Is Linux For Losers?
forbes.com
forbes.com
Nick Naylor: Whoa, Joey I'm never wrong.
Joey Naylor: But you can't always be right...
Nick Naylor: Well, if it's your job to be right, then you're never wrong.
Joey Naylor: But what if you are wrong?
Nick Naylor: OK, let's say that you're defending chocolate, and I'm defending vanilla. Now if I were to say to you: 'Vanilla is the best flavour ice-cream', you'd say...
Joey Naylor: No, chocolate is.
Nick Naylor: Exactly, but you can't win that argument... so, I'll ask you: so you think chocolate is the end all and the all of ice-cream, do you?
Joey Naylor: It's the best ice-cream, I wouldn't order any other.
Nick Naylor: Oh! So it's all chocolate for you is it?
Joey Naylor: Yes, chocolate is all I need.
Nick Naylor: Well, I need more than chocolate, and for that matter I need more than vanilla. I believe that we need freedom. And choice when it comes to our ice-cream, and that Joey Naylor, that is the defintion of liberty.
Joey Naylor: But that's not what we're talking about
Nick Naylor: Ah! But that's what I'm talking about.
Joey Naylor: ...but you didn't prove that vanilla was the best...
Nick Naylor: I didn't have to. I proved that you're wrong, and if you're wrong I'm right.
Joey Naylor: But you still didn't convince me
Nick Naylor: It's that I'm not after you. I'm after them.
- Thank You For Smoking
As far as the comparison between BSD, it is clear why most people prefer Linux, it works with more things than BSD.
I think that at least one of Theo's complaints is completely meritless -- that Linux developers are being exploited in some way. First Linux developers do their work on their own free will, nobody tricks them or forces them to do the work. Also, Linux developers are getting treated better than BSD ones, because Linux has the GPL. So IBM can use the work of Linux developers, but if IBM has to have something done in house, they cannot keep it closed, they have to submit their own code to the community just as any Linux developer would.
This of course is very different than BSD where companies like Microsoft and Apple and Sun routinely take code and use it for their own close sourced projects.
Please prove:
1. that Linux developers "are getting treated better"
2. that it's because of the GPL
I'm afraid I don't believe you.
> This of course is very different than BSD where companies like Microsoft and Apple and Sun routinely take code and use it for their own close sourced projects.
. . . and how Linux developers sometimes do the same thing to BSD code -- except that it's a different kind of "closed", and closed to a different set of people.
- support all my hardware including printers and external monitors?
- support all file and media formats?
- have equivalents for all the important software?
I am not much of a pioneer in such things. I prefer others to do the testing. If I see a significant amount of people using PCBSD without problems, I might give it a shot. Also I don't have much of an incentive to switch, as Ubuntu does OK for me.
> - support all my hardware including printers and external monitors?
That depends on what hardware you're using -- just as it does with Ubuntu, MS Windows, and MacOS X.
> support all file and media formats?
Applications do that -- not operating systems. Installing the win32-codecs port on FreeBSD (the basis for PC-BSD) works great for me, though.
> have equivalents for all the important software?
You are aware that Linux software generally runs on FreeBSD, right?
. . . and that FreeBSD has a built-in Linux emulation layer you can enable during installation?
. . . and that the FreeBSD ports system (which PC-BSD can use) has more software in it than any Linux distribution's software archives other than Debian?
. . . or: "You don't have to believe in the devil. He believes in you."
Not that it validates/invalidates the article, but just to keep in mind.
I can't imagine that Theo is going to take kindly to OS X and Solaris being cited as "two of the best operating systems in the world" in an article in which he was extensively quoted.
Theo's personality and opinions aside, I've been a fan of OpenBSD for a long time, and it really is a solid operating system.
OpenSSH - No explanation required
pf(4) - A fantastic stateful firewall.
CARP/pfsync - Zero downtime automatic router/firewall fail-over.
sasyncd - Zero-downtime failover of IPSec tunnels.
relayd - layer 3/7 pf-integrated load balancer
OpenBGPd, OpenOSPFd - Free BGP/OSPF implementations.
OpenSSH has had some issues in the past, but everyone uses it.
Also, don't forget their recent work with guard pages (I think?) that exposed heap problems in a bunch of userland software.
The kernel is a different matter. Speaking generally, the BSDs endeavor to not commit "revisit this later" code to the kernel at all. In the idealized development approach of the BSDs, it's either done right, fully analyzed for performance impacts across the usage spectrum, long-term API maintainability, security, et al, or it's not done.
"Is Linux For Lusers?"
(not that I agree with the content). Pretty much a meaningless rant. Seems like MS propaganda.
To summarise: Linux 2.6 scaled O(n) on almost every benchmark, and OpenBSD was called a "real stinker".
This sort of thing is complicated, depending on OS, workload, application design, etc, and your mileage may always vary.
You don't think the FreeBSD kernel engineers are capable and willing to objectively evaluate other software implementations?
False.
Now, if you want to take a deeper interpretation of the analogy of "clone", as in duplicate DNA, then Linux is less of a clone of "Unix" than BSD, since it was independently written, and a judge has pretty much declared that it contains no copyrighted content that originated with "Unix". Read the SCO travesty timeline at groklaw.com. Meanwhile, the Berkeley (Berkeley Standard Distribution - BSD) folks actually worked from a copy of the original Bell Labs Unix - but worked so hard on it that in a much earlier court case, a judge made a (sealed) declaration that BSD had only a handful of lines left from the original code, and that Berkeley could distribute their work, with the exception of that handful, on their own terms. Berkeley quickly replaced/removed any infringing content, and that became the basis for the *BSDs.
For the glorious history of Unix, read Peter Salus's "A Quarter Century of Unix".
FTFY
Being "UNIX" is about passing tests and paying money. Being "Unix" is about the ancestry of the code and the design philosophy. Being "a Unix clone" happens where the ancestry of the code diverges.
That's the way most of us who know something about the subject use the terms. If it is actually an exact duplicate of the original code, we tend to call that a "copy" -- not a "clone". For instance, "I have a copy of AT&T Unix at home."
btw, there are very important contributions from OpenBSD's team - openssh and pf (primary IP packet filter imported into FreeBSD).
FreeBSD itself is an example of very successful "niched" UNIX-like system - think about network interface's performance, removal of GKL from most used device drivers, SMP-aware scheduler and memory manager, kqueue, netgraph subsystems and so on.
OpenBSD is a great toy for hobbyist to resurrect some ancient SUN's 1U Netra server as firewall and proxy for home or very small business. I have several of them for pleasure and enjoyment. And of course it is a very good choice for some low-power single-core single-CPU system without heavy pthreads.
My opinion it cannot be compared to the "open source business ecosystem" that Linux[based systems] became.