OSNews

Wow, that's most funny comment I ever seen in OSNews. X-D

Nobody's stealing anybody's ideas in free software world. BSD is definitely _NOT_ some giant proprietary company like Microsoft. They both are _FREE_ in any common sense I can think of. (Forget about those pseudoreligiopolitical debates about GPLs and BSDL and such and such and such)

Also, Linux is not made by Linus: LINUX IS MADE BY ANYBODY. Linus is just one of Linux's whole bunch of great developers.

BTW, IMHO you should drop your star-worshipping habit and get back to the real world. Grow humans!

There is quite simply a huge gap between the amount of resources put into Linux versus the amount of resources put into FreeBSD. It's simply not possible to overcome.

Rephrasing:
"There is quite simply a huge gap between the amount of resources put into Windows versus the amount of resources put into Linux. It's simply not possible to overcome."

And about the "higher, lower and wider", I'd really suggest reading this recent article:
http://www.informit.com/articles/article.asp?p=421896
The fact that Linux still has a *slightly* better scalability than FreeBSD is very far from automatically making it the best solution. A seriously pondered choice would take into account many other factors - and, I would say, security is among those with the highest importance.

So, your claiming the "superiority" of Linux versus FreeBSD is as poorly grounded as would be claiming the "superiority" of Windows versus Linux.
Besides, it is an unambiguous sign of zealot talk.

Rephrasing: "There is quite simply a huge gap between the amount of resources put into Windows versus the amount of resources put into Linux. It's simply not possible to overcome."

First off, I don't know that this is true, if we limit ourselves to the actual kernel. The vast majority of MS-developers have nothing to do with the core OS.

Secondly, having a large number of developers is one thing, but you also have to make it scale, and the Linux development process has undergone many changes over the years to reflect the continuous increase in contributors.

Thirdly, my statement was a reflection of the fact that the Linux kernel has been adapted to run in far my diverse circumstances than the FreeBSD kernel. Where are the people working on putting FreeBSD on big iron, on handhelds, or on massively parallel supercomputers? The simple fact of the matter is that FreeBSD isn't even a viable alternative on PPC yet.


And about the "higher, lower and wider", I'd really suggest reading this recent article:

Excuse me? How exactly is that (somewhat silly) article relevant to what I've written?


The fact that Linux still has a *slightly* better scalability than FreeBSD

Slightly? Are you just making things up as you go along? Many of the fastest computers in the world run Linux. Four of the top five on the Top 500 Supercomputer Sites list run Linux:
http://www.top500.org/lists/plists.php?Y=2005&M=06

Hell, almost three years ago SGI demonstrated Linux scaling superbly with 64 CPUs on a single system image: http://www.sgi.com/company_info/newsroom/press_releases/2003/januar... 14 months later, they demonstrated the same with 256 CPUs: http://www.sgi.com/company_info/newsroom/press_releases/2004/march/...

Do you have any benchmarks for FreeBSD?


is very far from automatically making it the best solution.

That I agree with. But all the world is not a webserver, and there are many areas requiring heavy lifting where FreeBSD simply cannot compete.


So, your claiming the "superiority" of Linux versus FreeBSD is as poorly grounded as would be claiming the "superiority" of Windows versus Linux.

Linux' 'superiority' (not a word that I would use) has been amply demonstrated to anyone who has been paying attention for the past few years.

The fact that Linux still has a *slightly* better scalability than FreeBSD is very far from automatically making it the best solution.

Err, Linux scalability utterly blows FreeBSD out of the water.

The last scalability benchmarks I have seen from the FreeBSD camp were this:
http://marc.theaimsgroup.com/?l=freebsd-smp&m=111540468626257&w=2
Showing the VFS barely scales past 2 CPUs on this 12CPU system.


And this:
http://marc.theaimsgroup.com/?l=freebsd-sparc64&m=111120677323505&w...
Showing it doesn't even scale to 4 CPUs on the same system when doing a parallel compile workload.

Now out of interest, a parallel compile is about the easiest possible task imaginable that can be scaled in the kernel. 2 or 3 years ago, we had IBM folks demonstrate a Linux kernel compile on a 32-way system complete in about 3 seconds. SGI has posted benchmarks of parallel compile jobs on its 512 CPU systems.

But nobody is really interested in parallel compiles or rudimentry VFS operations anymore in Linux. They are basically solved problems.

And lastly:
[i]Rephrasing:
"There is quite simply a huge gap between the amount of resources put into Windows versus the amount of resources put into Linux. It's simply not possible to overcome." [i]

Err, the difference is that you are wrong and parent was right. Linux has people from IBM, SGI, SUSE, RedHat, HP, Intel, NEC, Fujitsu, Sony, Unisys, and others working on it full time. On just the kernel.
And that isn't counting all the embedded interests
that use the likes of ARM, MIPS, xtensa, cris, m32r, etc.

There is no huge gap at all, and if you actually have the Microsoft numbers, I'll try to come back with a reasonable estimate of kernel developers.