Weekly Edition
Return to the Press pageSponsored link
Vyatta –
Linux & Open Source
Alternative to Cisco –
Advanced Routing,
Firewall, VPN, QoS..
Free Download ->
| |||||||||||
Outside Looking In: The BSD Operating Systems (eWeek)
eWeek examines the
BSD variants. "BSD software, in any variety, is stable, extremely
flexible, arguably better tested, more secure. At the same time, those
things also mean that it tends to be less bleeding edge, slower to come out
with new features, and more difficult to initially install."
(Log in to post comments)
Outside Looking In: The BSD Operating Systems (eWeek) Posted Oct 31, 2003 22:04 UTC (Fri) by suckmebitch (guest, #16361) [Link] SCO owns Unix and Linux! Yeah!
Outside Looking In: The BSD Operating Systems (eWeek) Posted Oct 31, 2003 23:09 UTC (Fri) by s_cargo (guest, #10473) [Link] Take your thioridazine, Darl.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 1, 2003 2:27 UTC (Sat) by tjc (subscriber, #137) [Link] I'm guessing you didn't select the 16 month "project leader" subscription option.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Oct 31, 2003 23:24 UTC (Fri) by ninjaz (guest, #2083) [Link] "BSD software, in any variety, is stable, extremely flexible, arguably better tested, more secure. At the same time, those things also mean that it tends to be less bleeding edge, slower to come out with new features, and more difficult to initially install. I think this is a misguided statement. For instance, NetBSD first released USB support about 2 years before it landed in Linux. Similar for wireless networking support. And, OpenBSD has shipped a chrooted BIND for years. Also, when a new feature is announced for a BSD release, it really is there and ready to use. Eg., a few years ago when large file support (> 2G on 32 bit systems) was announced in the a new release of the Linux kernel, I rejoiced. But, that was short-lived, because it turned out that to get use the new large file support, new glibc support had to be added, every program on the system had to be recompiled, etc. Meanwhile, *BSD already had large file support and since the libc, kernel and the rest of the core OS are distributed together, the feature really was usable. No waiting a year for the new series of kernel to stabilize, the supporting tools to be brought up to date, and all packaged by a distro (or going it on your own ...). Another big win with BSD is that a specific release of BSD doesn't peg you to the specific release of all the applications that come with it. The ports system is continually updated, so you aren't stuck in a position of either installing a new version of, for instance, Mozilla by hand, breaking all the inter-dependencies of the packaged stuff, or waiting for the next version of the distribution to do it "the right way". So, the BSD package system really is more of a solution than a problem in my experience. It also helps them push out new core technology because the core OS doesn't get hung up waiting for all of the add-on applications to stabilize before releasing.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 1, 2003 0:39 UTC (Sat) by seanegan (guest, #15672) [Link] I looked into this once. Alan Cox made a comment once about how he chose Linux in large part because it had IDE disk support. My cursory inspection of "release" (as opposed to "developement") distros, found that the Linux kernel had IDE support about 18 months before FreeBSD.I can personally attest that IDE support for my chipset was what allowed me to switch to Linux (slackware) from OS/2. However, I didn't know about FreeBSD at the time. I would suggest that early IDE support was a critical element for spinning up the Adoption-Popularity virtuous cycle. I believe this "head start" explains the popularity-lead Linux currently enjoys. I would also agree with the article that kernel-wise FreeBSD is very competitive with Linux. I actually use FreeBSD on my laptop because it has better ACPI support.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 1, 2003 15:18 UTC (Sat) by davidl (guest, #12156) [Link] In view of the fact that 'BSD' and Linux based systems can run much of the same software on top, I think there is basically no difference between the two, particularly for the end-user. I have two machines, one is Linux based and the other based on FreeBSD and they both run the same desktop software, KDE et al, and I can't comment on which one performs better because they are both pretty identical. Yes, there are differences in getting different bits of software to compile and run, as can happen with different kernels, but these are all issues people are really trying to solve. I have found that ACPI support is undeniably good within FreeBSD, and with Kernel 2.6 (or will it eventually be 3?) Linux has had a little bit of catching up to do.People talk about Linux ending up being the standard base for an OS. Even as more of a Linux oriented person, I don't agree at all. Utilizing free software and open standards I think we'll get to a stage where all OSs will be customisable, compatible and interoperable with each other, to any degree, through these methods. Linux has merely become a focus for this, but the goal of Linux is not to destroy everything else. This raises the possibility of the long-held holy grail of having different OSs, even running on different hardware, being equally compatible, interchangeable and interoperable. That sounds like an Open Source roadmap to me! Want to buy a PC, Mac or even a Sun Workstation at your local Bainbridge's store to use as a dekstop? Fine. If the computer industry is to progress, compatibility, particularly with software, should never be an issue again. Too many people have made too much easy money out of it. So Open Source software is difficult to use is it? Taken to their logical conclusions you couldn't get systems that are more easy to use than this and can provide the base for all manner of commercial opportunities. Microsoft? Well they merely skip around these important issues by telling everyone to run WIntel. It's a total false dawn. Palladium or no Palladium, the advantages of the 'open' approach are undeniable. Well, that's my take on the future of OSs anyway. Slightly off-topic, but it should highlight that no OS, particularly the BSDs, should ever be dead. If you read another one of these <insert OS here> is dead statements, just ignore it and do your own thing. We're heading into a different world.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 1, 2003 19:24 UTC (Sat) by katsklaw (guest, #15643) [Link] BSD and Linux are about as similar as Windows95 and Windows 2000. There isno way they are "pretty identical". The only thing they share is alot of the software is interchangable. However, that is where the likeness ends. Under the skin they are 2 totally different OS's Linux is a compilation of efforts by many groups such as the kernel, most Linux distributions rely on the same kernel, therefore the kernel must be generic enough to allow for said usage. Not true with BSD. BSD's Ports Collection cannot be touched by any Linux distribution, again because Linux relies so much on the efforts of countless 3rd party efforts. This by nature will make BSD a more stable, better built, longer lasting, easier to use OS. Where BSD is behind in the power curve is it's adaptability to be used as a Desktop/multimedia OS. Nothing wrong with that since BSD has always been "Server Grade".
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 2, 2003 13:37 UTC (Sun) by davidl (guest, #12156) [Link] The only thing they share is alot of the software is interchangable. However, that is where the likeness ends. Under the skin they are 2 totally different OS'sDefine what an OS is please, because that's what we're talking about here. It's certainly a lot more than just a kernel. If the vast majority of the software is interchangeable between the BSDs and Linux I'd say that makes them pretty equal as whole, fully functioning systems. The differences occur when you are deciding what system to use for a particular job or when you are looking at low-level development aspects. Under the skin they are 2 totally different OS's Where does the skin begin and end? :). BSD and Linux are about as similar as Windows95 and Windows 2000. There is no way they are "pretty identical". How? For many people, particularly the end-user and even developers not involved in low-level development I think you'll find that there are precious few differences, if any. There are more differences between Windows 95 and 2000 than there are between a BSD and a Linux distro, as whole systems. As you've just mentioned and as I've outlined above, much of the software is interchangeable - which means that much of a Linux distro 'operating system' and a system like FreeBSD share common software. With Windows 2000 software could be ported and run through 'compatibility modes', but there was a lot of software that fundamentally had to change. That's not the case when porting from a Linux system to a BSD. In most cases the code is identical (compile and watch what happens) and the architecture of it does not need to be changed in any way. Linux is a compilation of efforts by many groups such as the kernel, most Linux distributions rely on the same kernel, therefore the kernel must be generic enough to allow for said usage. Not true with BSD. Certainly true, although I don't really see this as being an advantage for a BSD. You seem to be saying here that a Linux-based system is more modular :). BSD's Ports Collection cannot be touched by any Linux distribution, I'd beg to differ here. This comment is very subjective and represents your opinion. The Ports Collection could easily be ported to a Linux-based system and improved upon (as we've already agreed, much of the software itself is interchangeable :)) and this is what distributions like Gentoo and the Zynot fork are working on at the moment. It is the popularity of Linux-based systems that has been responsible for greatly driving forward development of much free software that has given FreeBSD the software to put into their Ports Collection :). again because Linux relies so much on the efforts of countless 3rd party efforts. Is that such a bad thing? I thought diversity was good? As we've all discovered there is more than one way to install software, and they all have their different merits :). Who says that a particular method or system is the correct way to install software? What exactly is the best 'standard' way? That's what all the 'third-party' efforts that surround Linux are about, by taking different methods out into the world, testing and improving them. We also have to recognize that two given systems may not ever have two 'totally compatible' ways of installing software (or doing anything), such as a computer and an 'embedded' device. How do we establish commonality between them? These are questions that the community seeks to answer. Although the Ports Collection and methods are undeniably good, it is not an answer to absolutely every situation. I don't think any BSD is in a situation to impose anything on anybody :). These differences are what drive people involved in more Linux-based software development. This by nature will make BSD a more stable, better built, longer lasting, easier to use OS. Which BSD would that be? There's more than one. Basically telling everybone that the Ports Collection is 'the' method for software installation and the more 'in-house' development of the BSDs is actually what has made them more scarce. The different community efforts clustered around Linux-based development is really what has driven it forward. Accepting that different systems may do things differently (as in the real world) and finding ways to establish commonality, communication and understanding between them will be a cornerstone of the killer success of not just free software and Linux, but the community effort as a whole. As usual, the various BSD projects will benefit from this process, seemingly without even noticing, but that's the nature of free software development and will ultimately be good for everyone. In the end the flexibility we will have with free software will mean that we can really choose anything we want without worrying about issues such as compatibility. The fact that we have agreed that much of the same software is interchangeable for BSD and Linux systems is testament to that. Think of the end-uses, not the technical differences that divide us. Remember, advocating a standard installation process/system, standard ways of doing things, criticizing third party efforts and advocating *insert OS here* everywhere is the sort of approach that Microsoft strongly follows :(. It's the easy way out. We don't want to be going there. I'm not going to get involved in Linux/BSD slanging matches. I've seen a lot of them and most of the stuff written and said on both sides is a total waste of time and effort as everyone is right, and everyone is wrong. I've really tried here to outline my reasons for thinking the way I do. I really don't want to have to ever worry about whether I use a BSD, Linux-based or any other system. Remember, think of where we are going not the technical, sometimes petty, differences that divide us.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 27, 2003 23:20 UTC (Thu) by katsklaw (guest, #15643) [Link] Actually there is alot more difference between BSD and Linux than you realize.Different kernels all together (not to mention compilation). BSD is Unix not Linux having packages for "Linux compatability" is a good indication that they aren't "nearly identical" hundreds of commands that are in BSD and not Linux and vice versa. As with all things, even "diversity" can be a bad thing, especially in the presence of lack of communication. It doesn't matter which version of BSD I refer to, even though they are different amongst themselves. For the generalistic approach of my point they are one in the same.
Outside Looking In: The BSD Operating Systems (eWeek) Posted Nov 1, 2003 18:11 UTC (Sat) by ccchips (guest, #3222) [Link] Back in about 1995, my first attempt at using FreeBSD failed, because the IDE controller in my computer wasn't supported. Not realizing that Walnut Creed didn't support FreeBSD directly, I wrote a complaint letter (I had installed Slackware on that machine, with no problems,) and they simply forwarded my complaint (which was rather abrasive) to Jordan Hubbard.He told me that the FreeBSD people weren't interested in supporting every darn piece of gimmickry out there, and that FreeBSD was really for big servers with SCSI drives and people who knew what they were doing, and that I had a bad attitude. Of course, my bad attitude came from having paid good money for the FreeBSD set, and paper manual, from Walnut Creek. I sent it back. A couple years later, I tried a newer copy of FreeBSD (with paper manual) from Walnut Creek, and it worked correctly with my controller. However, I had since been actively participating in Linux advocacy, had posted a few patches to software on Linux, and been tryng other OS's and distros (including Red Hat, Debian, etc.) One of the things I got from that delay was the sense that the development process on Linux was far more open. There were certainly flames going around, and personality conflicts, but by and large the process of getting things done felt a lot more confortable to me. I think that if my hardware had been supported in the first place, I might have kept FreeBSD running, and learned more about it. Maybe I would have seen the advantages of their development model, too. So be it.
|
|||||||||||