The KDE developers were ticked off at Apple not providing useful patches.
The KDE developers were ticked off by people flaming them for not picking up the Apple patches, because the Apple patches were against the Apple sources, and only posted for information. This doesn't mean they were particularly ticked off at Apple for not providing the sources they were patching against... they even noted that they couldn't use them even with that source because the code bases weren't compatible.
They were mostly ticked off at being blindsided by a bunch of complaints from people about something they had no control over.
The developers of KHTML felt that Apple was not being cooperative.
I'm sure some of them did, but that's not why the flame-war erupted, and developers of open source software (or, for that matter, non-open-source software) have to deal with people "not being cooperative" all the time. Not just companies, but other open source projects, and people in their own project. "Not being cooperative" isn't unethical, and it may not even be voluntary.
"Not being cooperative" includes (for example) open source projects dropping support for other open-source projects. It involves people making incompatible changes in interfaces. GCC has gone through splits because of people "not being cooperative". GCC developers have flamed Red Hat for "not being cooperative". There are open source developers with widely used projects that are notorious for "not being cooperative". When I submitted patches to Tcl back around Tcl 2.x, John Ousterhout completely reimplemented all my work. When Dan Bernstein does stuff like that, people flame him for "not being cooperative", when John did it waited, and I looked at his code and I was glad he'd done it... his code was much better than mine. He had a good reason for "not being cooperative".
People can "not be cooperative" for technical reasons. Apple's doing a lot of the rendering in Safari by passing the work off to their native graphics... none of that code could even potentially be used to the KHTML team EXCEPT as information about how someone else had approached the problem, and some of the patches that started this whole thing off were that kind of change.
You were asserting my events was pure conjecture.
No, I said your INTERPRETATION of the MOTIVES behind the events was pure conjecture. Because it was, and is.
You are asserting there is some sort of chip on my soldier when I've said nothing but that the problem is fully resolved.
The person I originally replied to (was that you?) obviously doesn't consider it fully resolved OR THEY WOULDN'T HAVE BROUGHT IT UP.
What does Caldera have to do with Godwin's law?
Peter's contraflow to Godwin's Law... every community has their own Hitler.
The GCC fork was IMHO an effort that was similar to their RedHat kernel
AFAICT it was an attempt to bring some GCC3 (IIRC) features in to a GCC2 code base to take advantage of them before they were ready to go to GCC3.
But even if it was a technically different fork, so is Webcore. You can't merge Webcore and KHTML without a huge amount of work that's probably not worth doing... because KHTML doesn't have Quartz and Aqua to do the work for it.
(goddamn/.... I tried to post this yesterday as I was leaving but it decided that posting a correction right after a message was abusive or something. It's also got 'issues' with cut-and-paste)
WSU is a lot more than Interix. Really, Interix itself *is* the core you're talking about... the updated POSIX subsystem (which did and does include the shell and tools, just as the BSD subsystem on OSX does... at one point, when it was called Rhapsody, Apple was considering leaving that stuff out as much as possible and only including the parts that were absolutely essential to running the system, but luckily they came to their senses). It started as an external product and only got bought by Microsoft and integrated with the rest of WSU later.
The developers were told by management that the bad PR was a result of their actions and they wanted a policy change.
That doesn't mean they thought it was unethical, that just means there was bad PR.
Because KHTML ain't under the business friendly BSD license. Different community different rules different attitudes.
Neither are many of the other components that are also released with the BSD licensed code on the same schedule. For that matter, there are actual open source groups that have maintained their own parallel forks of GPL code on the same kind of release basis. There's no reason for they or I to assume that there's any different "rules" or "attitudes" for GPLed codes. Because there isn't.
There's apparently a different attitude among certain outsiders to the KHTML team, and the KHTML team got irritated by the reaction of those outsiders... but the amount of steam they blew off was minimal. All the noise is coming from a very small group who still (as is proven by this exchange) got a chip on their shoulder about Apple and who don't, as you put it, consider Apple "in good standing" as a result.
BTW there have been companies with your attitude towards community development. Caldera [...]
Ah, Godwin's Rule already? Bravo. On the basis of I don't know what (you certainly can't tell what "my attitude towards community development" is from this thread) you've decided I must be a corporate shill.
To the best of my knowledge Microsoft doesn't violate copyright law.
What, apart from being found guilty in multiple suits of stealing code, technology, and trade secrets, and settling in more to avoid a finding?
The Free Software community doesn't care about companies using open source they care about them joining the open source community.
Yep, and Apple's maintained regular, timely, and increasingly complete releases of Darwin, they cooperate with and support the FreeBSD project, and they absolutely bent over backwards even where there was and is not even a hint why this ONE MAGIC SUPER SPECIAL ELITE open source product MUST be handled completely differently from EVERY OTHER open source product they're involved in... no matter what license they're using...
As for Red Hat... oh, man, you want to talk about a company that's got a bad attitude towards the community! Red Hat's forking of GCC, their magic installer-ofe-the-month and their "special status" causes no end of friction in the open source community. Their releases have often been "oh my god, what the hell are RH doing this time?", and RHN steers as close as they can to being "closed" as they think they can get away with... to the point where there's two separate projects doing nothing but maintaining a parallel distribution so Red Hat users can get the support they need despite Red Hat - I've got four copies of RHEL ES on the shelf and I'me running White Hat on the servers themselves because it's too much trouble to keep up with RHN through a hostile corporate firewall. The duplication of effort is insane.
Red Hat gets the support from the community because they buy it with market share and just plain cash. NOT because they're well-behaved.
Its an interesting argument, but you may have forgotten the upgrade revenue stream.
Why would you expect the upgrade revenue to change? What does Windows have to do with it? I'm not saying you're wrong, I just don't understand your reasoning.
You're looking at "the open source world" through rose-tinted glasses.
Also: "Evidentally Apple management didn't believe it was [ethical]" is pure speculation. Apple's internal source code management system wqsn't designed for open source work. Well, you know, neither is mine. I don't release my repository, I release complete snapshots. Am I being unethical?
Fuck no.
Apple has consistently done more than they had to with their open source components. They do regular releases of their BSD licensed code, *exactly* as they had been doing with the Safari code. That was... and still is... considered way more than reasonable. Why on earth do you think they should have expected ANYONE would complain about them doing the same thing with Safari?
Maybe they should have expected that the whacko wing of the open source movement would pounce on something they were doing and make up a bunch of complaints that nobody's ever made of other companies that use open-source products. Regardless, they came back and set up a repository just for the KHTML code. And in return for going beyond the call of duty you come back and argue that this was an admission of guilt.
Man, if you want to discourage companies from using open source code in the future, you couldn't have done a better job.
Are you really trying to argue that they shouldn't have fixed it.
Hell no. Until you started implying underhanded motives to that very act the thought wouldn't have entered my head. Keep trying, maybe you'll convince me I'm wrong and they shouldn't have bothered to release anything but what they were absolutely required to in the first place. It seems they would have gotten less flack for it.
Interex is similar in all but a technical sense to Cygwin.
Interix is not in any sense simlar to Cygwin. It is _precisely_ a replacement for the POSIX subsystem, it provides real usable POSIX semantics... not a crippled version... and serves the same purpose in Windows that what Apple calls "the BSD subsystem" does in Mac OS X.
Think of the NT kernel as Mach, and Interix as a BSD server, and you'll be closer.
Cygwin is just a bunch of compatibility libraries on top of Win32. It's not secure enough to be a real replacement for the BSD subsystem, but Interix is.
Apple _also_ already has OpenStep for Windows, so they could back-port Cocoa to that.
And for Carbon, they've got the compatibility library iTunes and Quicktime for Windows runs under.
No it wasn't. Apple was already doing more than the GPL required, the whole issue was caused by people outside BOTH Apple and the KHTML team complaining that KHTML wasn't picking up "Apple's Fixes" and hassling them over it. Demanding that a port to a *completely different window system* maintain detailed compatibility is letting the crack-pipe think for you.
That's because you are not accounting for history.
I was writing free software and porting the UNIX environment to other operating systems before there was a GPL, I was working on free UNIX before Linux existed. I'm one of the tens of thousands of little guys who created that history you're talking about... before the Free Software Foundation decided it belonged to them. I've written, ported, and run UNIX software on 5th, 6th, and 7th edition, on RSX-11 and VMS and RSTS, on CP/M and MS-DOS and TOS and AmigaOS and Windows NT, on more versions of UNIX than most peole should have to know about.
So don't bloody tell me about history.
The most important lesson of history is you don't depend on ANY platform, not OSX or Linux, VMS or Tru64, Solaris or SunOS, RSX or RTE-IV, CP/M or Cromix or Regulus or Lanetix or Idris. You write your software so it's portable, you PORT your software so you know what portable means, and you don't tie yourself to *anyone's* wagon.
That's how you stay free.
The point behind using Linux (and Free software in general) is because someday, for whatever reason, your favorite proprietary vendor can pull the rug out from under you.
Except they can't. Did you actually bother to read my message where I explain in detail why?
Apple knifing source code releases is a symptom of where their concerns lie.
Except that didn't happen. It may, I agree with the arguments about that, and I'm amazed AND impressed that it hasn't. But... it hasn't.
The people who really depend on Linux and its environment of hacker-as-customer-#1 mentality.
Man, I've been a hacker, in the original sense, since I taght myself programming as a kid in 1972 so I could cheat on the BASIC computer games on the mainframe at Dad's office.
Linux isn't all that interesting an environment to hack in. The thrust of Linux hacking is in file systems, getting games and Windows applications to run, grooving on what Reiser's going to do next, and trying to turn Linux into the successor to OS/360. BOR-ING.
What else is there? Applications? Gnome? KDE? That stuff isn't Linux specific, and Gnome and KDE are both hacks in the worst possible sense in any case. Mono? Let's set ourselves up for getting Gatesed (and Gatesing ain't no rumor)! Berlin/Fiesta? Dead. GNUstep? The "Linux Hackers" have mostly abandoned that. WHat's left. What do you actually depend on? Details, please!
Even an extremely optimistic projection would not allow OS X's sales to come close to generating the revenue Apple currently receives from hardware sales at anytime in the first few years after the transition.
Let's do the numbers.
Conventional wisdom is that Apple's margins on their systems are 40%. My experience with the Mac mini suggests that's probably about right, and their other prices are in line with that. Their margins on an Intel mini may actually be lower, based on Freescale's prices for the G4 and Intel's prices for the "Core", but let's go with 40%.
OK, for the Mini, if they sell the bundled software for $200 they'll break even.
For the iMac, they'd have to sell it for $500.
For the Macbook Pro or Powermac, $800.
For the iBook, $400.
They're bundling iLife, which they sell separately for $80.
Let's say their sales of all these systems are the same. Evidence is that the low-end systems are the bulk of the sales, but I'm not counting the higher end options on each system,...
So, the average value to Apple of the "software part" is a bit under $500.
$80 of that is iLife, leaving $420 to make up.
Now, let's say they were to sell OSX86 "generic" for the same price Microsoft sells Windows XP Pro 1-2 CPU version for, or about $400, and the "Mac only" version for the same price Microsoft sells an XP Pro upgrade, or about $130. This is a reasonable set of prices, I think they shouldn't have much trouble defending them.
Looks to me like their net lost profit per sale would be pretty small.
And they could (like Microsoft) sell the quad processor version for more, to make up the massive profits on the Powermac.
There's other issues, of course, but I don't think they'd be hurting for lost revenue.
If the BSDL wasn't there, they'd simply have gone with another kernel. Possibly BeOS, possibly NT, possibly they'd have continued to use the irrevocable UNIX licenses posessed by both Apple (for A/UX) and NeXT (for NeXTStep).
The chances of them using a GPLed kernel, though, are similar to the chances of Microsoft releasing the source to the NT kernel under the GPL.
Virgin reports that 93 million text messages are sent every day in the United Kingdom (U.K.). One estimate for the United States (U.S.), whose population is five times as large, is 700 million text messages a year.
That's odd... Texting has been practical in the UK much longer than in the US, where for the longest time it simply wasn't possible to send text between networks. There's so much more support for text in the UK and one sees so much more online evidence of a text 'culture' there that it seems unlikely that its per-capita text message use would be higher.
"Music Online" isn't really the market Apple's playing in, and they're only a few percent of the whole market.
I can take a Fairplay-protected song, burn it to a CD, rip it in Windows Media Player, and play it on a Rio. I can buy a CD with the music on it and rip it for any hardware. I can burn a song from Rhapsody and rip it for an iPod.
You're not locked in by music, at worst it's an inconvenience.
But I can't "rip" software from Windows and run it on UNIX or Mac OS, except by running Windows under emulation, or by emulating Windows (which is still far from practical for the average user, and enough of a moving target it probably will never be). THAT is why there's a difference - there's software you can't run at all without Windows, except through heroic measures. If you want to play music that's not available on Rhapsody on your Rio, you have dozens of easy options.
If it wasn't for the applications I'd still be using 'traditional' free UNIX as my primary desktop, and treating Windows as an X terminal that happened to run Office and games.
No matter how much you accelerate Linux or FreeBSD graphics, or how much eye candy you add, it won't matter if it's "a better OS" or not if it doesn't have the apps.
Apple isn't required to open-source Darwin, though there are components they are required to continue releasing. That doesn't mean this wouldn't be news, something to worry about, etc, etc, etc...
Mac computers are sold with a license to use Mac OS X. The retail box you can buy is an upgrade, which presupposes you already have a Mac (and thus a Mac OS license).
Interesting point, one I have made (albeit with one extra word... 'the retail box is effectively an upgrade') many times. The problem is that there's been people running retail copies of Mac OS X on machines never licensed for it for way too long for it to be called "piracy".
It may be against the EULA, it may be illegal, it may be many things, but "piracy" is too strong a term.
But even more interesting:
there is a way to avoid it being quote-piracy-unquote while still breaking the EULA: buy and wipe clean an old Mac.
Everyone with a legal copy of OS X Intel they can install on a non-Apple box got it with a Mac they can install Linux on to stay kinda kosher.
While I don't agree with the people who beable on about how "used Macs enable even the poor to productively use OS X, just look at eBay", it's worth noting that you CAN get Macs capable of running OS/X up to 10.2.8 without any tweaking for as little as $40... and get quite a good taste of the system from that.
Before we get into the digression I'm glad to hear that my prediction has not in fact come true, and a technical glitch has been blown way out of proportion (which is, now I think of it, what happened with the "Apple's ripping off KHTML" flamewar).
---
Onwards!
You can't actually buy a copy of OSX for x86, the only versions on the shelves are PPC.
That's only one of many reasons I haven't personally so much as downloaded any of the bits necessary to install any version of Mac OS X on Intel hardware. Don't make this about me, please, because it's not.
However...
and unless you wipe it from your new Mac
There's quite a few people working on and even successfully running Linux on the new intel-based iMac, as well as people working on running Windows on it.
Personally I think they're all nuts, but there are people who can install Mac OS X for Intel on a non-Mac without being "pirates".
It's booting through the BSD boot process, and the BSD kernel is running single-server, and they have increasingly moved away from using Mach messages because of the overhead. It's got more Mach in it than FreeBSD does, but FreeBSD was already using a lot of Mach code, and they've continued to re-import a lot of updated FreeBSD code (eg, FFS in Panther).
Like everything else, it's a mongrel. It's part Mach, part BSD, part NeXT, part Mac, something old, something new, something borrowed, something blue...
Exactly why I resisted the pressure to abandon Linux for MacOS X on the desktop
I'm not sure what your point is. Let's say the absolute worst thing happens and the next version of OS X is based on an NT kernel, all the UNIX-compatible stuff is supplied by Interix, and Bill Gates buys Apple.
All the commercial software I've got on my Mac will continue to run.
All the open source software I've got on my Mac will continue to run.
All the software I'd have been using under any other free UNIX will still be just as available as it is right now, and I can continue to use it on Linux or Mac OS.
If for some reason I want to run Linux on my Mac Mini, I will still be able to run all my Mac OS software under MOL.
If I want to run Linux on an Intel box I can, and all the software I would have been using on Linux and all the open source software I'm using on Mac OS X will still be available.
If for some reason I want to use my Mac software on an Intel box running Linux, I will be able to do so using Sheepshaver, under emulation, just as I woudl be able to use it using Rosetta, under emulation, on an Intel Mac.
What would have been the advantage of using Linux for the past three years instead of Mac OS X, even under the brutal worst-case regime I described above? I really don't get it.
Slashdot Mirror
Did you bother actually looking at the license on the code you're referring to?
What's GPL3 got to do with code that's never been under the GPL?
The KDE developers were ticked off at Apple not providing useful patches.
The KDE developers were ticked off by people flaming them for not picking up the Apple patches, because the Apple patches were against the Apple sources, and only posted for information. This doesn't mean they were particularly ticked off at Apple for not providing the sources they were patching against... they even noted that they couldn't use them even with that source because the code bases weren't compatible.
They were mostly ticked off at being blindsided by a bunch of complaints from people about something they had no control over.
The developers of KHTML felt that Apple was not being cooperative.
I'm sure some of them did, but that's not why the flame-war erupted, and developers of open source software (or, for that matter, non-open-source software) have to deal with people "not being cooperative" all the time. Not just companies, but other open source projects, and people in their own project. "Not being cooperative" isn't unethical, and it may not even be voluntary.
"Not being cooperative" includes (for example) open source projects dropping support for other open-source projects. It involves people making incompatible changes in interfaces. GCC has gone through splits because of people "not being cooperative". GCC developers have flamed Red Hat for "not being cooperative". There are open source developers with widely used projects that are notorious for "not being cooperative". When I submitted patches to Tcl back around Tcl 2.x, John Ousterhout completely reimplemented all my work. When Dan Bernstein does stuff like that, people flame him for "not being cooperative", when John did it waited, and I looked at his code and I was glad he'd done it... his code was much better than mine. He had a good reason for "not being cooperative".
People can "not be cooperative" for technical reasons. Apple's doing a lot of the rendering in Safari by passing the work off to their native graphics... none of that code could even potentially be used to the KHTML team EXCEPT as information about how someone else had approached the problem, and some of the patches that started this whole thing off were that kind of change.
You were asserting my events was pure conjecture.
No, I said your INTERPRETATION of the MOTIVES behind the events was pure conjecture. Because it was, and is.
You are asserting there is some sort of chip on my soldier when I've said nothing but that the problem is fully resolved.
The person I originally replied to (was that you?) obviously doesn't consider it fully resolved OR THEY WOULDN'T HAVE BROUGHT IT UP.
What does Caldera have to do with Godwin's law?
Peter's contraflow to Godwin's Law... every community has their own Hitler.
The GCC fork was IMHO an effort that was similar to their RedHat kernel
AFAICT it was an attempt to bring some GCC3 (IIRC) features in to a GCC2 code base to take advantage of them before they were ready to go to GCC3.
But even if it was a technically different fork, so is Webcore. You can't merge Webcore and KHTML without a huge amount of work that's probably not worth doing... because KHTML doesn't have Quartz and Aqua to do the work for it.
Oh!
/. ... I tried to post this yesterday as I was leaving but it decided that posting a correction right after a message was abusive or something. It's also got 'issues' with cut-and-paste)
Per day vs per year.
<emilylatella>nevermind</>
(goddamn
WSU is a lot more than Interix. Really, Interix itself *is* the core you're talking about... the updated POSIX subsystem (which did and does include the shell and tools, just as the BSD subsystem on OSX does... at one point, when it was called Rhapsody, Apple was considering leaving that stuff out as much as possible and only including the parts that were absolutely essential to running the system, but luckily they came to their senses). It started as an external product and only got bought by Microsoft and integrated with the rest of WSU later.
The developers were told by management that the bad PR was a result of their actions and they wanted a policy change.
That doesn't mean they thought it was unethical, that just means there was bad PR.
Because KHTML ain't under the business friendly BSD license. Different community different rules different attitudes.
Neither are many of the other components that are also released with the BSD licensed code on the same schedule. For that matter, there are actual open source groups that have maintained their own parallel forks of GPL code on the same kind of release basis. There's no reason for they or I to assume that there's any different "rules" or "attitudes" for GPLed codes. Because there isn't.
There's apparently a different attitude among certain outsiders to the KHTML team, and the KHTML team got irritated by the reaction of those outsiders... but the amount of steam they blew off was minimal. All the noise is coming from a very small group who still (as is proven by this exchange) got a chip on their shoulder about Apple and who don't, as you put it, consider Apple "in good standing" as a result.
BTW there have been companies with your attitude towards community development. Caldera [...]
Ah, Godwin's Rule already? Bravo. On the basis of I don't know what (you certainly can't tell what "my attitude towards community development" is from this thread) you've decided I must be a corporate shill.
To the best of my knowledge Microsoft doesn't violate copyright law.
What, apart from being found guilty in multiple suits of stealing code, technology, and trade secrets, and settling in more to avoid a finding?
The Free Software community doesn't care about companies using open source they care about them joining the open source community.
Yep, and Apple's maintained regular, timely, and increasingly complete releases of Darwin, they cooperate with and support the FreeBSD project, and they absolutely bent over backwards even where there was and is not even a hint why this ONE MAGIC SUPER SPECIAL ELITE open source product MUST be handled completely differently from EVERY OTHER open source product they're involved in... no matter what license they're using...
As for Red Hat... oh, man, you want to talk about a company that's got a bad attitude towards the community! Red Hat's forking of GCC, their magic installer-ofe-the-month and their "special status" causes no end of friction in the open source community. Their releases have often been "oh my god, what the hell are RH doing this time?", and RHN steers as close as they can to being "closed" as they think they can get away with... to the point where there's two separate projects doing nothing but maintaining a parallel distribution so Red Hat users can get the support they need despite Red Hat - I've got four copies of RHEL ES on the shelf and I'me running White Hat on the servers themselves because it's too much trouble to keep up with RHN through a hostile corporate firewall. The duplication of effort is insane.
Red Hat gets the support from the community because they buy it with market share and just plain cash. NOT because they're well-behaved.
Its an interesting argument, but you may have forgotten the upgrade revenue stream.
Why would you expect the upgrade revenue to change? What does Windows have to do with it? I'm not saying you're wrong, I just don't understand your reasoning.
*snort*
... and still is ... considered way more than reasonable. Why on earth do you think they should have expected ANYONE would complain about them doing the same thing with Safari?
You're looking at "the open source world" through rose-tinted glasses.
Also: "Evidentally Apple management didn't believe it was [ethical]" is pure speculation. Apple's internal source code management system wqsn't designed for open source work. Well, you know, neither is mine. I don't release my repository, I release complete snapshots. Am I being unethical?
Fuck no.
Apple has consistently done more than they had to with their open source components. They do regular releases of their BSD licensed code, *exactly* as they had been doing with the Safari code. That was
Maybe they should have expected that the whacko wing of the open source movement would pounce on something they were doing and make up a bunch of complaints that nobody's ever made of other companies that use open-source products. Regardless, they came back and set up a repository just for the KHTML code. And in return for going beyond the call of duty you come back and argue that this was an admission of guilt.
Man, if you want to discourage companies from using open source code in the future, you couldn't have done a better job.
Are you really trying to argue that they shouldn't have fixed it.
Hell no. Until you started implying underhanded motives to that very act the thought wouldn't have entered my head. Keep trying, maybe you'll convince me I'm wrong and they shouldn't have bothered to release anything but what they were absolutely required to in the first place. It seems they would have gotten less flack for it.
Interex is similar in all but a technical sense to Cygwin.
Interix is not in any sense simlar to Cygwin. It is _precisely_ a replacement for the POSIX subsystem, it provides real usable POSIX semantics... not a crippled version... and serves the same purpose in Windows that what Apple calls "the BSD subsystem" does in Mac OS X.
Think of the NT kernel as Mach, and Interix as a BSD server, and you'll be closer.
Cygwin is just a bunch of compatibility libraries on top of Win32. It's not secure enough to be a real replacement for the BSD subsystem, but Interix is.
Apple _also_ already has OpenStep for Windows, so they could back-port Cocoa to that.
And for Carbon, they've got the compatibility library iTunes and Quicktime for Windows runs under.
The apple ripping of KHTML was a real issue.
No it wasn't. Apple was already doing more than the GPL required, the whole issue was caused by people outside BOTH Apple and the KHTML team complaining that KHTML wasn't picking up "Apple's Fixes" and hassling them over it. Demanding that a port to a *completely different window system* maintain detailed compatibility is letting the crack-pipe think for you.
...something blue...
Yeah, but not for long. Bluebox is getting killed off with the release of the intel iMac (i^2Mac ?).
Har! MOD PARENT +1 FUNNY KTHX!
That's because you are not accounting for history.
I was writing free software and porting the UNIX environment to other operating systems before there was a GPL, I was working on free UNIX before Linux existed. I'm one of the tens of thousands of little guys who created that history you're talking about... before the Free Software Foundation decided it belonged to them. I've written, ported, and run UNIX software on 5th, 6th, and 7th edition, on RSX-11 and VMS and RSTS, on CP/M and MS-DOS and TOS and AmigaOS and Windows NT, on more versions of UNIX than most peole should have to know about.
So don't bloody tell me about history.
The most important lesson of history is you don't depend on ANY platform, not OSX or Linux, VMS or Tru64, Solaris or SunOS, RSX or RTE-IV, CP/M or Cromix or Regulus or Lanetix or Idris. You write your software so it's portable, you PORT your software so you know what portable means, and you don't tie yourself to *anyone's* wagon.
That's how you stay free.
The point behind using Linux (and Free software in general) is because someday, for whatever reason, your favorite proprietary vendor can pull the rug out from under you.
Except they can't. Did you actually bother to read my message where I explain in detail why?
Apple knifing source code releases is a symptom of where their concerns lie.
Except that didn't happen. It may, I agree with the arguments about that, and I'm amazed AND impressed that it hasn't. But... it hasn't.
The people who really depend on Linux and its environment of hacker-as-customer-#1 mentality.
Man, I've been a hacker, in the original sense, since I taght myself programming as a kid in 1972 so I could cheat on the BASIC computer games on the mainframe at Dad's office.
Linux isn't all that interesting an environment to hack in. The thrust of Linux hacking is in file systems, getting games and Windows applications to run, grooving on what Reiser's going to do next, and trying to turn Linux into the successor to OS/360. BOR-ING.
What else is there? Applications? Gnome? KDE? That stuff isn't Linux specific, and Gnome and KDE are both hacks in the worst possible sense in any case. Mono? Let's set ourselves up for getting Gatesed (and Gatesing ain't no rumor)! Berlin/Fiesta? Dead. GNUstep? The "Linux Hackers" have mostly abandoned that. WHat's left. What do you actually depend on? Details, please!
Yeah, that's why the (albeit confusing) statistics show that US users send far fewer texts than UK users.
Hrm? Not the ones I quoted.
UK: 93 million
US: 700 million / 5 = 140 million, accounting for population.
The population figures wrong, or did the story misquote something?
Even an extremely optimistic projection would not allow OS X's sales to come close to generating the revenue Apple currently receives from hardware sales at anytime in the first few years after the transition.
...
Let's do the numbers.
Conventional wisdom is that Apple's margins on their systems are 40%. My experience with the Mac mini suggests that's probably about right, and their other prices are in line with that. Their margins on an Intel mini may actually be lower, based on Freescale's prices for the G4 and Intel's prices for the "Core", but let's go with 40%.
OK, for the Mini, if they sell the bundled software for $200 they'll break even.
For the iMac, they'd have to sell it for $500.
For the Macbook Pro or Powermac, $800.
For the iBook, $400.
They're bundling iLife, which they sell separately for $80.
Let's say their sales of all these systems are the same. Evidence is that the low-end systems are the bulk of the sales, but I'm not counting the higher end options on each system,
So, the average value to Apple of the "software part" is a bit under $500.
$80 of that is iLife, leaving $420 to make up.
Now, let's say they were to sell OSX86 "generic" for the same price Microsoft sells Windows XP Pro 1-2 CPU version for, or about $400, and the "Mac only" version for the same price Microsoft sells an XP Pro upgrade, or about $130. This is a reasonable set of prices, I think they shouldn't have much trouble defending them.
Looks to me like their net lost profit per sale would be pretty small.
And they could (like Microsoft) sell the quad processor version for more, to make up the massive profits on the Powermac.
There's other issues, of course, but I don't think they'd be hurting for lost revenue.
If the BSDL wasn't there, they'd simply have gone with another kernel. Possibly BeOS, possibly NT, possibly they'd have continued to use the irrevocable UNIX licenses posessed by both Apple (for A/UX) and NeXT (for NeXTStep).
The chances of them using a GPLed kernel, though, are similar to the chances of Microsoft releasing the source to the NT kernel under the GPL.
Virgin reports that 93 million text messages are sent every day in the United Kingdom (U.K.). One estimate for the United States (U.S.), whose population is five times as large, is 700 million text messages a year.
That's odd... Texting has been practical in the UK much longer than in the US, where for the longest time it simply wasn't possible to send text between networks. There's so much more support for text in the UK and one sees so much more online evidence of a text 'culture' there that it seems unlikely that its per-capita text message use would be higher.
"Music Online" isn't really the market Apple's playing in, and they're only a few percent of the whole market.
I can take a Fairplay-protected song, burn it to a CD, rip it in Windows Media Player, and play it on a Rio. I can buy a CD with the music on it and rip it for any hardware. I can burn a song from Rhapsody and rip it for an iPod.
You're not locked in by music, at worst it's an inconvenience.
But I can't "rip" software from Windows and run it on UNIX or Mac OS, except by running Windows under emulation, or by emulating Windows (which is still far from practical for the average user, and enough of a moving target it probably will never be). THAT is why there's a difference - there's software you can't run at all without Windows, except through heroic measures. If you want to play music that's not available on Rhapsody on your Rio, you have dozens of easy options.
If it wasn't for the applications I'd still be using 'traditional' free UNIX as my primary desktop, and treating Windows as an X terminal that happened to run Office and games.
No matter how much you accelerate Linux or FreeBSD graphics, or how much eye candy you add, it won't matter if it's "a better OS" or not if it doesn't have the apps.
Apple isn't required to open-source Darwin, though there are components they are required to continue releasing. That doesn't mean this wouldn't be news, something to worry about, etc, etc, etc...
Mac computers are sold with a license to use Mac OS X. The retail box you can buy is an upgrade, which presupposes you already have a Mac (and thus a Mac OS license).
Interesting point, one I have made (albeit with one extra word... 'the retail box is effectively an upgrade') many times. The problem is that there's been people running retail copies of Mac OS X on machines never licensed for it for way too long for it to be called "piracy".
It may be against the EULA, it may be illegal, it may be many things, but "piracy" is too strong a term.
But even more interesting:
there is a way to avoid it being quote-piracy-unquote while still breaking the EULA: buy and wipe clean an old Mac.
Everyone with a legal copy of OS X Intel they can install on a non-Apple box got it with a Mac they can install Linux on to stay kinda kosher.
Sooner or later Steve will swollow his pride and create a subsystem consisting of a modernized POSIX [...]
...that already happened..
He doesn't have to.
While I don't agree with the people who beable on about how "used Macs enable even the poor to productively use OS X, just look at eBay", it's worth noting that you CAN get Macs capable of running OS/X up to 10.2.8 without any tweaking for as little as $40... and get quite a good taste of the system from that.
Before we get into the digression I'm glad to hear that my prediction has not in fact come true, and a technical glitch has been blown way out of proportion (which is, now I think of it, what happened with the "Apple's ripping off KHTML" flamewar).
---
Onwards!
You can't actually buy a copy of OSX for x86, the only versions on the shelves are PPC.
That's only one of many reasons I haven't personally so much as downloaded any of the bits necessary to install any version of Mac OS X on Intel hardware. Don't make this about me, please, because it's not.
However...
and unless you wipe it from your new Mac
There's quite a few people working on and even successfully running Linux on the new intel-based iMac, as well as people working on running Windows on it.
Personally I think they're all nuts, but there are people who can install Mac OS X for Intel on a non-Mac without being "pirates".
(and the vikings sing, "Mach Mach Mach Mach...")
It's booting through the BSD boot process, and the BSD kernel is running single-server, and they have increasingly moved away from using Mach messages because of the overhead. It's got more Mach in it than FreeBSD does, but FreeBSD was already using a lot of Mach code, and they've continued to re-import a lot of updated FreeBSD code (eg, FFS in Panther).
Like everything else, it's a mongrel. It's part Mach, part BSD, part NeXT, part Mac, something old, something new, something borrowed, something blue...
Exactly why I resisted the pressure to abandon Linux for MacOS X on the desktop
I'm not sure what your point is. Let's say the absolute worst thing happens and the next version of OS X is based on an NT kernel, all the UNIX-compatible stuff is supplied by Interix, and Bill Gates buys Apple.
All the commercial software I've got on my Mac will continue to run.
All the open source software I've got on my Mac will continue to run.
All the software I'd have been using under any other free UNIX will still be just as available as it is right now, and I can continue to use it on Linux or Mac OS.
If for some reason I want to run Linux on my Mac Mini, I will still be able to run all my Mac OS software under MOL.
If I want to run Linux on an Intel box I can, and all the software I would have been using on Linux and all the open source software I'm using on Mac OS X will still be available.
If for some reason I want to use my Mac software on an Intel box running Linux, I will be able to do so using Sheepshaver, under emulation, just as I woudl be able to use it using Rosetta, under emulation, on an Intel Mac.
What would have been the advantage of using Linux for the past three years instead of Mac OS X, even under the brutal worst-case regime I described above? I really don't get it.