multi-monitors?
The last time I tried Ubuntu Karmic it was still rubbish at doing heterogeneous multi monitor display - i.e. laptop (Sony Vaio) plus external monitor. It's the only thing stopping me from going fully Microsoft free...
Around 75 per cent of Linux developers raked in cash from their code crunching in the past year. It's a figure that in many ways comes as little surprise, given that Linux usage has become so widespread across industries, government and the public sector in recent years. Linux kernel contributor Jonathan Corbet told an audience …
What Apple products come even within a mile of Linux?
BSD, sure. But Linux?
BSD == Berkeley.
Mach == CMU
Big difference.
Mach was more focused on a microkernel.
Unless of course you're going to say BSD because Mach was initially based on BSD vs ATT?
C'mon, get real.
Mach -> NeXTStep -> Mac OSX.
If you want to get down to it, all of the flavors of Unix/Linux are either BSD based or AT&T based.
Some are a bastardization of both.
And I forget the name of the OS out of MIT which was the basis for both flavors of Unix. (Multics??) Go ask Stallman, he'd know.
But to your point, Yes its not fair to criticize Apple for not contributing to Linux.
Google? If they modded an OS, it would be to their advantage to keep mum about it.
And just before someone beats me to it: Apple sucks!
Mach is used by Mac OS at the lowest level of the kernel.
Higher up, Mac OS is essentially xnu (or Darwin) and that is derived (and contains a substantial number of lines of code) from BSD.
Multics originated at Bell Labs. As far as an OS originating from MIT, you might mean Exokernel.
You are skeptical that Google have modded an OS? ChromeOS is based on Linux as is Android. Have you heard of those two?
You want to appear knowledgeable, so try getting some knowledge before sounding off. If you do, you'll come across as less of an arsehole.
One of the most important subsystems of OSX is a complete BSD Unix userland
.
Even the kernel is a Mach/FreeBSD hybrid, in large part because pure Mach could never in a million years service the audio subsystem in a timely manner.
The last time I tried Ubuntu Karmic it was still rubbish at doing heterogeneous multi monitor display - i.e. laptop (Sony Vaio) plus external monitor. It's the only thing stopping me from going fully Microsoft free...
Yes, it isn't very good with device drivers and struggles with all my peripherals and attached equipment, but I think that's part of the Ubuntu culture. There is an expectation that drivers will be written by the peripheral and hardware vendors.
Wide enough hardware coverage. :-) In my experience Thinkpads and Acer lapdogs can do this pain-free, at least if you don't panic and try to configure two monitors or perform other hacks. Just use the function hotkeys provided and they work as expected, both with Karmic and Jaunty. I can't say anything about Vaio's, never had one to play with. Care to make a donation...? ;-)
Hopefully hardware vendors will discover that their own interests are best served by paying for Linux hackers to free them from MS, but as they operate on very narrow margins as it is it may take a long time.
Try OpenSUSE. Plugging my laptop (OpenSUSE 11.2 ) into my TV gave a combined desktop of ( a rather unequal) 1920x1080 plus 1280x800 with a few clicks.
I've found it to be much better with the .32 kernel, grab it from http://kernel.ubuntu.com/~kernel-ppa/mainline/
75% of fixes being contributed being from paid coders does not automatically equal 75% of coders being paid unless all coders contribute an equal amount.
It seems to me far more likely that the paid coders can contribute all day every day and so contribute more code.
I can get paid to write closed code for my company or I can get paid to write open source code for my company that gets reviewed and debugged by many many peers.
I get to to write FLOSS code and my company gets code that's 10 times better than I could write on my own, and 20 times better and almost infinitely cheaper than just arranging a contract with an outsourcing company.
So the 'army of coders' myth seems to have fallen flat on it's face. Most people don't want to work for free in their precious spare time on an OS. Who would have thunk it?
> Most people don't want to work for free in their precious spare time on an OS. Who would have thunk it?
How many of these started off doing this in their spare time and have since got the opportunity to do it for a living? Doing something they want and getting paid for it is a big bonus.
It's more of a mixed picture than that. Money speeds up the process but a lot of development would still happen without the dough, just with longer deadlines. Besides, we don't want "most people" to code on something as important as an OS kernel, do we? :-)
At what time did 'army of coders' ever equate 'army of free coders'?
According to the article -- and most of the other evidence around you -- the 'army of coders' model seems to be working splendidly. In fact, it's gone mainstream, so now that 'army' is being paid to do its coding. No wonder that FLOSS is scaring the crap out of Microsoft!
What version of Linux does Apple use and what products is it in? There's loads of OSS stuff they use like GCC, BSD, WebKit (which used to be Konqueror/khtml) and others, but I didn't think Apple had much to do with Linux.
> What version of Linux does Apple use and what products is it in? There's loads of OSS stuff they use like GCC.
GCC is not part of Linux, Linux is the kernal. FSF/GNU/OSS != Linux
> > What version of Linux does Apple use and what products is it in? There's loads of OSS stuff they use like GCC.
>
> GCC is not part of Linux, Linux is the kernal. FSF/GNU/OSS != Linux
Which is the same distinction that John 62 made, as I read it.
...someone kills a kitten. Then buys Windows.
Apple, re NeXT modified the GCC compiler to grok objective-c. (Think smallTalk wrappers around C.)
Of course they put the code changes out and anyone who has a gcc compiler, has an objective-c compiler. But they don't have the class libraries which are the secret sauce. ;-)
So Apple, actually NeXT gave back because they had to.
Sorry, I can't see into the future to that someone is making a comment along the same lines as mine. True the other post was submitted first but there is a time lag between submission and approval/appearing here.
Oops, hope no one else as pointed that out.
'ere, now, you can hardly be claiming that the post you quoted from was not visible when you posted. It's obvious that you misunderstood the post; accept it, and move on.
"code monkeys had a very busy 2009, working on projects for distributions such as Red Hat, Ubuntu, Debian and co."
Ubuntu? Heh. Pull the other one. This story's about the kernel, and Ubuntu notoriously contributes diddly squat to kernel development. Your source story gives a rather more accurate sampling: "Within that field, Red Hat topped that chart with 12%, followed by Intel with 8%, IBM and Novell with 6% each, and Oracle 3%". Ubuntu is well under 1%.
I thought Mac OSX was BSD based. Google is thought to be contributing open source code to the community in many other ways. However, both Apple and Google are probably losing out if they are using any significant Linux modifications internally which they are not contributing upstream. This is because they would be forgoing the benefit of community peer review of code and help with testing, and they are also having repeatedly to patch in their own changes whenever they no longer want to maintain an increasingly divergent base kernel target for their own development, in order to benefit from changes elsewhere appearing in mainstream kernel versions.
Jonathan Corbet has written about this in detail on lwn.net. This is a significant motivation for why the contributor companies described in the article are contributing upstream. It has very little to do with altruism, and has much more to do with the technical and commercial benefits being greater than the costs.
For one of JC's latest analysis, see
http://lwn.net/Articles/363456/
Someone working full-time have more available time to do stuff.
Also the chances are pretty good that if it's your full-time job, you're better at it than most enthusiastic amateurs. This might be different if there was no money in the job and the work required significant natural ability - think music or dancing - for example. But in a job like software where ability largely comes through experience and there's a decent payscale for all experience levels, if you're good at it then chances are you'll already be doing it for money.
And then there's management. Geeks naturally like to sit around gold-plating their software, bcos to them that's the important part. But managers just want something that works, so that the rest of their guys can get on with their jobs more effectively. With management in place, full-time employees are more likely to be doing useful stuff and less likely to be gold-plating some unimportant element. And team-leaders/architects are also on hand to help ensure that the code is well-designed, not the tangled mess that a lot of hobbyists end up creating.
Anyway, the intention of OSS was never for it to be solely provided by hobbyists. RMS, ESR and others have always been clear that one of the key points of OSS is that multiple organisations can pool their coding resources and come up with a viable alternative to buying a licensed copy of commercial software. And after that, other organisations can simply reuse the existing software for free, or they can contribute resources to develop it further on the principle of "standing on the shoulders of giants". This doesn't always hold true - for software which has limited potential for reuse, it may well make more sense to buy it - but that's the theory, and it certainly works for easily-reusable software like word processors, web browsers and the like.
Graham Bartlett wrote:
"And then there's management. Geeks naturally like to sit around gold-plating their software, bcos to them that's the important part. But managers just want something that works, so that the rest of their guys can get on with their jobs more effectively. With management in place, full-time employees are more likely to be doing useful stuff and less likely to be gold-plating some unimportant element. And team-leaders/architects are also on hand to help ensure that the code is well-designed, not the tangled mess that a lot of hobbyists end up creating."
Half a lifetime of real-world experience shows that to be a flight of fancy.
Most managers just want something out of the door when it was promised. When push comes to shove, most don't care that much about whether software works well, is well designed etc.
Apple uses Linux (kernel)?
Further to Raumkraut's and John 62's comments I tend to believe -- from the Apple employees I know -- that the vast majority of them use OS X.
Mr. Fiveash probably thinks that Mac OS X is based on FreeBSD too, and wondering where he can download that Linux distro.
Could we get this straight, once and for all.
The Mac OS X kernel is Mach. The userland is a mix of the various BSD userlands.
Linux code in the OS X kernel? Not bloody likely. Can you say GPL?
Please hand in your Geek card at the kiosk on your way out.
It's a bit less clear cut than that, there's a fair amount of BSD stuff in what you could realistically call the Kernel:
"In Mac OS X, however, the kernel environment contains much more than the Mach kernel itself. The Mac OS X kernel environment includes the Mach kernel, BSD, the I/O Kit, file systems, and networking components. These are often referred to collectively as the kernel. Each of these components is described briefly in the following sections. For further details, refer to the specific component chapters or to the reference material listed in the bibliography."
From:
http://developer.apple.com/mac/library/DOCUMENTATION/Darwin/Conceptual/KernelProgramming/Architecture/Architecture.html#//apple_ref/doc/uid/TP30000905-CH1g-CACDAEDC
Oh, and it appears Apple may have contributed at least something to linux: :)
http://www.mklinux.org/
There's no mention of Apple in the original article. and the article and statistics are referring specifically to Linux the kernel - not "Linux" the GNU ecosystem.
Original article: "Those figures also don't include companies like Google which rely on Linux-based systems for their own technology, but tend not to contribute code back to the kernel."
Register article: "The likes of Google and Apple were notably absent from Corbet's list as, while both companies use Linux-based systems to build their tech, neither have a tendency to contribute code back to the kernel."
So whilst Apple do indeed use lots of components of the Linux environment and that's a valid point the register article could be discussing it's nothing to do with the kernel which is what this report was referring to.
It's called Free Software, not free software. Libre not Gratis. It's not a surprise that devels are paid for their work and so they/we should be.
"What has been written is usually given away gratis, paid for but also decided for. What has yet to be is expensive to develop but comes with control."
I have contributed more money to the Free Software Foundation than I have ever paid for non-free software, combined. (I am a subscription member at $10 a month, and have been for years)
Its the freedom I like, not the cheapness.
Sony's are awful to try and install an OS on if you don't have a driver, and that includes Windows. They use a lot of proprietary tech on them
Lots of big and small companies have joined the linux community. Lots of people who contribute to the linux kernel work for companies that have not joined the community.
Perhaps there are unemployed people too who contribute.
I suppose there is a number of students too who contribute.
So what is the problem, or is it a problem that big and small companies find it useful to contribute.
You mean they actually get PAID for their work? Whodathunkit.
Good, innit?
There's a lot of people being paid for contributing to a wide variety of open source projects, and you'd be quite surprised at the volume of contributions from some places.
I don't think the idea of GPL was to have programmers not able to earn a living. Perhaps it's to have programmers able to not pay a tithe to Microsoft. Or, today, to the app store of your choice...
So corporations, computer builders especially (IBM), pay Linux developers? Fine. We get OS support for new hardware features. Also new interoperability standards - neither a priority for Microsoft. They missed the Internet, missed USB, pretty much killed some technologies by stt!arving them of support...
Never mind that if the world had embraced OS/2, IBM's ardour for Linux today would be very much cooler...
Why are these devs being paid for their work ? To get what the company they work for and pays for work so that company can then make a profit. I doubt there are many companies who are altruistically paying their devs to contribute simply for the good of the world.
So, apart from getting to fiddle with the innards of kernels ( which is limited to device driver writing for Windows devs ), what's the real difference between writing code for Linux and writing drivers and apps for Windows and getting paid for it ?
In the Windows and Mac world there is equal opportunity for free software producers, just different limitations on what they can hack about.
"I get to to write FLOSS code and my company gets code that's 10 times better than I could write on my own,"
That either speaks of a lack of confidence in your own ability, or you really suck and need to find a job you're competent at.
I hear Burger King are hiring,
Google contributes a lot of code to the Linux kernel. A good chunk of what's left I'd imagine. I have friends that work at google that do pretty much nothing but and they're all told to submit there code anonymously or under pseudonyms as google doesn't want people coming to them for tweaks and help on it, they just do it as they need it and release it and don't plan on maintaining it or have deadlines.
Testers and debuggers generate very very few lines of code, but they're not an insignificant part of the dev process.
Can we see a follow-up that examines paid-for testing. If Red Hat et al aren't putting doing their own thorough testing, they're setting themselves up for liability....