It’s been a week since FreeBSD 5.0 got released. Many argue that this release is the most important and most technically advanced, ever, for FreeBSD. I am truly disapointed though on the way the media have responded to this release.Except the *BSD news sites, OSNews, Slashdot, a few Mac news sites (mostly because of the OSX-BSD connection) and a small number of Linux news sites, none of the big “pro” news sites reported on the release. It was a negative surprise to not see eWeek or ZDNews and News.com report on the release. BusinessWeek, InfoWorld and the rest of the “high-end pro” news sites didn’t report about it either.
This situation really saddens me, and I am not particularly a FreeBSD user (I don’t claim to reboot to it more than once or twice a month on my personal box – however our home router/server box does run happily FreeBSD). But no matter how I look at it, this release was really important, FreeBSD 5.0 features SMP and MT technology not found on other OSes.
Instead, we see Linux making headlines for every little thing that’s happening to it. And if you think that “FreeBSD is not as popular”, then think again. FreeBSD is as important as other Unices out there, which do make headlines from time to time, plus it has more userbase/installbase than Lindows, Lycoris (and even BeOS back in the day). Heck, even Bitstream donating 10 puny fonts made headlines around the web, while the 5.0 release didn’t.
The difference here is the PR/marketing people. There is no doubt in my mind that the news regarding FreeBSD should have been big on all these tech sites, just because FreeBSD does have the required installbase to look important. The problem, I believe, does not only lie in the editors of these news sites who aren’t able to distinguish the important news from hype. The blame is also on the FreeBSD project which didn’t release a registered press release via newswire. At least I didn’t see one (no, this is not a PR). Even the KDE, Gnome and other OSS projects have “PR people” who roll out press releases. In fact, KDE is issuing press releases even for betas, because this greatly help with the project looking responsible to the media and to its users.
I remember that in the past the FreeBSD project had a warmer reception from the media, I believe that the project needs to work on their communication skills a bit. The fact that “people work for free, for fun” doesn’t excuse the fact that they could do better and ask for some exposure. More exposure, means more users and developers, which eventually means more life and evolution for the project itself (more explanation here regarding this).
Since Jordan Hubbard left to work for Apple, the project seems to many people a bit isolated. FreeBSD has lost its “leader” and its main spokeperson. I believe that this is also one of the reasons behind this “black out of FreeBSD news”.
No, FreeBSD is not dead (TM). It just needs to work on its communication skills and get the word out. And of course, it also needs to overturn somehow the whole Linux hype engine, which could be blocking it from “looking as important” in the eyes of the net population. And these editors at these news sites, should re-evaluate as to “what sells and what not”.
FreeBSD has NEVER been about promoting itself, ever. Yes it doesn’t get a lot of press but it doesn’t need it. FreeBSD just rolls on without hype.
As for Jordan K. Hubbard, he’s still very much involved with FreeBSD and FreeBSD advocacy. I talk to him often on #opendarwin.
Jordan K. Hubbard does send out press releases but he finds their generally ignored until days later, so he doesn’t see the point.
>FreeBSD has NEVER been about promoting itself, ever.
Well, in the past, previous FreeBSD releases _WERE_ reported by these sites. THIS release is very important and a technological treasure, and it went unoticed. IMO, this is unfair. And I believe that every project needs promotion. It is the only way it can go on and get many new users, devs and companies interested enough to donate hardware or code. The way I see it, being “popular” is imperative when your project “employes” more than 2 coders and needs support on specs, hardware and drivers.
My beef here is that companies like Be, which had less userbase than FreeBSD actually were making headlines with the help of their marketing/PR dpt. Today, we see that with Lindows. FreeBSD is used by more people, therefore, it should have its representation on the media when we are talking about such an important release.
And as I explained, love it or hate it, it is imperative to have articles on news sites. People read news sites so they can get the news. By isolating yourself, you gain _nothing_. It is not about “promoting” if you must. It is about not “isolating”. Think about it this way.
I misunderstood your remarks. I see what you are saying. I stand corrected, thank you.
What the heck is FreeBSD? Why do we need another Linux distro?
J/K , FreeBSD is like a middle child. The oldest (Unix) and the youngest (Linux) gets all the attention.
Excellent writeup, Eugenia. And you definately hit the nail on the head on one issue: FreeBSD and PR. Linux has many large corporations/entities surrounding it (like VA) who can handle the PR for them. Furthermore, Linux has tremendous mindshare compared to FreeBSD, to the point that your average joe on the street may have heard of it or will recognize it when you say it.
And of course, it also needs to overturn somehow the whole Linux hype engine, which could be blocking it from “looking as important” in the eyes of the net population.
The problem here is that FreeBSD users don’t tend to be quite as zealous as Linux users. I certainly don’t see anyone out there preacing about FreeBSD taking on Windows on the desktop.
Despite its almost complete lack of PR, FreeBSD has garnered some commercial support, both from NVIDIA who released FreeBSD native drivers and Opera who made a native release of their browser. I hope there’s more of this to come in the future, particularly from database companies. Until then, you’ll see places like distributed.net running Sybase through FreeBSD’s Linux compatibility.
One thing that I am sick of to no end is the population of Linux zealots who seek to downplay every major accomplishment FreeBSD makes. The 5.0 release was met with a slurry of comments about how SMPng is still nowhere as good as Linux. Well first, not everything is a competition with Linux, and regardless of how it stacks up against Linux it’s still a major technical accomplishment, and I think the FreeBSD developers deserve respect for that if nothing else.
Perhaps the worst is when features developed by the FreeBSD developers on their spare time are compared to features developed by corporations. I see many comments along the lines of “UFS2 is good, but it’s nowhere near as good as XFS” Well obviously, SGI has poured a considerable amount of money into developing XFS, so I would hope it’s better.
So, I hope all the Linux zealots can respect FreeBSD as an impressive technical accomplishment if nothing else, created in the same open source spirit which spawned your beloved Linux.
It’s true, I think FreeBSD (or all *BSD’s for that matter) do not get the recognition and headlines they deserve even though I am not currently a FreeBSD user myself (ran it for “educational purposes” for a few months and plan on doing so again after 5.x becomes more stable). One of the reasons could be that the Linux people and the *BSD folk are usually quite different. With Linux, we often see this “Linux World Domination !” etc headlines, because many Linux users tend to actually think Linux SHOULD / MUST be used everywhere. The *BSD folk just silently hack their code and use what they think is best for the job, be it FreeBSD, Linux, MacOS or Windows.
GetOutofHere: the first line of your post made me jump
I just installed 4.7 on an old box because I had some trouble with 5.0. It installed quickly and didn’t take up too much space. I managed to configure X after about an hour (which is better than my previous experiences using xfconfig). The box responds great and has all the utilities I need for a while. My redhat8.0 box at home OTOH is acting weird and rpm isn’t working. I had to install some software from tarballs because of it. When my freebsd chops are better I will use it as my main box. More stability and less unpredictabilty.
Still MS free after 2 years and counting!
“I am truly dissapointed though on the way the media have responded to this release.”
That’s because FreeBSD 5.0 is truly dissapointing. The project is actively discouraging people from using it on production systems. I’m using it right now and it is pretty buggy. It was released way too early. It simply isn’t ready yet. But like Microsoft and Windows 95, they didn’t dare delay it again, so they released it even though it really isn’t ready.
I think there is a reason the media hasn’s responded much to it. And I think the FreeBSD project wants it that way. It’s a disapointment. They want a nice quiet release that people don’t use for production so that they can hopefully get all the bugs worked out for 5.1.
If I were to ask the sales guys in my company about Linux, they’d have heard of it. Same can’t be said for FreeBSD – or any BSD for that matter.
The only thing they’d be able to tell you about Linux, however, is that it is competition to Microsoft, and maybe that it’s free. This is what Linux is known for to the general public. It’s the anti-Microsoft.
That’s not how FreeBSD or its supporters have come across with FreeBSD. Without having a frame of reference, the “average joe” has no clue what the heck the difference between BSD, Linux, IRIX, HPUX, Solaris, or any other *NIX would be. The only reason they can even grasp Linux is that everything they read/hear about it is that it’s free and it isn’t Windows.
You see BSD news on sites like OSNews and Slashdot because those sites are populated by developers and tech-heads that actually USE technology. Sites like ZD have a readership that is much broader, and less tech-centric. Trying to explain the difference isn’t what they’re about. They want readers, and in order to get them, they need to throw around the latest buzzwords.
Reminds me of something a professor told me back in college in my TV production class. He told the class that TV programming was simply the filler between the commercials. TV news wasn’t different, it was a program.
I think web news is the same – sites like ZD don’t want to take time to *teach* their readers anything. They just want to throw them what they think will keep people around.
>I think there is a reason the media hasn’s responded much to it.
Trust me, these guys at these news sites aren’t going to run FreeBSD 5.0 and see if it is buggy or not. They would only care to report the news.
>They want a nice quiet release
That was wrong. They shouldn’t have release then.
> Without having a frame of reference, the “average joe” has no clue what the heck the difference between BSD, Linux, IRIX, HPUX, Solaris, or any other *NIX would be
The part you are wrong is that these sites DO report on Solaris, AIX etc. Simply because these companies have PR dpts that do their job. FreeBSD just doesn’t play the game exactly right. But as I said, it is not just BSD’s blame, the journalists are also to blame in this instance.
” Trust me, these guys at these news sites aren’t going to run FreeBSD 5.0 and see if it is buggy or not. They would only care to report the news.”
Well, what I am saying is that I think the FreeBSD project intentionally didn’t promote the release very well since I don’t think they want people using it for production yet. I think they wanted it to be relatively quiet.
I’ve been running 5.0 since 2001, and I had a lot of hopes for it. But the Release really failed to meet my expectations. I’m just not impressed with it at all.
That’s because FreeBSD 5.0 is truly dissapointing. The project is actively discouraging people from using it on production systems.
Perhaps you’re unfamiliar with the FreeBSD release cycle, but that’s how it always works. It will remain tagged -RELEASE until a concensus is reached that it is ready for production use, at which point it will be tagged -STABLE.
I’m using it right now and it is pretty buggy.
What bugs are you experiencing? My only ones have been lockups involving background fsck on a UFS1 vinum array. Upgrading the filesystem to UFS2 solved the problem.
hmm… speaking of lycoris/Lindows, i wonder if it is possible to implement the FreeBSD OS to make something easy to use as lycoris? perhaps that could bring in some people
“Perhaps you’re unfamiliar with the FreeBSD release cycle, but that’s how it always works. It will remain tagged -RELEASE until a concensus is reached that it is ready for production use, at which point it will be tagged -STABLE.”
A RELEASE is supposed to be relatively stable though. In my opinion, the RELEASE is less stable than some of the cvsup 5.0’s I was running in 2001.
As far as the bugs I have experienced, the biggest one is a ppp bug that I haven’t been able to track down yet. Basically, uploading a file trashes my entire Internet connection. It only occurs when I upload a file, and even after canceling the upload, I have to kill and restart ppp for it to recover. It happens with both kernel ppp and user ppp. I haven’t filed a bug report about it yet because I haven’t been able to isolate it. Sometimes it seems to happen 30 second into the file upload. Othertimes not for a couple minutes. I’m trying to isolate it better before I file a bug report about it.
The other ones are minor. Certain GNOME applets crash as soon as you start them (why are they installed by default and placed in the menu if they are broken?), and so on.
Simba:
There is truth to what you say. 5.0 is not production ready and has some serious bugs. I don’t think they want to publicize it too highly beyond their standard userbase until 5.1 or later.
Eugenia:
You say they shouldn’t have a released if they want it quiet. 5.0 was originally supposed to be released in November of 2001, and was already way, way behind. It has been made clear that this is an “early adopters” release. In other words, it’s a beta that’s been labelled as a release in order to get more people using it and more bug reports. It’s certainly a non-standard way to go about releasing a piece of software, but they were also in a bind, and I think what they did was necessary.
A RELEASE is supposed to be relatively stable though. In my opinion, the RELEASE is less stable than some of the cvsup 5.0’s I was running in 2001.
Well, I too have been running it since mid-2001, namely for the ACPI support (I have an Athlon MP system… without CPU throttling it tends to produce an inordinate amount of heat)
While I can’t attest to your particular problems, I can certainly attest to the locking bugs. 5.0 was plagued by the locking bugs, and I watched them slowly diminish. By 5.0-RC1, I wasn’t seeing a single lock order reversal notice, and haven’t since. I’d say for an open source fine-grained locking project they did an excellent job debugging it, although it did result in a several month delay.
What you wrote doesn’t much sense for my arguments. You see, if the FreeBSD project needed more testers they should have actually write up the PR and caution the users about its stability. By calling something “-RELEASE” and not creating a PR is stupid. Just doesn’t make sense. Forget the fact that it shouldn’t have been labeled as -RELEASE if it is indeed unstable. But from the moment they did a -RELEASE release, it SHOULD have accompanied by the PR. This way they should get more testers anyway if the news were going to be picked up from the big news sites. And then add the caution statement in the PR and you are all set. But by not creating the PR, and by not trying to get more recognition, to me that shows isolation and bad management. And from the other side, we have the journalists who don’t get it, and as a result we have FreeBSD losing momentum. And this is bad in the long run for the project as I explained (less users/devs/specs/drivers/companies/code)
Most of what Simba is saying is actually true. There’s a reason why we’ll see a 4.8 “Production Release” some time next month, even if the “New Technology Release” 5.0 is out.
5.0 isn’t considered stable, but the decision was made to release it anyway to solve the chicken and egg problem (they need more testers, but to get more testers they have to make a release). They explicitly tell you to not use it on production systems, and are very honest about it.
I think they made the right decision, and I think they’re handling it very well so far.
“Well, I too have been running it since mid-2001, namely for the ACPI support (I have an Athlon MP system… without CPU throttling it tends to produce an inordinate amount of heat)”
Ah… Thanks for reminding me. That’s another one. The ACPI support is pretty buggy.
Example: shutdown -p will turn off my system, but then I can’t turn it back on again without turning off the power supply for like 15 seconds. The first time that happened, I thought my power supply had fried.
What’s sad is that my Java instructor asked everyone in the class what they run on their computer at home. The usual answers were Windows, Mac, Linux. When I mentioned that I had FreeBSD and Windows he had never even heard of it. Hmm…
In fact, if what they wanted to do is get more testers but not too many testers, this release should have been called FreeBSD 5.0-RC4 and *make a PR about it*. This way, they will get more people, but not too much.
The way they did it now: -RELEASE without a PR, is just not right for the recognition of the project. In fact, I see this as a mismanagement. When stable FreeBSD 5.1 will be released, the news journalists will think “oh, it is just a point release, we won’t report on it”, even if that 5.1 is the “real” release. There is definately mismanagement here by the FreeBSD project regarding its recognition.
“But by not creating the PR, and by not trying to get more recognition, to me that shows isolation and bad management.”
Well, possibly, this is to be expected at this point. After all, ownership of FreeBSD was jacked around 3 times when 5.0 was being developed. First BSDi, then Wind River, and now no one. And of course, Jordan Hubbard got a lot less active during this as well. I think the project is still in some turmoil and still trying to adjust to all of that.
“And from the other side, we have the journalists who don’t get it, and as a result we have FreeBSD losing momentum. And this is bad in the long run for the project as I explained (less users/devs/specs/drivers/companies/code)”
I don’t agree that FreeBSD is losing momentum. The FreeBSD IRC channels and such are more active than ever–even more active than the Linux channels at many times. And we just got a native Opera port for FreeBSD, and nVidia released GeForce drivers for FreeBSD. We got SUN’s blessing on volunteers doring a native port of the Java SDK for FreeBSD, and Open Office, Abiword, and some other OSS projects have started including FreeBSD binary packages along side their Linux ones. The project is gaining momemtum, despite some very poor management at the present time.
why is it that linux userbase can’t shut up about it? this has been happening for years back when linux had no technical merit but the userbase kept going “linux is so great linux is so great”
to anyone who says linux is a UNIX you are WRONG linux does not share the same internal design and interfaces that are common to all UNIX OS. if u don’t believe me look at XFS and SGI “page buffer” they add to linux to make XFS code compile against linux kernel. linux filesystem code was too crappy and they couldn’t link XFS code against it directly. this would not be the case for freebsd, since it uses similar filesystem code to IRIX.
now linux is finally getting better than the alternatives, but that’s because linux users couldn’t shut up about it so everyone eventually heard about it and how great it was. even though it wasn’t great companies decided to come along and fix it up like SGI and IBM and now it is much better OS, not because of open sorce but because of corporate support! freebsd is still going strong with no big companies helping to write it, it is better open sorce posterboy than linux!
simba why u use shutdown instead of halt isn’t you on a single user system? halt is faster
Another tell-tail sign that I think the project is gaining momentum despite poor managment at the current time…
Publishing companies have taken notice of FreeBSD. In 2000 there were only two books in print on FreeBSD. Today there are 7 and an 8th one is on the way. FreeBSD has gained enough momentum that it can support a book market. That’s a good sign.
“simba why u use shutdown instead of halt isn’t you on a single user system? halt is faster.”
Because I admin multi-user systems. And using “halt” is not even a habit I want to get into. If I do it on my home desktop system often enough, sooner or later I will probably accidently do it on one of my servers that currently has 100 users logged into it or something.
Basically, “halt” is just a bad habit to get into unless you exclusively use FreeBSD on a single user system.
Example: shutdown -p will turn off my system, but then I can’t turn it back on again without turning off the power supply for like 15 seconds. The first time that happened, I thought my power supply had fried.
The exact opposite of this bug appeared for me in RC1 (i.e. I couldn’t power my system back up from hard off, only soft off) It was gone after RC2.
In fact, if what they wanted to do is get more testers but not too many testers, this release should have been called FreeBSD 5.0-RC4 and *make a PR about it*. This way, they will get more people, but not too much.
I don’t think FreeBSD even considers PR, they’re focused on releasing the best operating system possible. I think a big problem is they don’t have anyone to handle PR on their behalf.
linux is finally getting better than the alternatives, but that’s because linux users couldn’t shut up about it so everyone eventually heard about it and how great it was. even though it wasn’t great companies decided to come along and fix it up like SGI and IBM and now it is much better OS
Linux has a long history of corporate support, not just from SGI and IBM, but all of the commercial distributions such as RedHat and Mandrake, and also the entire VA powerhouse. They have multiple companies with PR departments to field questions from the media. FreeBSD, lacking any real corporate support at all, has nothing like this.
Solaris and AIX get PR because they’re the products of large corporations. What press agency is going to take FreeBSD seriously when it’s an open source project no one has even heard of?
when pitched against M$, linux would be described as a unix like thing – refering to the proven stability
when compared to other *nix – linux isn’t unix at all
[to anyone who says linux is a UNIX you are WRONG ]
hehe, is Linux POSIX compliant ???? does linux offer fork, pipe, tty line discipline , SYSV IPC, like shm ?
I guess part of the bsd trouble is that major unix out there is aligned along SVR4 – Sun is the last commercial unix house moved to SVR4 – I could be wrong though
FreeBSD primarily aims to be a server OS. How much interest can you generate with that? Trust me, those who need FreeBSD already know about the new release.
Linux is going everywhere -servers, mainframes, embedded, and, most notably, to the desktop. Truly or falsely, Linux is generally seen, and aims to be, a potential competitor to MS, on the the server and the desktop. Its like David against Goliath. The press love that kind of story.
>> And of course, it also needs to overturn
>> somehow the whole Linux hype engine, which
>> could be blocking it from “looking as important”
>> in the eyes of the net population
Why linux, why not Windows, or OS X, the otherBSD?
Eugenia, this part of the article is just flame-bait. How did you manage to drag linux into it? How is linux “blocking” FreeBSD from looking important, and why does FreeBSD need to “overturn the whole linux hype engine” to make itself heared? It doesn’t need to overturn nothing, it can damn well make itself heared alongside every other OS. You are just giving all these anti-linux zealots another flame-bait to bite.
“simba why u use shutdown instead of halt isn’t you on a single user system? halt is faster”
My name isn’t simba but I’ll answer your question. Halt shuts down immediately without warning for other users. If you have a multiuser system, it is paramount that you use shutdown to give the users some warning.
Eugenia,
It’s funny that you wrote this editorial, I went looking for RC2 and found the 5.0 Release iso’s. Ever Since Walnut Creek went down with the merger of BSDi then wind whatever, FreeBSD has lost an efficent delivery channel. I do use FreeBSD on my servers and on my laptop. 5.0 will have the bugs worked out, turn in to stable, current will go to 6x and then 4x will cease to be developed. This is how it has been ever since I downloaded 2.2.2 and replaced my NT3.5.1
“i wonder if it is possible to implement the FreeBSD OS to make something easy to use as lycoris?”
You know everyone talks about the version of OS X hidden away in cuppertino running on X86 hardware. I wouldn’t be suprised if Microsoft had a version of FreeBSD with a Windows like API and GUI stashed away in Redmond. Think about it, they don’t know how to deal with Linux right now, but what if they could give away a forked BSD system for users to hack on, and sold a GUI that gave the ability to run win32 binaries? Got to love the BSD License!!
Why linux, why not Windows, or OS X, the otherBSD?
Eugenia, this part of the article is just flame-bait. How did you manage to drag linux into it? How is linux “blocking” FreeBSD from looking important, and why does FreeBSD need to “overturn the whole linux hype engine” to make itself heared? It doesn’t need to overturn nothing, it can damn well make itself heared alongside every other OS. You are just giving all these anti-linux zealots another flame-bait to bite.
Linux has effectively won the battle for Unix mindshare, and it is now seen as the platform to standardize upon. It won this battle not on technical merit, but through hype. I think you can see how this might upset Unix hardliners, especially considering Linux’s non-pedigreed roots.
but I just don’t get it. Every time there’s something new about *BSD or Linux on OSnews, a lot of posters start to whine about how much “better” *BSD is, and how all Linux users are zealots, and how much better *BSD users are since none of them are zealots and none of them ever whine about anything. It’s just that all of those *BSD users (if they really use *BSD that is) sound like moronic twats, no less zealous than newbie Linux gits. Choice of OS is not a competition of who’s got the biggest and fastest VM, is it?
But on to the subject of the thread, as this isn’t meant to be just a flamebait. FreeBSD isn’t quite as much a hobbyist OS as Linux. When Linux reached 2.4.0 it was buggy as hell (and 2.4.20 still has it’s bugs, although not as many, and some are replaced by new bugs), but it meant all the hobbyists could upgrade without risking their FS (hopefully), and start doing some real testing. Of course, many of the hobbyist FreeBSD users are going to do the same with 5.0, but it’s not a big deal for the serious users (Yahoo et al). It’s not just a question of mind share, but also where that mind share is. Linux is used by people who install Mandrake and admire the nice colours of the latest KDE (in addition to those who actually need a unix or server of some sort), and is therefore news to the average nerd. FreeBSD isn’t as often used as a plaything, in my experience.
I use FreeBSD 4.7 on my laptop, because it’s slow (133 MHz), and I prefer something stable that just works well. I’m not going to upgrade it to 5.0, and probably not 4.8 or 4.9. As long as it does what it’s supposed to do, and is free from glaring security holes, I’m happy. If I was just fooling around with it, I’d upgrade to 5.0 already.
“I guess part of the bsd trouble is that major unix out there is aligned along SVR4 – Sun is the last commercial unix house moved to SVR4 – I could be wrong though”
FreeBSD has kernel support for SysV IPC and such. So this generally isn’t an issue. I believe the support is enabled by default. But if it isn’t, it is a simple matter of adding an options line.
FreeBSD 5.0 also made major strides with the POSIX compliant. For example, it now supports p-threads natively.
And as far as POSIX compliance, IMHO, it’s overrated. Windows NT/2000/XP is also POSIX compliant. But that hasn’t made porting applications between Windows and UNIX all that much easier.
“simba why u use shutdown instead of halt isn’t you on a single user system? halt is faster”
My name isn’t simba but I’ll answer your question. Halt shuts down immediately without warning for other users. If you have a multiuser system, it is paramount that you use shutdown to give the users some warning.
I think he/she was aware of this. That is why he/she said, “[aren’t] you on a single user system?”
The people who use BSD; know about it. BSD isn’t going to adopt new users for the desktop or any other place. I don’t see the need for news to have been reported. The major technological advances aren’t for anyone on the desktop; just the need for PR in this situation is not needed.
Some big server advances were made in Freebsd considering the way the project has been going. Things are starting to evolve and the team seems to be trying new and different things. This is good but again it’s definitely not worth a PR release. The type of people who are gonna use freebsd still will use it and find out about it etc etc. You seem to forget that before all of these news sites people found out about it and will continue to do so.
[And as far as POSIX compliance, IMHO, it’s overrated. Windows NT/2000/XP is also POSIX compliant. But that hasn’t made porting applications between Windows and UNIX all that much easier.]
POSIX has over a dozen levels, NT was only at the bare minimum level to be gov. friendly.
Personally, I have a lot of respect for all open software developers, be it GNU, freebsd/netbsd/openbsd or linux – POSIX or not, pthread or not doesn’t matter – as there are limitations, there will be workaround, future improvements; What I don’t like is the fact that some linux users just can’t admit any existing limitations with their holy OS/platform.
1) Windows 2000/NT had POSIX libraries available, however, from what I have read, they weren’t too savoury. In Windows XP and .NET, POSIX compliances has been removed and instead Services For Unix is offered as an “add on package”. I assumed the move was because POSIX was used by such a small amount of people there was no justification of maintaining it.
2) IRIX is a BSD based kernel, IIRC, based around 4.3BSD, however Linux is mainly pointed in the SYSV direction as with the case of Solaris, UnixWare and HP-UX. As for the differences between BSD and Linux, the issues with Linux are more to do with the fact that it is still moving towards UNIX compliance. Sure, glibc is UNIX 98 compliant, POSIX compliant etc etc, however, in the kernel level there are a still areas that need to be fixed.
3) IIRC, KSE and SMP are still not finished yet. Now, unless the page http://www.freebsd.org/kse is horribly out of date, native posix threads have not been implemented yet. So, unless you can point out otherwise, KSE and POSIX threads are still no in there.
4) I use FreeBSD 4.7p3 and personally I don’t really care want the Linux fan club want to do. Yes, the Linux community is filled with zealots, just like the FreeBSD, BeOS and OS/2 communities. If you talk to the *real* users and developers of the OS, they barely say anything about Microsoft. When I refer to the *real* users, I am talking about the people who actually USE the operating system, not those idiots who sit behind their Windows XP computer spitting and cursing on IRC about “how bad Microsoft is” and how “[operating system] roolz” even though they have never used the bloody operating system in their life.
is it such a bad thing that it didn’t make as much noise as the big players do these days?? Those who care about freeBSD certainly were aware of the news. In my opinion, it’s okay to fill that particular niche, make that much noise and just deliver the quality product you’ve been known to always deliver. FreeBSD kicks ass, but it’s more focused than linux. therefore I would recommend it to anybody who wants the best OS for [insert server need here], but I wouldn’t install it on my girlfriend’s desktop. Maybe I’m a little conservative here, and maybe freebsd really needs a boost, but I like to think my ideal of small but efficient is still a reality in a world where the hype is all about what the corporate world wants to hear.
Best release ever my arse, X doesnt work on my laptop, how come all the way up until now it has but 5.0 doesnt, best release ever… LoL
I remember the first time I come acress FreeBSD as about 8 years ago, and that was because our school had it as a mail and internet sharing server. That was also the first time I come across a unix-like OS (*BSD OSes can’t call themself UNIX OS because of some trademark issues, as far as I know, which is why they’re always called BSD). But then I stopped using it when I left school and switched to Linux.
Now, I’m a RHCE, but also what to get a feel of FreeBSD again. I actually downloaded the ISOs on Saturday (the first day of release)…but it truely is said that not many sites had talked about it. It’s such a big step for 4.7 to 5.0 in terms of SMP technologies. I truely believe that it deserves some proper reviews…
Personally, I also get the feeling that FreeBSD is suffering from the Gnu/Linux hype.
Everybody is jumping on the Gnu/Linux train, even Big Blue and Sgi although it’s own Irix is awesome.
BSD is a far better server OS than Gnu/Linux is, despite of what all the Gnu/Linux gurus are yelling.
Besides, if you install a Unix like os, why not
install a -real- Unix like BSD ! 😉
Concerning using it as a desktop OS, I think both Gnu/Linux and BSD are ok.
But what about all the other OS’s mentioned here on OsNews ??
I see quite a lot good OS’s around here and they don’t get media coverage at all.
(Except on OsNews.com ofcourse 😀 )
Oh, and to come back to the desktop OS :
I still think BeOS was one of the most promissing desktop OSes ever.
It’s a shame that Be went out of business. Luckily http://www.openbeos.org is doing everyting to keep the good things alive.
I thinks the Fonts-Issue is very important, because in the end i really decide if i like a distro by how good the standard fonts look like.
To be honest, this boggles me…
I’ve used a number of Linux flavours (Debian, Mandrake, Redhat and Slackware) but for the life of me, I find
FreeBSD alot more user friendly, easier to navigate around
and not so bogged down…
HOWEVER, This isn’t my issue as to Why Linux?, the thing
that makes me truely wonder, is that companies like
Mandrake, Redhat, Lindows(now), IBM, HP etc.
and others are using Linux in the first place and not
BSD.. I mean, compare the licenses for goodness sakes, which license is more commercially viable in the world
of “trade secrets” and wanting to keep your products
“closed source”… No wonder Linux companies come and go like the weather.
I remember two or three years ago, there was an article on one of the main news site (ZDNet? Cnet? Can’t remember) about FreeBSD, the best kept secret of the Internet.
Apparently, this is not going to change anytime soon!
FreeBSD 5.0 features SMP and MT technology not found on other OSes.
Hardly. Not only are SMPng and KSE not done yet, but even when they eventually get there, what performance they will bring to the table is still a bit up in the air. I regularly skim almost all the FreeBSD mailing-lists and I have yet to see any benchmarks.
FreeBSD is as important as other Unices out there, which do make headlines from time to time, plus it has more userbase/installbase than Lindows, Lycoris (and even BeOS back in the day).
In the server space, certainly. On end user desktops, I’m not so sure. It was never something they focused on in the first place (“FreeBSD: The Power To Serve”, remember), and most of the things that has made FreeBSD a viable desktop system has been things that has happened because of the end user focus of the Linux camp (KDE, Gnome, MPlayer, and so on) and spilled over to FreeBSD.
Since Jordan Hubbard left to work for Apple, the project seems to many people a bit isolated. FreeBSD has lost its “leader” and its main spokeperson.
Indeed. Not only was JKH really their only public face, he was also very good at it.
“Many argue that this release is the most important and most technically advanced, ever, for FreeBSD.”
That’s great.
“…the whole Linux hype engine, which could be blocking it from “looking as important” in the eyes of the net population.”
Of course, blame it on Linux, not to FreeBSD. The evil Linux zealots keep FreeBSD from breaking through, everybody knows that.
“I don’t claim to reboot to it more than once or twice a month on my personal box.”
Why not?
FreeBSD isn’t user freindly, Linux is at least almost user freindly. If FreeBSD could boot people to a GUI right off the bat, people would use it more, but it doesn’t, so people don’t keep it on there system long. That is my stumbling point EVERY time, its just a bitch to set up… I don’t want to spend 4-5 hours after an install JUST setting stuff up, I don’t mind tweaking over time, but after just spending an hour or so setting up an OS, I don’t want to spend more time staring at text based config files, and I don’t want to see Xfree86’s GUI at any point, I want to see the system boot up to KDM or GDM, and I want the system to be usefull right off the bat. Power to customize is great, but hype comes from users experiances, FreeBSD’s user experiance is UGLY!
And another thing, the BSD SGI license allows companies to STEAL from each other, look at DirectX and Quartz, both are based on OpenGL. Why would companies want to invest in such a license? I know personally I am more likely to release the soruce if I know it will make MY code better then knowing someone could just take my code and change it around a little, then make MORE money then I did!
.
Why can’t people see this? Why would anyone think that people want others to make more money then them for THEIR code? This is stupid, and so are BSD and SGI type licenses. Sure they are better for proprietary OS’s, but it sure isn’t better for the BSD’s or OpenGL or X Window System etc. I really don’t understand people that think its better… stealing code is the very reason people hold back code, not the reason people would release it…
.
The GPL is the EXACT reason why Linux is so popular, its not usability, its not features, its not anything else that people have said, its the GPL, and what that means for developers and corporations alike. Its the possiblility that maybe, just maybe, the Computer industry can be FORCED to work together, cuz it sure wasn’t happening when they were given the choice…
.
DirectX is probably the one reason MS is so liked, now MacOS x has proprietary code based on it too, they were both able to STEAL, this is wrong, people shouldn’t be able to steal from others in any society, borrowing, then giving something back in return (better code), the computer industry can gain alot more from.
Well first, not everything is a competition with Linux
In what space is Linux not a competitor to FreeBSD? Does it currently occupy some niche that Linux isn’t even trying to get into? If not, everything _is_ a competition with Linux. The same isn’t true the other way around though, as the current Linux space is a superset of the FreeBSD space. The superset being the low low end (though there are other BSD players there), the high end, and more supported platforms.
Perhaps the worst is when features developed by the FreeBSD developers on their spare time are compared to features developed by corporations. I see many comments along the lines of “UFS2 is good, but it’s nowhere near as good as XFS” Well obviously, SGI has poured a considerable amount of money into developing XFS, so I would hope it’s better.
Yes, looking at the amount of resources put into each of the systems, it’s not really a fair fight. But again, this is what FreeBSD has to compete against. Are you suggesting that people should use FreeBSD out of pity?
hmm… speaking of lycoris/Lindows, i wonder if it is possible to implement the FreeBSD OS to make something easy to use as lycoris?
Of course it’s possible. Problem is, nobody has done it. Also, such a system would not be allowed to call itself ‘FreeBSD’. Heck, the FreeBSD Foundation doesn’t even own the FreeBSD trademark. It’s currently owned by Wind River.
the german newsticker heise.de has a comment on FreeBSD 5.0 beeing out:
http://www.heise.de/newsticker/data/jk-19.01.03-000/
and it even made it into the category of the four “currrently most important” subjects!
Hiya
You don’t seem to understand the idea of the BSD license. If you use BSD code in your project and you give the required notice, then you are not stealing! Rather than everyone develop their own tcp/ip stack, system libraries, daemons etc they can use a tried and tested implemetation. Code I’ve written is in 5.0 and I’m completely happy for people to reuse it as they see fit.
On the subject of Linux hype, I’m not sure why it attracts so much but I think it does play a big part in Linux’s rise to fame. In my experience, the less technical kmowledge a linux user has, the more he hypes its abilities. For instance, I used to work with a couple of real Linux fanatics and I used to cringe at some of the preposterous claims they would make about it 🙂
–Jon
why Linux – probably somebody knows there are lot people hate M$, and the very idea of the BSD license would turn those guys off in considering writing code.
But personally, I like the BSD idea better – if say M$ made a lot of money on DirectX, they obviously did some work that people want to pay for it – after all, M$ doen’t bundle major games with Windows; Game companies can write software based on DirectX, instead of writing their own DirectX equivalent and the games – consumers are able to enjoy these technology advances faster.
BSD ppl go on and on about how their source is sharing. The GPL is not about that sort of sharing. It is about keeping the sharing pure. This is the difference that makes Linux more popular.
Why should the popularity of Linux be so encumbering to BSD if BSD is better? Maybe there are different sorts of better like a better liscence.
The easiest concept in this world, when we fail to do something, we can just simply blame others. Many company blame M$ when they failed. Now FreeBSD ZEALOT seem to blame Linux hype that lead to their some sort of failure.
For me, if FreeBSD (or other OS) can at anytime overtake Linux if they can become a better alternative. If the target is the end user, the technical advance alone is valueless but the user friendliness is more important. And if the Linux hype is strong, just create a stronger hype which require what was highlighted byEugenia in the articles, PR works!!!
You mean Linux
WON’T run on multiple networked Cray T3E computers?
WON’T save our computers from “evil Micro$haft”?
ISN’T as easy to use as MacOS or Windows?
WON’T wash my dishes, fold my clothes, and take out the trash?
WON’T stop the US from going to “war” with Iraq?
WON’T save Earth from a a collision course with a rogue meteor?
WON’T enlarge your penis or increase your breast size (safely and naturally)?
Man, Linux sux 😉
–JM
I wish I wish I wish someone would do the knowledge and do a technical comparison of FreeBSD 5.0 and Linux 2.4. Compare threads, compare SMP, compare net IO, disk IO, extra features along those ip stacks…?
i wish someone had done it with FreeBSD 4…
there are plenty of people out there who are intelligent and want to know WHY people say FeeBSD/linux is more stabel or faster or better for a task…. than be told to just believe it.
taz
why is it that people are talking about a user friendly FreeBSD/Lunux… sureley they actaully mean desktop environments and the associated apps (email, browser)… nothing to do with Linux or *BSD?
Ease of installing software – again thats the role of a software package mgt tool thingy – again something that doesn’t depend on FreeBSD or Linux?
Hardware detection/compatibitlity? …well why don’t these folks just say so… instead of pouring out yet more volumes of “user-friendly desktop-ready corporate-desktop task-oriented”…
(personally i like the UNIX permissions thing – it stops errant employees adding palmpilots, cameras and silly hardware toys against company policy … the only things that work are are set out bny the sysadmin)…
t
I went to try out freebsd for the first time today, however I couldn’t install it because the installer locked once it loaded the “gui”. Ahh well, i will have to try to get a hold of 4.7 i guess…
BSD ppl go on and on about how their source is sharing. The GPL is not about that sort of sharing. It is about keeping the sharing pure. This is the difference that makes Linux more popular.
Not really. Do you think TCP/IP and The X Window System would be de-facto industry standards today if they were released under GPL? Would SUN and DEC have used them if they had to release the sources to their X-servers and kernels? Hardly.
The answer is:
Why the articles was written to highlight about the lack of news coverage on FreeBSD 5.0 release?
Maybe not many have interest in FreeBSD, except advance user like you. Maybe that is why not many news site wrote something about it since it didn’t get to their user attention.
User friendly aspect will attract more less advance user to FreeBSD, meaning more will get interested in any news regarding FreeBSD.
Hey, you’ve worked with those guys too? 😀
–Jon
When there’s mention of “media” attention, are you regarding the tech sites/magazines or television and newspaper? There’s a huge difference here in coverage and Linux gets both.
I think most users who are interested in using FreeBsd already know about it and look regularly enough for updates ( incl. existing users and “newbies” ). In this sense it doesn’t really need tech media coverage – enough popular sites talk about it – OSnews and Slashdot for instance.
Linux, however, gets the added benefit of coverage in newpapers and business/informational magazines ( Fortune & Time for instance ) when new kernel relases happen, Red Hat is updated, or some “big change” is imminent. FreeBSD doesn’t get this coverage. Big difference audience wise.
“FreeBSD isn’t user freindly.”
This is a myth. FreeBSD is more user friendly than Linux. It’s easier to install software, easier to rebuild the kernel, easier to use USB devices, and so on. The only basis for this claim is that it might not be quite as easy to install. But once it is installed, admin and upkeep is easier than in Linux.
“f FreeBSD could boot people to a GUI right off the bat, people would use it more, but it doesn’t, so people don’t keep it on there system long.”
Um…. How hard is it to type “startx”?
“That is my stumbling point EVERY time, its just a bitch to set up… I don’t want to spend 4-5 hours after an install JUST setting stuff up”
It took me 30 minutes… INCLUDING waiting for the install to finish. I don’t know how people can honestly claim FreeBSD installation is difficult unless they are basing the claim on their experiences with 2.x or something. Sysinstall gives good prompts, it walks the user through it, and explains concepts that might be foreign. It really is not that difficult.
“I don’t want to see Xfree86’s GUI at any point, I want to see the system boot up to KDM or GDM.”
What do you consider XFree86’s GUI to be. It doesn’t have one.
And once again, how hard is it to type “startx”? Not having KDM or GDM available as soon as you finish the install is not a very good reason for throwing out the entire system. And besides, it’s not that hard to make FreeBSD start KDM or GDM on system boot. (In fact, its easier than on Linux because the XFree86 docs tell you how to do it. XFree gave up on giving instructions for tellng you how to do it in Linux because it is different in different distros.)
“Power to customize is great, but hype comes from users experiances, FreeBSD’s user experiance is UGLY!”
Hype also causes short lived fads. In the end, power and stability are what stick around.
“BSD ppl go on and on about how their source is sharing. The GPL is not about that sort of sharing. It is about keeping the sharing pure. This is the difference that makes Linux more popular.”
Sure… By implementing a Stalinist agenda that forces people to use the GPL. That scares a lot of commercial software vendors away. Why do you think SUN, IBM, and HP went with GNOME instead of KDE? Despite the fact that KDE is a vastly superior desktop? (Easier to use, better widgets, more stable, Qt toolkit is easier to work with)? They did it because the Qt toolkit is licenced under the GPL unless you want to shell out over $2,000 per license. There is a reason GPL’d libraries are called “virus like”, because they behave exactly like a virus. If you link against a GPL library, even to use only a SINGLE class or function, your entire source code automatically inherits the GPL.
Why do you think Apple chose to base OS X on FreeBSD and not on Linux? Once again, the GPL. Because the GPL would have forced them to open source and GPL ALL of Mac OS X.
I am a programmer, and personally I won’t touch a GPL library with a 10 foot poll. (note distcintion between GPL and LGPL though. I don’t have a problem with LGPL libraries. Only GPL ones.)
“Victim” is the exact opposite. FreeBSD is being dragged kicking and screaming out of its isolation _by_ Linux.
There are people who use FreeBSD on their desktops. Would that be possible if not for Linux? Not really. Which companies have funded the development of KDE, GNOME, XFree, gcc and more. A hint: none connected with FreeBSD.
I always think it’s laughable that FreeBSD is supposed to be more commercially friendly, yet all the free software companies out there today are basically Linux based. Redhat, Ximian, theKompany, CodeWeavers.
nVidia released Linux drivers because their customers wanted them, and then a FreeBSD port came later because a few of the programmers thought it’d be cool. If the Linux drivers didn’t exist, neither would the FreeBSD drivers.
This isn’t because FreeBSD doesn’t promote itself, or because Linux is full of zealots. That’s utter BS. It’s because they have different attitudes. FreeBSD is full of contradictions – it’s supposed to be more “open” than Linux, yet they have private core teams, invitiation only. It’s supposed to be more commercial friendly, yet the only company that’s used it is Apple, and how much did FreeBSD get out of that? A few test suites and some minor patches.
Their attitude is all wrong. I once got into an argument with Jordan Hubbard – I wondered, what was the point of FreeBSD? It clearly wasn’t to ensure software freedom, otherwise they would use a license that allowed commercial hijacking. If the goal was to make a kickass OS, then why simply refine a technology that was state of the art 20 years ago? UNIX isn’t that great.
He replied that it didn’t matter, he believed everybody should be using UNIX, because it was better. Putting UNIX onto peoples machines wasn’t a means to an end, it was THE end, which is why he asked for an interview with Apple after they thoughtlessly didn’t hire him.
Linux is going places, whereas FreeBSD is being dragged behind it. That isn’t because “Linux is good at promotion”. At the end of the day, they are basically competing kernels, perhaps a few user land tools. They still use the same desktops, the same X server, the same compiler. It’s because actually in terms of attitude and organisation, Linux is better.
I have thought of trying FreeBSD, but I quite like nicities like graphical setup programs. I remember installing Slackware in ’95 and don’t really want to go through all that hassle again. I’m just not l33t enough, most people aren’t. Why bother? I couldn’t care less about VM performance, or the elegance of the kernel. Linux is good enough for me, and it has the support and userland stuff I like. Maybe when FreeBSD addresses these issues and opens up, then maybe it’ll take off in the same way Linux did (but maybe it won’t, why bother with 2 free operating systems?).
Sure… By implementing a Stalinist agenda that forces people to use the GPL. That scares a lot of commercial software vendors away.
Man, that’s rich. I’ll simply quote something I posted right here on OSnews a few weeks back:
“You know, it’s really quite comical how proponents of the BSD license keeps harping on about how business friendly it is, while out in the real world, Linux is getting all the contributions from companies like IBM, SGI, Veritas and so forth, almost all of it under a GPL license (some of it cross-licensed). I know that for instance FreeBSD-people inside of IBM has made a lot of effort trying to get JFS released under a BSD license in addition to the GPL, but to no avail. The GPL allows IBM and other businesses to make open source contributions without letting anyone who doesn’t want to play ball grab it and run with it. It’s obvious that the GPL, not the BSD license, is the one that best looks after their interests.”
Why do you think Apple chose to base OS X on FreeBSD and not on Linux? Once again, the GPL. Because the GPL would have forced them to open source and GPL ALL of Mac OS X.
Utter rubbish. First of all, it’s wrt the unix userland that FreeBSD is the reference platform (anyone claiming anything else, _please_ provide references). Second of all, they _do_ include a fair amount of GPL userland. For instance, from what I’ve read, the default shell in the latest OS X release is GNU Bash.
i’m having a blast with many of the unofficial ports. mainly the Hitachi SuperH (SEGA DreamCast) one. right now FreeBSD is confined to things that run in PC cases, and i’m still discovering all the goodness in other machines that can compute. NetBSD runs on a lot of things but won’t scale as high or as low as linux. so if i’m going to put a lot of effort into learning something, shouldn’t it be something i can take ANYWHERE(linux)?
in the end the *BSDs are just Unixes that hopefully will be made irrelevant by linux in the future. they have their uses, just like Amigas and Commodores do 🙂
I have never understood why people use numbers to form words on the internet. What the hell does “l33t” mean and how is it pronounced? Is this a new special language for people that refused to learn proper english?
elite -> eleet -> el33t (hax0r language) -> l33t.
dunno why, just one of those things net culture spits out every so often.
l33t (leet), is kind of difficult to explain. Hardcore is about the closest I can think of. We say, man that hacker is elite. Well “we” being used rather flexibly. I say it because I think it’s funny, and maybe even post-ironic, but that’s because i’m a pretentious git
More misinformation from the Linux crowd. FreeBSD is not limited to PCs. Look at the platforms it is actively developed for besides Intel-based computers: sparc64 and alpha, both serious computing platforms. Please get your information right before making yourself and the Linux community look foolish.
Is it me, or all that thread is more and more a perfect illustration of the topic?
“Man, that’s rich. I’ll simply quote something I posted right here on OSnews a few weeks back:”
It’s also absolutely true. The GPL forces programmers to GPL their source code, even if they use only a single function from a GPL library. That’s BS, and that’s why I and many other programmers won’t touch GPL libraries. It’s probably why SUN, HP, and IBM, woudln’t touch KDE and went with GNOME instead. Releasing your code as GPL is one thing. But being forced to release your code as GPL is another. (Side note: This is exactly why FSF came up with the LGPL license. Because they knew that licensing GCC’s libraries under the GPL would never fly since any application compiled with GCC would have been required to be GPL. They knew that commercial vendors would never go for it.)
“Utter rubbish. First of all, it’s wrt the unix userland that FreeBSD is the reference platform (anyone claiming anything else, _please_ provide references). Second of all, they _do_ include a fair amount of GPL userland.”
Sure. But the point is Apple didn’t want to be forced into having to release their own code under the GPL. If they had used Linux, they would have had to open source any part of OX that linked against any part of the Linux subsystem. Apple didn’t want to play that game. Neither did SUN, IBM, or HP.
Once again, If you WANT to GPL your code that is one thing. But being FORCED to GPL your code because it inherits the virus like GPL license from a library that you called a single function from is quite another. GTK uses the LGPL, which is acceptable to SUN, IBM, and HP, (and me). Qt uses the GPL, which is not acceptable to vendors or programmers that might want to write closed source extensions.
Once again, if you WANT to GPL your code that is one thing. But being FORCED to GPL your code because it inherits the virus like GPL license from a library that you called a single function from is quite another
What a load of crap. If you’ve got this application, which requires a single function from a GPL library, rewrite the damn function! If the GPLed function is so damn valuable that you can’t do without it (and can’t rewrite it), then it’s important enough for you to gladly accept the function, along with whatever license it’s using.
“And of course, it also needs to overturn somehow the whole Linux hype engine, which could be blocking it from “looking as important” in the eyes of the net population”
Boy you REALLY do hate linux don’t you? Why does the linux “hype engine” need to be fought?
“What a load of crap. If you’ve got this application, which requires a single function from a GPL library, rewrite the damn function!”
Yeah.. And that way I can contribute further to the dependancy hell that already exists for Linux and FreeBSD. “I don’t like this library, so I am going to write my own.” Now people have to install my library to use my application. And pretty soon, other people start using my library to write their applications because they happen to like it better than other ones that already exist. Now the end user gets even more confused. Because now they have to have:
GTK
QT
MOTIF
SIMBA-TK
all installed on their system. Linux dependancy hell keeps getting worse with time. I’d rather not contribute to the problem.
Reinventing the wheel completely negates the point of standard libraries.
BTW… I would be extremely surprised if that clause of the GPL is legally enforcable. It’s never been tested in court, but I suspect that it probably is not legally enforcable since all of the code the programmer wrote is original, and therefore, does not violate any copyright. The GPL as applied to a library basically tells me what I can and cannot do with my own intelectual property that I wrote. That’s legally dubious and may not fly in a court of law.
Actually you are free to change the license of your own stuff anytime. It just will be pretty useless when those are just some lines using further existing GPL’ed code to work. And when you use GPL’ed code you automatically accept and take over its license since that’s the license all programmers prior to you agreed on mutually for their code. So just don’t write code using existing code if you can’t accept its specific license.
“Why do you think SUN, IBM, and HP went with GNOME instead of KDE? Despite the fact that KDE is a vastly superior desktop? (Easier to use, better widgets, more stable, Qt toolkit is easier to work with)? They did it because the Qt toolkit is licenced under the GPL unless you want to shell out over $2,000 per license.”
Qt is triple-licensed under the GPL, QPL, and a proprietary license. GPL apps can use a GPL’d Qt; apps under other free software licenses can use a QPL’s Qt; and yes, as you said, the owners of proprietary Qt apps must pay for Qt. However, Trolltech, the purveyor of Qt, had long charged for the proprietary use of Qt, long before they even applied the GPL to its work.
Further, there are other issues to consider. GNOME used CORBA, a technology that Sun, IBM, HP, et al., were invested in. KDE did not. KDE is written in C++, which does not have a stable, compiler-independent ABI, so everything in KDE has to be compiled with the same compiler, and AFAIK, apps built against KDE libraries have to be compiled against that same compiler, which could be a problem when an ISV wants to support multiple versions of the same Unix.
“Why do you think Apple chose to base OS X on FreeBSD and not on Linux? Once again, the GPL. Because the GPL would have forced them to open source and GPL ALL of Mac OS X.”
For one, that statement isn’t even close to true, even on its face. Proprietary GUIs can legally run on top of the Linux kernel. Take Xi Graphics’ Accelerated-X, for example, or the CDE.
Apple’s use of BSD has more to do with history than the GPL, anyway. OS X is basically NeXT reborn. NeXT was a combination of BSD, Mach, and a GUI based on Display PostScript. OS X has inherited most of that; the main change was the shift from Display PostScript to Display PDF. (PDF and PostScript are close cousins.)
As long as it’s the brightest 1%.
I don’t need to see FreeBSD on all desktops to be happy.
And I’d rather see FreeBSD do better than the press we get than disappoint because we’ve been hyped.
Poul-Henning
“Actually you are free to change the license of your own stuff anytime. It just will be pretty useless when those are just some lines using further existing GPL’ed code to work.”
Incorrect. Suggest you read the terms of the GPL again. Once you have GPL’d your code, it is permanent. You cannot “unGPL” your existing code. I can unGPL future versions of the program, but I can’t change the license of the code I wrote that I already GPLed.
“And when you use GPL’ed code you automatically accept and take over its license since that’s the license all programmers prior to you agreed on mutually for their code.”
According to a clause in the GPL license. But like I said, it sounds to me that this clause is on very shakey legal ground since it is basically telling the programmer what he or she can and cannot do with intelectual property that they wrote. And if you look at the source code for such intelectual property, you won’t find any copyrighted GPL code actually in it. It just links to a GPL library. I question the legal enforcability of this clause. It will be interesting to see what happens if it ever gets tested in court. So far, it hasn’t been tested.
As I am sure you know though, there are many legal experts who question whether the GPL itself is enforcable. I’m not going to go there. But I do seriously question whether this particular clause is enforcable.
“For one, that statement isn’t even close to true, even on its face. Proprietary GUIs can legally run on top of the Linux kernel. Take Xi Graphics’ Accelerated-X, for example, or the CDE.”
OS X isn’t that simple. For example, it isn’t using a FreeBSD kernel. It’s using the MACH kernel. And Aqua isn’t running on top of the UNIX subsystem like X Windows does. It is an inseperable part of it just like the Windows GUI is inseperable from Windows. So yes, various parts of OS X that are proprietary ARE linked BSD libraries. If they had done that with Linux, they would have had to GPL those parts of OS X, as well as any dependancies on those parts of OS X. GPL is like a bad gene that gets passed on from generation to generation. Even when the original GPL code is 50 generations diluted, new code that links against that 50th generation still needs to be GPL’d.
And another thing, the BSD SGI license allows companies to STEAL from each other, look at DirectX and Quartz, both are based on OpenGL. Why would companies want to invest in such a license? I know personally I am more likely to release the soruce if I know it will make MY code better then knowing someone could just take my code and change it around a little, then make MORE money then I did!
[snip]
DirectX is probably the one reason MS is so liked, now MacOS x has proprietary code based on it too, they were both able to STEAL [snip]
You can’t steal BSD code. How do you steal air?
This is the difference between closed-source, GPL and BSD.
Closed-source is Coca-cola. You have to pay for it. And they won’t give you their trade secret, the secret recipe.
GPL is free beer. But if someone buys you a beer at the pub, you’re expected to return the favour and pay a round for all the other blokes later on in the evening. Otherwise you’re not a good mate.
BSd is air. Air is always free. You may breathe as much of it as you want. Also, you may put up a factory which takes that air and bottles it (as in Japan, a BSD stronghold). You may sell your bottled air for as much as you want. There is still enough air for all of us to breathe as much as we would ever want for our entire lives. OTOH, all civilised countries have environmental agencies which forbid you from polluting the air, since it’s a basic resource we all need to survive. You may take the air and do whatever you want with it, but there are strict regulations as to what kind of air may be released in the open. The BSD equivalent is the engineering team and the board of governors. They’ll let you do whatever you want with the BSD code, but not all alterations are let into the main code base, just as you may not let all kinds of exhaust fumes pollute the open air.
No matter how much you breathe, even if you’ve got an industrial bottling plant (Apple et al), you can never steal the air. It’s a public resource, and arguably infinite.
You can’t steal that which is gratis. And BSD is an infinite resource.
I fail to understand why GPL upset so many people here. As someone said before, if you need a function in a GPL library and don’t want to GPL your own code, then rewrite your own function. Remember that is the GPL what allowed you to see those libraries anyway.
Also, if you agreed in use/modify/redistribute GPL code, you must agree with its terms. The people who donated their code to you expect this. I think it is fair enough.
Cheers,
DeadFish Man
Not really. Do you think TCP/IP and The X Window System would be de-facto industry standards today if they were released under GPL? Would SUN and DEC have used them if they had to release the sources to their X-servers and kernels? Hardly.
And this is advertising or contributing back into FreeBSD how?
I understand the liscence. I don’t think you understand its community implications. Sharing code is no good unless you can get something back.
I am a programmer, and personally I won’t touch a GPL library with a 10 foot poll. (note distcintion between GPL and LGPL though. I don’t have a problem with LGPL libraries. Only GPL ones.)
So are there more BSD liscence apps out there or GPL ones? Do you think that no-one writes GPL code?
It is the virus like nature of the GPL that fosters development in Linux.
Look at the history of both BSD and Linux at what point in time did Linux become more hypeable than BSD. When was it better. Where did all this hype come from. Think real hard about it.
i’m having a blast with many of the unofficial ports. mainly the Hitachi SuperH (SEGA DreamCast) one. right now FreeBSD is confined to things that run in PC cases, and i’m still discovering all the goodness in other machines that can compute. NetBSD runs on a lot of things but won’t scale as high or as low as linux. so if i’m going to put a lot of effort into learning something, shouldn’t it be something i can take ANYWHERE(linux)?
NetBSD runs the same code on VAXes from the 70s, on 68010-based Sun 2 workstations and on several 64-bit platforms.
That’s scalable.
If you ever find yourself faced with some kind of hardware with an MMU, look to http://www.netbsd.org/ for a port.
If there isn’t a port, the NetBSD codebase is without any doubt the best place to start. It’s more modular and proven than any other OS when it comes to porting across architectures. Linux is by comparison a lot more convoluted, with ill-defined concepts of buses, chips or processors.
This is one effect of the Linux steamroller. People port it not because it seems like a sound approach, but because it seems cool.
True, Linux has been ported to run on misc. MMU-less devices, but how much Linux that is is questionable. NetBSD is one source tree, Linux is spread thin with one source tree for every architecture, and when it comes to porting drivers or even programs it’s far from easy. Linux ports are also often little more than a page describing the work-in-progress, or they’re based on already existing BSD ports, which is possible due to the friendly BSD licence. The opposite is not possible.
“It is the virus like nature of the GPL that fosters development in Linux.”
There is no development in Linux. There is only copying of existing ideas. Nothing in Linux is original.
The GPL does not foster innovation it inhibits it. People want to capitalize on their great ideas. Human nature (Read: Selfish Gene – Richard Dawkins). People can’t do that with the GPL.
“Look at the history of both BSD and Linux at what point in time did Linux become more hypeable than BSD. When was it better. Where did all this hype come from. Think real hard about it.”
The GPL had nothing to do with this. The lawsuit over the Berkeley Net/2 tape did though. FreeBSD’s distribution was resticted for a long time because of this.
Sure. But the point is Apple didn’t want to be forced into having to release their own code under the GPL.
No, the point is you’re making things up/don’t know what you’re talking about.
If they had used Linux, they would have had to open source any part of OX that linked against any part of the Linux subsystem.
What exactly are you talking about here? The Linux kernel or the GNU userland? Anyway, as I’ve already told you, GNU tools are distributed as part of OS X.
This is the classic discussion – the generousity of FreeBSD vs. the so-called “World domination” tactics of Linux. Personally, I think FreeBSD deserves a much more focus in the media, — but the history and curriculum vitae of FreeBSD simply isn’t enough to make it more interesting than Linux. The press wants to write about sensations — and Linux was the software to fill the gap.
From a hacker’s (and economist’s) point of view, FreeBSD is far more interesting than any Linux distribution. Primarily, the license is non-GNU, and it’s developed and maintained by a dedicated crew of hackers. You can actually trust the software, which you install onto your computer, because the FreeBSD dev-team have no version-deadlines or release dates needed for acquiring profite. It’s like an effective factory (compared to the ‘bazaar’-model’), which produces goods for free.
Now, stop dreaming of global GNU-domination and start realizing the situation – Linux will never reach the desktop user’s heart. FreeBSD – the power to be pragmatic.
“No, the point is you’re making things up/don’t know what you’re talking about.”
Who’s making things up? Your statement implied that you didn’t even know that that OS X does not use a FreeBSD kernel given that you made reference to “if they had used the Linux kernel”.
I suspect you have no clue about the archetechure of OS X. I’m talking about Linux kernel subsystems, which OS X would have used had it gone with Linux and not FreeBSD.
“Anyway, as I’ve already told you, GNU tools are distributed as part of OS X.”
So what? That doesn’t restrict Apple from writing proprietary code that links against FreeBSD libraries. Sun OS ships with the GNU utilities on on a CD. FreeBSD used the GCC compiler. So what? There is no point to what you are sayign about Apple shipping GNU utilities with OS X. It has nothing at all to do with the issue at hand.
Also, if Apple had used Linux subsystems, they would have had to GPL Darwin. Apple did want to open source Darwin, but not under the GPL.
*gasp for air* Your general lack of half a clue and inability to keep the context in mind is simply astounding. I’m out of words.
I will never respond to any of your posts again. Holy moly!
“OS X isn’t that simple. For example, it isn’t using a FreeBSD kernel. It’s using the MACH kernel.”
I had already acknowledged that OS X inherited from both BSD and Mach.
“And Aqua isn’t running on top of the UNIX subsystem like X Windows does. It is an inseperable part of it just like the Windows GUI is inseperable from Windows. So yes, various parts of OS X that are proprietary ARE linked BSD libraries.”
If Aqua is that inseperable from the underlying kernel, then how come Darwin, the BSD/Mach stuff minus Aqua, exists at all? Indeed, there is a way to run OS X without Aqua running. AFAIK, you log in as “console” or “>console”. BTW, the Windows GUI isn’t quite as inseperable from the Windows kernel as you might think. Most of it is encapsulated in the executable explorer.exe.
“If they had done that with Linux, they would have had to GPL those parts of OS X, as well as any dependancies on those parts of OS X.”
One does not link to the Linux kernel the way one links to a library. One runs stuff on top of it. Apple could have used Linux without any legal problems, easily. As I pointed out already, OS X evolved from NeXT, and Apple’s decision to use BSD had less to do with GPL issues and more to do building from the NeXT framework, which was BSD and Mach.
You’ll find that all licenses on code you write cannot be changed afterwards. You simply cannot retroactively change the license of some code, even MS can’t do that. If you write some code that’s BSD licensed, that code is always BSD licensed.
Now you can dual-license it. So you can make it GPLd, but you can also release that code under another license if you want (because you hold the copyright). Again though, if you dual license it once somebody has used the code under that license you can’t simply change the terms afterwards.
“I will never respond to any of your posts again. Holy moly!”
Translation: “I dug myself a hole and tried to argue about something I don’t know anything about, so I am just going to bow out and say I’m never responding to your posts again.”
Like I said, your argument about Apple including GNU utilities is completely irrelevant.
“One does not link to the Linux kernel the way one links to a library. One runs stuff on top of it. Apple could have used Linux without any legal problems, easily.”
OS X IS NOT running on top of FreeBSD! Just like it would NOT be running on top of the Linux kernel! That’s the point I can’t seem to get through to you.
“Now you can dual-license it. So you can make it GPLd, but you can also release that code under another license if you want (because you hold the copyright). Again though, if you dual license it once somebody has used the code under that license you can’t simply change the terms afterwards.”
I know. And that is the point. So basically, once you decide to GPL your software, you better decide that you don’t want to maintain any form of control over your software whatsoever. Because once you have gone down that route, you have basically given the entire world a license to do whatever they want with your code provided that they don’t restrict anyone else from doing the same thing.
Obviously, this is not a debate that is going to be solved here though. People have debating the merits / drawbacks of the GPL for years.