Um... How do you get everyone in the crowded room to enter their PIN and authorize the "charge"?
TFA says nothing about a PIN. Indeed, I would imagine having to enter a PIN would pretty much remove any of the additional convenience over a Credit or Debit Card.
The usual defence against the idea of hijacking payments is that you need to be within a few centimetres for NFC to work, to which I typically respond "you don't catch the [subway|tube|metro|etc] very often, do you ?".
Finally, those things can sometimes activate from a surprisingly long distance. I've seen an Oyster or SUICA card trigger the reader from a good 15-20cm away on more than one occasion (conversely, I've had to slide it across the top of the reader multiple times on many occasions, as well).
Correction. We will be dealing with it so long as we keep using a heavily outdated rights system where programs have the same rights as the user executing the program.
Limiting the rights of applications won't help, because lazy developers will just configure the system during installation so their application can do anything it wants.
Exhibit A: Windows applications needing Administrator privileges. It's been over a decade since the necessary infrastructure was in place to deprecate this, yet *still* applications are released today that must be run as Admin for no good reason.
Blaming the ignorant end users is just bad form [...]
I'm not blaming them - "ignorant" is an assessment of their knowledge, not their intelligence. So long as you want the ignorant end users - ie: end users who lack the knowledge and/or experience to make good decisions - to be able to decide what a given program can and can't do, then malicious programs have a trivially-exploited entry point into the system.
"We" will still be dealing with it, and have been dealing with it for decades, until certain companies start designing their software for their intended audience, rather than some fictional perfect being that never makes a mistakes.
I'm not aware of any company designing its software to be resistant to end user ignorance.
Fail. Although Linux users are indeed generally more educated on the finer points of computing, there seems to be this persistent myth that Linux doesn't get viruses because it has such a small user base.
That is an important factor. However, by far the biggest reason is - as I said - because Linux users don't represent anything like the exploitability as Windows users.
It's not because there are fewer of them - although that certainly plays a non-trivial part - it's because Linux users are far (*FAR*) less likely to let a virus into their system, either by leaving known security holes unpatched, or by the more common method of being socially engineered into executing it.
Linux servers control a major portion of www. If those aren't prime targets then what is?
Desktop PCs run by ignorant end users outnumber Linux servers tens of thousands to one. Why would you try to aim a virus whose success is largely predicated on a low level of knowledge and experience from the victim, at systems run by seasoned professionals ?
Or, to put it another way, not only will you be able to exploit something like 50,000+ desktop PCs to every web server, when you do exploit the average desktop PC, chances are extremely low it will ever be detected, so you basically have the run of the machine. However, if you exploit the average web server, chances are extremely high your intrusion will be detected and fixed within a matter of hours or days.
Plain and simple, the Linux security model is superior.
The classic UNIX security model (as used by most Linux installs) is demonstrably inferior to Windows NT's.
We get the best of both worlds. efficiency for high power draw items, and cord ends smaller than bricks. Also we have more than 3 receptacles in a room.
Here in Switzerland, standard power is 220v, but even the 3-prong grounded plugs are essentially the same size as US plugs. Typically there are two sets of 3 sockets per room, with one socket wired into the room's light switch, and additional sockets integrated into the panel of any light switches that might be present (which I've decided is an awesome idea).
Back home in Australia (240v), two-prong plugs are about the same size as US plugs, and it would be unusual to find less than 4 outlets in a room (two sets of two).
If you were truly a Linux power user, then you'd know that the Linux/UNIX security model is not conducive to the spread of viruses since any program attempting to modify system files would require root access first.
There's not much the average virus needs to do that requires "modifying system files".
It's not the "security model" that's non-conducive to viruses spreading in Linux, it's the users.
In Denmark we have a fixed grace period, however foreign workers do have a hard time getting a job since there is a minimum required pay for keeping the green card (which in effect puts the foreign worker in the top 50% payment), this severely reduces the gain for companies when hiring foreigners.
That's actually one of the better solutions to the problem I've heard.
I don't care how you try to word it, that is the major difference between the OSX Dock and the Windows Taskbar.
Er, no. The major difference between the Dock and the Taskbar is that icons on the Dock represent *applications* and buttons on the Taskbar represent *windows*.
As differences go, that's a fairly fundamental one, because it reflects the fundamental difference between the applicate-centric OS X UI and the window-centric Windows UI.
The fact they've (finally) merged the Quicklaunch toolbar and the window list is, at most, a minor evolutionary change.
I don't know how clear that is to some of us, but regardless of how one switches windows or applications using hotkeys, the Mac windowing system (as the article makes clear) is essentially document-centric - each window corresponds (with some exceptions) to a document, which is sort of why closing the last document window doesn't terminate the application - i.e. it doesn't make this assumption, since your next action might be to open a new document.
No, it's application-centric, and your example demonstrates one of the (relatively minor) reasons why (even with no documents, the application stays open). In Windows - generally speaking - the application is only running if there are documents open, because Windows is document- (more accurately, window-) centric.
Wow. Do you even read the front page? From Slashdot yesterday: 'experience has shown that multiprocessing across discrete CPUs is not the same thing as multiprocessing across integrated cores within the same CPU.' Here's the article [slashdot.org].
I've read the article. The atrocious use of percentages, alone, makes it suspect (to say nothing of some of the questionable numbers).
Further, "Not the same" does not mean "completely different". The majority of the improvements for multicpu systems (eg: locking), are just as valid for multicore systems.
Finally, if you think Microsoft haven't been concentrating on multicore systems at least as long as Apple, you're just silly. For example, Windows has been supporting NUMA-capable platforms (some similarities to multicore) nearly as long as OS X has even existed.
But then anyone who says, "Microsoft has never released a version of Windows that performed that badly." has go to be trolling.
They haven't. When OS X was first released, and for a good couple of years afterwards, you quite simply could not buy hardware that would run it well. No version of Windows has *ever* been that slow.
OSX has SMP support in-line with Windows XP (maybe 2003 at a stretch),
NeXT never had an SMP machine so your point there is moot.
No, it's not, because the claim was "OS X was far more advanced than NT/2000 or Win9x upon release", which its lack of basic SMP capabilities (let alone fine-grained locking and heavy use of multithreading) demonstrate as false.
OS X has always had decent SMP support.
It has not. Early versions were awful, with a single big kernel lock and other inefficiencies. It has only been since 10.4 (roughly on par with NT4) that it has moved into the realms of "not bad".
Grand Central in 10.6 is going to be awesome as well.
It'll be basically equivalent to Windows 7.
Quartz (especially after Quartz Extreme in 10.2), [...]
Is now at least half a generation behind Windows.
[...] a better windowing system, [...]
Not sure exactly what you mean here, but the overall sluggishness of OS X's window operations (eg: when resizing) and the lack of certain capabilities (resize from any side, maximise) suggest otherwise. Expose looks cool, but ultimately doesn't provide any more functionality than the Taskbar.
[...] more accurate font rendering, [...]
Can't say anything about this.
[...] better stability, [...]
Bollocks. At best, it's equivalent.
[...] better scripting environments (TCL and AppleScript vs VBScript and CMD), [...]
You can do a LOT with VBScript. Not to mention PowerShell.
[...] less security holes, [...]
Not according to the people who keep track of them.
[...] a much more rich command line environment that isn't feeble and borderline useless, [...]
PowerShell.
[...] API's that aren't near as annoying as Win32, free development tools, etc. I'll take XCode/Objective C over Visual C++ any day.
And I'm sure there are developers who prefer Windows.
NT has always been a joke and even though it was touted as a "unix killer", it has NEVER been able to do the deed.
It certainly replaced a lot of professional workstations and low-end servers that used to run UNIX.
Possibly, but who cares? The world is going multi-core, not multiple CPUs and I would suggest that Apple has the lead there, and will strengthen it with Grand Central, Grand Central Dispatch, and OpenCL.
The difference between multiple cores and multiple CPUs is not significant, and Apple is, in raw timeframe terms, a solid decade behind everyone else when it comes to optimising for SMP.
Read this if you don't believe me.
I'm sorry, but anything that opens with something like:
Apple is clearly a leader in implementing multi-core support, beginning with the first dual processor Power Macs 5 years ago, while the DayStar multi-processor Macs date back to the mid-90s.
Is so laughably biased and/or misinformed, it's not even worth reading.
ok, a pre-emptive multi-tasking os OS like the early 90s OS/2 had 16 bit drivers. so it is not considered a "real OS" and should be categorized in the same list as Windows 95? right.
That's not what I said at all. You were complaining about 16-bit code in Windows 95, so I pointed out that there was also 16-bit code in OS/2.
and I liked the part in that link where Intel didn't allow the people putting up the article to do their own testing. I was there, I ran it, it ran OS/2 apps much faster, it ran windows apps much slower and Intel had to come out with a new design because Windows 95 was not optimized for 32bit and that was all Microsoft had.
Rubbish. Microsoft had Windows NT (which was technically superior to OS/2 in every way).
Like I said, I was there and it was yet another example of Microsoft's marketing driving yet another poor OS experience for anyone who ran other OS's and new what an OS was supposed to be. What and OS should be.
I was also there, and an ex-OS/2 user, and Windows 95 did everything it was supposed to do and did it quite well.
However, getting back to the actual argument, the point is that claims of Windows 95 'running more slowly on a PPro than a Pentium because of 16-bit code' are flat-out wrong. If you benchmark a 32-bit application on a properly configured Windows 95 system, with both a PPro and Pentium of equivalent clock speed, the PPro will be faster.
Mac OS X used to be called NeXTstep, and NeXTstep had a dock which Windows 95 copied to create the task bar.
If you had actually used NeXTSTEP, you would know that its Dock and the Windows 95 Taskbar behave very differently. Much like the taskbar and the OS X Dock behave differently.
The Windows 95 look which came to be called the Windows classic look which was in fact a shameless but inferior copy of the NeXTstep look from 1988.
Rubbish. Application launching, task switching, menu interaction, window management - all these things were quite different in NeXT compared to Windows 95. Indeed, you'd struggle to find ways they were similar, that weren't also shared by every other GUI.
hrmmm... the Ars article gave me the impression that one of the benefits of OS X (and shortcomings of Windows' MDI model) [...]
One of the problems with the article is it talks about MDI as if it were still "the way", yet MDI has been deprecated since the release of Windows 95 (there was quite a hoo-ha about it at the time).
Now, certain applications have managed to keep the MDI fire burning, for whatever reason (eg: Excel is the poster child for this silliness), but developers are *supposed* to have been writing to a "one document:one window" model for nearly a decade and a half now.
There's also tabbed browsing, which is a sorta-but-not-really implementation of MDI. From a technical UI viewpoint it's atrocious - it takes all the problems inherent to MDI and adds a couple more besides - but for something like a browser which is mostly "read-only" use and the need for "cross-window-interaction" is low, it actually works OK.
Normally Ars stuff is pretty good, but that article is *very* ordinary, with a lot of conceptual, functional and historical errors.
The main thrust is correct, however, the Windows 7 Taskbar is clearly a descendant of its Windows 95 Great-great-grandfather, not the bastard child of NeXT and MacOS.
OSX and BeOS continue(d) to get faster with each new release.
OSX was so slow at release it had nowhere to go but up. Microsoft has never released a version of Windows that performed that badly. Relatively speaking, Vista was greased lightning at release compared to OSX 10.0 (10.1, 10.2, maybe even 10.3).
Slashdot Mirror
That's a mighty bold claim. The only country I've been to where cash is used more often that credit or debit cards, is Switzerland.
Um... How do you get everyone in the crowded room to enter their PIN and authorize the "charge"?
TFA says nothing about a PIN. Indeed, I would imagine having to enter a PIN would pretty much remove any of the additional convenience over a Credit or Debit Card.
The usual defence against the idea of hijacking payments is that you need to be within a few centimetres for NFC to work, to which I typically respond "you don't catch the [subway|tube|metro|etc] very often, do you ?".
Finally, those things can sometimes activate from a surprisingly long distance. I've seen an Oyster or SUICA card trigger the reader from a good 15-20cm away on more than one occasion (conversely, I've had to slide it across the top of the reader multiple times on many occasions, as well).
I can't wait to be able to steal money just by walking through a crowded room and "charging" each person's phone $5.
Correction. We will be dealing with it so long as we keep using a heavily outdated rights system where programs have the same rights as the user executing the program.
Limiting the rights of applications won't help, because lazy developers will just configure the system during installation so their application can do anything it wants.
Exhibit A: Windows applications needing Administrator privileges. It's been over a decade since the necessary infrastructure was in place to deprecate this, yet *still* applications are released today that must be run as Admin for no good reason.
Blaming the ignorant end users is just bad form [...]
I'm not blaming them - "ignorant" is an assessment of their knowledge, not their intelligence. So long as you want the ignorant end users - ie: end users who lack the knowledge and/or experience to make good decisions - to be able to decide what a given program can and can't do, then malicious programs have a trivially-exploited entry point into the system.
"We" will still be dealing with it, and have been dealing with it for decades, until certain companies start designing their software for their intended audience, rather than some fictional perfect being that never makes a mistakes.
I'm not aware of any company designing its software to be resistant to end user ignorance.
Given the lacklustre security history of NT servers and desktops, the world eagerly awaits your demonstration.
Per-user ACLs vs User/Group/Other.
All OS objects have ACLs, vs applying permissions only via filesystem abstractions.
Superuser vs none.
Fail. Although Linux users are indeed generally more educated on the finer points of computing, there seems to be this persistent myth that Linux doesn't get viruses because it has such a small user base.
That is an important factor. However, by far the biggest reason is - as I said - because Linux users don't represent anything like the exploitability as Windows users.
It's not because there are fewer of them - although that certainly plays a non-trivial part - it's because Linux users are far (*FAR*) less likely to let a virus into their system, either by leaving known security holes unpatched, or by the more common method of being socially engineered into executing it.
Linux servers control a major portion of www. If those aren't prime targets then what is?
Desktop PCs run by ignorant end users outnumber Linux servers tens of thousands to one. Why would you try to aim a virus whose success is largely predicated on a low level of knowledge and experience from the victim, at systems run by seasoned professionals ?
Or, to put it another way, not only will you be able to exploit something like 50,000+ desktop PCs to every web server, when you do exploit the average desktop PC, chances are extremely low it will ever be detected, so you basically have the run of the machine. However, if you exploit the average web server, chances are extremely high your intrusion will be detected and fixed within a matter of hours or days.
Plain and simple, the Linux security model is superior.
The classic UNIX security model (as used by most Linux installs) is demonstrably inferior to Windows NT's.
We get the best of both worlds. efficiency for high power draw items, and cord ends smaller than bricks. Also we have more than 3 receptacles in a room.
Here in Switzerland, standard power is 220v, but even the 3-prong grounded plugs are essentially the same size as US plugs. Typically there are two sets of 3 sockets per room, with one socket wired into the room's light switch, and additional sockets integrated into the panel of any light switches that might be present (which I've decided is an awesome idea).
Back home in Australia (240v), two-prong plugs are about the same size as US plugs, and it would be unusual to find less than 4 outlets in a room (two sets of two).
It's 2009... I can't believe we're still dealing with this crap in 2009.
"We" will be dealing with it so long as ignorant end users can execute arbitrary code.
If you were truly a Linux power user, then you'd know that the Linux/UNIX security model is not conducive to the spread of viruses since any program attempting to modify system files would require root access first.
There's not much the average virus needs to do that requires "modifying system files".
It's not the "security model" that's non-conducive to viruses spreading in Linux, it's the users.
That's kind of unfair isn't it? Where multiple CPU workstations even available during NextStep's lifetime?
Yes.
In Denmark we have a fixed grace period, however foreign workers do have a hard time getting a job since there is a minimum required pay for keeping the green card (which in effect puts the foreign worker in the top 50% payment), this severely reduces the gain for companies when hiring foreigners.
That's actually one of the better solutions to the problem I've heard.
I don't understand why do they not just sell the whole team and game to somebody else?
In case they want to resurrect it later.
I don't care how you try to word it, that is the major difference between the OSX Dock and the Windows Taskbar.
Er, no. The major difference between the Dock and the Taskbar is that icons on the Dock represent *applications* and buttons on the Taskbar represent *windows*.
As differences go, that's a fairly fundamental one, because it reflects the fundamental difference between the applicate-centric OS X UI and the window-centric Windows UI.
The fact they've (finally) merged the Quicklaunch toolbar and the window list is, at most, a minor evolutionary change.
I don't know how clear that is to some of us, but regardless of how one switches windows or applications using hotkeys, the Mac windowing system (as the article makes clear) is essentially document-centric - each window corresponds (with some exceptions) to a document, which is sort of why closing the last document window doesn't terminate the application - i.e. it doesn't make this assumption, since your next action might be to open a new document.
No, it's application-centric, and your example demonstrates one of the (relatively minor) reasons why (even with no documents, the application stays open). In Windows - generally speaking - the application is only running if there are documents open, because Windows is document- (more accurately, window-) centric.
Wow. Do you even read the front page? From Slashdot yesterday: 'experience has shown that multiprocessing across discrete CPUs is not the same thing as multiprocessing across integrated cores within the same CPU.' Here's the article [slashdot.org].
I've read the article. The atrocious use of percentages, alone, makes it suspect (to say nothing of some of the questionable numbers).
Further, "Not the same" does not mean "completely different". The majority of the improvements for multicpu systems (eg: locking), are just as valid for multicore systems.
Finally, if you think Microsoft haven't been concentrating on multicore systems at least as long as Apple, you're just silly. For example, Windows has been supporting NUMA-capable platforms (some similarities to multicore) nearly as long as OS X has even existed.
But then anyone who says, "Microsoft has never released a version of Windows that performed that badly." has go to be trolling.
They haven't. When OS X was first released, and for a good couple of years afterwards, you quite simply could not buy hardware that would run it well. No version of Windows has *ever* been that slow.
OSX has pretty solid SMP support.
OSX has SMP support in-line with Windows XP (maybe 2003 at a stretch),
NeXT never had an SMP machine so your point there is moot.
No, it's not, because the claim was "OS X was far more advanced than NT/2000 or Win9x upon release", which its lack of basic SMP capabilities (let alone fine-grained locking and heavy use of multithreading) demonstrate as false.
OS X has always had decent SMP support.
It has not. Early versions were awful, with a single big kernel lock and other inefficiencies. It has only been since 10.4 (roughly on par with NT4) that it has moved into the realms of "not bad".
Grand Central in 10.6 is going to be awesome as well.
It'll be basically equivalent to Windows 7.
Quartz (especially after Quartz Extreme in 10.2), [...]
Is now at least half a generation behind Windows.
[...] a better windowing system, [...]
Not sure exactly what you mean here, but the overall sluggishness of OS X's window operations (eg: when resizing) and the lack of certain capabilities (resize from any side, maximise) suggest otherwise. Expose looks cool, but ultimately doesn't provide any more functionality than the Taskbar.
[...] more accurate font rendering, [...]
Can't say anything about this.
[...] better stability, [...]
Bollocks. At best, it's equivalent.
[...] better scripting environments (TCL and AppleScript vs VBScript and CMD), [...]
You can do a LOT with VBScript. Not to mention PowerShell.
[...] less security holes, [...]
Not according to the people who keep track of them.
[...] a much more rich command line environment that isn't feeble and borderline useless, [...]
PowerShell.
[...] API's that aren't near as annoying as Win32, free development tools, etc. I'll take XCode/Objective C over Visual C++ any day.
And I'm sure there are developers who prefer Windows.
NT has always been a joke and even though it was touted as a "unix killer", it has NEVER been able to do the deed.
It certainly replaced a lot of professional workstations and low-end servers that used to run UNIX.
Possibly, but who cares? The world is going multi-core, not multiple CPUs and I would suggest that Apple has the lead there, and will strengthen it with Grand Central, Grand Central Dispatch, and OpenCL.
The difference between multiple cores and multiple CPUs is not significant, and Apple is, in raw timeframe terms, a solid decade behind everyone else when it comes to optimising for SMP.
Read this if you don't believe me.
I'm sorry, but anything that opens with something like:
Apple is clearly a leader in implementing multi-core support, beginning with the first dual processor Power Macs 5 years ago, while the DayStar multi-processor Macs date back to the mid-90s.
Is so laughably biased and/or misinformed, it's not even worth reading.
ok, a pre-emptive multi-tasking os OS like the early 90s OS/2 had 16 bit drivers. so it is not considered a "real OS" and should be categorized in the same list as Windows 95? right.
That's not what I said at all. You were complaining about 16-bit code in Windows 95, so I pointed out that there was also 16-bit code in OS/2.
and I liked the part in that link where Intel didn't allow the people putting up the article to do their own testing. I was there, I ran it, it ran OS/2 apps much faster, it ran windows apps much slower and Intel had to come out with a new design because Windows 95 was not optimized for 32bit and that was all Microsoft had.
Rubbish. Microsoft had Windows NT (which was technically superior to OS/2 in every way).
Like I said, I was there and it was yet another example of Microsoft's marketing driving yet another poor OS experience for anyone who ran other OS's and new what an OS was supposed to be. What and OS should be.
I was also there, and an ex-OS/2 user, and Windows 95 did everything it was supposed to do and did it quite well.
However, getting back to the actual argument, the point is that claims of Windows 95 'running more slowly on a PPro than a Pentium because of 16-bit code' are flat-out wrong. If you benchmark a 32-bit application on a properly configured Windows 95 system, with both a PPro and Pentium of equivalent clock speed, the PPro will be faster.
OS X was far more advanced than NT/2000 or Win9x upon release.
In a couple of ways, yes. In most ways, no.
NextStep was FAR more advanced than WinNT or the classic MacOS which is funny considering the first version was released in 1988.
False. Support for multiple CPUs, is but one area where Windows NT was (and remains) superior.
Mac OS X used to be called NeXTstep, and NeXTstep had a dock which Windows 95 copied to create the task bar.
If you had actually used NeXTSTEP, you would know that its Dock and the Windows 95 Taskbar behave very differently. Much like the taskbar and the OS X Dock behave differently.
The Windows 95 look which came to be called the Windows classic look which was in fact a shameless but inferior copy of the NeXTstep look from 1988.
Rubbish. Application launching, task switching, menu interaction, window management - all these things were quite different in NeXT compared to Windows 95. Indeed, you'd struggle to find ways they were similar, that weren't also shared by every other GUI.
hrmmm... the Ars article gave me the impression that one of the benefits of OS X (and shortcomings of Windows' MDI model) [...]
One of the problems with the article is it talks about MDI as if it were still "the way", yet MDI has been deprecated since the release of Windows 95 (there was quite a hoo-ha about it at the time).
Now, certain applications have managed to keep the MDI fire burning, for whatever reason (eg: Excel is the poster child for this silliness), but developers are *supposed* to have been writing to a "one document:one window" model for nearly a decade and a half now.
There's also tabbed browsing, which is a sorta-but-not-really implementation of MDI. From a technical UI viewpoint it's atrocious - it takes all the problems inherent to MDI and adds a couple more besides - but for something like a browser which is mostly "read-only" use and the need for "cross-window-interaction" is low, it actually works OK.
Normally Ars stuff is pretty good, but that article is *very* ordinary, with a lot of conceptual, functional and historical errors.
The main thrust is correct, however, the Windows 7 Taskbar is clearly a descendant of its Windows 95 Great-great-grandfather, not the bastard child of NeXT and MacOS.
OSX and BeOS continue(d) to get faster with each new release.
OSX was so slow at release it had nowhere to go but up. Microsoft has never released a version of Windows that performed that badly. Relatively speaking, Vista was greased lightning at release compared to OSX 10.0 (10.1, 10.2, maybe even 10.3).
So network performance wasn't degraded severely when audio was played black like this [zdnet.com], right?
Not by DRM, it wasn't.