Linux and Mac OS X

William J writes "Here is an article with an interesting slant on the relationship between the Mac OS and Linux. The author suggests that Gnome and KDE developers can learn from the Mac GUI. Worth quoting: 'It is amazing to me that an OS which was developed largely by volunteers (and which is essentially free) can run with unprecedented stability on the same hodgepodge of PC hardware on which another company has spent billions of dollars in R&D costs and is still unable to produce a product which can run for more than a few days without crashing -- and it costs hundreds of dollars.'"

Comments

Proposal

[__past__]

Could our beloved Editors perhaps consider not to post links to stories without any content?

What the heck is the point of this article? OS X is more polished than KDE/Gnome, Windows is not so stable, you can use Linux as a server for Macs... News at eleven!

Re:Proposal

[sydb]

This story is a melange of Linux, Mac, Windows and *BSD and hence is likely to fan the religious flames, generate page hits, advertising views and revenue.

That's why it's here.

KDE and Linux

[TheReverand]

Why oh why does everyone think KDE has something to do with Linux? Sure, it runs on Linux, but it works just fine on loads of other platforms as well (I've been running it on FreeBSD for a while now with no problems). I wouldn't be surprised if there was an OS X port of KDE sometime in the future as well...

That's the whole thing about Linux, KDE and Gnome... You're not limited to one combination of hardware and software. All these articles which put KDE and Linux together are missing that point.

Re:KDE and Linux

[DrSkwid]

anyway KDE is a MS Windows App

Re:KDE and Linux

[Arandir]

We just need to get the Gnome and KDE developers to realize this as well. Most of Gnome and KDE run just fine under my FreeBSD boxen, but there's always an occasional linuxism so that something doesn't work or works oddly. There's this big huge standard out there called "POSIX", and another one called "X11R6". It's a shame there is still the occasional developer who doesn't know about them.

Re:KDE and Linux

[Arandir]

What the hell are you talking about? One of the goals of FreeBSD-5.0 is "to implement all requirements of the ISO 9899:1999 (C99) and IEEE 1003.1-2001 (POSIX) standards." I don't have time to do a POSIX audit right now, but my -4.5 system has a man page for POSIX.1e, every random man 3 page I've tried yet states compliance for C89, C99, POSIX.1 or SUSv2 under the STANDARDS section. Having done a bit of pthread programming, I know from first hand experience that FreeBSD us *much* more compliant with the standards than Linux.

Is FreeBSD fully compliant? No. No free unix implementation is. But FreeBSD (and I assume the other *BSDs as well) is a lot more compliant than some commercial unices I've used.

Re:KDE and Linux

[stevew]

This set me to wondering - and I'm not trying to lite off a BSD vs Linux debate with this, but another thread entirely. Of unices & unix like systems... Linux is probably the most wide spread at this point in time...(maybe it's solaris, but I'm betting that the total number of installs of linux is greater than solaris now...) So let's assume that assumption is correct. When does Linux become the standard as oppossed to Posix? As a practical matter it might be argued that Linux has already achieved that status when you consider that both BSD and Solaris have Linux compatibility libraries!

Thoughts?

Re:KDE and Linux

[Arandir]

When does Linux become the standard as opposed to Posix?

There are two kinds of standards: informal de facto standards and formalized official standards.

Linux certainly is not the formal official standard for anything, not even for itself, since you are not allowed to define a thing in terms of itself. POSIX is the formal standard because it went through a formal standardization process. There's an actual document reviewed and approved by experts after much discussion that says what POSIX is.

Is Linux an informal de facto standard? Maybe, maybe not. It depends on what your definition is. It may be a standard for a kernel, but your system is much more than a kernel, it's an amalgamation of software from several different projects. So you end up with software that says it needs this version of a kernel and that version of a libc and you're still not sure you won't have to crowbar it into place unless you're running the exact same distro version that the packager used.

The aim of POSIX is to get beyond all of this. If you have a POSIX compliant system, and the software claims POSIX compliance as well, you are virtually guaranteed that the software will work. That's great! (POSIX is actually a set of standards, so my referal to it as a single standard is merely semantic shorthand)

But I want to respond to your unwritten question. If my powers of telepathy still operate sufficiently, I can tell that your real question is "when will Linux finally become a standard?" Okay, may telepathy is a bit rusty, it might not be you thinking this, but I'm definitely picking it up from someone. The answer is, "Linux should never become a standard." The reason is simple: standards and implementations of the standards are two very different things. Eventually Linux may implement POSIX so well that it becomes a "reference" implementation, but it will never be the standard itself.

Think about the web for a bit, and you'll understand. Because of a variety of bizarre circumstances, including certain recessive genes in most web developers, implementations have gained the status of standards. And the result is chaos. Web sites aren't written to standards, they're written to specific implementations of the standards. Unless you're using one particular browser released by a company in Washington, you will *never* be able to access every web site claiming standards conformance. The situation is even worse in word processing land where MSWord is the standard.

I don't want to see those situations occur in Unix. So please write your software to the POSIX standards instead of to the Linux implementation.

Can't connect to Windows?

[sg3000]

> I was unable to get a Windows 2000 PC to see any
> Macs (OSX or 9) on the network or vice-versa.

I don't want to be tough, but I've had no problems getting my Mac to load Windows servers via SAMBA. I do it every day, and I'd hardly say I'm an expert at SAMBA. So I'd say that rather than this being an OS X or Windows limitation, this guy may just "lack the skills to pay the bills". What was the point of the article again?

People accomplish things

[Drone-X]

Worth quoting: 'It is amazing to me that an OS which was developed largely by volunteers (and which is essentially free) can run with unprecedented stability on the same hodgepodge of PC hardware on which another company has spent billions of dollars in R&D costs and is still unable to produce a product which can run for more than a few days without crashing -- and it costs hundreds of dollars.'"

For software development you only need some brains, time and a computer. Because of this and also thanks to a certain global communication tool, it's hardly suprising people accomplish grand things without coorporations backing them.

Really, people don't need management to accomplish something. Given the resources (money and time) people can and will do productive things for society.

Re:People accomplish things

[__past__]

Aaah, it's nice to once again hear this, after all this focusing on how to make money with open source!

Before the Linux hype, it was quite common knowledge (or, at least, opinion) that Free Software is not only great because of its unbeatable price or even the "philosophic" implications (that come down to "be kind to your customers" when restricted to licenses), but because of its quality, which in turn is directly related to the freedom of programmers - at least as important as the freedoms of software users.

Not only do people not need management, they can build better things without worrying about deadlines (leading to "good enough" solutions), corporate politics, marketability etc. Free programmers can focus on doing the Right Thing, which is often not possible in a corporate environment. The results of this are where the pride of the free software movement should come from, not the sympathy of venture capitalists or IBM or Apple.

Re:People accomplish things

[MaxQuordlepleen]

Agreed, and to illustrate your point, look at what happened when Linus was under heavy pressure to release 2.4. We got an unstable mess that clearly would have benefited from a few more months of development...

Re:People accomplish things

[sydb]

Linus released 2.4 because he wanted more testers. He got them. 16 or so releases later and we have a stable kernel.

Porting Aqua

[darkov]

Why isn't someone undertaking a port of Aqua to Linux? It is build on top of a Unix system, after all. It wouldn't have to be a perfect copy, maybe just a set of APIs mostly compatible with Carbon APIs.

The advantages of this are :
(a) you have sound user interface design for free
(b) you have an instant installed base familiar with the user interface
(c) you have many applications which can be ported possibly with a minimum of effort.

Something like this would definately put it up Apple. But it does make sense. If people can rush off and build .Net clones, why not do someting actually useful?

Re:Porting Aqua

[base3]

Hit it on the head. Every time I think of going over to the Mac to avoid giving money to the Evil Empire, I remember things like

o Apple licensed the nefarious Amazon one-click patent, giving Amazon a precedent with which to bludgeon smaller companies.

o Apple crippled their DVD writing software to disallow mastering for replication.

o Apple used legal threats on non-for-profit skinners.

o Apple screwed over the clone vendors.

And that's just off the top of my head. I'm not thrilled about buying Microsoft, but I wouldn't feel particularly good about supporting another company with a monopoly (can you get Mac OSX for a clone? No. Then they have a monopoly) which leverages their software to sell overpriced hardware. It's obvious that if Apple were sufficiently competent, they would be another Microsoft. But while Apple is ruthless, predatory, and sells out its users just like MS, they're not as good at it.

Re:Porting Aqua

[Drone-X]

(a) you have sound user interface design for free

GNOME and KDE will get there, it will just takes a bit more time. Looking at the GNOME usability project with a.o. the work Sun is putting into it, I'm confident that by version 3.0 GNOME will be a killer.

MacOS X OTOH has had a lot of critism.

(b) you have an instant installed base familiar with the user interface

Not all that much people are familiar with Aqua.

The number of people familiar with GNOME and/or KDE is probably larger? I admit I don't have any numbers to back this up except for the fact that there are more machines out there running GNU/Linux rather than Macintosh, add to this that MacOS X was released not so long ago and I may just be right.

Either how, the number is probably not going to be worth the bother.

(c) you have many applications which can be ported possibly with a minimum of effort.

GnuStep should allow for this without actually porting Aqua. The advantage of this strategy is that you get to keep X :). (For those that for whatever reason believe X is bad/bloated/whatever, think about the drivers.)

But it does make sense. If people can rush off and build .Net clones, why not do someting actually useful?

Gimme a break. .NET seems like a sound development platform and it's almost garantueed to be a huge success (it's Java with lots of extras plus MS is backing it).

Re:Porting Aqua

[pressman]

Just because you can't get OS X on anything other than a Macintosh does not mean that Apple has a monopoly in any meaningful sense. They own 5% of the personal computer market. To be a monopoly a compnay must own at least 51% of their respective market. M$ has a monopoly in operating systems, office suites and web browsers. Apple makes their own hardware and OS and competes with M$, SGI, Linux and Sun. They don't have a monopoly in anything.

Re:Porting Aqua

[alex_ant]

That's a really funny definition of "monopoly." By your logic, Sun has a monopoly on Solaris, SGI has a monopoly on IRIX, HP has a monopoly on HPUX, and IBM has a monopoly on AIX. Monopolies = bad, so Sun, SGI, HP, and IBM are all evil, and will be until they port their big iron OSes to your peecee. "I want 4Dwm! Open-source it, SGI! Give it to me free, or else yer nothin' but a dirty monopolist!"

Aqua is a work of art. Believe it or not /., some people in the world actually believe in intellectual property.

Apple is not predatory. It's too small to be predatory. Its attack of the clones happened only after a radical shift in management. I think the term there would be "non-suicidal," not "predatory."

How does Apple sell out its users? I've had a mac.com email address for the longest time (Mac owners get them for free - how evil of Apple to offer such nasty tie-ins!), even though I've rarely used it, and I've not gotten a SINGLE piece of spam to it. Ever.

As has been covered so many times here before, more expensive hardware != overpriced hardware. You get what you pay for. This is a myth that really needs to get shot down - I don't see why so many obviously smart geeks have such a terrible time understanding this. Some people in the world are actually not content with cheap-ass high-MHz beige commodity boxes built by soulless vendors like Dell, Gateway, etc. who just don't give a shit about their product and who WOULD sell their customers out to gain any edge they could in the cutthroat Wintel market.

I'm not an Apple apologist, but I am a Mac/Linux user and I will go to certain lengths to defend the company against the heaps of obvious bullshit piled upon it. I agree that a large and powerful Apple would not be a pretty sight. I would be most content with Apple at around 10-15% market share.

Re:Porting Aqua

[pressman]

What the /. Linux crowd fails to see is that Apple customers don't mind paying a little extra money for the time and care put into developing Macintosh systems. Most Mac users don't want to get into the guts of their computer except to maybe install an extra harddrive or some extra RAM. We generally don't care about getting under the hood because the people who made the computer engineered it so we don't have to do that if we don't want to.

For Linux users I can understand why getting to the guts of the computer and the OS is so important. It's part of the computing experience for that market. Linux users LIKE getting to the very core of their computers. I don't understand why they have to bash Apple and it's users just because Apple doesn't consider them part of their target market demographic. Why would Apple market to people who don't want to spend money on anything? They are a company whose goal is to make money and they can't make money off a free OS and low margin computer components.

I don't go around bashing Linux because it doesn't meet all of my computing needs. It's a good OS for what it's intended to do, but it doesn;t come close to meeting my needs or the needs of millions and millions of other computer users... users being the operative word.

Macs just work out of the box. Ceratin people want that.

Linux only works if you configure it to work the way you want it to and have the technical knowledge to do that. Certain people want and need that from their OS and computer.
Windows has lots of games and is ubiquitous. Very few people really want Windows but it is the only option they know. It seems to meet their needs reasonably, but then again their standards and expectations of a computer might be a bit lower than Mac or *NIX users.

So, if the macintosh doesn't fit your criteria for a computing environment, DON'T BUY ONE OR USE MAC OS! Stop complaining about the price of their hardware and buy the system you need. You're just wasting energy and the time of other people.

Re:Porting Aqua

[Pathwalker]

You said:
Apple crippled their DVD writing software to disallow mastering for replication.

I reply:

I fail to see any problematic "crippling" of DVD production in Apple's current hardware.

If you were referring to the fact that iDVD will not export a master image (to a DLT tape for example) - Apple has to differentiate DVD Studio Pro from their free software.

If you were referring to the fact that the SuperDrive can only produce DVD-R 5-general disks, and not authoring disks, I also fail to see that as a problem, as all of the DVD manufacturing services I have checked will accept a general disk as a media source.

It is true that disks produced in this way can not be protected with CSS and macrovision, (an author would need to make a master on a DLT, or an authoring DVD-R rather than on a superdrive to add these protections ) but I feel that many on Slashdot would see this as an advantage in that it increases the amount of unprotected, legally viewable under Linux (or [Free,Open,Net]BSD ) available in the world ;-)

Re:Porting Aqua

[pressman]

BTW, I second most of what you said. For example, there are so many turnkey PCs out there, that mentioning it as a Mac selling point is laughable. But you'll see a lot of Mac sales points that ceased being valid 10 years ago. The people still using them are really that out of touch.

Ok, please provide a list as to these out of date selling points and reasons why Mac users are so out of touch.

Re:Porting Aqua

[pressman]

Ok, so you're not going to buy a Mac. That's fine. Based on your criteria, a Mac does not fit your computing needs.

What bothered me about your original post was the tired, old "Mac users are out of touch" comment.

I have 4 Macs. They work exceedingly well for me. They're getting a little old, but they're still good for the web design that I do. Video editing, well, not so great by current standards.

I do graphic design and video editing for a living. I need Photoshop and Final Cut Pro running at peak efficiency. Photoshop has always been more stable and in most (though not all) cases faster than in Windows. There is no Final Cut Pro for Windows, so there goes any compelling reason for me tu purchase a Wintel machine.

Do you see the point here? Purchase the system that best matches your professional and personal needs. You apparently REALLY don't want a Mac. Fine. Don't get one. I have no need for a Wintel box except to mess around with Linux and teach myself some new things.

Re:Porting Aqua

[pressman]

I realize I'm being totally trolled here, but it's fun.

To quote the original post:

BTW, I second most of what you said. For example, there are so many turnkey PCs out there, that mentioning it as a Mac selling point is laughable. But you'll see a lot of Mac sales points that ceased being valid 10 years ago. The people still using them are really that out of touch.

Mac users are, I'm assuming, the people who use these no longer valid sales points. Is that not what your sentence implied? You don't hear many Wintel people defending the Mac with ANYTHING let alone ten year old sales rhetoric.

Also, 10 years ago there were no PowerPC Macs, 486's were the top of the line consumer offering from Intel and Windows was still in it's infancy. The Mac had a much larger marketshare back then. They didn't really need to make any dubious claims as to the performance benefits of the Mac over Wintel.

You have proven nothing other than you like to be argumentative. It's fine with me. I actually enjoy playing semantics.

Re:Porting Aqua

[Drone-X]

You're right, the "OTOH" was misplaced. What I wanted to say was that MacOS X lost sacrifised some usability for eye-candy. If GNOME with its usability project avoids that, then it can end up more user-friendly long-term.

But of course MacOS X may change its appearant philosophy in the future.

Re:Porting Aqua

[pressman]

Aw, man. It is so tempting to just keep this going, but alas I have work to do on this computer.

It would be much more fun if you weren't an AC also.

Has WIlliam J even used OSX?

[LeapingGnomeArs]

"unable to produce a product which can run for more than a few days without crashing"

Has the submitter even used OSX? I've been using it daily since October and it has only crashed on me once. The majority of OSX users do not rebooted their Macs, they just put them to sleep. Remeber, Macs have instant wake-up from sleep, unlike Windows or Linux.

OSX uptime is typcially measured in weeks, not days.

The submitter made a misleading quote...

[Daniel Wood]

Worth quoting: 'It is amazing to me that an OS which was developed largely by volunteers (and which is essentially free) can run with unprecedented stability on the same hodgepodge of PC hardware on which another company has spent billions of dollars in R&D costs and is still unable to produce a product which can run for more than a few days without crashing -- and it costs hundreds of dollars.'"

This was in fact a reference to Windows, not Mac OSX.

On the subject of Windows stability. If you're not using that crap VIA puts out, but instead use tested Intel solutions, it is not an issue. Since my migration to Windows 2000, I have had a total of eight memory dumps. That is since my initial use of Windows 2000, RC2. As a desktop OS in the Intel world, nothing comes close(available software versus stability). Eight memory dumps over twelve systems in a period of Three years seems like a good track record to me.

I will probably be modded to hell for posting anything positive about Windows, but these are the facts.

Re:The submitter made a misleading quote...

[pressman]

I don't know what Macs you've used or who set it up, but my 4 Macs all have uptimes that range into months rather than weeks. Generally if there is a crash it's because of a poorly written application rather than the OS. Lirterally, I have 4 Macs that have uptimes in the 4-6 month range. The laptop gets shut down a lot, but then again, that's a battery and portability issue. Though I'm sure someone could find a way to make that Apple's fault also. "Can you believe it? A one button mouse AND I have to shut it down when I'm carrying it around? What a load!"

Personally I think it's MS that has fooled everyone all of the time. They have actually convinced the general public that they are creating innovative products when in fact they are either stolen or copied from from smaller competitors who are soon out of business once their "air supply has been cut off".

Re:The submitter made a misleading quote...

[pressman]

Sheesh. I should have known that! Regardless, my macs almost never crash and when applications do, generally, I'm right back at the Finder.

What is a monopoly?

[Dictator For Life]

This is nonsensical. The fact that no one else makes an OS for the Mac in no way makes Apple a monopolist, unless of course you think that Ford is a monopolist because no one else makes engines for Fords.

In a commodity market, differentiation is a major way to lure customers. "Computers" is the commodity here, and the Mac is just one part of the market for computers. Apple protects those things by which they attempt to differentiate their products. Love it or hate it, that does not constitute monopolist action.

Re:What IS the fucking deal with Apple?

[Melantha_Bacchae]

An AC wrote:

> you're claiming that >>OSX has "a large application base"? ROFL!!!
> What are you smoking?

Okay, you asked for it ;):

1) Most Mac applications ever written. I've found very few that don't run well under the Classic environment. This includes popular commercial packages and tons of shareware and freeware.

2) New and ported Carbon and Cocoa Mac OS X applications. This is increasing in number daily, especially since the development tools are free. Again, this includes popular commercial packages and tons of shareware and freeware.

3) Most Java J2SE applications, and J2EE applications if you obtain the necessary libraries. The first Airport utilities that were in use on OS X were Java apps originally created for Windows. OS X has the best Java 2 on the desktop.

4) Various Unix applications, many Open Source or GNU, ported for OS X/Darwin. Some of these beloved programs come with OS X, such as the Apache web server and Emacs. Others (like an X server or postgresSQL) are available for the download.

5) If you care to plunk down around $200 (ranges from about $99 for DOS to $249 for Windows NT, per operating system) for Virtual PC 5.0, you can run most any Windows, DOS, or Linux application.

6) Go to "http://www.versiontracker.com/macosx/index.shtml" and search for "emulator". I found about 29 entries for different computers and video game consoles.

In short, OS X can pretty much run whatever you want it to run. I've been using it for nearly a year now. It is great, and getting better all the time.

OS X: the Apple of Mothra's Aqua eye.

Re:What IS the fucking deal with Apple? (readme)

[DavidRavenMoon]

Bzzzzt!!! Wrong. OSX is a product that includes a desktop GUI, a baffling array of APIs and subsystems, and no UNIX whatsoever.

BZZZZZZ!! Wrong! From Getting Started With Darwin

Q: What is Darwin, and how does it relate to Mac OS X?

A: Darwin is an open source, UNIX-based operating system built on BSD 4.4 and Mach 3.0 which forms the core of Mac OS X. Darwin is primarily what is called the "core operating system" (i.e, the kernel, drivers, and command-line utilities common to UNIX distributions), but a Darwin release includes several other pieces, including the compiler toolchain, a security framework based on CDSA, and parts of the Mac OS X "Core Foundation" framework. When we say a "Darwin system", we usually mean one built only using Open Source code, though technically every "Mac OS X" system is also a Darwin system, since it is built on top of Darwin.

Q: How hard is it to port BSD or Linux applications to OS X?

A: Given Mac OS X's strong BSD roots, this is actually very easy. Thousands of existing BSD and Linux applications (as well as Solaris, SCO, etc...) have already been ported to Mac OS X. Our dedicated Darwin developers are constantly striving to simplify portability, since they use these applications themselves and frequently are the first to encounter any problems. With their help, portability will surely get easier over time.

If you are interested in porting BSD or Linux applications to Mac OS X yourself, here are some common gotchas:

• The latest GNU configure supports Darwin, so check to see if your package is using an up-to-date version (currently version 1.2). Usually, it's just a matter of typing "./configure ppc"

• On Mac OS X, "GCC" is called "CC," and some common libraries and headers (e.g. "-lm", "stdio.h") are implicitly included in the System.framework, which can confuse hard-coded Makefiles. You can always create a symbolic link from GCC to CC (i.e. "ln -s /usr/bin/cc /usr/local/bin/gcc")

• Our dynamic library mechanism (dylib) and executable format (Mach-O) differs considerably from other UNIX implementations, so applications that require detailed knowledge of runtime and user loadable modules may need to be modified.

• We currently offer limited support for POSIX threads, so some thread-intensive applications may encounter problems. We are working to address this over time.