Personally, I believe this line to be pure bull - not only in the Mac community, but also in the Linux community.
I know many people who started using Linux, and couldn't write, without help, a single, simple, syntax-error-free line of C code.
Now that may sound very harsh against Linux users - and it is very well meant to be. All in all, the technical competence of most Linux users IS above average. But that doesn't mean that they all have the competence of commercial quality programmers, or even mediocre programmers. I doubt the majority of them could fix a simple buffer overflow bug given adequate debugging information that points the bug out directly.
The claim that there are "40,000" contributors to "Linux" is very misleading. First off, Linux is more than the kernel. It is all the programs that it takes advantage of or requires to get the job done. Linux would be NOWHERE without programs like 'make', 'gcc', 'gdb', and all those similiar tools. However, those tools WOULD exist without Linux. So, that number (which is pure estimate on the part of anyone claiming it as fact anyways) most likely includes the developers of all the programs that Linux takes advantage of... and in this case, that is like saying a printing company (Which can be run by a handful of people - a manager, a couple secretarys, a couple graphic artists, and a few printing press operators) employees "hundreds of people", by counting the foresting company, the people that operate the paper-making machines, and the people who produce the ink.
IF, on the other hand, said number was restricted to the heart of Linux - for TRUE "Linux" is nothing more than the KERNEL - I wouldn't be surprised if, over the course of Linux's 8+ years of evolution, 4,000 (not 40,000) people have contributed something to the source code for some program for some platform or another - but even then Im wondering how many of those contributions were very minor ones - such as a bug fix - and how many contributions overlapped. How many people provided moderate (more than a few dozen lines of code) contributions, and how many provided SIGNIFICANT conributions (on the lines of a few hundred or thousand lines). I'd guess more along the lines of 400 and 40, respectively. And how many Linux kernel contributors are well known by any kernel hacker? I think the number falls around 4 (to 8).
In the end, the true core work of a project is done by just a few people (scaled to the size of the project, of course). While they may take advantage of existing libraries (such as a PNG or JPEG library, ZLIB compression, a communications archiecture, or some such), that does not truly increase the size of the development team - that actually gives them less of an excuse to have large amounts of people. You can't claim the authors of these libraries on your dev team (tho giving them credit for their work and it's use in the project is something else entirely different).
And finally, what if the project is commercial? The problem these days is that Free Software (Free as in Freedom of Speech/Open Source) is equated these days with Free Software (Free as in Free "Beer"/No Cost). This is a very dangerous equation that is being made, and even promoted by so-called experts such as Richard M. Stallman. Eric S. Raymond has a clearer view, but it is still distorted by the belief that everything can and should be source-available. In many cases, especially Games, this option isn't available while the product is still commercially viable. Hence why Doom and Quake were not Open Sourced until years after they were off the shelves and replaced by better products (in the cases of Doom and Quake, 2 generations of products later).
"Hundreds of Thousands of Developers" my ass (Score:5)
I know many people who started using Linux, and couldn't write, without help, a single, simple, syntax-error-free line of C code.
Now that may sound very harsh against Linux users - and it is very well meant to be. All in all, the technical competence of most Linux users IS above average. But that doesn't mean that they all have the competence of commercial quality programmers, or even mediocre programmers. I doubt the majority of them could fix a simple buffer overflow bug given adequate debugging information that points the bug out directly.
The claim that there are "40,000" contributors to "Linux" is very misleading. First off, Linux is more than the kernel. It is all the programs that it takes advantage of or requires to get the job done. Linux would be NOWHERE without programs like 'make', 'gcc', 'gdb', and all those similiar tools. However, those tools WOULD exist without Linux. So, that number (which is pure estimate on the part of anyone claiming it as fact anyways) most likely includes the developers of all the programs that Linux takes advantage of... and in this case, that is like saying a printing company (Which can be run by a handful of people - a manager, a couple secretarys, a couple graphic artists, and a few printing press operators) employees "hundreds of people", by counting the foresting company, the people that operate the paper-making machines, and the people who produce the ink.
IF, on the other hand, said number was restricted to the heart of Linux - for TRUE "Linux" is nothing more than the KERNEL - I wouldn't be surprised if, over the course of Linux's 8+ years of evolution, 4,000 (not 40,000) people have contributed something to the source code for some program for some platform or another - but even then Im wondering how many of those contributions were very minor ones - such as a bug fix - and how many contributions overlapped. How many people provided moderate (more than a few dozen lines of code) contributions, and how many provided SIGNIFICANT conributions (on the lines of a few hundred or thousand lines). I'd guess more along the lines of 400 and 40, respectively. And how many Linux kernel contributors are well known by any kernel hacker? I think the number falls around 4 (to 8).
In the end, the true core work of a project is done by just a few people (scaled to the size of the project, of course). While they may take advantage of existing libraries (such as a PNG or JPEG library, ZLIB compression, a communications archiecture, or some such), that does not truly increase the size of the development team - that actually gives them less of an excuse to have large amounts of people. You can't claim the authors of these libraries on your dev team (tho giving them credit for their work and it's use in the project is something else entirely different).
And finally, what if the project is commercial? The problem these days is that Free Software (Free as in Freedom of Speech/Open Source) is equated these days with Free Software (Free as in Free "Beer"/No Cost). This is a very dangerous equation that is being made, and even promoted by so-called experts such as Richard M. Stallman. Eric S. Raymond has a clearer view, but it is still distorted by the belief that everything can and should be source-available. In many cases, especially Games, this option isn't available while the product is still commercially viable. Hence why Doom and Quake were not Open Sourced until years after they were off the shelves and replaced by better products (in the cases of Doom and Quake, 2 generations of products later).
- Chris Jacobson
(MaineCoon)