(unless you want to randomly skip through them and hope you get close, but even then, you need to memorise your playlist...)
Only if you care what the next song is.
I typically stick a few songs I think are interesting at the beginning, then fill it up with random good stuff. That way I can play the interesting songs (even if I haven't memorized them, I *can* recognise them), or just let it be random.
The Shuffle is like a good radio station with no chatter, no bad music, and a decent sized playlist.
Listen to up to 1 GB of songs, but never know what you're listening to!... and never be able to skip through songs at a rate of more than 1 per second!... and skip through them only randomly, or in sequential order, no playlists!
Yep, that's exactly what I wanted from an MP3 player. It's basically what my previous MP3 player did, before my daughter broke it the third time (managed to resuscitate it the first two)... about 3 years before the shuffle came out. After the iPod Mini came out and Jobs denigrated flash MP3 players I actually called Apple and bent their ears over it. I was so glad that the Shuffle came out a year later.
1GB - more than I had, which is good, but not out of control. that's about 8 hours of music, a decent day's playlist.
Don't know what you're listening to - I don't care, if I don't recognise it I'll catch the name some other time, or won't... if I don't like it I'll skip it.
Slow skip... takes me a few seconds to tell if the song is one I don't like, and the next one is random, so... meh.
No playlists? I choose my playlist for the day, change it once a week or so.
Shuffle is perfect for "background music for my life".
I mean, that's one of SCO's claims, the argument that code that is similar because it's functionally equivalent must have been copied. Which is ludicrous. I've written code and then found open source equivalents that had the same function names, the same variable names, almost the same code, because some things just are natural and obvious.
To argue that similarities imply copyright infringement would like claiming that if two stories have a dog named "spot" that rides a firetruck, that's copyright infringement. Software contains elements much more complex and much more constrained than narrative, so even large sequences of common code are not automatically evidence of copyright violation.
Run an operating system developed in a hostile multiuser environment instead of one developed for standalone PCs?
At Berkeley, we had professors and CS students using the same computers. There were hardly any "personal computers" back then: tests, assignments, and exams were sitting on the same machine as student accounts. The students were intelligent and highly motivated, and yet the security worked.
So, today, you have a choice of running your business on the descendants of that OS, or the OS that turned the science fantasy idea of an automatically executing 'worm' from "if you can even get someone else's computer to run untrusted code, that's a bug... nobody would be dumb enough to treat it as a feature" to the most common and disruptive network security problem in the world.
If you're going to cry about how the mini has no slots or additional internal drive bays, I'd like to point out the large number of nice, used Power Mac G4 and G5 models available on eBay
Used Powermac G5s still cost significantly more than comparably equipped PCs, and any G4 that's going to give you comparable performance to even a mini are almost as much. The affordable G4s are... well... not going to excite anyone. When I moved from a G4/466 to the Mini I got a HUGE boost in speed.
If you just want a machine that'll run OS X, you can buy a used G3 and pimp it out for under $100. My first machine running OS X was a used *7500* with a G3 upgrade card.
If you're not going to get something vaguely contemporary, you might as well drop back to 1998 and forget about it.
Yeh, I bought a Mac mini. For $800 with all the options. And I'm glad Apple made it available, someone must have hit Steve with a clue bat. But... damn... it's pretty trailing edge... I think Steve needs another whack (don't worry, it's made of foam, it won't hurt him).
Radeon 9200 with 32MB VRAM. If you luck out you might get one of the 64MB versions, but either way you can't use Quartz Extreme 3d... and you can't buy a 9200 with as little as 32M or even 64M for a PC... the smallest ones I found new have 128M. For $40.
ONE DIMM slot, max 1GB RAM.
Laptop drive, low power, low performance. Plugging in an external Firewire drive gets you a performance boost, but not as much as the people who hack their minis with a new case an 3.5" drive get.
The USB ports are below spec, and won't even charge an iPod Shuffle without an external powered hub.
It's quiet... unless you do something CPU-intensive like running a Flash animation (!) whereupon the fan goes into overdrive.
Making it a little bigger would have made it REALLY comparable to a $300 PC in hardware, but I can see why people who already have a PC are reluctant to go for the mini. If it ran Windows you'd be lucky to find a taker at $200.
I already paid hundreds dollars premium for the privilege of running OS X on an unexpandible desktop because Apple finally, after years of denial, realised that there was a market for desktop computers that didn't have a monitor bundled. I would pay that much of a premium just for the software, even if I had to buy a new PC to run it on, of that's what it took to get something between a Mac mini and the Powermac G5.
I don't want an all-in-one. Monitors are fragile and Apple's monitors and LCDs are lower resolution (yes, they have lots of pixels, but they take up too much screen area delivering them) than I want, so it's an external for me.
Apple has FINALLY started delivering laptops with decent resolution, but they are so unpleasant to use, with their flat "stylish but sloppy" keyboards and single-button touchpads, that I'd much rather buy a Thinkpad and OS X than a Macbook, even if the Thinkpad solution costs more.
Apple: I'll take a slab, like the NeXTstation, with 2 drive bays and 2 PCIE slots, or another IBM-uh-Lenovo collaboration, or OS X 86, and I'll pay you MORE money than you got in "Mac Tax" on my mini for the privilege... but I'm not spending 2 grand on a Powermac G5 I can't afford to use during the summer, or a laptop that I have to carry an external keyboard to use.
If the free and open software community was capable of producing Mac OS X, then it would have happened by now. And I say this as someone who is an active member of the free and open software community.
It's not just the consistent user interface (though we haven't managed that), or the slick user interface (though the open source "flashy" interfaces that I've seen so far are pretty nasty), it's the standard ABI that's consistent across versions and releases that allows vendors to distribute software in binary format and the reputation that allows them to trust that it's going to be supported long enoug for them to develop a product line and a business.
If it was just a "cool, flashy" OS, then Linux would have been a hit since Enlightenment was ready for use... and the latest versions of KDE and Gnome are pretty spiffy looking. But there's KDE and Gnome and there's 14 variants of Linux and BSD and Linux has had enough Red Flag Days to kill ten operating systems. Even if it had the market share, the only commercial software for Linux would still be limited to back-office products and companies doing it for altruistic and promotional purposes.
The problem is that Apple's refused to sell a decent ordinary desktop computer since they dumped the Beige G3, and their laptops favor style over comfort. I bought a Mac mini because the Powermac G5 is ludicrous for my purposes (I have to limit how much I use my computer in the summer as it is because the A/C can't handle the heat), but I'd have paid a few hundred more for something with a couple of PCI-E slots and drive bays, even if it wasn't a barnstormer. Or, for a copy of OS X I could use on a PC with those characteristics.
So, the open source software community can't produce the OS, and Apple refuses to take my money to run theirs on hardware I find more than barely acceptable. Me, I'm putting up with the mini and hoping they get their act together over 2006 and come up with a REAL desktop... but I don't have any problem understanding why other people are less patient.
BTW, which Apple-built PCI PowerMac was it that featured an Intel x86 cpu?
I didn't say it used an Intel x86 CPU. I said it was compatible with clones that ran Windows. This was back when Microsoft was pushing Windows on Power PC, Alpha, and MIPS, and you could get clones running Windows using all of those processors.
the MacIntel, which is literally Windows-compatible ROOB.
And you think this is a good idea from a security perspective?
What, that a program shuld come in an installer? No, I didn't say that. I said that when installation is required, it already comes in an installer.
I'd much rather the OS is hard wired to put dashboard widgets only in their proper place than that users are conditioned to run arbitrary "installer" applications
I didn't say that you would use "installer" applications for things like dashboard widgets. I said that, just like right now, Finder would call LaunchServices to find the right application to open (in this case, install) the widget.
The only difference is that Safari wouldn't be asking Finder to do it at the request of a potential attacker, but rather at the request of the user.
All you are adding is making it harder to do things because you think it will make users think more.
It's the same number of steps. It's not harder, it's a slightly different process that changes the operation from one that's initiated by the attacker to one that's initiated by the user.
And it's not making users think more, it's giving them a chance to think, and it really does work.
They are all based upon the premise that dialogue boxes are bad.
No, they are based on the premise that in this context dialog boxes do not add enough to the security of the system. They're not "bad", they're just not the right tool to use here.
So, adding a user interface designer to the picture is going to keep the customer from getting poo? I don't believe it for a second. Next year, they're going to add an application specialist because the problem is that none of the people in the chain know what the customer actually wants.
So instead of getting pretty poo, he'll get pretty *square* poo. At least it's easier to stack.
The real problem is that back in the "Golden Age", the customer was involved in the software development. In all the rest of the pictures, the customer is this unhappy blue guy off to the side... all the happy guys are on the LEFT side of the picture: if you're only on the right side, you know you're going to get poo.
So, take that blue guy, and stick him on the right.
That way, at least if he gets poo it's his own damn fault.
How can they explicitly request a program be installed if they don't know it needs to be installed.
If the program needs to be installed, what they will download will be an installer, so when they explicitly request it by opening it in Finder it will be installed. Similarly, other files that need to be installed already have handlers that Finder can use after the user requests it to install.
The only difference is that instead of facing the user with a "choose now" dialog box, the user makes an explicit request to open the object when they decide to. That small difference, between "responding to a dialog requesting action" and "making a request for action" makes a huge difference.
Good security lets the user do what they want and informs them what is going on.
I agree. Good security lets the user do what they want. Good security doesn't let a potential attacker do what they want, subject to a snap decision by the user.
I am not proposing making it hard for the user at all. For the normal case, the user will even end up performing the same number of actions: instead of clicking "Open" in the dialog box, they will select "Open" in the download manager.
Look users don't know they have to "install" a dashboard in a specific location.
I didn't say they should have to. I said they should explicitly request an install.
Please try reading for content.
If you have to click "infect me" to get a virus, then you are a lost cause or you probably really want a virus.
I didn't say I thought the people who clicked the equivalent button at work were the brightest bulbs in the box, but that didn't mean I didn't have to go in and clean up after them.
THe Mac guy who insisted he didn't need to set WEP security on his Wifi access point because Macs couldn't get viruses wasn't any brighter.
The people who provide the pool of infected machines on any platform are the source of all kinds of great tech support stories, but you can't ignore them, and when it comes to security you have to design your software with these guys in mind.
The fragmentary evidence there is available to non-insiders suggests that Apple is selling in spite of OSX, not because of it, and that many more people would buy the hardware if they could get it with the OS they prefer.
People CAN get comparable hardware with Windows XP, and they can get it for less.
Apple would have to cut hundreds of dollars from their prices to go head-to-head with a generic Windows solution.
Apple has already switched to 100% Windows-compatible hardware... This has never been true of the Mac hardware environment before.
Not so. Remember back when Microsoft was pushing Windows NT on multiple processors? The PCI-based Powermacs were pretty much as "100% compatible" as these machines are.
It wasn't particularly useful, and AFAIK nonody implemented the missing drivers, and people were more interested in buying Power PC clones to run Mac OS on, but...
It should read "Safari should NEVER have an option for any application that displays untrusted content to open files automatically after downloading."
No.
"No application (such as Safari, but not limited to Safari) should have an option to automatically open foreign untrusted content automatically after downloading.
Excepting of course that Apple does not use Yes/No dialogue boxes
It doesn't matter if "Yes" is marked "Yes", "Open", "OK", or "Whack me".
It doesn't matter if "No" is marked "No", "Cancel", "Abandon", or "Go away".
The point isn't what the dialog says, the point is that the dialog presents and accept/deny choice, that accept is the normal response, and the system as a whole presents accept/deny choices on a regular enough basis for it to become a reflex.
People click those buttons before they are consciously aware of what the dialog says or what labels are on the buttons. Windows applications don't always label the buttons "yes" and "no" either, and people still get caught.
Do you really think it is better to get users used to having to take the downloaded file and put it in some random path?
No, and I didn't say they should.
I said that it shouldn't do it automatically without an explicit request.
Clicking "yes", "install", "ok", or "infect me" on a dialog box is not a request, it's approving an action that the untrusted content has requested.
Re:Time to get into the habit of read-only filesys
on
First Mac OS X Virus?
·
· Score: 1
1. This wouldn't provide ANY more protection than what Mac OS X does already. 2. This is easily bypassed by storing the malware under $HOME.
I much prefer my browser not to download ANYTHING unless I tell it to.
Unfortunately, if it hits a link it doesn't know how to render, it downloads it. You can't turn that off... but you can keep it from using LaunchServices to open it after it's downloaded.
I turn off "Mount Disk Images" and "Continue to Expand" in Stuffit Expander, for the same reason.
Yes, if I install a program on my computer that abuses the "safe files" then bad things could happen.
You're thinking backwards.
We're not talking about an installed program deliberately abusing "open safe files".
We're talking about an installed program having a bug in it that allows it to be exploited, like the "help" or "man" bugs that have already shown up and been patched, individually, or like the "chm" bug on Windows.
If you turn off "open safe files", none of those bugs can be automatically exploited.
I think it's much more important to fix that exploit than to just tell Safari not to open them for you
If you have a *real* JPEG file, Safari will open it itself. That's not the problem, it's the next ".foo" file that is run by "BuggyApplication" that happens to be widely installed. If you turn off "open safe files" you won;t be one of the people who are exploited by the ".foo hole".
So in this case how is OS X encouraging you to hose your home dir?
I'm talking about cases where "Open safe files after downloading" is relevant.
I already said that if this exploit doesn't take advantage of that, it's not a problem, but there have been in the past and will be in the future applications registered with LaunchServices that ARE exploitable. Safari should not act as an enabler for those exploits.
As for the botnet, that's true, but it'll get turned off when they log out.
Not if they know about fopen, fwrite, fork, setsid, and execl.
[example code and other comments deleted lest some idiot think it's a big secret and flame me for "posting an exploit to slashdot"]
Re:You own a Mac MINI ?!? I'm so confused...
on
OSx86 Cracked Again
·
· Score: 1
I'd ask Apple to expand its hardware offerings long before I'd ask it to support another company's hardware
If I believed Apple was likely to even consider offering the kind of laptop or desktop I want, I'd be 100% in agreement.
An iSlab, with two 3.5" and one 5.25" drive bays, one or two PCI-E slots (one 16x), choice of single or dual core (or even an MPC8641D, but that's crazy talk these days), at least two RAM slots up to 2GB, and a bundled 5200fx (minimum requirement for QE3d, IIRC).
It shouldn't be much more than $600 for a solo, maybe $800 for a duo.
Slashdot Mirror
Safari will not automatically open things anyway, and will ask you before decompressing anything with an executable in it.
Oh, Really?
Yes I can bloody well say "I told you so".
(unless you want to randomly skip through them and hope you get close, but even then, you need to memorise your playlist...)
Only if you care what the next song is.
I typically stick a few songs I think are interesting at the beginning, then fill it up with random good stuff. That way I can play the interesting songs (even if I haven't memorized them, I *can* recognise them), or just let it be random.
The Shuffle is like a good radio station with no chatter, no bad music, and a decent sized playlist.
Listen to up to 1 GB of songs, but never know what you're listening to! ... and never be able to skip through songs at a rate of more than 1 per second! ... and skip through them only randomly, or in sequential order, no playlists!
Yep, that's exactly what I wanted from an MP3 player. It's basically what my previous MP3 player did, before my daughter broke it the third time (managed to resuscitate it the first two)... about 3 years before the shuffle came out. After the iPod Mini came out and Jobs denigrated flash MP3 players I actually called Apple and bent their ears over it. I was so glad that the Shuffle came out a year later.
1GB - more than I had, which is good, but not out of control. that's about 8 hours of music, a decent day's playlist.
Don't know what you're listening to - I don't care, if I don't recognise it I'll catch the name some other time, or won't... if I don't like it I'll skip it.
Slow skip... takes me a few seconds to tell if the song is one I don't like, and the next one is random, so... meh.
No playlists? I choose my playlist for the day, change it once a week or so.
Shuffle is perfect for "background music for my life".
So you think they should go back to a 2d top- or side- scrolling ortho view for Zelda Revolution, or do you have something else in mind?
I wish people would look up the original press release instead of advertising the physorg tarpit.
Here.
(yes, all the stupid "teleportation" stuff was in the original)
How's IBM doing it in the SCO lawsuit over Linux?
I mean, that's one of SCO's claims, the argument that code that is similar because it's functionally equivalent must have been copied. Which is ludicrous. I've written code and then found open source equivalents that had the same function names, the same variable names, almost the same code, because some things just are natural and obvious.
To argue that similarities imply copyright infringement would like claiming that if two stories have a dog named "spot" that rides a firetruck, that's copyright infringement. Software contains elements much more complex and much more constrained than narrative, so even large sequences of common code are not automatically evidence of copyright violation.
Run an operating system developed in a hostile multiuser environment instead of one developed for standalone PCs?
At Berkeley, we had professors and CS students using the same computers. There were hardly any "personal computers" back then: tests, assignments, and exams were sitting on the same machine as student accounts. The students were intelligent and highly motivated, and yet the security worked.
So, today, you have a choice of running your business on the descendants of that OS, or the OS that turned the science fantasy idea of an automatically executing 'worm' from "if you can even get someone else's computer to run untrusted code, that's a bug... nobody would be dumb enough to treat it as a feature" to the most common and disruptive network security problem in the world.
Pick your poison.
If you're going to cry about how the mini has no slots or additional internal drive bays, I'd like to point out the large number of nice, used Power Mac G4 and G5 models available on eBay
Used Powermac G5s still cost significantly more than comparably equipped PCs, and any G4 that's going to give you comparable performance to even a mini are almost as much. The affordable G4s are... well... not going to excite anyone. When I moved from a G4/466 to the Mini I got a HUGE boost in speed.
If you just want a machine that'll run OS X, you can buy a used G3 and pimp it out for under $100. My first machine running OS X was a used *7500* with a G3 upgrade card.
If you're not going to get something vaguely contemporary, you might as well drop back to 1998 and forget about it.
New Mac Minis are $500.00.
Yeh, I bought a Mac mini. For $800 with all the options. And I'm glad Apple made it available, someone must have hit Steve with a clue bat. But... damn... it's pretty trailing edge... I think Steve needs another whack (don't worry, it's made of foam, it won't hurt him).
Radeon 9200 with 32MB VRAM. If you luck out you might get one of the 64MB versions, but either way you can't use Quartz Extreme 3d... and you can't buy a 9200 with as little as 32M or even 64M for a PC... the smallest ones I found new have 128M. For $40.
ONE DIMM slot, max 1GB RAM.
Laptop drive, low power, low performance. Plugging in an external Firewire drive gets you a performance boost, but not as much as the people who hack their minis with a new case an 3.5" drive get.
The USB ports are below spec, and won't even charge an iPod Shuffle without an external powered hub.
It's quiet... unless you do something CPU-intensive like running a Flash animation (!) whereupon the fan goes into overdrive.
Making it a little bigger would have made it REALLY comparable to a $300 PC in hardware, but I can see why people who already have a PC are reluctant to go for the mini. If it ran Windows you'd be lucky to find a taker at $200.
I already paid hundreds dollars premium for the privilege of running OS X on an unexpandible desktop because Apple finally, after years of denial, realised that there was a market for desktop computers that didn't have a monitor bundled. I would pay that much of a premium just for the software, even if I had to buy a new PC to run it on, of that's what it took to get something between a Mac mini and the Powermac G5.
I don't want an all-in-one. Monitors are fragile and Apple's monitors and LCDs are lower resolution (yes, they have lots of pixels, but they take up too much screen area delivering them) than I want, so it's an external for me.
Apple has FINALLY started delivering laptops with decent resolution, but they are so unpleasant to use, with their flat "stylish but sloppy" keyboards and single-button touchpads, that I'd much rather buy a Thinkpad and OS X than a Macbook, even if the Thinkpad solution costs more.
Apple: I'll take a slab, like the NeXTstation, with 2 drive bays and 2 PCIE slots, or another IBM-uh-Lenovo collaboration, or OS X 86, and I'll pay you MORE money than you got in "Mac Tax" on my mini for the privilege... but I'm not spending 2 grand on a Powermac G5 I can't afford to use during the summer, or a laptop that I have to carry an external keyboard to use.
If the free and open software community was capable of producing Mac OS X, then it would have happened by now. And I say this as someone who is an active member of the free and open software community.
It's not just the consistent user interface (though we haven't managed that), or the slick user interface (though the open source "flashy" interfaces that I've seen so far are pretty nasty), it's the standard ABI that's consistent across versions and releases that allows vendors to distribute software in binary format and the reputation that allows them to trust that it's going to be supported long enoug for them to develop a product line and a business.
If it was just a "cool, flashy" OS, then Linux would have been a hit since Enlightenment was ready for use... and the latest versions of KDE and Gnome are pretty spiffy looking. But there's KDE and Gnome and there's 14 variants of Linux and BSD and Linux has had enough Red Flag Days to kill ten operating systems. Even if it had the market share, the only commercial software for Linux would still be limited to back-office products and companies doing it for altruistic and promotional purposes.
The problem is that Apple's refused to sell a decent ordinary desktop computer since they dumped the Beige G3, and their laptops favor style over comfort. I bought a Mac mini because the Powermac G5 is ludicrous for my purposes (I have to limit how much I use my computer in the summer as it is because the A/C can't handle the heat), but I'd have paid a few hundred more for something with a couple of PCI-E slots and drive bays, even if it wasn't a barnstormer. Or, for a copy of OS X I could use on a PC with those characteristics.
So, the open source software community can't produce the OS, and Apple refuses to take my money to run theirs on hardware I find more than barely acceptable. Me, I'm putting up with the mini and hoping they get their act together over 2006 and come up with a REAL desktop... but I don't have any problem understanding why other people are less patient.
an experimental OS written to support multiple cpus
What are you talking about?
Windows NT 3.51? An experimental OS? Written for a limited number of hardware environments?
I think you're pretty fundamentally misunderstanding something, but I'm not going to try and figure it out.
BTW, which Apple-built PCI PowerMac was it that featured an Intel x86 cpu?
I didn't say it used an Intel x86 CPU. I said it was compatible with clones that ran Windows. This was back when Microsoft was pushing Windows on Power PC, Alpha, and MIPS, and you could get clones running Windows using all of those processors.
the MacIntel, which is literally Windows-compatible ROOB.
Except that it isn't.
And you think this is a good idea from a security perspective?
What, that a program shuld come in an installer? No, I didn't say that. I said that when installation is required, it already comes in an installer.
I'd much rather the OS is hard wired to put dashboard widgets only in their proper place than that users are conditioned to run arbitrary "installer" applications
I didn't say that you would use "installer" applications for things like dashboard widgets. I said that, just like right now, Finder would call LaunchServices to find the right application to open (in this case, install) the widget.
The only difference is that Safari wouldn't be asking Finder to do it at the request of a potential attacker, but rather at the request of the user.
All you are adding is making it harder to do things because you think it will make users think more.
It's the same number of steps. It's not harder, it's a slightly different process that changes the operation from one that's initiated by the attacker to one that's initiated by the user.
And it's not making users think more, it's giving them a chance to think, and it really does work.
They are all based upon the premise that dialogue boxes are bad.
No, they are based on the premise that in this context dialog boxes do not add enough to the security of the system. They're not "bad", they're just not the right tool to use here.
So, adding a user interface designer to the picture is going to keep the customer from getting poo? I don't believe it for a second. Next year, they're going to add an application specialist because the problem is that none of the people in the chain know what the customer actually wants.
So instead of getting pretty poo, he'll get pretty *square* poo. At least it's easier to stack.
The real problem is that back in the "Golden Age", the customer was involved in the software development. In all the rest of the pictures, the customer is this unhappy blue guy off to the side... all the happy guys are on the LEFT side of the picture: if you're only on the right side, you know you're going to get poo.
So, take that blue guy, and stick him on the right.
That way, at least if he gets poo it's his own damn fault.
How can they explicitly request a program be installed if they don't know it needs to be installed.
If the program needs to be installed, what they will download will be an installer, so when they explicitly request it by opening it in Finder it will be installed. Similarly, other files that need to be installed already have handlers that Finder can use after the user requests it to install.
The only difference is that instead of facing the user with a "choose now" dialog box, the user makes an explicit request to open the object when they decide to. That small difference, between "responding to a dialog requesting action" and "making a request for action" makes a huge difference.
Good security lets the user do what they want and informs them what is going on.
I agree. Good security lets the user do what they want. Good security doesn't let a potential attacker do what they want, subject to a snap decision by the user.
I am not proposing making it hard for the user at all. For the normal case, the user will even end up performing the same number of actions: instead of clicking "Open" in the dialog box, they will select "Open" in the download manager.
Look users don't know they have to "install" a dashboard in a specific location.
I didn't say they should have to. I said they should explicitly request an install.
Please try reading for content.
If you have to click "infect me" to get a virus, then you are a lost cause or you probably really want a virus.
I didn't say I thought the people who clicked the equivalent button at work were the brightest bulbs in the box, but that didn't mean I didn't have to go in and clean up after them.
THe Mac guy who insisted he didn't need to set WEP security on his Wifi access point because Macs couldn't get viruses wasn't any brighter.
The people who provide the pool of infected machines on any platform are the source of all kinds of great tech support stories, but you can't ignore them, and when it comes to security you have to design your software with these guys in mind.
The fragmentary evidence there is available to non-insiders suggests that Apple is selling in spite of OSX, not because of it, and that many more people would buy the hardware if they could get it with the OS they prefer.
People CAN get comparable hardware with Windows XP, and they can get it for less.
Apple would have to cut hundreds of dollars from their prices to go head-to-head with a generic Windows solution.
Apple has already switched to 100% Windows-compatible hardware ... This has never been true of the Mac hardware environment before.
Not so. Remember back when Microsoft was pushing Windows NT on multiple processors? The PCI-based Powermacs were pretty much as "100% compatible" as these machines are.
It wasn't particularly useful, and AFAIK nonody implemented the missing drivers, and people were more interested in buying Power PC clones to run Mac OS on, but...
It should read "Safari should NEVER have an option for any application that displays untrusted content to open files automatically after downloading."
No.
"No application (such as Safari, but not limited to Safari) should have an option to automatically open foreign untrusted content automatically after downloading.
Excepting of course that Apple does not use Yes/No dialogue boxes
It doesn't matter if "Yes" is marked "Yes", "Open", "OK", or "Whack me".
It doesn't matter if "No" is marked "No", "Cancel", "Abandon", or "Go away".
The point isn't what the dialog says, the point is that the dialog presents and accept/deny choice, that accept is the normal response, and the system as a whole presents accept/deny choices on a regular enough basis for it to become a reflex.
People click those buttons before they are consciously aware of what the dialog says or what labels are on the buttons. Windows applications don't always label the buttons "yes" and "no" either, and people still get caught.
Do you really think it is better to get users used to having to take the downloaded file and put it in some random path?
No, and I didn't say they should.
I said that it shouldn't do it automatically without an explicit request.
Clicking "yes", "install", "ok", or "infect me" on a dialog box is not a request, it's approving an action that the untrusted content has requested.
1. This wouldn't provide ANY more protection than what Mac OS X does already.
2. This is easily bypassed by storing the malware under $HOME.
I much prefer my browser not to download ANYTHING unless I tell it to.
Unfortunately, if it hits a link it doesn't know how to render, it downloads it. You can't turn that off... but you can keep it from using LaunchServices to open it after it's downloaded.
I turn off "Mount Disk Images" and "Continue to Expand" in Stuffit Expander, for the same reason.
Yes, if I install a program on my computer that abuses the "safe files" then bad things could happen.
You're thinking backwards.
We're not talking about an installed program deliberately abusing "open safe files".
We're talking about an installed program having a bug in it that allows it to be exploited, like the "help" or "man" bugs that have already shown up and been patched, individually, or like the "chm" bug on Windows.
If you turn off "open safe files", none of those bugs can be automatically exploited.
I think it's much more important to fix that exploit than to just tell Safari not to open them for you
If you have a *real* JPEG file, Safari will open it itself. That's not the problem, it's the next ".foo" file that is run by "BuggyApplication" that happens to be widely installed. If you turn off "open safe files" you won;t be one of the people who are exploited by the ".foo hole".
So in this case how is OS X encouraging you to hose your home dir?
I'm talking about cases where "Open safe files after downloading" is relevant.
I already said that if this exploit doesn't take advantage of that, it's not a problem, but there have been in the past and will be in the future applications registered with LaunchServices that ARE exploitable. Safari should not act as an enabler for those exploits.
As for the botnet, that's true, but it'll get turned off when they log out.
Not if they know about fopen, fwrite, fork, setsid, and execl.
[example code and other comments deleted lest some idiot think it's a big secret and flame me for "posting an exploit to slashdot"]
I'd ask Apple to expand its hardware offerings long before I'd ask it to support another company's hardware
If I believed Apple was likely to even consider offering the kind of laptop or desktop I want, I'd be 100% in agreement.
An iSlab, with two 3.5" and one 5.25" drive bays, one or two PCI-E slots (one 16x), choice of single or dual core (or even an MPC8641D, but that's crazy talk these days), at least two RAM slots up to 2GB, and a bundled 5200fx (minimum requirement for QE3d, IIRC).
It shouldn't be much more than $600 for a solo, maybe $800 for a duo.