your temprament to fit in, the fake DJ, the social pressure to keep an animal (even if its fake) the entire fake police station...
For me the biggest thing missing from the film was the details of the "Mercerism" religion, which make it clear that the VK test used to identify "andys" is mainly a test of religious orthodoxy, and the significance of replicant's inability to take part in the empathic "communion" that is central to the religion (and later revealed as electronic trickery).
OK, so all DRM is broken, because while cryptography is the advanced science of getting a message from Alice to Bob without Eve being able to intercept it, DRM is a futile attempt to stop Bob then handing the message to Eve in return for a good seeing-to. That's a given. I wasn't defending it.
My point was that a cryptography-based system can use an open, stable protocol which a hypervisor could "proxy" - with clean hands - and let OSX satisfy itself that it is being hosted on a Mac. Anybody found reverse-engineering the ID chip or hacking the software can then be carted off to Guantanamo under the DMCA, so that's OK. If its a kludge, the hypervisor would have to know the secret and if its just a collection of hardware characteristics that Mac hardware happens to have it could be (a) very difficult to emulate and/or (b) hard to keep secret.
Apple's main concern is, presumably, to stop people running OSX on PC hardware. Currently, any OSX running virtualized must have had the hardware limitation hacked out. That would require Apple, VMWare & Parallels to get their heads together to sort out a solution - which also assumes that they are using some sort of strong cryptography-based system with a key held on the motherboard somewhere that the hypervisor could securely "proxy" - rather than some security-by-obscurity kludge.
True-ish, and Apple certainly can't chuck any bricks in that particular greenhouse. However, there are a couple of mitigating factors:
If Apple doesn't sort this out soon they're going to start hacking off developers - virtualization is so darn useful. This will come to a head when 10.5 is released and betas of 10.6 go out and developers have to juggle past, present and future major versions of the x86 based OS - but the initiative will have to come from developers, via Apple - Parallels and VMWare have no strong incentive to break a sweat over it.
P.S. Also bear in mind that the last thing Apple want is, officially or otherwise, a "try-before-you-buy" route for OSX: even if the implementation was non-flakey, the first impression of playing with a new OS is always frustration because of the differences and the fact that your instinct is to plunge into "clever stuff" rather than work through the basics. Better if you are sold on the idea by an evangelist, part with cash, and have a $2000 incentive to get over having to press the fricking pretzel key instead of "ctrl".
I think we're talking security risks of the "customers might be able to circumvent DRM" using VMs which (e.g.) make it easy to change the MAC on your ethernet kind.
Hey, that would certainly make MS feel insecure. Why do you persist in assuming that they're talking about our security? :-)
Unfortunately - in many fields (particularly school/college) the test is also the de-facto definition of the curriculum. It would be nice if 'twere not so, but its pragmatically inescapable. Writing a psychometrially valid test is one thing - writing a psychometrically valid test that also exemplifies how the subject should be taught is a bit of a toughie - and might mean sacrificing some of the psychometric rigour.
The real problem occurs when you have the people who understand the subject in silo (A), the people who understand teaching in silo (B), the psychometricians in silo (C), and - for computer based testing - the programmers in silo (D) all passing messages via the Administration. I really, really, really wish that I'd never seen that happen.
As long as you realise that the relationship between "reliable" and "good" is one of correlation, not causation.
Ensuring that a test is measuring some factor in a stable and reliable fashion has, indeed, become a pretty exact science. The danger is that all the impressive statistics this produces will give a false impression of validity and detract from the massive inference that this some factor is the thing you are actually purporting to measure - and, of course, if the test is vital, it will Heisenberg the educational process into one that teaches some factor.
TFA mentioned bar exams. Now, IANAL but I'm pretty sure that representing someone in court does not consist of answering a battery of multi choice questions covering the entire spectrum of jurisprudence. How reliable the tests are - year on year or against other (probably similar) tests does little to support the assertion that the test measures what it claims to measure.
I've encountered math questions where, when you actually observe and question schoolchildren working on them, the actual thought processes involved have nothing to do with the what the question purported to test - however, the bright kids cope better and the psychometric analyses see no problem. Best example was a computer-based question for young kids about telling the time where, on observation, sucess or failure clearly depended more upon fathoming out the question's user interface than telling the time. Looking at the psychometrics gave the question a clean bill of health. Other questions on high stakes assessment test math but in a inappropriate way - e.g. probability questions that require you to multiply two probabilities that, in context, are clearly interdependent; statistics questions that require you to say that statistically insignificant data prove a hypothesis; plus loads and loads of little isolated techniques tested separately with no requirement to "synthesize" them into a credible application.
Psychometric analysis is a valuable tool - when combined with other measures, acknowledgement of the underlying axioms and a healthy dollop of judgement and experience - but what it offers is so seductive in today's pseudo-accountability, management-by-algortithm environment that it can obscure the truth: testing is an inexact science for reasons that go far deeper than the statistics and we should not place so much cultural emphasis on the differnce between 49% and 51%.
...QED :-)
He should have stuck with the more useful observation that almost* any test with a very low pass rate will be unreliable.
All tests have a margin of error, although its a rather taboo subject - when did you ever get a test result that stated the 95% confidence interval? If only a small proportion pass, there is a danger that these errors will dominate.
There are:
Now, an issue with multi-choice is the "guessing" problem, but there are (as TFA points out) work-arounds. TFA misses out the most important way of reducing guessing - which is designing the questions carefully so that each alternative is seductive and/or represents a common error. The real problem with multi-choice is the last two bullets above - it really is the most artificial and superficial form of test possible. Done well, its a good way of quickly romping through a large domain to offset the "sampling" problem, but it should never be the totality of a test. The depressing problem is that its so easy to mark and administer - and is cheap to deliver on computer (c.f. more ambitious computer-based testing, which is expensive to develop).
*I'm sure its possible to contrive a counter-example.
If you want a quad- or octo- core xeon workstation-class computer that doesn't sound like a helicopter taking off, and you either mainly do 2D work or at all or are in the market for a $1k+ pro graphics card then Apple is good value. However, if you want a 1 or 2 core mini-tower with a gaming-quality 3D card then, not so much.
If you want a tiny but reasonably powerful and very quiet desktop "micro PC" then waffle waffle Mac Mini However, waffle waffle..
(Rinse and repeat across the whole Apple range)
Get the idea? Apple have a small-ish range c.f. Dell, and concentrate on a few not-for-Walmart niches. Whether they are good value depends on exactly what you want, how you rate Apples design and quality and how much value you place on OSX.
Oh, and they don't update their product range nearly as frequently as big box-shifters, so the picture varies monthly...
and the statement that some people appear to be reading:
...which is obviously complete horse droppings.
In future, I promise to make any slashdot postings completely literal and will avoid hyperbole on pain of deat... oh, blast!
(* Mainly from threads on the Parallels support forums from UK and other EU Mac users asking why they can't type a "@" in windows...)
Nope. The dust lies thick on my num lock key. Having several commonly used (especially when coding) punctuation keys juggled - and one hidden completely - is rather more annoying. Biggest problems with this are (a) Mac Mini: bring your own keyboard - then scour the internet for a correct keyboard definition and find that some applications (Carbon?) helpfully switch back to the Apple layout and (b) switching OSs with Parallels, Boot Camp etc. which have tools for re-mapping alt/apple/command etc. but don't accommodate other differences.
By "native" I mean Aqua apps that live in Mac OS land rather than Unix land. Examples that spring to mind are GUI text editors and GUI file archive utilities (that let you browse inside archives). I'm not saying they don't exist - but they're harder to find (jEdit is swing-based and isn't really native, and bbEdit is very definitely non-free).
Yes, but to a "hardcore BSD or Linux or Solaris geek" a GUI has always been just a device for running 16 copies of vim side-by-side while looking as much like NeXTStep as possible, so they're in hog heaven with MacOS - and won't cry if their save as dialog looks a bit unixy. Anybody who has ever tried to get anything else done on a *nix desktop environment will quickly deduce that *nix developers never actually use them :-)
Seriously, there's no doubt that OS X is very well served for *nix shell-based and X11 software, both bundled (emacs! gcc! grep!) and via macports/fink and that these are better integrated than (say) Cygwin on windows. That's a big deal if that's what you want, and one of the reasons I'm using a Mac.
I think that - although biassed and with some suspect statements (NeoOffice requires X11 and is impossible to install? WTF??) there is a grain of truth in what he says. OSX is not the worlds greatest platform if your main requirment is GUI-based Open Source.
I certainly share the impression that there is less native free (beer) ware than Windows - and that although most of the FOSS stalwarts have been ported they do often rely on X11. This is rather second-best, Since the Unique Selling Point of OSX over Linux/FreeBSD is its GUI, many of the advantages of which disappear under X11, this does rather defeat the object. There's quite a lot of reasonably priced shareware though, and I get the impression that things are stepping up a bit post-Intel.
So, basically, if you want a totally free ride, use Linux or FreeBSD - its no great revelation that OSX is aimed mainly at people who are either going to use iLife + (maybe) Office or shell out $$$HOW MUCH!? for professional creativity gear.
I've been using OSX for web development (targetting Linux servers) using Eclipse, PHP PostgreSQL and its largely great - proper unix filesystem (unlike Windows - what's ths point of living if you don't have symlinks?) better/more responsive GUI than Gnome/KDE, easy testing on Firefox & Safari & fire up parallels for testing on (multiple versions) of IE. However - I've had a few issues with the PostgreSQL/MySQL GUI tools not being up to snuff on the Mac.
PS - last time I looked there were multiple sources for Apple-compatible RAM - which isn't so much non-standard as not-the-cheapest (e.g. SODIMS instead of regular sticks). Crucial and Kingston will even arrange - for far less than Apple's price - for the traditional seventeen virgins to journey to the summit of Mount Fuji laden with gold and crushed lotus blossums and obtain from the ancient and venerable hermit therein the rare and valuable FB-DIMM chips coveted by the Mac Pro.
Yeah, but Apple don't really do non-US keyboards, do they? They just replace the # key with the local currency symbol (presumably because USAians call it the "pound" sign). What? You still need to type a # sometimes? This is a minor pain when using Parallels/Bootcamp especially if you do something really deviant like plugging in an actual, non-Apple non-US keyboard.
When I first tried it, on a Mac Mini G4, NeoOffice was a bit of a dog.
However, on a decently powerful machine (e.g. Mac Pro) NeoOffice is eminently usable - haven't done any timings (pointless unless you reboot between each test), but I'd say it feels faster than the "release" Mac Oo running under Apple X11. Main issue is that Neo tends to be a point release or two behind Oo so you get a slightly different bugset - and the Neo people are disinclined to waste valuable porting time patching known bugs in the Oo codebase (very sensible, but still annoying if you are being bugged by one of them).
Of course, the experience of Mic^H^H^H SCO vs IBM is that it takes four years of expensive legal shennanigans before the judge is allowed to even speculate about the possibility of considering making such a judgement.
Anyway, you seem to be mistaking this for something that is intended to be legally enforcable.
The idea is to imply that lots of products are in violation, without making any concrete claims that could be tested and possibly disproved. If the open source community make the GPL less acceptable to industry in a well-meaning but doomed effort to nail this jelly to a tree then that's just a bonus.
The difference is that certain large traditional software companies have a motive to burn some of their spare cash - or risk or having a few patents invalidated - in order to cripple the pesky open source industry. Patent trolls - sock-puppet shenanigans aside - are only in it for the direct profit.
At worst, trolls are an equal threat to the whole software industry, not just open source. At best, the open source industry should be less attractive to them - attack an open source company with a plausible patent case and there is a risk that they'll go titsup.com before you get your damages. You certainly won't see any continuing royalties. Worse, lots of big players who would just sit back and eat peanuts while you went after a commercial competitor, have a vested interest in the same bits of FOSS and might gang up on you while every geek on the internet searches for prior art. Best stick to closed-soruce companies who have a budget for patent extortion.
The real glass-half-full aspect is that these clowns are helping discredit the patent system, and upsetting the Mutually Assured Destruction status quo that keeps the big players on the pro-patent side.
So much for mobile phone radio frequencies interfering with saftey-critical avionics! I guess milirtary helicopters don't have the most vulneable equipment (namely the credit card readers in seatback phones).
In other news: President stung to death by bees driven into a frenzy by mobile phone radiation... (Yes, yes I know the mobile phones affect bees thing has been debuinked).
If someone wants to write a conventional wordprocessor they can choose to give it away for free without taking on any costs or liability. If someone wants to use that wordprocessor they can test it to their satisfaction and be fairly certain that it will then continue working. Worst case scanario - it doesn't keep up with some OS update in the future - and obviously they've checked that it uses an open file format, so that they won't lose their data.
OTOH, a software-as-a-service wordprocessor is as much use as an inflatable dartboard unless someone is going to provide that service and make some minumum level of service guarantee (including data backups). That costs.
Now, it would be great if there were Free/Open Source software-as-a-service SERVERS for people or companies who wanted to run their own "personal" centralised system - but as the main source of applications for a "Free" OS it just ain't gonna fly.
I can't believe this guy's 1969 maths book literally said:
(a / b) x (c / d) = a c / b d
as opposed to writing it out properly with fractions - which kinda makes the problem go away. This is simply an artefact of not bothering to fire up Equation Editor/LaTeX/etc. Substituting "/" for a vinculum (the line in a fraction) loses information - you need extra brackets to remove the ambiguity. Film at 11. Move along.
OK, you still get an ambiguity when even the traditional division sign rubs shoulders with implied multiplication - which is probably why algebra traditionally uses fractions (and computer languages rarely implement implied multiplication 'cos it really complicates the parser).