Fortunately, Sun realized that mistake with ZFS - the byte order is determined by the machine that formatted the drive (zpool), but the ZFS driver will swap bytes if needed.
Even without extending UFS this can be implemented in UFS drivers... there are a plethora of elements in the superblock that can be used to determine the byte order and figure out other compatibility problems. Going forward, the superblock can be extended to explicitly identify the variants.
Extending one implementation to support read-write access of the others is the first step. Given how many implementations are open-source, there's nothing technically in the way.
It's not Linux, but who cares? UNIX is UNIX is UNIX, whether it's called AIX or Interix or HPUX or Linux or FreeBSD or Solaris or OSX, anyone who can write portable UNIX or UNIX+X11 or UNIX+Tk or UNIX+Gtk software can 'write one, run anywhere'.
Even on Windows.
Don't get hung up on one brand. It's bad for your blood pressure.
The really crazy part of this is that if you don't want to buy Office or CheapOffice, but instead want to by SmallFastWord instead, you can't because SmallFastSoftware is out of business, the open source community is focussed on coming up with a replacement for Office, and when you try out Pages on your Mac (thinking it might be a NotReallySmall, OrFast, ButAtLeastNotMessedUp Word) you discover that it's got the same problems shared by Word and Open Office... because everyone's trying to be compatible with Office. The competition is afraid to go in any direction that the piracy-supported monopoly leads.
Which gives said monopoly yet another advantage that should pay for an AWFUL lot of putting-up-with-piracy.
The distribution of copies of Microsoft products that have, shall we say, "uncertain provenance" has frequently been a major factor in Microsoft's ability to infiltrate new markets. At times Microsoft has allowed users to use at home... on their own computer... the Microsoft software they use at work, if not explicitly then with rules and guidelines so worded that people could be forgiven believing it was legal.
When people have said that piracy was "hitting back at Microsoft" that's always been a sure way to elicit a response like this from me... because by using a pirated copy of Word or Excel instead of a cheaper product that you could actually afford, that helped cut off the oxygen to Microsoft's competitors. Microsoft's only taken a strong stand and employed strong copy protection after they've become so dominant in a market that there was no chance of a significant number of people bailing out and buying a cheaper product rather than paying Microsoft's fees.
I've been hoping that Microsoft would continue their crackdowns and push people away from Windows. Hopefully the won't read this article.:)
Steve hates docking stations, colored LEDs, and ergonomic keyboards as well. Steve hates ugly monitors on nice Macs. Steve hates user-upgradable and user-repairable devices. Steve hates DRM.
Basically the judge also found that it is not sufficient to only refer to a URL to access the source code and license text . This would be fine if the software product itself was also distribited over the internet, but in any other case that is not sufficient.
Moore's law is not about exponential increases in absolute performance, it's about exponential increases in performance PER UNIT COST. The original formulation was based on the fact that the number of transistors in a chip using the CHEAPEST transistors was doubling every 24 months.
It doesn't matter whether you get twice the performance for the same price, or the same performance for half the price (and half or less the power usage), you're still following Moore's Law.
The really interesting thing is that Moore's Law applies to everything we make. The doubling time depends on the technology, but the best performance-per-unit-price for every technological product from oxcarts and clay tablets to rockets and ebooks can be shown to follow an exponential curve back as far as we have hard enough figures to plot meaningful points.
The actual problem is that they did not include the text of the GPL with the phone. The summary here in Slashdot didn't mention that, and had me wondering what the problem with the provided URL was.
If a politician votes against me... they have to be corrupt.
That's because the second part of the sentence is all too likely true, no matter what the first part says.
There is a probably unavoidable conflict of interest in politics that makes it hard to imagine someone making it to Washington with their soul intact and unpledged to their funding. It's probably not completely true that an honest politician is one who stays bought, but there's more than enough to it to make us regular joes uneasy.
(Mostly this is a criticism that Apple doesn't include a simple and transparent preference for disabling Bonjour.)
I agree, and the many well-popularized techniques for disabling Bonjour on public LANs are only a mitigating factor... it's not an excuse for Apple to have left this out of the preferences.
But that's why I wrote "no routed open ports by default". Bonjour/Rendezvous/Zeroconf only talks on link-local addresses. It *is* technically possible for those addresses to be routed, but it would take an unlikely level of misconfiguration for them to be routed beyond a local LAN, let alone even a couple of hops past your ISP's access point. Unless you're deliberately routing it over a tunnel (in which case you know you're doing it, or you're already owned) it really can't be attacked over the Internet like LAN Manager and SMS can.
The biggest exposure is in combination with insecure home WiFi, and if you've got that you're already skiing naked through a briar patch.
As long as you don't mind giving up on-demand video, and having to queue your downloads for later viewing, you don't need a cable bill.
Well, you're either downloading that over cable modem, or you're downloading it over DSL. You're watching TV over the airwaves, over cable, or (increasingly) over DSL. You're paying for it either way... and they will without question adjust the relative costs of data and video over broadband to match demand and usage.
All they need to do is to get the advertising back in, by doing something like offering high quality seeds of better-than-ripped-quality un-stripped versions of the shows, and we're back where we started.
Windows CE as used in 'Windows Powered' devices is pretty much a desktop OS. The WinCE API is derived from Win32, cleaned up and modularized and with its own set of libraries and a real-time kernel. It does support a traditional embedded OS model where code is executed in place from whatever file system is wrapped around it, but the "Windows Powered" handhelds don't work that way.
The first WinCE-based handhelds were pretty much "laptop replacements" with stripped down versions of Windows applications that run by copying them from a file system to RAM, explicitly use an open/read/write/seek/close mechanism to access files, and so on. There's a set of database calls for PDA applications that run on top of this. Microsoft subsequently stripped down the applications, removed the more desktop-like ones, and repurposed Windows Powered handhelds as "Palm Killers". By the time Palm lost the plot and sent haring off trying to port BeOS to the Palm in a quixotic attempt at fighting Microsoft on a field Microsoft was already abandoning they'd done a good enough job at "cloning" enough of Palm's look and feel that their current character recognizer is a better clone of Graffiti than Palm's current offering... but under the hood they're very "desktop-like".
I haven't worked with Symbian devices, so I can't say if they're more like a stripped down desktop or a classical embedded system like Palm, but the older high-end devices certainly looked more like a desktop.
The OS isn't the problem, here. It's Safari. The comments about finding crashes in Safari make me suspect that this is probably a stack/buffer overflow attack. If it's easy to crash Safari on the iPhone then they've got problems in the implementation of Safari on the iPhone... especially in the extensions to webcore that are unique to the device. If Pocket Internet Explorer had the same problems, then Windows CE would have the same exposure (luckily for Windows CE users, Pocket IE seems to be the most secure version of IE out... probably due to the fact that it doesn't include the same kind of "active content" support as the desktop version).
And the original article is right, the presence or absence of an official dev kit has very little to do with this... it just makes it harder to switch from Safari to another browser while Apple is easing Safari on the iPhone through its birth trauma.
It's more likely to introduce new 976-style exploits, through autoloading form using "tel:" URIs. The iPhone is not a real smartphone because it has no native API.
The silver lining in the dark cloud of the iPhone's lack of a native API is that there's no mechanism to install any kind of worm, rootkit, or spyware on the device.
But of course it's silly to worry about the iPhone, not when there's so many REAL smartphones out there that actually have the technical capabiilty of supporting the kind of viral ecosystem you're talking about... and have been for several years. We haven't seen a mass flowering of malware for Symbian- Palm- and Windows-powered devices yet, and they're SO much more credible targets for attack.
You hear a lot about Lockheed Martin's "Skunk Works", but Northrop Grumman keeps a lower profile in the "Crazy Ideas That Just Might Work" department. Perhaps they're looking to change that.
It's not really different from what happened to RMS at MIT! He too felt so harassed in his freedom to share information that he finally came up with a fantastic alternative. There would be no OSS today, had they been more liberal at MIT AI-lab back then.
Open Source Software owes at least as much to Berkeley's liberal attitude, *and* MIT's liberal attitude, as to RMS. RMS has effectively and unfairly demonized the AI lab, and he's been given too much credit for being one of the more visible rocks in an avalanche of open source software that was already in motion well before he penned the GNU manifesto.
The idea that MIT-AI, MIT-MC, and the rest... running an OS that let any user be "root" and giving accounts to outsiders (including Larry Niven and Jerry Pournelle)... were insufficiently "liberal", or that there would be no FOSS without RMS... is distressingly common today, and completely at odds with what the growing open-systems and free-source software community was actually like at the time.
"Most of the machines I see pitched in catalogs are in the $700 range, certainly under $1,000," said Cherry. "Computers with that amount of hardware are a better fit for XP. With Vista's requirements, people may be thinking about sticking with XP, and putting less money into the hardware."
OK, let's put this into perspective. We're not talking about some processor-hungry application here, we're talking about the overhead of teh OS itself.
Several years ago I was dual-booting FreeBSD and Windows on my main computer. It had 128M of RAM and an unaccelerated video card. It was getting time to upgrade, so I got a used Powermac G3 and added a G4 upgrade card and installed Jaguar on it. Jaguar was specced for a minimum of 256MB RAM and really required about 128M more than that to be happy. I was able to get 768MB of old RAM for that box cheap, and that was fine. I transferred all my UNIX stuff to it, it was faster than my old dual boot box, but I had that as a backup. I upgraded to a Mac mini when that came out, with 512MB of RAM running Panther, and it was so much faster even though it had less RAM. By this time my old dual boot box was only ever running Windows. With 256MB and a geforce 4 Ti-200 and perfectly reasonable for most of the Windows apps I needed to run.
I was a little annoyed that OSX required more memory than Windows. I'd been accustomed to mainstream UNIX that had gone from being "bloated" (since it required several hundred KILObytes to run and OVER A MEGABYTE to be happy) to being "lightweight" (Kids today, you tell them about punch cards and drum memory and they believe you, you tell them about running a desktop with a GUI in less than a megabyte and they think you're kidding), with Windows being the memory hog... but Windows was working well on a system that I knew OSX would balk at if it was a Power PC.
But I had applications that needed more than that.
So a year or so ago I put together what I thought was a pretty nice computer, with 2GB of RAM, a dual core Athlon 64 CPU, a gf7600 video card, and it was under $600. If anything, I thought I was going pretty gold-plated.
So I'm completely boggled at the idea that an "under $1000" computer would be considered inadequate for Vista. He's just talking about the *extra overhead* of the OS here, the applications (which for me had been what was driving my hardware requirements) would be the same on XP and Vista... what the HELL is Vista doing that makes this kind of machine "inadeqate"?
It's NOT just the eye candy. I already took the "eye candy" hit when I switched from FreeBSD/Windows to OSX. My first generation Mac mini with a 1.42 MHz PPC and 512M RAM and a 32M Radeon 9200 GPU handles all the eye candy I can stand _just_fine_. The GPU isn't good enough for the Quartz Extreme effects, but it's one that was 'trailing edge' when I got the computer _three_years_ ago. I can't imagine what kind of eye candy they could be including that any desktop or laptop HOWEVER CHEAP sold in the past three years (let alone on sale today) would have a problem with.
So just what is it that Vista requires that brings _the_OS_requirements_ up from the 256M RAM and a geforce 4 to something that my "cheap and nasty" six-hundred-dollars-over-a-year-ago 2GB plus geforce 7600 box would supposedly balk at? Or is this guy just making excuses for Microsoft?
The question shouldn't be "will MySpace disrupt TV?", it's "How much more will MySpace disrupt TV than streaming video already has?". The current leader in the "most televisions staked" vampire hunter contest has to be YouTube...
It's got nothing to do with whether it's a video game.
If you're playing Pacman in an arcade, and you discover that the machine lets you run through the ghosts if you move through the tunnel from the right side of the screen to the left just as they're moving in the other direction, that's you against the house.
Similarly, if you're playing Poker and you notice the deck is marked, that's PvP, not you against the house.
Slashdot Mirror
Fortunately, Sun realized that mistake with ZFS - the byte order is determined by the machine that formatted the drive (zpool), but the ZFS driver will swap bytes if needed.
Even without extending UFS this can be implemented in UFS drivers... there are a plethora of elements in the superblock that can be used to determine the byte order and figure out other compatibility problems. Going forward, the superblock can be extended to explicitly identify the variants.
Extending one implementation to support read-write access of the others is the first step. Given how many implementations are open-source, there's nothing technically in the way.
Mac users want high-quality applications.
Applications to theme the iPhone?
That sounds more like a Linux thing.
I wonder who'll port Enlightenment to the iPhone first?
(HHOS)
It's not Linux, but who cares? UNIX is UNIX is UNIX, whether it's called AIX or Interix or HPUX or Linux or FreeBSD or Solaris or OSX, anyone who can write portable UNIX or UNIX+X11 or UNIX+Tk or UNIX+Gtk software can 'write one, run anywhere'.
Even on Windows.
Don't get hung up on one brand. It's bad for your blood pressure.
So you think this is just a way to get you to buy "The White Album" again?
The really crazy part of this is that if you don't want to buy Office or CheapOffice, but instead want to by SmallFastWord instead, you can't because SmallFastSoftware is out of business, the open source community is focussed on coming up with a replacement for Office, and when you try out Pages on your Mac (thinking it might be a NotReallySmall, OrFast, ButAtLeastNotMessedUp Word) you discover that it's got the same problems shared by Word and Open Office... because everyone's trying to be compatible with Office. The competition is afraid to go in any direction that the piracy-supported monopoly leads.
Which gives said monopoly yet another advantage that should pay for an AWFUL lot of putting-up-with-piracy.
The distribution of copies of Microsoft products that have, shall we say, "uncertain provenance" has frequently been a major factor in Microsoft's ability to infiltrate new markets. At times Microsoft has allowed users to use at home... on their own computer... the Microsoft software they use at work, if not explicitly then with rules and guidelines so worded that people could be forgiven believing it was legal.
:)
When people have said that piracy was "hitting back at Microsoft" that's always been a sure way to elicit a response like this from me... because by using a pirated copy of Word or Excel instead of a cheaper product that you could actually afford, that helped cut off the oxygen to Microsoft's competitors. Microsoft's only taken a strong stand and employed strong copy protection after they've become so dominant in a market that there was no chance of a significant number of people bailing out and buying a cheaper product rather than paying Microsoft's fees.
I've been hoping that Microsoft would continue their crackdowns and push people away from Windows. Hopefully the won't read this article.
SSH is easy. You can run SSH on *Windows* for god's sake.
Steve hates docking stations, colored LEDs, and ergonomic keyboards as well. Steve hates ugly monitors on nice Macs. Steve hates user-upgradable and user-repairable devices. Steve hates DRM.
Well, he's not all bad. I hate DRM too.
Something that's got good Cocoa bindings, anyway, so you can write native apps in them...
Basically the judge also found that it is not sufficient to only refer to a URL to access the source code and license text . This would be fine if the software product itself was also distribited over the internet, but in any other case that is not sufficient.
Note text in boldface.
Furrfu.
The Internet is a "medium customarily used for software interchange". And a URL is a "written offer to provide the source code".
And it's distributed on a device that needs an internet connection to work.
Moore's law is not about exponential increases in absolute performance, it's about exponential increases in performance PER UNIT COST. The original formulation was based on the fact that the number of transistors in a chip using the CHEAPEST transistors was doubling every 24 months.
It doesn't matter whether you get twice the performance for the same price, or the same performance for half the price (and half or less the power usage), you're still following Moore's Law.
The really interesting thing is that Moore's Law applies to everything we make. The doubling time depends on the technology, but the best performance-per-unit-price for every technological product from oxcarts and clay tablets to rockets and ebooks can be shown to follow an exponential curve back as far as we have hard enough figures to plot meaningful points.
The actual problem is that they did not include the text of the GPL with the phone. The summary here in Slashdot didn't mention that, and had me wondering what the problem with the provided URL was.
If a politician votes against me... they have to be corrupt.
That's because the second part of the sentence is all too likely true, no matter what the first part says.
There is a probably unavoidable conflict of interest in politics that makes it hard to imagine someone making it to Washington with their soul intact and unpledged to their funding. It's probably not completely true that an honest politician is one who stays bought, but there's more than enough to it to make us regular joes uneasy.
Only too happy to be of service!
(Mostly this is a criticism that Apple doesn't include a simple and transparent preference for disabling Bonjour.)
I agree, and the many well-popularized techniques for disabling Bonjour on public LANs are only a mitigating factor... it's not an excuse for Apple to have left this out of the preferences.
But that's why I wrote "no routed open ports by default". Bonjour/Rendezvous/Zeroconf only talks on link-local addresses. It *is* technically possible for those addresses to be routed, but it would take an unlikely level of misconfiguration for them to be routed beyond a local LAN, let alone even a couple of hops past your ISP's access point. Unless you're deliberately routing it over a tunnel (in which case you know you're doing it, or you're already owned) it really can't be attacked over the Internet like LAN Manager and SMS can.
The biggest exposure is in combination with insecure home WiFi, and if you've got that you're already skiing naked through a briar patch.
As long as you don't mind giving up on-demand video, and having to queue your downloads for later viewing, you don't need a cable bill.
Well, you're either downloading that over cable modem, or you're downloading it over DSL. You're watching TV over the airwaves, over cable, or (increasingly) over DSL. You're paying for it either way... and they will without question adjust the relative costs of data and video over broadband to match demand and usage.
All they need to do is to get the advertising back in, by doing something like offering high quality seeds of better-than-ripped-quality un-stripped versions of the shows, and we're back where we started.
Windows CE as used in 'Windows Powered' devices is pretty much a desktop OS. The WinCE API is derived from Win32, cleaned up and modularized and with its own set of libraries and a real-time kernel. It does support a traditional embedded OS model where code is executed in place from whatever file system is wrapped around it, but the "Windows Powered" handhelds don't work that way.
The first WinCE-based handhelds were pretty much "laptop replacements" with stripped down versions of Windows applications that run by copying them from a file system to RAM, explicitly use an open/read/write/seek/close mechanism to access files, and so on. There's a set of database calls for PDA applications that run on top of this. Microsoft subsequently stripped down the applications, removed the more desktop-like ones, and repurposed Windows Powered handhelds as "Palm Killers". By the time Palm lost the plot and sent haring off trying to port BeOS to the Palm in a quixotic attempt at fighting Microsoft on a field Microsoft was already abandoning they'd done a good enough job at "cloning" enough of Palm's look and feel that their current character recognizer is a better clone of Graffiti than Palm's current offering... but under the hood they're very "desktop-like".
I haven't worked with Symbian devices, so I can't say if they're more like a stripped down desktop or a classical embedded system like Palm, but the older high-end devices certainly looked more like a desktop.
The OS isn't the problem, here. It's Safari. The comments about finding crashes in Safari make me suspect that this is probably a stack/buffer overflow attack. If it's easy to crash Safari on the iPhone then they've got problems in the implementation of Safari on the iPhone... especially in the extensions to webcore that are unique to the device. If Pocket Internet Explorer had the same problems, then Windows CE would have the same exposure (luckily for Windows CE users, Pocket IE seems to be the most secure version of IE out... probably due to the fact that it doesn't include the same kind of "active content" support as the desktop version).
And the original article is right, the presence or absence of an official dev kit has very little to do with this... it just makes it harder to switch from Safari to another browser while Apple is easing Safari on the iPhone through its birth trauma.
It's more likely to introduce new 976-style exploits, through autoloading form using "tel:" URIs. The iPhone is not a real smartphone because it has no native API.
The silver lining in the dark cloud of the iPhone's lack of a native API is that there's no mechanism to install any kind of worm, rootkit, or spyware on the device.
But of course it's silly to worry about the iPhone, not when there's so many REAL smartphones out there that actually have the technical capabiilty of supporting the kind of viral ecosystem you're talking about... and have been for several years. We haven't seen a mass flowering of malware for Symbian- Palm- and Windows-powered devices yet, and they're SO much more credible targets for attack.
You hear a lot about Lockheed Martin's "Skunk Works", but Northrop Grumman keeps a lower profile in the "Crazy Ideas That Just Might Work" department. Perhaps they're looking to change that.
It's not really different from what happened to RMS at MIT! He too felt so harassed in his freedom to share information that he finally came up with a fantastic alternative. There would be no OSS today, had they been more liberal at MIT AI-lab back then.
Open Source Software owes at least as much to Berkeley's liberal attitude, *and* MIT's liberal attitude, as to RMS. RMS has effectively and unfairly demonized the AI lab, and he's been given too much credit for being one of the more visible rocks in an avalanche of open source software that was already in motion well before he penned the GNU manifesto.
The idea that MIT-AI, MIT-MC, and the rest... running an OS that let any user be "root" and giving accounts to outsiders (including Larry Niven and Jerry Pournelle)... were insufficiently "liberal", or that there would be no FOSS without RMS... is distressingly common today, and completely at odds with what the growing open-systems and free-source software community was actually like at the time.
"Most of the machines I see pitched in catalogs are in the $700 range, certainly under $1,000," said Cherry. "Computers with that amount of hardware are a better fit for XP. With Vista's requirements, people may be thinking about sticking with XP, and putting less money into the hardware."
OK, let's put this into perspective. We're not talking about some processor-hungry application here, we're talking about the overhead of teh OS itself.
Several years ago I was dual-booting FreeBSD and Windows on my main computer. It had 128M of RAM and an unaccelerated video card. It was getting time to upgrade, so I got a used Powermac G3 and added a G4 upgrade card and installed Jaguar on it. Jaguar was specced for a minimum of 256MB RAM and really required about 128M more than that to be happy. I was able to get 768MB of old RAM for that box cheap, and that was fine. I transferred all my UNIX stuff to it, it was faster than my old dual boot box, but I had that as a backup. I upgraded to a Mac mini when that came out, with 512MB of RAM running Panther, and it was so much faster even though it had less RAM. By this time my old dual boot box was only ever running Windows. With 256MB and a geforce 4 Ti-200 and perfectly reasonable for most of the Windows apps I needed to run.
I was a little annoyed that OSX required more memory than Windows. I'd been accustomed to mainstream UNIX that had gone from being "bloated" (since it required several hundred KILObytes to run and OVER A MEGABYTE to be happy) to being "lightweight" (Kids today, you tell them about punch cards and drum memory and they believe you, you tell them about running a desktop with a GUI in less than a megabyte and they think you're kidding), with Windows being the memory hog... but Windows was working well on a system that I knew OSX would balk at if it was a Power PC.
But I had applications that needed more than that.
So a year or so ago I put together what I thought was a pretty nice computer, with 2GB of RAM, a dual core Athlon 64 CPU, a gf7600 video card, and it was under $600. If anything, I thought I was going pretty gold-plated.
So I'm completely boggled at the idea that an "under $1000" computer would be considered inadequate for Vista. He's just talking about the *extra overhead* of the OS here, the applications (which for me had been what was driving my hardware requirements) would be the same on XP and Vista... what the HELL is Vista doing that makes this kind of machine "inadeqate"?
It's NOT just the eye candy. I already took the "eye candy" hit when I switched from FreeBSD/Windows to OSX. My first generation Mac mini with a 1.42 MHz PPC and 512M RAM and a 32M Radeon 9200 GPU handles all the eye candy I can stand _just_fine_. The GPU isn't good enough for the Quartz Extreme effects, but it's one that was 'trailing edge' when I got the computer _three_years_ ago. I can't imagine what kind of eye candy they could be including that any desktop or laptop HOWEVER CHEAP sold in the past three years (let alone on sale today) would have a problem with.
So just what is it that Vista requires that brings _the_OS_requirements_ up from the 256M RAM and a geforce 4 to something that my "cheap and nasty" six-hundred-dollars-over-a-year-ago 2GB plus geforce 7600 box would supposedly balk at? Or is this guy just making excuses for Microsoft?
The question shouldn't be "will MySpace disrupt TV?", it's "How much more will MySpace disrupt TV than streaming video already has?". The current leader in the "most televisions staked" vampire hunter contest has to be YouTube...
It's got nothing to do with whether it's a video game.
If you're playing Pacman in an arcade, and you discover that the machine lets you run through the ghosts if you move through the tunnel from the right side of the screen to the left just as they're moving in the other direction, that's you against the house.
Similarly, if you're playing Poker and you notice the deck is marked, that's PvP, not you against the house.
One argument I've always had over the years in various networked video games I've played [...]
Those are PvP games. Completely different environment. You're playing against your peers, not the house.