Happy 20th Birthday, FreeBSD

mbadolato writes "FreeBSD celebrates its 20th birthday this week. On 19 June 1993, David Greenman, Jordan Hubbard and Rod Grimes announced the creation of their new fork of the BSD 4.3 operating system, and its new name: FreeBSD." And in the time since then, FreeBSD hasn't exactly stood still; it's spawned numerous other projects (like DragonFly BSD and PC-BSD), as well as served as the basis for much of Mac OS X; there's even a Raspberry Pi build.

Comments

Well...

[grub]

Given enough time, Netcraft will confirm...

Re:Well...

[Plunky]

Honest question, who uses NetBSD?

Well I do, and moreover I personally have written ~30 thousand lines of code for NetBSD which has been used in other OS projects (the other BSDs, and OpenSolaris at least - see Bluetooth code) in varying amounts, and I am certainly not the only one to have had code re-used. The NetBSD libc is being used for Android now, I believe.

Also, many companies do use it, though they don't always advertise that fact.

Seriously, after 25 years in the business I've never seen or heard about anyone using NetBSD in production ever.

The licence is liberal, and companies are not obligated to mention their usage.

It just works

[approachingZero]

I've been using it since about 1998 to serve web pages. Solid product, thanks for all the hard work people.

Re:It just works

[mlts]

It works without issue. I've used a BSD on and off since the days of Jolitz's 386BSD which came with a compressed image with a number of utilities on a 1.44MB floppy disk. Before this, if a person wanted source code to look at, they would have to pay a good mint to BSDI or a company like that... and of course, if you wanted SVR4 source... good luck with that.

Ahh... memories.

Re:It just works

[houstonbofh]

It is also the core of several appliances like m0n0wall, pfSense, FreeNAS, Nas4free, and Askozia. (even if Askizia ported to Linux later) Not bad for a little OS no one uses... ;)

Re:It just works

[steg0]

I recall frequent kernel panics while booting that were related to the Intel Ethernet chipset on a SuperMicro H8SGL-F board (not exactly the least common hardware) in a released version (I think it was 8.2 or 8.3), which was probably this. Rather annoying.

There have been other problems, too (off the top of my head), like

• the mediocre PAE support,
• and the in my eyes rather ungracefully handled transition to Xorg 7.2 in the 6.x releases, which for me didn't work at all like the documentation said, although this was not a problem of the base system, but the ports collection.
• Then there's stuff like some guys arbitrarily deciding to reimplement the system installer and on top of that, to remove the old one in the time window between 9.0 RC 3 and 9.0-RELEASE, see (along with some elitist Linux bashing going on:) here and here
• or the transition to Clang at a time when it wasn't even ready for the non-x86 architectures!

So sometimes I ask myself whether this OS is really ready for prime time

But enough of the rant. I've been sticking to it since 2000 and for most of the time it just runs and does its job. It's got some nice coherent documentation too.

Congrats FreeBSD

[laffer1]

FreeBSD is a great example of open source working. Not only has it been successful, but it has spawned a lot of other open source projects such as GhostBSD, PC-BSD, DesktopBSD, DragonFly, pfsense, freenas, nanobsd, and my own MidnightBSD.

There are a lot of people who have donated a lot of time to FreeBSD. This wouldn't have happened without all the committers and folks offering patches to the project. FreeBSD and all the other projects I mentioned wouldn't be here without the. Thanks!

Re:Congrats FreeBSD

[smash]

It is also the basis of JunOS, Netapps Data OnTap, and various other commercial products. FreeBSD is really under-rated and works very very well.

Re:Congrats FreeBSD

[jbolden]

Develop a "standard" under the GPL and it won't become standard at all, as no commercial OS will be able to use a starting reference implementation as a base.

Absolutely. And you can see that by noting that:

gcc compatibility C++ source
gnu make's -o extension
Qt API's as a standard for mobility cross platform
Linux kernel API as a standard for emulation (mainframe, supercomputing, mini...)
Wordpress as a blogging standard
Sword standard for bibles
Guile as a standard Scheme
Blender API for 3D modeling
etc...

aren't standards. Oh wait.

Re:Congrats FreeBSD

[Arker]

There is a little bit of truth hiding behind your words but your statement is still very misleading.

GPL is much more 'truly open' precisely because no 'proprietary' implementation of a standard with a GPL reference implementation will be able to simply lift the code (legally.) *Proprietary* being the keyword here - you said commercial, and that is simply false. You can make a commercial implementation of a standard with a GPL reference implementation, and furthermore you can simply copy that reference implementation to do it!

Proprietary != commercial. Slackware, RedHat, Ubuntu, etc. are all commercial. GPL is perfectly fine with commercial. It's only proprietary that it objects to (and for good reason!)

Had AT&T not sued BSDi

[stox]

We might be talking about FreeBSD as we do Linux these days.

FreeBSD's developers CHOSE to not be popular

[jphamlore]

Let's explode that myth. Here's what actually happened. Linux distributions such as Slackware back then supported booting from a floppy into the OS so that one could run the rest of the userland from a hard drive. That meant one could preserve Microsoft Windows booting yet run Linux at the same time with no risk. I cannot stress how important a feature that was back then to someone like me, back when PCs were very expensive and had to be shared among family members. The FreeBSD developers took a different tack. Their OS was for grown-ups, for servers. They openly mocked on their mailing lists the feature of being able to boot into the OS from a floppy drive. (Note this is different from being able to INSTALL from floppy, everyone back then could do that.) The FreeBSD developers CHOSE to not be popular.

Re:FreeBSD's developers CHOSE to not be popular

[camperdave]

They openly mocked on their mailing lists the feature of being able to boot into the OS from a floppy drive... The FreeBSD developers CHOSE to not be popular.

... and here it is umpteen years later and NOBODY boots from a floppy. Sounds to me like they were just ahead of their time.

Re:FreeBSD's developers CHOSE to not be popular

[stox]

I disagree, as far as real adoption goes. Yes, booting Linux from a floppy using a MSDOS filesystem did enable a lot of people to get exposed, but the race was lost before those people made a difference. Had BSD development not stalled for two years, many of the early commercial and big site adoptions would have gone to BSD instead. Many started with BSD and then jumped to Linux because that is where the momentum was. Red Hat's IPO sealed the deal.

BTW, I introduced Pat Volkerding to the Church of the SubGenius, and pioneered a lot of the early work with Linux at Fermilab. I know a little about these things.

Re:FreeBSD's developers CHOSE to not be popular

[Anonymous Coward]

It wasn't just something specific like floppy boot, it was the entire attitude that Linux was for "Peecees" or "Windoze" users, while *BSD was for the Sun Workstation Master Race (who couldn't actually personally afford a sun workstation). Just as an example, *BSD thought "real workstations use SCSI (period)". While Linux had all sorts of workarounds for your buggy IDE chipset and support for your proprietary Soundblaster CD-ROM drive.

And while the Net/Open/Free factions were flaming each other on the maillists, there was this persistant attitude that Linux was vastly inferior thing, even after the "the battle was over", and Linux had clearly won. When the history is really written, the story of *BSD has little do with AT&T and is more about how arrogance and personal politics alienated a entire genration of users.

Re:FreeBSD's developers CHOSE to not be popular

[jbolden]

Red Hat's IPO was in 1999. The Linux :: BSD ratio was already massive by then.

I don't agree with you about 1994. I don't think it was losing time. I think the BSD community was hostile to people like me (Windows Power Users and Unix users -- non admins) who became the people who pushed Linux into corporate America during the 1990s and early 2000s.

Re:What's the difference with Linux ?

[real-modo]

Hmmm.. last used for any length of time: FreeBSD 2.2.2 with FVWM95 (serial mouse!), back in '00 or thereabouts. (At the time, FreeBSD was reliable; Linux was not. Corel Linux--remember that, anyone? Urg. Red Hat? flaky, at the time. Mandrake Linux: slightly less flaky.)

But the culture hasn't changed much, from a recent scout-round. I'd say your impression is correct. Here are some random thoughts:-

  *BSDers will say *BSD is more like "real Un*x", but as far as I can tell the OS has been riddled with schisms since the '70s. The "real Un*x" is a nostalgic fantasy (or an artefact of Stockholm syndrome, take your pick).

*BSDers will say *BSD is reliable. That hasn't been a problem for Linux for a decade. (Except for Intel's video drivers...grrr.)

Differences...apart from being behind the times hardware-wise (which you can do with Centos 4, if you want), the main difference is: only one "distro". (Although there are a few derivatives of FreeBSD and NetBSD, only their creators use them, pretty much.) BDSM submissives enjoy OpenBSD; no-one'd dare fork it.

The FreeBSD man pages were better. Way better, as I recall. That's in part because they tried to avoid all that dubious GNU stuff. Can't say they were wrong about info(1), but I can say they were wrong about make(1).

Filesystems. Linux and *BSD have *FAT*, NTFS, and ZFS in common. That's about it. FreeBSD has had ZFS for a couple of years longer than Linux.

Culture. For a long time the *BSDs' attitude was "compile it from source, and fix the dependencies yourself". Like combining the bad parts of old-time Slackware and Gentoo. Might be better now; I've only tried Live CDs.

Startup: I like the rc.conf startup configuration approach. (Way better than System Five initscripts. "Fragile" hardly begins to describe that approach.) I used Arch Linux for a long time because it had the closest approximation to rc.conf, but it also had drivers for USB and stuff. You know, the hardware I had attached to my PC. Not much, back in the day; but I wanted to use it. Arch was a pretty good compromise.

Now, Arch Linux has moved away from an rc.conf-ish approach to using systemd. I've been getting progressively more annoyed with all the Sieved Poots appearing in linux, so I recently tried PC-BSD, which is supposed to be an end-user friendly porcelain on top of FreeBSD. Unfortunately, it's dire. Bug after glitch after missing object. On both my PCs, the typography is eyewatering. Worse than Windows 3.11.

You're better off with FreeBSD. I might be going back there soon. Probably, though, it won't have support for my USB wifi stick. If you never see me comment again, you'll know what's happened.

Re:What's the difference with Linux ?

[TheRaven64]

The reason I switched for FreeBSD around 2001 is that in-kernel sound mixing Just Worked. Two apps could write to /dev/dsp and get working sound. I'm now writing this while watching a DVD on the FreeBSD box connected to my projector and 5.1 surround sound system and I didn't need to do any configuration.

The other big difference is ZFS, which makes a huge difference to how you manage storage. Creating a new volume is as easy (and fast) as creating a new directory. You get compression, deduplication, constant-time snapshots, and a load of other things via a very easy administration interface.

If you're doing development work or running servers, jails give you a way of deploying a complete system that's got almost the same isolation as a VM but with much lower overhead. With ZFS, you can create one stock install and then clone it into a new jail in a few seconds.

The base system is maintained as a whole and the developers take the principle of least astonishment (POLA) seriously. User-visible changes are minimised and new configuration utilities are expected to follow the pattern of existing ones.

For firewalling, there are a number of choices, but the most sensible is probably the fork of OpenBSD's pf, modified to have better SMP scalability.

For security, there's the MAC framework, which is roughly equivalent to SELinux, that the sandboxing on OS X and iOS are based on, and also Capscium, which provides a capability model that is better suited to application compartmentalisation. An increasing number of the system daemons use Capsicum for privilege separation.

That's probably most of the user-facing things. You'll notice that GPU drivers (except for the nVidia blobs) tend to lag Linux somewhat. For Intel it's not so bad, for AMD it's quite a way behind (catching up, but not there yet).