Slashdot Mirror
Is FreeBSD really 'The Other Linux'
Why2K writes "This article provides some interesting comments on the differences between FreeBSD and Linux, and the reasons why Linux is more popular. " Its sorta an outsider fluff piece, but its a nice summary. Talks broadly about the technical differences in the code,
the sociologies, and of course, the fact that BSD has a cooler mascot ;)
It's funny up to a point. Seeing FreeBSD bigots flame away, saying how much superior their kernel, development system, method of distribution, networking, (insert issue here) is than Linux' implementation of the same feature, or Linux bigots saying how FreeBSD is dead, that it can't, won't, hasn't, will never go anywhere, that it doesn't support hardware, (insert isssue here) - It sort of grinds on you.
Why people feel the need to make their system the best is obvious. Why, in the process, they have to put down, flame, or actively disrupt other systems and their development is beyond me, though. I'm sure FreeBSD is great. I'm sure BeOS is fast and it scales pretty well. What people fail to realise is that every single person on earth thinks differently. That's why there's such diversity.
I've come across this sort of mentality before - in children. For children, it's hard to comprehend that there are a lot of other possiblities other than the one that they can come up with between their two ears. They'll become devoted to a rock band, or a television show, or perhaps something a little more extreme, and suddenly everything else in the world isn't an option anymore, and anyone who chooses it is not only misguided, but stupid - and they must be put on the right path, or they'll surely perish.
FreeBSD is developed in a different manner than Linux, and BeOS is developed in yet another way. Each has their own goals and methods. Each is better at some things than other systems are. You know the saying "Jack of all trades, master of none." - It applies.
The simple fact is that systems have been chugging along quite nicely, doing what they're doing, and lots of people like it they way it is. Sure, you might be disappointed in the rapid changes of Linux' kernel - but a lot of people like it that way. You might be mad that BSD doesn't christen new versions often enough for you, but a whole whack of people don't even want the option of updating often.
If everyone could try, not to understand, but just to live and let live, I'm sure that flames would go down, the signal to noise ratio would go up, and fewer rants like this would have to be written. (I may just have to write something substantial next time)
I fail to see why people continue to compare FreeBSD to Linux. It is like comparing apples to oranges.
The advantages I see being campaigned by the FreeBSD community tends to sounds something like this: (correct me if I left something out)
1) Less people are allowed to modify FreeBSD (both userland and kernel land). And thus it is more stable.
2) Less people are allowed to modify FreeBSD (both userland and kernel land). And thus it is more secure.
3) The FreeBSD group controls both userland and kernel land and thus they are able to support a superior change control management scheme.
4) FreeBSD has superior developers. (over linux) since everyone knows linux developers are a bunch of rebels hacking elaborate "fixes" rather then implementing sound designs. (Resulting in unmanageabke code)
5) FreeBSD was based on source code provided by the LORD himself *BSDlite* and thus FreeBSD is UNIX which makes it superior.
I will attempt to rebuttal each point stated above.
1) Less people are allowed to modify FreeBSD resulting in less people looking at source code. Resulting in fewer new features and new design (possibly better) concepts from being implmented in a timely manner. The fact that adding new code tends to create a possibility of it being unstable merely creates an opportunity for a better CVS like solution to be created (case in point:BitMover)
2) Less people are allowed to modify FreeBSD and thus less people look at the code, resulting in less bugs being discovered. Having a smaller user base also has the advantage of being less satisfying to "crack". I wouldn't brag to my supervisor saying I found this bug in FreeBSD. I could brag saying I found a bug in linux.
3) This is one of the most interesting arguements because it seems to be the one that has the LEAST effect on the success of Linux/Gnu. However it seems to be the one that is campaigned most by the FreeBSD community (at least the vocal part of it).
I see this as flawed thinking, the very reason why the GNU/LINUX model works is because it follows (IMHO) the model of creating that has the most longevity. I would call this a Natrual process of creating. I say that because I see the GNU/Linux process the same way I see nature. It evolves based on the energies put into it. Now this is true for all things in my opinion however it is my experince that when you attempt to "control" evolution (or nature) it has undesireable results. (fragmentation, stagnation, death, corruption ect..).
4) For the most part core FreeBSD developers and core Linux developers are of the same breed. This arguement is silly. Research the names in the Linux CREDITS list and this whole theory gets blown out of the water.
5) FreeBSD was based on the efforts of many before them. A very solid well designed system. Linux was created with these same good designs in mind. But it was created with the BAD designs in mind too. Linux is not UNIX, nor has it had the luxury of it's 40 year old code base and thus it has taken a number of years for it to mature to the point of most UNIX's. It has just recently (last year or two) reached a point where it can claim as such. If you put this in perspective, you will see that Linux is growing at a rate that is truly amazing.
The Linux/GNU process (in it's entirety) is so radically differnt then other development processes that judging it based on standard methodologies leaves you in the position of supporting an arguement that holds no water (i.e. the position/arguements you hold against something has no noticeable effect on it (leaving you always saying (just wait, it will "REAL SOON now")).
-- You can be a geeklord too
There's not much competition per se between Linux and *BSD.. they're both great OSes. BSD is incredibly powerful and stable but it's got a learning curve that's steep enough to require a safety line and supplmental oxygen. Linux, believe it or not, is a more "friendly" UNIX-like OS.
Not to say that the BSD guys wouldn't like to be the darling of the world's attention as we dig ourselves out of the soft-inspired rut we're in now but they seem content with the fact that they're not. Their user base is growing (partially, at least) as a residual effect of the popularity of Linux. I know that when I was going to get into UN*X like systems, I saw how cool Linux was that encouraged me to try BSD on some REALLY old hardware I had laying about. Found myself one hell of a small server and can only imagine how my NetBSD install would run on a processor/logic board that was minted after 1993!
Seriously, BSD is a really cool system, fast, stable and in some ways even elegant. But the turnaround times, back then, were HORRIBLE! That's about the time I migrated, once and for all, to Linux. Linux was behind, then, but was gaining ground. BSD was playing the hare, and sleeping away it's huge lead.
These days, it seems to take less than an eon before the next BSD updates come out, which is a huge improvement, but as I see it, it's still not got the pace it could - or should - have. I've been tempted to put a partition aside for one of the *BSD systems, but I'm holding off to see if it'll be overtaken by Linux on the few advantages it has left. If it is, what's the point? On the other hand, if it makes real progress, I want to have a copy to play with.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
But that's incompatible in a different way. Nothing there says that you can't take BSD'd stuff, slap the GPL on it (remember, you can do anything you want with it), and put it with the GPL'd stuff (except with that advertising clause, I guess; don't know about that). What you said just means that you can't take GPL'd stuff and put it in a BSD program. And aren't you using "commercial" incorrectly? I don't think you're allowed to make proprietary yet non-commercial products out of GPL'd code. As long as you intend to distribute it, it matters.
The picture used to represent the FBSD daemon is clearly the Christian interpretation of demons, not from the Greco-Roman Daimon. The Christian demon is described as a spirit with no body which has a nasty habit of posessing other people's bodies. They aren't even top dog in Christian mythology, they are inferior to archangels.
The Greek daimon was not exactly something I'd want as a mascot anyway. A daimon is a supernatural being of the lowest order, inferior to other supernatural beings. It sits at a level between the gods and man.
There are earlier definitions of demons being called daimones or "divine powers". Later however daimones began to refer more towards the spirits of the dead (Roman) and to the above daimon. They also refer to spirits which each person is assigned to watch over.
--
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
I've read over the GPL several times and can't find any reason why the BSD license is incompatible with the GPL.
The GPL states the BSD requirements in it (copyright notice) and the only clause that comes close to saying BSD is incompatible with GPL is the sublicensing restrictions clause, but it clearly is not incompatible with the BSD license if you read it carefully.
--
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
Ok, so that was a bad example. The point still holds though. Most FreeBSD programs have up-to-date manpages, while most Linux programs seem to be following gcc's lead in abandoning their man pages in favor of other types of documentation.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
In case you did not notice, RedHat is funding a security audit of their distribution. As a result there are many announcements of security holes - but this is a side-effect of improving security.
OpenBSD did this a while ago you say? Sure. But read those announcements because if you are using the same applications that the Linux folks are, then you are vulnerable. Security holes are things that tend to build up over time.
In fact security is a constant problem with the *nix organization. If you want to see a fundamental design which can lead to far more security than either OpenBSD or Linux, take a look at EROS.
(Solution to root-exploits, get rid of the possibility!)
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
That's because, by definition, UID 0 is root. Some other UID may be called "root," and UID 0 may have a different name, but the UID 0 is still the superuser with all the "root" privilages. Read up on your UNIX specs.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Once linux has mainstreamed and everybody is using it, geeks won't feel so special anymore. I think alot of people jumped ship off of Microsoft simply because they wanted to differentiate themselves from everybody else. At the time, linux was the obvious choice. Now that linux has mainstreamed, geeks need a new toy that appeals to just geeks.
The answer, it would seem, comes in the form of a small stuffed devil with a pitchfork.
--
While they went to great lengths to say that one of the differences was how code was added to the base, they didn't really say why that matters that much.
:)
Before, when I did a lot of custom drivers for Linux, I was continually annoyed with kernel interfaces changing, and code that was 'current' a week ago, suddenly becoming legacy.
With FreeBSD, I've had amazing longevity with my code. Not because they're slow to change, but because, in my opinion, (flame proof clothes ready) it was designed 'right' the first time.
The works of people like Kirk McKusick and David Greenman are quite possibly some of the best designs I've ever seen. Even if you don't plan on using FreeBSD, I think all developers should at least take a look at how it works, and I promise you'll learn a thing or two.
I also hold 'good design' as the reasons for why I've had systems with 400+ day uptimes, that were sometimes under nearly constant attack and/or intrusion attempts.
Finally, the BSD license. My 'day job' is designing an embedded product, which is using FreeBSD for it's OS. Why? The license. Many companies are hesitant, or even contractually prohibited, from giving out changes made to the system, which the GPL rather insists on. The BSD license is very open, which I think may become more important in the near future.
Go download a boot floppy, or buy a CD from Walnut Creek or even Cheap Bytes and give it a try. Even if you don't end up keeping it, if you're a hacker, you'll learn something.
Anyone thinking FreeBSD (in particular) doesn't "release" often enough... well, just doesn't get it. FreeBSD's releases are just points at which it goes into cds.
If you want to know at which pace FreeBSD progresses, send a message to majordomo@freebsd.org subscribing to cvs-all.
(8-DCS)
I think the choice of mascots is actually a fairly good indicator of the basic difference between Linux and xxxBSD.
The BSD mascot is an in-joke, a play on the ubiquitous (sp?) demons in a UNIX system. For those who already are familiar with unix-like systems this is fairly obvious and somewhat amusing. By contrast, outsiders are likely to see the mascot in its original setting, as a mythological demon/devil. Insiders get it, outsiders are pushed away.
Tux the Penguin, on the other hand, needs no inside knowledge to appreciate, but is as accessible a symbol for the neophyte as for long-time users. It's not an in-joke.
This is of course a lot of symbolic baggage to overload these poor mascots with, but in a small way I do think this reflects a part of the culture of the respective systems.
Trust the Computer. The Computer is your friend.
Yeah, but Linux doesn't keep its man pages updated (try doing a "man gcc" to see what they have to say about it). FreeBSD does. What good does RTFM do if the FM is out of date?
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
FreeBSD is the most stable kernel on the planet!
The distro and environment are not that great, which is why I use slackware & Linux 2.0.3x to
run a major internet operation.
However-- check out this uptime for a BUSY FreeBSD box we have (we even moved it when we changed offices, kept it connected to its UPS in the car!)
% uptime
10:43AM up 871 days, 1:07, 1 user, load averages: 0.31, 0.30, 0.31
Aside from the system design, one thing that really impresses me with FreeBSD is its spectacular documentation. Nearly every aspect of the OS is documented in manpages or in The Complete FreeBSD book.
For MS, the mascot would have to be a dollar sign, or of a large boot stomping on a computer (with apologies to George Orwell). Unless, of course, MS would like to remind us of "Bob" =)
and, off-topic though it is (well, it concerns which mascot is "cooler"): anybody remember Thin Lizzy's "The Boys are Back in Town?"
So, in summation, the situation with which mascot is "cooler" is inevitably muddled =)
"Oh, I hope he doesn't give us halyatchkies," said Heinrich.
Now, there are many complicated reasons for this, but what I am trying to say is that the number of people who have a big problem with the GPL is relatively small. Most would be developers who wish to use others' work in their own proprietary work.
You may define freedom as including the ability to restrict the freedom of others, but not everyone does.
I use software that is GPLd, BSDd, X11d, and proprietary. I even am currently maintaining a program that has a (ugh) non-commercial-distribution license. But, I prefer the GPL, whereever suitable, because it discourages forking, and keeps software free. And quite a number of people seem to think they can make money from GPLd software, regardless of the views of a few rabid proprietary developer advocates.
I guess my view would agree with Linus'. If you didn't write the software, you have no right to bitch about the license.
But what it really boils down to is this: the user/customer will, in the end, determine what licensing schemes are successful. If the user prefers GPLd software, that is how things will tend to be licensed in the future. All the pissing and moaning in the world isn't going to change that reality.
--
Interested in XFMail? New XFMail home page
The thing is, FreeBSD and Linux are like twins, except that FreeBSD does not have a new libc every day that breaks every program in existence, and FreeBSD is an operating system, not a kernel.
I finally switched to FreeBSD because I got tired of Netscape exhausting virtual memory and crashing my machine under Linux. FreeBSD kills Netscape (the proper behavior) and keeps running. FreeBSD runs all my commonly-used commercial Linux programs (Applix, WordPerfect, etc.), and the 'packages' system has a lot of packages that most Linux distributions don't come with (with the exception of SuSE!), so it's not like I had to give anything up.
That said, FreeBSD has some disadvantages -- it supports fewer hardware peripheral devices than Linux, system administration is a pain in the rear (almost everything is done via rc.conf, which is almost totally undocumented), and you're right, all the "cool" stuff is being done for Linux (stuff like neat new filesystems, etc.) and FreeBSD continues to lag. Still, if you want an OS that doesn't have a new libc library every fifteen minutes, FreeBSD is the one.
-E
Send mail here if you want to reach me.
Here's that exact clause:
Note how it says you may not impose further restrictions on the rights granted. The copyright notice does *NOT* impose upon the rights as it is in paragraph 1 and paragraph 2 section c.
As far as I can tell, it is perfectly legal to include BSD licensed code in a GPL program. This is a big loss to *BSD folks because Linux kernel types can integrate their code but the reverse can not happen unless you segregate it or switch the entire kernel source over to GPL.
--
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
- Linux
- Exim
- ProFTPd
- KDE (A number of KDE programs could make up this whole list, actually, if you want to get picky about what is and isn't "enduser software.")
- XEmacs
- ncftp
- MiniVend
- PHP
- GNUTar
- samba
IMO, of course. And, yes...I also use a Mac occasionally.And, because all this stuff is GPLd, I can hack on it, or ask others to hack on it for me. And I need not worry about someone else's "business decision" orphaning my software (happened to me with OS/2; not gonna happen again because I refuse to chain myself to proprietary software when free alternatives exist).
--
Interested in XFMail? New XFMail home page
Heh. The difference is the *BSD crowd points you to other websites instead of answering your question first. Therefore, Linux is more popular.
:)
No, really, think about this before replying to it. I just spent an hour trying to figure out why the hell some people call the "BSD Daemon" Beastie, and others call him Chuck, while his creator is very careful not to name him.
Apparently it's a *BSD style ego-clash, in the same tradition as forking the main source tree. Under Linux, we don't care. Some people don't even care if the mascot is a penguin, much less if he's called "Tux".
...and if you ask me to elaborate, I won't refer you to a web page.
pb Reply or e-mail; don't vaguely moderate.
As a person who uses both Mandrake 6.0 and FreeBSD 3.1, they are both kick ass. Without them, I would be running 98 on my laptop and NT on my desktop and I would be a sad person.
That said, Linux is a huge boon to BSD because:
1. Linux makes people aware of BSD. I first heard about BSD here at slashdot. Linux enlarges the PC Unix user base.
2. BSD benifits from development of all of the Linux oriented applictions. I have not had any trouble running any Linux apps on BSD.
3. Linux inspires innovation and provides competition, forcing BSD to provide what its users want.
In the end, if you like to tinker and contribute, Linux is a good choice. If you just want to to work and not worry about it, Free BSD might be a better choice.
Or, do like I do, use them both.
Actually, as far as "media performance" goes Linux has recently made some great strides. mingo released a patch that reduces scheduling latencies to under 5 milliseconds, and David Olofson has ported the linux sound drivers to RT-Linux so that linux can now do sound processing with a signifigantly lower max latency than BeOS (~1 ms). NT and macos both need external DSP boards to be able to beat this (pricey, but they beat the hell out of anything you can do on a general purpose PC).
If you want to follow a faster track I suggest you try out FreeBSD-current. Many more features are available at the expense of the thorough testing the the -stable branch provides.
m l
Here's a link that describes -current:
http://www.freebsd.org/handbook/cutting-edge.ht
Also, it should be noted that "-current" is not for newbies or people that aren't used to helping themselves or even providing fixes to problems that they encounter.
One of the benifits of developing applications for FreeBSD is the long term binary compatibility offered, it is almost unheard of to see a change go into FreeBSD-stable that would break current applications that are programmed correctly.
This gives ISVs the advantage of not having to worry about binary compatibility between minor (and usually major) versions of FreeBSD.
I also don't see why one would _want_ thier OS to change under thier feet almost daily, it sounds more like a punishment than a good thing imo.
-
- Alfred Perlstein - Programmer and Administrator, Wintelcom.