"There are half a dozen different kinds of executables, with entirely different behaviors."
Half a dozen? Don't you mean two? There's CFM and dlyd, or PEF and Mach-O if you prefer those names.
Those are the two binary formats (well, there is Java, too). But they can be packaged in different ways as well. Some executables come as directory trees (but show up as single applications in the Finder), others come as files. Some may be relying on resource forks, while others come as plain UNIX files. Some files appear to be associated with applications through extensions, others through their resource fork. It's certainly a much more complex set of possibilities than on Linux or Windows.
You say schizophrenic, I say rich. All of the BSD, Carbon, and Cocoa APIs can be called from the same Objective C program;
The problem isn't that you can call all three of them from a single Objective-C program (why wouldn't you be able to?), but that you have to: there is no single, consistent API that gives you complete access to the machine and OS. It's not a fatal flaw, but it certainly makes programming the Macintosh much more complex than, say, programming Linux and X11.
your comments on Quartz are just way off the mark.
What? That it's comparatively slow and resource intensive? Come on, do your own measurements, you'll see. Quartz has a nice imaging model, but that costs, and it isn't a particularly efficient implementation either.
I predict that X11 toolkits and desktops like Gnome and KDE will have all the flash and visual appeal of Aqua within 6-12 months and require much less memory and CPU to do so. You are welcome to disagree. Fortunately, we'll get the answer to that question pretty soon.
Yes, for desktops, almost everybody I know has thrown out their Apple mouse and replaced it with a three button mouse. But for laptops, you don't have a choice, and that's a shame.
Apple could conceal two or three buttons under that one button and let people choose in software how many distinct functions they want.
"Please understand that Federal Regulations prohibit. . . all other air carriers from sharing specific information regarding this program with the public."
As the transition to a police state progresses further, officials don't even bother making up such lies anymore, they just tell you to shut up or get arrested.
Firms (operating airports) should be made totally subject to competitive pressures to perform and should be held fully accountable for any breaches in the civil-liberty rights of passengers.
As if private entities in the US are ever held responsible for violating people's privacy rights. The US government out-sources such violations to companies, after all.
The system will feed the results to an analysis application that mathematically ranks travelers' potential as security threats.
It may do so "mathematically", but that doesn't mean "reliably": garbage-in, garbage-out. In this case, the few dozen terrorists we have had do things with planes over the last few years simply aren't enough to establish reliable criteria for who is a security risk.
What will actually happen is that police make wild guesses on what seems reasonable to them. Once programmed into the computer, stereotyping, racial profiling, and discrimination become "mathematical", and at that point, you effectively lose your right to complain or sue. "Sorry that every check-in takes 8 hours, but the computer insists YOU are a security risk; it's not our fault--WE aren't prejudiced." Overall, this system will result in lots and lots of false interrogations and arrests, and the real terrorists will likely not fit the profile anyway. Eventually, some people will just have to give up flying altogether.
I'm sure that there are quite a few people who will love the combination of the BSD command line with a very pretty GUI, as well as commercial applications.
But under the surface, OS X also has some pretty big issues. It is quite schizophrenic about APIs: the BSD, Carbon, and Cocoa APIs really aren't all that well integrated. There are half a dozen different kinds of executables, with entirely different behaviors. Many applications see a Mac file system, others see a UNIX file system. OSA scripting doesn't work for the majority of applications. And Carbon applications ignore Cocoa preferences. Some devices are accessible through BSD-like APIs, others are only available through Carbon, some have Cocoa wrappers.
And the crown jewel of OS X, the GUI, is also a bit iffy under the covers. Quartz is an enormous resource hog and rather sluggish. The Cocoa API requires lots of manual storage management and manual layout management. Objective-C is getting rather long in the tooth and will not take the world by storm anymore (it was a nice idea in 1985, now we have better systems). In terms of usability, OS X is better than Windows, but it is still far from "intuitive" (all current GUIs, including Apple's, commit some grave sins), as you will quickly find out if you try to explain how to use it over the phone to non-computer users.
I like my Macs (and am typing from a Mac right now). But they are not replacements for UNIX workstations or Linux machines--they are replacements for Windows desktop machines. And Apple has their work cut out for them. Let's hope they'll clean up some of the mess under the covers. I think the more open source software they can use, the better for them. In the medium term, they might even be well advised to drop Quartz and Objective-C and adopt technologies more widely used in the open source world--I think Apple won't be able to keep up with Gnome, KDE, Ximian, and other efforts like that.
The biggest advantage of Mac OS X are probably still the hardware/software integration, brand, distribution channels, and surrounding infrastructure. Those, rather than amazing technical differences, are what make the Mac a good choice for many non-technical users.
People shouldn't be able to accumulate $2.1m before demonstrating their complete incompetence. Obviously, the Nigerian Money Scam didn't reach these people in time to demonstrate their complete inability to deal with money before substantial amounts were involved. If the spam had reached the woman when she was in college, or the lawyer's roomate when he was in law school, maybe neither of them would have made it through school, and they would be doing something safely removed from handling large amounts of money. The clear antidote to that is more spam, not less; give people the opportunity to remove themselves from the economy (or gene pool, as the case may be) early, before they can do too much damage. It's safer for the rest of us. Really.
"Night of the Lepus". Giant, mutant bunny rabbits terrorize the southwest. Unfortunately, despite the silly premise, this one wasn't quite bad enough to be good.
Thanks for the pointer. Sounds like they went with 1:1 for a good reason. I always thought of M:N threading as kind of a kludge and not entirely trustworthy anyway (scheduling and I/O become rather iffy).
Re:What's with the attitude?
on
New MP3 Portables
·
· Score: 3, Informative
I have both FireWire and USB2 on several machines; FireWire's architectural superiority makes no difference in practice. If Apple also supported USB2, I'd switch over to USB2 devices completely.
And a big advantage of USB2 is that it also works on USB1 ports--slower, but it works, and just about every computer has USB1.
It's nice that the Linux kernel can handle that many threads. But user level threads generally are even more lightweight, and high performance implementations like those on Solaris provide both user level and kernel level threads and map the former onto the latter. Is Linux going to get something similar? Is Sun perhaps donating their implementation? Or are these new kernel threads so lightweight and quick that they are competitive with Solaris on their own, without the mess and complication of adding user level threads?
Basically, the company isn't obligated to pay you anything. They are offering you the severance pay in return for not having to worry about you anymore.
You have to decide whether the deal is worth it to you. If you don't plan on sueing and if the rest of the conditions don't prohibit you from doing something you might reasonably want to do, why not agree?
The idea of using a process to perform authentication on your behalf is pretty old (as Cox's paper points out). Most of Cox's paper is concerned with how to integrate this idea into Plan 9's file system and file-server-based architecture. It's "this is how we integrated a key agent into Plan 9" paper, not "here is a completely new idea for how to manage keys and authentication" paper.
But UNIX and Linux have SSH and ssh-agent. It's not as elegant as Plan 9's file servers, but it is just as flexible. SSH is built around the idea of establishing secure and authenticated tunnels. And SSH with ssh-agent has become, for many purposes, the separate entity into which cryptography has been factored on Linux and UNIX: SSH gives you secure, authenticated remote system administration, the ability to transfer large amounts of data securely, the ability to create secure communications channels, and it is used by systems like rsync as its secure and authenticated transport protocol.
Maybe rather than reinventing the wheel, we should figure out how to extend what is already used and works. For that, we need a clearer idea of what problem "single sign-on" is supposed to solve that ssh and key agents/keychains aren't already solving, and then to figure out what we can do about it. And there isn't a whole lot I can think of that ssh isn't solving, at least in principle. Of course, wide, practical deployment for something like web services would require a set of UIs to be developed for Windows users and a lot of salesmanship. But, then, the same is true for whatever Sun cooks up.
"Giga" is so 20th century. It has the ring of Dr. Evil's "One Million Dollars" to it (imagine backwards pinky to corner of mouth). The new marketing-compliant prefix is "Peta". Please take note.
Well, I was careful to say "too many open source projects are ported to Windows", not that open source software should never get ported to Windows.
Large database servers are often run on a dedicated machine, so for something like PostgreSQL, I think there is no need to port it to Windows--bringing up a dedicated database server under Linux is an excellent way to start switching to open source.
Often, the cross-platform compromises of supporting something on both Windows and Linux can be harmful to the software in question, making it more complex or limiting features. I think Apache 2.0's thread support is a good example: it causes a lot of extra work and is arguably completely unnecessary for Linux.
Oh! Whoops! Postgres doesn't run natively on Windows. [postgresql.org] This is COMPLETELY unacceptable.
Well, tough, either port it yourself or you buy Microsoft SQL Server, or you switch to Linux. The world doesn't owe you a free database server, much less one for Windows.
If anything, I think too many open source projects are ported to Windows. That eats up a lot of effort, supports Microsoft, and the users would be better off switching to a free OS in the long run anyway.
The way things seem to be working out is that you end up being able to synchronize pairs of devices (not all pairs, but at least a spanning tree). That way, you don't need a centralized server. You configure what goes where for each pair. Some of those synchronizations happen automatically when devices are within range of each other (phone to PDA), others happen over the Internet.
That, rather than the Soviet-style centralized identity management that Microsoft and Sun envision, is likely to be how identity information gets passed around: from a variety of source on a variety of devices.
Companies give software away for many reasons: PR, establishing standards, driving competitors out of the market, and hurting competitors financially are among them. Sharing development efforts may be as well, but usually is not. Sometimes such strategies are combined with "dual licensing schemes", where open source is used to gain a foothold in a commercially meaningless part of the market to prop up a product that otherwise wouldn't be competitive.
Not all such gifts are useful for the recipient, and some are genuinely harmful to the interests of open source users. So, do look a gift horse in the mouth, or you may be stuck with large vet bills otherwise.
This one seems harmless if it is on unpatented technology, or if the patents are free for use by open source.
I can still run 16 bit DOS programs on windows. I can still use DDE for IPC. COM and COM+ haven't gone away with.NET.
Yup, you can still run all that stuff, and I didn't claim otherwise; Microsoft would be foolish to alienate their customers that blatantly. Your 16bit programs will run in all their 16bit glory.
What you are missing is that those APIs are dead or dying--programs using them can't take full advantage of the system, Microsoft doesn't want you to use them anymore, people usually don't develop for them anymore, and there are fewer and fewer tools. That's what "changing paradigms" means.
UNIX and Linux provide continuity and growth that goes beyond merely backwards-compatible support. You should find out about it some time.
US homes are usually constructed by companies just for selling them, or by people who don't expect to live in them for more than a decade. That means that they usually cut a lot of corners: the quality of the materials may be good (though often it is not), but any frill gets cut.
Europeans often still build homes for themselves and expect to live in them for a lifetime or generations. Then, it makes sense to pay for extra conveniences.
I'm surprised they didn't call it "piracy" or "terrorism". Those are so much more dramatic sounding, and so much more popular with people who like to taint behavior that's legal or cover up their own technical ineptness.
In this case, if the wireless standards had incorporated decent security, warchalking wouldn't exist. 802.11b, and the various products based on it, are the most incompetently designed products from a security point of view: lousy user interface (you have to type lots of weird numbers into dialog boxes before your system becomes secure), and even if you go through that, it still isn't secure.
They can ask anybody to sign a non-disclosure agreement as long as they still give the sources of GPL'ed programs to people who ask and allow them to redistribute those. The GPL is only about ensuring obtaining and redistributing source.
So, they are probably not violating the GPL by having a non-disclosure agreement or not permitting redistribution of the entire distribution. And their installers can check for the presence of non-GPL'ed software and refuse to install. The GPL does not guarantee that entire distributions are redistributable in binary format. All they need to do is provide sources to GPL'ed programs on request, and only to people they distributed the binaries to, who then have the right (but not necessarily inclination) to redistribute those sources further.
Whether that's a good thing or not I can't say. I think a user would not be very smart to start relying on such a company for their Linux distribution, since there are so many better alternatives out there.
Using Linux as the server OS is not about getting the OS for free--as you observe, the cost of the OS is pretty much lost in the noise.
The reason to use Linux is all aspects of its openness and compatibility with other systems. With Linux, you aren't locked into a single vendor. You use tools and APIs that have been around for nearly two decades and are available, in multiple implementations, from dozens of vendors. And you control how you upgrade, when you upgrade, and what path you follow with the software. And if you don't like Linux anymore, you can switch to any of a dozen other, compatible platforms.
With Windows, you are locked into a single, proprietary implementation and Microsoft has you by the proverbial precious body parts; there is no other vendor you can get a compatible implementation of Windows or all the Windows libraries from. Every couple of years, Microsoft completely changes their computing paradigms to ape what they perceive is a threat from some other company, and when the threat is gone, they just drop the initiative and move on to the next thing.
You can get stability buy paying a premium to a company like IBM, which is committed to providing it, or through open systems available from multiple vendors or open source, where you control your future. But building a large, long-lived infrastructure on Microsoft platforms is a costly folly--the company has proven that they will change approach every couple of years and that they will force their customers to move along.
A honeypot is a machine that looks suspectible to break-in but is monitored. It's used by sys admins and security "experts" to find out what techniques people use to break into machines.
You can crack software legally all you want (well, in most of the world) if you don't distribute the result, and as a practical matter, are unlikely to get caught anyway. There are also plenty of machines you can practice breaking into with no or limited recriminations. But cracking software or breaking into machines are not complex skill--you can become as skilled as most of the "experts" in a few weeks.
Personally, I think your time would be better spent getting involved in open source software development.
Slashdot Mirror
Those are the two binary formats (well, there is Java, too). But they can be packaged in different ways as well. Some executables come as directory trees (but show up as single applications in the Finder), others come as files. Some may be relying on resource forks, while others come as plain UNIX files. Some files appear to be associated with applications through extensions, others through their resource fork. It's certainly a much more complex set of possibilities than on Linux or Windows.
You say schizophrenic, I say rich. All of the BSD, Carbon, and Cocoa APIs can be called from the same Objective C program;
The problem isn't that you can call all three of them from a single Objective-C program (why wouldn't you be able to?), but that you have to: there is no single, consistent API that gives you complete access to the machine and OS. It's not a fatal flaw, but it certainly makes programming the Macintosh much more complex than, say, programming Linux and X11.
your comments on Quartz are just way off the mark.
What? That it's comparatively slow and resource intensive? Come on, do your own measurements, you'll see. Quartz has a nice imaging model, but that costs, and it isn't a particularly efficient implementation either.
I predict that X11 toolkits and desktops like Gnome and KDE will have all the flash and visual appeal of Aqua within 6-12 months and require much less memory and CPU to do so. You are welcome to disagree. Fortunately, we'll get the answer to that question pretty soon.
Apple could conceal two or three buttons under that one button and let people choose in software how many distinct functions they want.
As the transition to a police state progresses further, officials don't even bother making up such lies anymore, they just tell you to shut up or get arrested.
As if private entities in the US are ever held responsible for violating people's privacy rights. The US government out-sources such violations to companies, after all.
The system will feed the results to an analysis application that mathematically ranks travelers' potential as security threats.
It may do so "mathematically", but that doesn't mean "reliably": garbage-in, garbage-out. In this case, the few dozen terrorists we have had do things with planes over the last few years simply aren't enough to establish reliable criteria for who is a security risk.
What will actually happen is that police make wild guesses on what seems reasonable to them. Once programmed into the computer, stereotyping, racial profiling, and discrimination become "mathematical", and at that point, you effectively lose your right to complain or sue. "Sorry that every check-in takes 8 hours, but the computer insists YOU are a security risk; it's not our fault--WE aren't prejudiced." Overall, this system will result in lots and lots of false interrogations and arrests, and the real terrorists will likely not fit the profile anyway. Eventually, some people will just have to give up flying altogether.
But under the surface, OS X also has some pretty big issues. It is quite schizophrenic about APIs: the BSD, Carbon, and Cocoa APIs really aren't all that well integrated. There are half a dozen different kinds of executables, with entirely different behaviors. Many applications see a Mac file system, others see a UNIX file system. OSA scripting doesn't work for the majority of applications. And Carbon applications ignore Cocoa preferences. Some devices are accessible through BSD-like APIs, others are only available through Carbon, some have Cocoa wrappers.
And the crown jewel of OS X, the GUI, is also a bit iffy under the covers. Quartz is an enormous resource hog and rather sluggish. The Cocoa API requires lots of manual storage management and manual layout management. Objective-C is getting rather long in the tooth and will not take the world by storm anymore (it was a nice idea in 1985, now we have better systems). In terms of usability, OS X is better than Windows, but it is still far from "intuitive" (all current GUIs, including Apple's, commit some grave sins), as you will quickly find out if you try to explain how to use it over the phone to non-computer users.
I like my Macs (and am typing from a Mac right now). But they are not replacements for UNIX workstations or Linux machines--they are replacements for Windows desktop machines. And Apple has their work cut out for them. Let's hope they'll clean up some of the mess under the covers. I think the more open source software they can use, the better for them. In the medium term, they might even be well advised to drop Quartz and Objective-C and adopt technologies more widely used in the open source world--I think Apple won't be able to keep up with Gnome, KDE, Ximian, and other efforts like that.
The biggest advantage of Mac OS X are probably still the hardware/software integration, brand, distribution channels, and surrounding infrastructure. Those, rather than amazing technical differences, are what make the Mac a good choice for many non-technical users.
People shouldn't be able to accumulate $2.1m before demonstrating their complete incompetence. Obviously, the Nigerian Money Scam didn't reach these people in time to demonstrate their complete inability to deal with money before substantial amounts were involved. If the spam had reached the woman when she was in college, or the lawyer's roomate when he was in law school, maybe neither of them would have made it through school, and they would be doing something safely removed from handling large amounts of money. The clear antidote to that is more spam, not less; give people the opportunity to remove themselves from the economy (or gene pool, as the case may be) early, before they can do too much damage. It's safer for the rest of us. Really.
"Night of the Lepus". Giant, mutant bunny rabbits terrorize the southwest. Unfortunately, despite the silly premise, this one wasn't quite bad enough to be good.
Thanks for the pointer. Sounds like they went with 1:1 for a good reason. I always thought of M:N threading as kind of a kludge and not entirely trustworthy anyway (scheduling and I/O become rather iffy).
And a big advantage of USB2 is that it also works on USB1 ports--slower, but it works, and just about every computer has USB1.
It's nice that the Linux kernel can handle that many threads. But user level threads generally are even more lightweight, and high performance implementations like those on Solaris provide both user level and kernel level threads and map the former onto the latter. Is Linux going to get something similar? Is Sun perhaps donating their implementation? Or are these new kernel threads so lightweight and quick that they are competitive with Solaris on their own, without the mess and complication of adding user level threads?
You have to decide whether the deal is worth it to you. If you don't plan on sueing and if the rest of the conditions don't prohibit you from doing something you might reasonably want to do, why not agree?
But UNIX and Linux have SSH and ssh-agent. It's not as elegant as Plan 9's file servers, but it is just as flexible. SSH is built around the idea of establishing secure and authenticated tunnels. And SSH with ssh-agent has become, for many purposes, the separate entity into which cryptography has been factored on Linux and UNIX: SSH gives you secure, authenticated remote system administration, the ability to transfer large amounts of data securely, the ability to create secure communications channels, and it is used by systems like rsync as its secure and authenticated transport protocol.
Maybe rather than reinventing the wheel, we should figure out how to extend what is already used and works. For that, we need a clearer idea of what problem "single sign-on" is supposed to solve that ssh and key agents/keychains aren't already solving, and then to figure out what we can do about it. And there isn't a whole lot I can think of that ssh isn't solving, at least in principle. Of course, wide, practical deployment for something like web services would require a set of UIs to be developed for Windows users and a lot of salesmanship. But, then, the same is true for whatever Sun cooks up.
"Giga" is so 20th century. It has the ring of Dr. Evil's "One Million Dollars" to it (imagine backwards pinky to corner of mouth). The new marketing-compliant prefix is "Peta". Please take note.
Large database servers are often run on a dedicated machine, so for something like PostgreSQL, I think there is no need to port it to Windows--bringing up a dedicated database server under Linux is an excellent way to start switching to open source.
Often, the cross-platform compromises of supporting something on both Windows and Linux can be harmful to the software in question, making it more complex or limiting features. I think Apache 2.0's thread support is a good example: it causes a lot of extra work and is arguably completely unnecessary for Linux.
Well, tough, either port it yourself or you buy Microsoft SQL Server, or you switch to Linux. The world doesn't owe you a free database server, much less one for Windows.
If anything, I think too many open source projects are ported to Windows. That eats up a lot of effort, supports Microsoft, and the users would be better off switching to a free OS in the long run anyway.
That, rather than the Soviet-style centralized identity management that Microsoft and Sun envision, is likely to be how identity information gets passed around: from a variety of source on a variety of devices.
Not all such gifts are useful for the recipient, and some are genuinely harmful to the interests of open source users. So, do look a gift horse in the mouth, or you may be stuck with large vet bills otherwise.
This one seems harmless if it is on unpatented technology, or if the patents are free for use by open source.
Yup, you can still run all that stuff, and I didn't claim otherwise; Microsoft would be foolish to alienate their customers that blatantly. Your 16bit programs will run in all their 16bit glory.
What you are missing is that those APIs are dead or dying--programs using them can't take full advantage of the system, Microsoft doesn't want you to use them anymore, people usually don't develop for them anymore, and there are fewer and fewer tools. That's what "changing paradigms" means.
UNIX and Linux provide continuity and growth that goes beyond merely backwards-compatible support. You should find out about it some time.
Europeans often still build homes for themselves and expect to live in them for a lifetime or generations. Then, it makes sense to pay for extra conveniences.
In this case, if the wireless standards had incorporated decent security, warchalking wouldn't exist. 802.11b, and the various products based on it, are the most incompetently designed products from a security point of view: lousy user interface (you have to type lots of weird numbers into dialog boxes before your system becomes secure), and even if you go through that, it still isn't secure.
Nokia: fix your systems, don't blame others.
So, they are probably not violating the GPL by having a non-disclosure agreement or not permitting redistribution of the entire distribution. And their installers can check for the presence of non-GPL'ed software and refuse to install. The GPL does not guarantee that entire distributions are redistributable in binary format. All they need to do is provide sources to GPL'ed programs on request, and only to people they distributed the binaries to, who then have the right (but not necessarily inclination) to redistribute those sources further.
Whether that's a good thing or not I can't say. I think a user would not be very smart to start relying on such a company for their Linux distribution, since there are so many better alternatives out there.
The reason to use Linux is all aspects of its openness and compatibility with other systems. With Linux, you aren't locked into a single vendor. You use tools and APIs that have been around for nearly two decades and are available, in multiple implementations, from dozens of vendors. And you control how you upgrade, when you upgrade, and what path you follow with the software. And if you don't like Linux anymore, you can switch to any of a dozen other, compatible platforms.
With Windows, you are locked into a single, proprietary implementation and Microsoft has you by the proverbial precious body parts; there is no other vendor you can get a compatible implementation of Windows or all the Windows libraries from. Every couple of years, Microsoft completely changes their computing paradigms to ape what they perceive is a threat from some other company, and when the threat is gone, they just drop the initiative and move on to the next thing.
You can get stability buy paying a premium to a company like IBM, which is committed to providing it, or through open systems available from multiple vendors or open source, where you control your future. But building a large, long-lived infrastructure on Microsoft platforms is a costly folly--the company has proven that they will change approach every couple of years and that they will force their customers to move along.
I see... someone had too many moderator points again.
A honeypot is a machine that looks suspectible to break-in but is monitored. It's used by sys admins and security "experts" to find out what techniques people use to break into machines.
Personally, I think your time would be better spent getting involved in open source software development.