as someone who worked in ordering huge lots of Apple machines, I know you can get Apple to load whatever you want on them at the factory
Interesting.
That doesn't mean that part of the cost isn't the right to install a legally licensed copy of whatever Mac OS you might later buy on it.
When I buy a Dell, it is not tied to Windows upgrades forever and ever.
The retail Windows license is a complete license to run Windows, it's not tied to Dell or any other vendor's hardware. The retail OS X license is tied to Apple branded hardware only, it is not by itself a license to run OS X the way a Windows license is a license to run Windows.
When you buy a Windows upgrade, however, you do not get a complete Windows license. You get a package that treats the previous version of Windows as a dongle that must be present to perform the install. It's an upgrade to a license you already have.
When you buy a copy of Mac OS X, you get a package that treats the presence of an Apple-branded computer as a dongle that must be present to perform the install. The only difference between this and the Windows license is that Microsoft calls theirs an upgrade explicitly.
What is Apple doing that is so different here?
What Apple's doing that's different is building part of the revenue they receive for the operating system into the sale of the computer. Dell doesn't do that, the OS is a cost for them. Microsoft doesn't do that, the only revenue they receive from the hardware sale is the license fee Dell pays them.
and you can still install it on your clone too.
Not legally, unless you're being really careful in your interpretation of what an Apple computer is.
Apple just wants to sell hardware, thats it.
If that was it, then they would sell the hardware for a significantly lower price, and the OS for a significantly higher one. Because it's the right to install Mac OS on it that allows them to charge hundreds (and in some cases pushing four figures) more for a Mac than for a comparably equipped computer that won't legally run Mac OS.
Its the same thing as Red Hat one day revoking Dell license to provide Linux on their machines
Since Red Hat doesn't recieve any money from the sale of Dell machines, they have no incentive to do that. In addition, they don't have the legal right to do that under the terms of the GPL.
Apple won't give a damn. They sell hardware. They don't care what OS you use.
Exactly what Apple sells depends on who you talk to.
To someone developing Power-PC based embedded systems, Apple sells relatively cheap development and test systems for Power-PC software... because industrial quality hardware and enclosures are MUCH more expensive than the ones Apple builds.
To some Mac users, Apple sells style and mystique. These people consider the pretty boxes worth the money.
To other Mac users, Apple sells computers that are well enough made to justify their price.
To still other Mac users, Apple sells computers running Mac OS. These people consider Mac OS to be worth enough to justify the premium they pay for a Mac.
My claim is that this last category constitutes the majority of the Macintosh market. The sales decline that resulted from the creation of a Mac clone market, and the poor quality of many Macs (including, in my opinion, all the laptops... but that is my opinion and I am aware that many people disagree), as well as my own reasons for buying around a dozen Macs over the years, is the basis for this claim.
So, if Mac OS is the driver behind the sale of most Macs, what Apple sells is software. Their business model involves selling hardware (Macintosh computers) that their software requires to run, but it's the software that makes these sales possible.
If you buy a Macintosh, and you don't subsequently install Mac OS on it, Apple has still received most (or all, do they actually
I love Linux, and as such, I feel pain every time someone says "just fix it yourself".
What I sometimes find frustrating about BSD and more often consider a major advantage is that it's one of those systems (or three of those systems) where you get a lot less "just fix it yourself" and a lot more "we already thought of that, we're not doing it that way because of X, Y, and Z, and the right way to do what you want to do is...".
It's frustrating when the official right way sucks, but great when it doesn't. And the official right way sometimes changes to something closer to what you wanted, and sometimes it's even better.
Did you write your software for you and then give it away?
Yes.
Now its in use by a million people, do you care about their opinions?
Yes.
That doesn't mean I don't say both "just fix it yourself" and "no I didn't do it that way because...". Also, more and more, "I'm not the maintainer any more, tell so-and-so...".
What I really like is when someone sends me a fix that fixes something I haven't gotten around to yet and ALSO works well with what I HAVE gotten around to.
If Linux just "happens to run", how come it knocks out OpenBSD when it comes to performance?
Taking shortcuts is often a very effective way to get better performance. OpenBSD is particularly notorious for NOT taking shortcuts even in the BSD world.
I used to say that, and I got tired of it, so I set up Linux and moved some of the software I was running on BSD over to it. I won't go into the variations and distributions and versions and things that worked and things that didn't work, but the thing that was most amazing is how hard it was to get Linux users to see why I thought some of them were problems.
I'm not using Linux any more, when I have an alternative. Apple took away the last incentive I had for giving it another shot on my desktop.
Historically, putting "Open" in a product name or service has been a useful warning sign, similar to "Honest" on a used card dealer's lot.
For example, "OpenVMS", "MVS Open Edition", "Open Desktop" or "Open Server",...
Unfortunately this is no longer a reliable guide, as some open systems and open source organizations have muddied the waters by using it in the previous (and, for a time, obsolete) sense.
Re:I only wish BeOS was a member of the UNIX famil
on
Zeta Goes Gold
·
· Score: 1
Practically speaking, sometime before 1985 or so with POSIX, and after Berkley first developing TCP for their Unix, there was no UNIX in any current sense of the term.
OK, so the PDP-11s and VAXes that I was working on at Berkeley in 1980 weren't UNIX. The software I wrote that shipped in some of the 4.1BSD tapes wasn't UNIX software.
Would you say Unix style job control is a part of your definition for a member of the UNIX family?
That would have to be "no", I'm afraid, because that was actually developed at Berkeley while I was an undergraduate there, and it didn't become a normal part of "Trademark UNIX" until well after the 1985 date you referred to, since AT&T used a quite different mechanism called "layers" based on the BLIT terminal model and what became Plan 9.
Is VMS a member of the UNIX family? No, it isn't, but it too can be treated as a Unix alike with the right functionality installed, and so can meet the Single Unix Spec and be POSIX complaint,
That sure sounds like you agree with me that having a subsystem that meets the POSIX specification is actually unrelated to whether the OS that hosts that subsystem is a member of the UNIX family. So, really, there's no point even bringing POSIX into the question... POSIX is a definition that's useful for writing and supporting portable software, but it doesn't fundamentally differ in kind from (for example) the "Software Tools" virtual operating system and other schemes for running software developed for UNIX on other operating systems.
as far as BeOS being Unix goes, it depends on which sense of Unix one means
Precisely:
One can easily treat BeOS as an early, pre-X11, pre-POSIX single-user Unix, as long as only externally presented interfaces are used
Are you familiar with the Turing Test? Consider the thought experiment Turing created: when you're deciding whether there's an intelligent enity on the other end of the teletype link, you don't have any way to examine that entity's internal design, all you have to go on are the externally visible interfaces. If you can't tell by using those interfaces whether the entity at the other end is a machine or a human, then you have to treat it as if it's intelligent... whether it's a human, a martian, a DEC VAX, a dog with an Internet account, a Turing machine, or a thousand literate monkeys reading responses from books called "how to be human".
UNIX is like Intelligence in this respect.
Since the only thing that you can interact with an operating system through are the externally presented interfaces, that's all you can use when you're talking about whether it's a member of the UNIX family. As in the Turing test, the internals of an operating system are irrelevant.
I would love to concede every point you made, and smugly bash Windows as inferior in every technical and practical sense.
I'm sorry, we seem to be having unrelated converstions here. It's not my intention in this discussion of UNIX to "smugly bash Windows", and I certainly wouldn't describe Windows NT (at least) as inferior in every technical sense. There's a lot of really interesting and useful solutions in the NT kernel. I'm also not equating (as you seem to think I am, unless I'm badly misunderstanding you) the concepts "is a member of the UNIX family" with "is good" or "is not a member of the UNIX family" with "is bad". There's quite a lot of operating systems and environments I've used that I consider very good indeed that aren't in any way UNIX-like, and quite a lot of members of the UNIX family that I've used that I'm absolutely sure you wouldn't consider "UNIX-like", based on the kinds of facilities you've said you expect to be in any UNIX system.
The only aspect of being a member of the UNIX family that's really relevant to being "good" or "bad" here is that it's a lot easier to get along with the creature or machine at the other end of a teletype link if one can find common ground with i
When you buy a Macintosh part of the cost is the actual physical hardware, and part of the cost is the license that makes it legal to install Mac OS and upgrade to future versions of Mac OS. It is not legal to install Mac OS, any version of Mac OS, on a computer that isn't a Macintosh. What makes a piece of physical hardware a Macintosh rather than a PREP or CHRP based PC... even if it's a Macintosh that conforms to the CHRP or PREP standard... is that license that entitles you to install Mac OS and to upgrade to new versions of Mac OS. When you buy a retail copy of Mac OS, any version, you are not buying the right to install it on any computer, which would be the case if it was a "full install", you are buying the right to install that version of Mac OS on a computer that's already licensed to run Mac OS. When you recieve a copy under the "upgrade program", you can only upgrade Mac OS on an actual Macintosh to the new version. Apple explicitly excludes the licensed Mac clones, for example, from being upgraded from Mac OS to Mac OS X.
And the value of that license to Apple is the difference between the cost of the computer and what they can sell it for. That difference is what a true "full license" for Mac OS X would have to cost, if Apple were to sell such a thing.
If this distinction is too subtle for you, I'm sorry. I don't know how to make it clearer.
Um, my main machine up until a year ago never came with OS X. Ever.
No version of Mac OS at all? It's an actual Mac, not a clone? I call shenanigans: I have never heard of ANY Mac ever that shipped without Mac OS and a license for the same.
the current (aluminum) PowerBooks have the most outstanding laptop keyboard that I've ever used.
You've never used a Thinkpad, then. It's got the closest thing to a real keyboard I've found on any laptop, with textured keys that have a deeper "throw" than any other laptop I've seen since the Toshiba Sattelite.
Compared to the Thinkpad, even the latest aluminum powerbook is uncomfortable to use. It may be better than previous 'books in this form factor, but they're still flat and square with poor feedback.
The Apple desktop keyboards are pretty unpleasant, for that matter. The older keyboards they used on the Beige Powermacs were much better.
mini: Slabs are cool, but what's wrong with the mini's form factor?
It's so small that it's got more of a heating problem than the G5 iMac, which means it's shortchanged on disk (low speed 2.5" disk instead of a cheaper, faster, and more reliable 3.5" disk) and power (both the video and USB have less power than specced, I can't even spin up the external 2.5" bus-powered USB disk I use for file transfers using without sticking a powered hub in the way, and video problems on the mini are legendary).
A slightly larger unit based on the eMac instead of the iBook would have been much more versatile.
So by your dumbfoundingly twisted logic, it seems you argue that if I buy that mini then I can use it's included discs to install OSX on something other than a Mac?
No, no, the full version is only licensed for a particular machine. I wouldn't dream of implying otherwise.
You can't just install OS X on any old Power-based machine because it doesn't have drivers for all the hardware to boot and run natively.
Ah, but that's the interesting bit. See, Apple provides us with all the source code to boot the kernel on generic intel hardware and, if it were Mac OS X for the x86, it could even run natively. And, you see, since PREP and CHRP motherboards are so well documented and have Linux drivers available, running Mac OS X on them is really only prevented by the licensing.
The cost of the mini hardware has been calculated elsewhere, and it was over $300 the last time I saw it.
I've seen many calculations of it, and these calculations are always higher when they're made by a Mac fan, and lower when made by a PC fan. Some of them, like the $100 figure one person posted, are clearly ridiculous. I believe it's closer to $200 than $300... but whatever it is, it's a useful guideline as to the lowest possible price Apple would have to sell a full licensed version of Mac OS X Intel for generic PC hardware and still make the kind of profit they need.
If you pick a more expensive Mac as your baseline, that value is higher. For a top of the line Powermac G5 it's close to four digits. I suspect that Apple would need to sell SMP and non-SMP-capable versions for different amounts.
But even after it opens the "safe" file, apps get the "This is the first time you are running this program, are you sure you want to run it?" prompt if something tries to just auto start itself just after download.
Yes, that's a clever trick, isn't it? Microsoft does it too, and it gets people to feel safe about going ahead and clicking on links and downloading files like a wild monkey. They tell themselves "It's OK, it'll ask me before it does something stupid"... but of course because they're used to clicking OK so enough of them go ahead and click OK anyway, without thinking, to keep the exciting virus ecosystem flourishing. It's this kind of extra attention from Microsoft that system administrators everywhere really appreciate and enjoy.
Unfortunately Apple left people the option of turning off "open safe files after downloading" so instead of having a dialog they can reflexively click OK on, they just get a boring icon on the desktp that they can let sit there, and think about, and have second thoughts about and throw in the trash. What a bunch of spoilsports Apple are, denying people the full impact of reflexively hitting the wrong button like that.
Show me real evidence that a full version of OSX costs $300, even though Apples sells it for $129.
The full version can cost a lot more than that if you buy it with a higher end machine. $300 is the smallest amount you can get away with if you want to buy the right to install the OS. It's never before been so inexpensive.
As stated, when you pay $129, you can install it on any supported Mac, with or without an OS.
That's true. All supported Macintosh computers come with the right to install Mac OS X, along with the copy of Mac OS that they shipped with. If you don't buy a Macintosh computer, you don't get that right... even if it's technically possible (as with the 'Mac Clones').
Apple doesn't currently sell a separate "full version", of course, it's only available as a bundle with a Macintosh computer. If they did (as some people have speculated that they might at some point in the future) it's likely that would cost as least as much as the "Mac Tax" that Apple imposes when one buys a Macintosh rather than a similarly specced computer without that license.
Apple only explicitly labels their software an upgrade when it's a special reduced cost version for people who bought a previous version just before the new upgraded version comes out, but in practice every version is an upgrade that is only licensed for use on a computer that was originally shipped by Apple with Mac OS included.
I realise that this is a popular myth, and one that keeps people feeling comfortable that nothing Microsoft might do will damage their rich and exciting viral ecosystem, but unfortunately it's not true. Microsoft would merely have had to back down under the tyrannical pressure of the US Department of Justice and unbundled IE and the refreshing flood of viruses and spyware would have collapsed into a disappointing trickle.
How lucky we are that they showed such wisdom and determination!
Microsoft had just about as much marketshare before they integrated IE and the Desktop as after it, but the difference to fans of malicious software was startling. What kicked viruses into high gear was the ability of the Microsoft HTML control to run native code delivered by an untrusted website or mail message. It was the killer application for virus delivery, because it turned the previously unthinkable idea of a virus that could spread if you just opened an email message into reality. Sites like ours that banned IE and Outlook and later Windows Media Player and Realplayer and any other internet-enabled applications that used the HTML control had to put up with completely dismal virus levels.
Even refusing to use any antivirus protection other than telling people not to open unexpected attachments didn't help. It wasn't until we were forced to switch to IE by our enlightened corporate management that our virus load reached normal levels.
You feel that spam is so important that websites which offer to sell spam software should be blacklisted, along with many other innocent websites hosted at the same ISP.
Anyone who uses a blacklist in such a way that they or anyone at their location or any of their customers can't get to a website because it shares the address of a server that's in a blacklist is (a) a fool, and (b) the person on whom responsibility for this abuse of that blacklist should fall.
Pardon, I thought "you're a fucking idiot" was referring to me.
The anonymous coward who called you one was a sorry bastard, but he was right in saying that you're going to have to find a static smarthost for your outgoing mail.
you're making the assumption that we all have static IPs.
If you want to send mail, you need a static IP.
If you have a dynamic IP, then my mail server won't accept mail from you regardless of whether you're part of a botnet or not, because not only is one of the BLs I use simply a list of dynamic IPs, I also explicitly block hosts whose reverse lookup indicates they're in an un-RBLed dynamic IP space.
my home machine isn't my mail server, it's simply my home machine. Sometimes it needs to send out mail for various reasons.
Yep, so does mine.
I pay £400 a year and have a server hosted somewhere.
I don't pay quite that much, but it's in the same range. I'm also about to set up a backup virtual server for another $70 a year.
So when my home machine wants to send mail. it does it over an encrypted tunnel to my colo.
I should not be forced into using smart relaying
Entirely true. You shouldn't be forced to deal with the results of spammers pissing in the pool. But until spammers are fined or otherwise sanctioned sufficiently to make spam the minor problem that it should be, having to arrange for a reliable smarthost is part of the cost of being on the net.
Please consider what you're saying before you hit submit.
The point is that it will always be impossible to keep an accurate map of infected vs. uninfected systems when dealing with numbers this large.
So? You don't need to maintain an accurate map of infected vs. uninfected systems. You just need to keep track of systems that have been infected "recently", for values of "recently" that can be days or even weeks long. Since SBL expires botnet entries in 48 hours they're not contributing to abuse from this cause.
In fact you've given me a good idea for expanding my current bot tracker.
The only solution is to transition to a new protocol that enables authentication so that it won't happen.
The trick is to come up with a scheme that works, that people actually uses, and that doesn't put Microsoft or Verisign or someone similar in charge. Otherwise you wouldn't be able to convince people to turn off their old SMTP servers.
No, the behaviour he's referring to would occur if one applied filters late in the chain. The thing is, he assumes that you HAVE to apply filters late in the chain. It's a common misunderstanding about qmail.
Think about it: what you're ultimately doing is give some complete stranger near-complete control over what email is or isn't accepted by your system.
I'm already doing that. I'm trusting my registrar, I'm trusting my ISP, and I'm trusting the people running my secondary DNS, and I'm trusting the people "near" me on the Internet.
Blacklists are something that might seem like a good idea in theory
Blacklists are something that is absolutely essential in practice. Without blacklists I wouldn't be able to run a mailserver at all.
The thing about the Mac mini is that it has a large value associated with A, the industrial design.
If you say so. The design may have some cost to Apple, but it's shared with the iBooks that use largely the same components. In fact Apple could have produced a better computer by not using that design: building it to laptop specifications means they have to reduce the power draw significantly, and (coincidentally) the USB and video interfaces are underpowered, it's undersupplied with video RAM, and the 2.5" disk is lower performance than the cheaper higher-power 3.5" disk they could have otherwise used.
Anyway, I think that $300 is a reasonable estimate of the low end of the range Apple would need to charge for a generic Mac OS where they don't share the cost between the retail package and the computer it runs on.
You can purchase a FULL version of OS X, not the upgrade, for $129.
You can only legally install that on genuine Apple Machintosh: a computer that was originally purchased from Apple with a copy of Mac OS (OS 9 or OS X). You can not legally install that on a generic PREP or CHRP based Power PC, or under a Power PC emulator on an x86 computer. Part of the cost of OS X is the license that's bundled with a genuine Apple Macintosh.
I realize you are referring to the price of a Mac Mini.
The price of a Mac mini, with the Apple OS license bundled, is $500, not $300.
Slashdot Mirror
as someone who worked in ordering huge lots of Apple machines, I know you can get Apple to load whatever you want on them at the factory
Interesting.
That doesn't mean that part of the cost isn't the right to install a legally licensed copy of whatever Mac OS you might later buy on it.
When I buy a Dell, it is not tied to Windows upgrades forever and ever.
The retail Windows license is a complete license to run Windows, it's not tied to Dell or any other vendor's hardware. The retail OS X license is tied to Apple branded hardware only, it is not by itself a license to run OS X the way a Windows license is a license to run Windows.
When you buy a Windows upgrade, however, you do not get a complete Windows license. You get a package that treats the previous version of Windows as a dongle that must be present to perform the install. It's an upgrade to a license you already have.
When you buy a copy of Mac OS X, you get a package that treats the presence of an Apple-branded computer as a dongle that must be present to perform the install. The only difference between this and the Windows license is that Microsoft calls theirs an upgrade explicitly.
What is Apple doing that is so different here?
What Apple's doing that's different is building part of the revenue they receive for the operating system into the sale of the computer. Dell doesn't do that, the OS is a cost for them. Microsoft doesn't do that, the only revenue they receive from the hardware sale is the license fee Dell pays them.
and you can still install it on your clone too.
Not legally, unless you're being really careful in your interpretation of what an Apple computer is.
Apple just wants to sell hardware, thats it.
If that was it, then they would sell the hardware for a significantly lower price, and the OS for a significantly higher one. Because it's the right to install Mac OS on it that allows them to charge hundreds (and in some cases pushing four figures) more for a Mac than for a comparably equipped computer that won't legally run Mac OS.
Its the same thing as Red Hat one day revoking Dell license to provide Linux on their machines
Since Red Hat doesn't recieve any money from the sale of Dell machines, they have no incentive to do that. In addition, they don't have the legal right to do that under the terms of the GPL.
Apple won't give a damn. They sell hardware. They don't care what OS you use.
Exactly what Apple sells depends on who you talk to.
To someone developing Power-PC based embedded systems, Apple sells relatively cheap development and test systems for Power-PC software... because industrial quality hardware and enclosures are MUCH more expensive than the ones Apple builds.
To some Mac users, Apple sells style and mystique. These people consider the pretty boxes worth the money.
To other Mac users, Apple sells computers that are well enough made to justify their price.
To still other Mac users, Apple sells computers running Mac OS. These people consider Mac OS to be worth enough to justify the premium they pay for a Mac.
My claim is that this last category constitutes the majority of the Macintosh market. The sales decline that resulted from the creation of a Mac clone market, and the poor quality of many Macs (including, in my opinion, all the laptops... but that is my opinion and I am aware that many people disagree), as well as my own reasons for buying around a dozen Macs over the years, is the basis for this claim.
So, if Mac OS is the driver behind the sale of most Macs, what Apple sells is software. Their business model involves selling hardware (Macintosh computers) that their software requires to run, but it's the software that makes these sales possible.
If you buy a Macintosh, and you don't subsequently install Mac OS on it, Apple has still received most (or all, do they actually
I love Linux, and as such, I feel pain every time someone says "just fix it yourself".
What I sometimes find frustrating about BSD and more often consider a major advantage is that it's one of those systems (or three of those systems) where you get a lot less "just fix it yourself" and a lot more "we already thought of that, we're not doing it that way because of X, Y, and Z, and the right way to do what you want to do is...".
It's frustrating when the official right way sucks, but great when it doesn't. And the official right way sometimes changes to something closer to what you wanted, and sometimes it's even better.
Did you write your software for you and then give it away?
Yes.
Now its in use by a million people, do you care about their opinions?
Yes.
That doesn't mean I don't say both "just fix it yourself" and "no I didn't do it that way because...". Also, more and more, "I'm not the maintainer any more, tell so-and-so...".
What I really like is when someone sends me a fix that fixes something I haven't gotten around to yet and ALSO works well with what I HAVE gotten around to.
Mac OS computers don't come with any license for Mac OS X built in.
Go back and read what I actually wrote. That's the second time you've demonstrated the same misunderstanding about it.
And yes, there are Macs that come with Linux that are blessed by Apple.
They didn't ship with Linux installed. They shipped with Mac OS installed and it was replaced with Linux.
If Linux just "happens to run", how come it knocks out OpenBSD when it comes to performance?
Taking shortcuts is often a very effective way to get better performance. OpenBSD is particularly notorious for NOT taking shortcuts even in the BSD world.
"I don't know. I've never used Linux"
I used to say that, and I got tired of it, so I set up Linux and moved some of the software I was running on BSD over to it. I won't go into the variations and distributions and versions and things that worked and things that didn't work, but the thing that was most amazing is how hard it was to get Linux users to see why I thought some of them were problems.
I'm not using Linux any more, when I have an alternative. Apple took away the last incentive I had for giving it another shot on my desktop.
I think you're conflating Open Desktop or Open Server with UNIXware?
Historically, putting "Open" in a product name or service has been a useful warning sign, similar to "Honest" on a used card dealer's lot.
...
For example, "OpenVMS", "MVS Open Edition", "Open Desktop" or "Open Server",
Unfortunately this is no longer a reliable guide, as some open systems and open source organizations have muddied the waters by using it in the previous (and, for a time, obsolete) sense.
Practically speaking, sometime before 1985 or so with POSIX, and after Berkley first developing TCP for their Unix, there was no UNIX in any current sense of the term.
OK, so the PDP-11s and VAXes that I was working on at Berkeley in 1980 weren't UNIX. The software I wrote that shipped in some of the 4.1BSD tapes wasn't UNIX software.
Would you say Unix style job control is a part of your definition for a member of the UNIX family?
That would have to be "no", I'm afraid, because that was actually developed at Berkeley while I was an undergraduate there, and it didn't become a normal part of "Trademark UNIX" until well after the 1985 date you referred to, since AT&T used a quite different mechanism called "layers" based on the BLIT terminal model and what became Plan 9.
Is VMS a member of the UNIX family? No, it isn't, but it too can be treated as a Unix alike with the right functionality installed, and so can meet the Single Unix Spec and be POSIX complaint,
That sure sounds like you agree with me that having a subsystem that meets the POSIX specification is actually unrelated to whether the OS that hosts that subsystem is a member of the UNIX family. So, really, there's no point even bringing POSIX into the question... POSIX is a definition that's useful for writing and supporting portable software, but it doesn't fundamentally differ in kind from (for example) the "Software Tools" virtual operating system and other schemes for running software developed for UNIX on other operating systems.
as far as BeOS being Unix goes, it depends on which sense of Unix one means
Precisely:
One can easily treat BeOS as an early, pre-X11, pre-POSIX single-user Unix, as long as only externally presented interfaces are used
Are you familiar with the Turing Test? Consider the thought experiment Turing created: when you're deciding whether there's an intelligent enity on the other end of the teletype link, you don't have any way to examine that entity's internal design, all you have to go on are the externally visible interfaces. If you can't tell by using those interfaces whether the entity at the other end is a machine or a human, then you have to treat it as if it's intelligent... whether it's a human, a martian, a DEC VAX, a dog with an Internet account, a Turing machine, or a thousand literate monkeys reading responses from books called "how to be human".
UNIX is like Intelligence in this respect.
Since the only thing that you can interact with an operating system through are the externally presented interfaces, that's all you can use when you're talking about whether it's a member of the UNIX family. As in the Turing test, the internals of an operating system are irrelevant.
I would love to concede every point you made, and smugly bash Windows as inferior in every technical and practical sense.
I'm sorry, we seem to be having unrelated converstions here. It's not my intention in this discussion of UNIX to "smugly bash Windows", and I certainly wouldn't describe Windows NT (at least) as inferior in every technical sense. There's a lot of really interesting and useful solutions in the NT kernel. I'm also not equating (as you seem to think I am, unless I'm badly misunderstanding you) the concepts "is a member of the UNIX family" with "is good" or "is not a member of the UNIX family" with "is bad". There's quite a lot of operating systems and environments I've used that I consider very good indeed that aren't in any way UNIX-like, and quite a lot of members of the UNIX family that I've used that I'm absolutely sure you wouldn't consider "UNIX-like", based on the kinds of facilities you've said you expect to be in any UNIX system.
The only aspect of being a member of the UNIX family that's really relevant to being "good" or "bad" here is that it's a lot easier to get along with the creature or machine at the other end of a teletype link if one can find common ground with i
The Mac OS X license only allows you to install on computers that are Apple branded.
You just repeated what I said using slightly different words.
I can install that OS on any Apple branded machine no matter what OS it came with.OS 9, Linux, etc.
No Apple has ever shipped with Linux.
When you buy a Macintosh part of the cost is the actual physical hardware, and part of the cost is the license that makes it legal to install Mac OS and upgrade to future versions of Mac OS. It is not legal to install Mac OS, any version of Mac OS, on a computer that isn't a Macintosh. What makes a piece of physical hardware a Macintosh rather than a PREP or CHRP based PC... even if it's a Macintosh that conforms to the CHRP or PREP standard... is that license that entitles you to install Mac OS and to upgrade to new versions of Mac OS. When you buy a retail copy of Mac OS, any version, you are not buying the right to install it on any computer, which would be the case if it was a "full install", you are buying the right to install that version of Mac OS on a computer that's already licensed to run Mac OS. When you recieve a copy under the "upgrade program", you can only upgrade Mac OS on an actual Macintosh to the new version. Apple explicitly excludes the licensed Mac clones, for example, from being upgraded from Mac OS to Mac OS X.
And the value of that license to Apple is the difference between the cost of the computer and what they can sell it for. That difference is what a true "full license" for Mac OS X would have to cost, if Apple were to sell such a thing.
If this distinction is too subtle for you, I'm sorry. I don't know how to make it clearer.
Um, my main machine up until a year ago never came with OS X. Ever.
No version of Mac OS at all? It's an actual Mac, not a clone? I call shenanigans: I have never heard of ANY Mac ever that shipped without Mac OS and a license for the same.
the current (aluminum) PowerBooks have the most outstanding laptop keyboard that I've ever used.
You've never used a Thinkpad, then. It's got the closest thing to a real keyboard I've found on any laptop, with textured keys that have a deeper "throw" than any other laptop I've seen since the Toshiba Sattelite.
Compared to the Thinkpad, even the latest aluminum powerbook is uncomfortable to use. It may be better than previous 'books in this form factor, but they're still flat and square with poor feedback.
The Apple desktop keyboards are pretty unpleasant, for that matter. The older keyboards they used on the Beige Powermacs were much better.
mini: Slabs are cool, but what's wrong with the mini's form factor?
It's so small that it's got more of a heating problem than the G5 iMac, which means it's shortchanged on disk (low speed 2.5" disk instead of a cheaper, faster, and more reliable 3.5" disk) and power (both the video and USB have less power than specced, I can't even spin up the external 2.5" bus-powered USB disk I use for file transfers using without sticking a powered hub in the way, and video problems on the mini are legendary).
A slightly larger unit based on the eMac instead of the iBook would have been much more versatile.
So he's running on the same machine as this spammer?
So by your dumbfoundingly twisted logic, it seems you argue that if I buy that mini then I can use it's included discs to install OSX on something other than a Mac?
No, no, the full version is only licensed for a particular machine. I wouldn't dream of implying otherwise.
You can't just install OS X on any old Power-based machine because it doesn't have drivers for all the hardware to boot and run natively.
Ah, but that's the interesting bit. See, Apple provides us with all the source code to boot the kernel on generic intel hardware and, if it were Mac OS X for the x86, it could even run natively. And, you see, since PREP and CHRP motherboards are so well documented and have Linux drivers available, running Mac OS X on them is really only prevented by the licensing.
The cost of the mini hardware has been calculated elsewhere, and it was over $300 the last time I saw it.
I've seen many calculations of it, and these calculations are always higher when they're made by a Mac fan, and lower when made by a PC fan. Some of them, like the $100 figure one person posted, are clearly ridiculous. I believe it's closer to $200 than $300... but whatever it is, it's a useful guideline as to the lowest possible price Apple would have to sell a full licensed version of Mac OS X Intel for generic PC hardware and still make the kind of profit they need.
If you pick a more expensive Mac as your baseline, that value is higher. For a top of the line Powermac G5 it's close to four digits. I suspect that Apple would need to sell SMP and non-SMP-capable versions for different amounts.
But even after it opens the "safe" file, apps get the "This is the first time you are running this program, are you sure you want to run it?" prompt if something tries to just auto start itself just after download.
Yes, that's a clever trick, isn't it? Microsoft does it too, and it gets people to feel safe about going ahead and clicking on links and downloading files like a wild monkey. They tell themselves "It's OK, it'll ask me before it does something stupid"... but of course because they're used to clicking OK so enough of them go ahead and click OK anyway, without thinking, to keep the exciting virus ecosystem flourishing. It's this kind of extra attention from Microsoft that system administrators everywhere really appreciate and enjoy.
Unfortunately Apple left people the option of turning off "open safe files after downloading" so instead of having a dialog they can reflexively click OK on, they just get a boring icon on the desktp that they can let sit there, and think about, and have second thoughts about and throw in the trash. What a bunch of spoilsports Apple are, denying people the full impact of reflexively hitting the wrong button like that.
You'd think they didn't care.
Show me real evidence that a full version of OSX costs $300, even though Apples sells it for $129.
The full version can cost a lot more than that if you buy it with a higher end machine. $300 is the smallest amount you can get away with if you want to buy the right to install the OS. It's never before been so inexpensive.
As stated, when you pay $129, you can install it on any supported Mac, with or without an OS.
That's true. All supported Macintosh computers come with the right to install Mac OS X, along with the copy of Mac OS that they shipped with. If you don't buy a Macintosh computer, you don't get that right... even if it's technically possible (as with the 'Mac Clones').
Apple doesn't currently sell a separate "full version", of course, it's only available as a bundle with a Macintosh computer. If they did (as some people have speculated that they might at some point in the future) it's likely that would cost as least as much as the "Mac Tax" that Apple imposes when one buys a Macintosh rather than a similarly specced computer without that license.
Apple only explicitly labels their software an upgrade when it's a special reduced cost version for people who bought a previous version just before the new upgraded version comes out, but in practice every version is an upgrade that is only licensed for use on a computer that was originally shipped by Apple with Mac OS included.
I realise that this is a popular myth, and one that keeps people feeling comfortable that nothing Microsoft might do will damage their rich and exciting viral ecosystem, but unfortunately it's not true. Microsoft would merely have had to back down under the tyrannical pressure of the US Department of Justice and unbundled IE and the refreshing flood of viruses and spyware would have collapsed into a disappointing trickle.
How lucky we are that they showed such wisdom and determination!
Microsoft had just about as much marketshare before they integrated IE and the Desktop as after it, but the difference to fans of malicious software was startling. What kicked viruses into high gear was the ability of the Microsoft HTML control to run native code delivered by an untrusted website or mail message. It was the killer application for virus delivery, because it turned the previously unthinkable idea of a virus that could spread if you just opened an email message into reality. Sites like ours that banned IE and Outlook and later Windows Media Player and Realplayer and any other internet-enabled applications that used the HTML control had to put up with completely dismal virus levels.
Even refusing to use any antivirus protection other than telling people not to open unexpected attachments didn't help. It wasn't until we were forced to switch to IE by our enlightened corporate management that our virus load reached normal levels.
You feel that spam is so important that websites which offer to sell spam software should be blacklisted, along with many other innocent websites hosted at the same ISP.
Anyone who uses a blacklist in such a way that they or anyone at their location or any of their customers can't get to a website because it shares the address of a server that's in a blacklist is (a) a fool, and (b) the person on whom responsibility for this abuse of that blacklist should fall.
Pardon, I thought "you're a fucking idiot" was referring to me.
The anonymous coward who called you one was a sorry bastard, but he was right in saying that you're going to have to find a static smarthost for your outgoing mail.
you're making the assumption that we all have static IPs.
If you want to send mail, you need a static IP.
If you have a dynamic IP, then my mail server won't accept mail from you regardless of whether you're part of a botnet or not, because not only is one of the BLs I use simply a list of dynamic IPs, I also explicitly block hosts whose reverse lookup indicates they're in an un-RBLed dynamic IP space.
my home machine isn't my mail server, it's simply my home machine. Sometimes it needs to send out mail for various reasons.
Yep, so does mine.
I pay £400 a year and have a server hosted somewhere.
I don't pay quite that much, but it's in the same range. I'm also about to set up a backup virtual server for another $70 a year.
So when my home machine wants to send mail. it does it over an encrypted tunnel to my colo.
I should not be forced into using smart relaying
Entirely true. You shouldn't be forced to deal with the results of spammers pissing in the pool. But until spammers are fined or otherwise sanctioned sufficiently to make spam the minor problem that it should be, having to arrange for a reliable smarthost is part of the cost of being on the net.
Please consider what you're saying before you hit submit.
I have, at length, and I stand by it.
The point is that it will always be impossible to keep an accurate map of infected vs. uninfected systems when dealing with numbers this large.
So? You don't need to maintain an accurate map of infected vs. uninfected systems. You just need to keep track of systems that have been infected "recently", for values of "recently" that can be days or even weeks long. Since SBL expires botnet entries in 48 hours they're not contributing to abuse from this cause.
In fact you've given me a good idea for expanding my current bot tracker.
The only solution is to transition to a new protocol that enables authentication so that it won't happen.
The trick is to come up with a scheme that works, that people actually uses, and that doesn't put Microsoft or Verisign or someone similar in charge. Otherwise you wouldn't be able to convince people to turn off their old SMTP servers.
He may be referring to an older qmail version
No, the behaviour he's referring to would occur if one applied filters late in the chain. The thing is, he assumes that you HAVE to apply filters late in the chain. It's a common misunderstanding about qmail.
Think about it: what you're ultimately doing is give some complete stranger near-complete control over what email is or isn't accepted by your system.
I'm already doing that. I'm trusting my registrar, I'm trusting my ISP, and I'm trusting the people running my secondary DNS, and I'm trusting the people "near" me on the Internet.
Blacklists are something that might seem like a good idea in theory
Blacklists are something that is absolutely essential in practice. Without blacklists I wouldn't be able to run a mailserver at all.
The thing about the Mac mini is that it has a large value associated with A, the industrial design.
If you say so. The design may have some cost to Apple, but it's shared with the iBooks that use largely the same components. In fact Apple could have produced a better computer by not using that design: building it to laptop specifications means they have to reduce the power draw significantly, and (coincidentally) the USB and video interfaces are underpowered, it's undersupplied with video RAM, and the 2.5" disk is lower performance than the cheaper higher-power 3.5" disk they could have otherwise used.
Anyway, I think that $300 is a reasonable estimate of the low end of the range Apple would need to charge for a generic Mac OS where they don't share the cost between the retail package and the computer it runs on.
You can purchase a FULL version of OS X, not the upgrade, for $129.
You can only legally install that on genuine Apple Machintosh: a computer that was originally purchased from Apple with a copy of Mac OS (OS 9 or OS X). You can not legally install that on a generic PREP or CHRP based Power PC, or under a Power PC emulator on an x86 computer. Part of the cost of OS X is the license that's bundled with a genuine Apple Macintosh.
I realize you are referring to the price of a Mac Mini.
The price of a Mac mini, with the Apple OS license bundled, is $500, not $300.