Slashdot Mirror
Debian FreeBSD Distro?
antigen fiend wrote in to note that Debian Weekly News has a bit about a recent debate about a Debian port to the FreeBSD Kernel. There are several comments relating to licensing, ease of porting, and other relevant topics. Its definitely an interesting project, with a lot of political and technical snags. Any thoughts?
Like allways, I am in support of anything DEBIAN FREEBSD. LONG LIVE OPEN SOURCE!
*************************************************
Porting Debian to FreeBSD seems like a good way of getting software to be portable to BSDs as well as Linux. Furthermore it would increase the distribution of dpkg and .deb packages, which can't be a bad thing. I'd like .deb to be the "default" package format instead of RPM. :-)
There's already work being done on a Debian/GNU Hurd distribution, which in some ways is more different than FreeBSD (although it does use glibc, which helps).
I hope to see more disto based on lots of kernel, but *BSD isn't GPL. I think GPL make the difference.
--
Does this mean this distro will be uder the BSD license? If so it could be pinched as sold as a commercial product, which would suck. While I am not agianst FreeBSD itself I do not agree with the lisence.
---
Just because life sucks, it doesnt mean you have to care.
I use both debian and FreeBSD,
Debian for it's excellent desktop/workstation capabilities with their wonderful autoupdating website/dependencies thing and the total hackability of it all. The packages always work, and their dependencies aren't all funny like they often are with SuSE and RedHat,
However I use FreeBSD for my servers, yes it's not the easiest of distros and it's very hard to get anything real working (for a relative newbie like me anyway), but it works, very well in fact. It never ever crashes and it servers my websites just fine. There are very few bugs very few updates, and I can just leave it alone totally.
To combine both? Well - heaven - what else can it be. Add to it KDE 2.0 and you have the complete MS competitor for the desktop/low end server market.
Bring it on!!
Sure this is an interesting project, but I'm a bit doubtful concerning the impact this will have. Currently the Linux distros, as different as they are, all use the same kernel. Given that the issue of compatability/fragmentation is a real concern to any potential corporate (Linux) adopters, I'm wondering wether this won't just turn out to be food for MSs propaganda machine ... after all we're taking Linux software and porting it to a different (yet still Unix based) kernel ... technically though, this is probably good news as it will uncover some new bugs and make the software more portable ...
This is not intended as a flame. I've read the appropriate links and could not see a good reason for doing this. FreeBSD is already a complete system with its own kernel, filesystem, utils, documentation etc. As is Debian GNU/Linux. I'm curious to know what advantages there would be in this system that aren't present in Debian or FreeBSD. Is this a step forward for Free Software? Or another example of UNIX fragmentation?
-- Liquor up front, poker in the rear.
I've got nothing against BSD, and IMHO Debian is the nicest distro to admin. (Ever try to trace your way thru RedHat's initscripts, each one sourcing 18 different files each sourcing yet others based on variable set in yet another bunch... ) But anyway, much as I like Debian, they are way behind, this current delay over boot floppies of all things, and no disrespect to the people doing them, but how much really needs to be changed from the last set? They're BOOT floppies, they're just supposed to have the bare minimum to kickstart an install. That's it. I know, I know the old argument, ours is the best because we take our time, well there's a point of diminishing returns. I'm running potato now and it's pretty damn solid. I really have to wonder if the Debian's just gotten too big and bureaucratic for its own good.
Getting the libraries ported would make supporting applications on both systems a dream. Also, the FreeBSD group could then take advantage of the ported glibc to increase interoperability.
We talk about freedom, yet all the major Free Software vendors are using the same kernel. If Debian pushed GNU/Linux, GNU/Hurd, and now GNU/FreeBSD, then we will eventually have the day that we can choose which kernel is the best for our purposes.
Few people use Linux for the kernel, they use it for the capabilities. I have a friend who loves his GNU/Solaris machine. Don't laugh, he administered Solaris machines for years, so his home PC has the Solaris x86 stuff on it, but all it runs is Free Software (and StarOffice).
Linux advocacy is silly. Advocating *nix, or even Free Software makes sense, but obsession with a kernel is silly. As Linus says, (I might be paraphrasing) "Linux sucks, it just sucks less than anything else out there."
The Linux kernel is fast, it is stable, but it isn't revolutionary. As I understand it, it is Microkernel-esque, although it still probably has remnants from its days as a monolithic kernel. I mean, when someone comes out with a solid exokernel, are we going to scream and yell about how Linux is still better?
Support free software, but support choice. In a free software world, we could pick our kernel without worrying about our apps breaking. This project has both technical and political merit. Although, if the BSD license allows you to do whatever you want with the code, can Debian release the GNU/FreeBSD system under the GPL? If they made the best FreeBSD distro and put it under the GPL, the license issues would be over. I'm not sure of the specifics of the BSD license, but if you can include it in proprietary licensed systems, I don't see why Debian's system couldn't be GPLed.
First, Debian is also being ported to the HURD, which I am eagerly awaiting.
On a more general note, the more people that do porting projects like this, the more pluggable the whole system will get. Wouldn't it be great if you could select your kernel, libc, and packages independent of each other? This is what UNIX is about... choice.
It would also be great to have glibc ported to FreeBSD. I sure hope that is the route they decide to take. This would allow FreeBSDers to have better access to new applications (although I'm not sure if they care - BSD is usually about stability, not newness). Also, given the user base of Debian, this would give more competition to Linux, and give us a reason to still be competitive.
Finally, this will expose more bugs in all systems, which is always a good thing.
Engineering and the Ultimate
I'm a bit confused about some issues. Maybe someone can set me straight.
How much difference will it make which kernel is used? (I mean technical, not licencing, issues). How much variation is there in the performance of different kernels? Is it noticeable in daily usage?
Also, how different are the kernels that the various *BSD's use? For example, OpenBSD has a reputation for being very secure. Is that due to kernel design, or more to the way the whole distribution and its packages are configured on top of that?
11.0010010000111111011010101000100010000101101000
Sure, as a desktop FreeBSD user this would make me happy, as it would boost FreeBSD acceptance. However, I'd rather see HURD ready first (this is one of Debian's projects right now). It seems like a very interesting and worthwhile alternative to current kernels, and might show the world that the OSS movement can produce a true, quality microkernel-based system.
I have always wanted to try BSD, but never had the time to. I would probably run OpenBSD as it is most like Debian... but now I can just run Debian. I think. Do I understand correctly that they want to make a FreeBSD distro? What is the difference between the FreeBSD kernel and other BSDs? If so, will they stop making their wonderful (please no distro wars!) Linux distro? I sure hope so, because if it turns out that I dislike BSD I want to have good ole' Debian to come back to.
If you think you know what the hell is really going on you're probably full of shit.
If you think you know what the hell is really going on you're probably full of shit.
jdube is who I am.
It seems the Debian guys go from one internal fight to anyother. Just give it up guys, Debian BSD is pointless. It'd be just the same as Debian Linux only 20% slower and 5% more stable.
This "project" is currently not official. It is being worked on by a few developers that are doing this without the support of Debian as a whole. By readin ghte lists it is very hard to tell wether it's going to be official or not.
The idea of more FreeBSD distributions (with a different userland, different package management) sounds like a horror to me. "Sorry, our product only works with Debian/FreeBSD, not FreeBSD." Shrug.
Intosi
Intosi
Now, this is extremely weird.
/. reader could explain this whole situation to me, I would be definitely grateful!
Why on earth port Debian (an excellent system, BTW) to FreeBSD (another excellent system IMHO)? FreeBSD already has its own (outstanding) ports/package system and I totally fail to see what kind of interest is at play here.
And the absolutely worst thing is that it raises all kind of licensing questions (BSD vs GPL) and library porting (glibc vs libc) -- this thing is totally beyond me. Again: what is the interest of having Debian ported under FreeBSD?
I can understand Debian/HURD, but Debian/FreeBSD is the weirdest thing I have heard in ages. If a gentle
Ah well. As long as it's open source... =)
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
Please, if they wish to attach the GNU name to anything, they should call themselves GNU/Debian or Debian/GNU. There is NO justification for calling it GNU/FreeBSD.
Tired of politics? View a nude.
Debian is the best linux dist imho.
FreeBSD is the best x86 os imho.
KDE 2.0 looks _very_ promising.
And them together sounds lika a real threat to MS Win* as the desktop alternative.
The 4 s:
Speed, security, stability and style.
Its my lucky day!
Lost Carrier
Lost Carrier
http://www.geekboys.org
Not knowing anything about the BSD package system, I can see this as a good thing. I agree with the post above about porting the newest glibc2 to it as necessary too. BSD is nice and stabile. My ISP runs it, and the only problem I have with it (at least in their case with BSDI) is that it gets a bit kvetchy with the connections through telnet. Otherwise, I consider it a safe "alternative" OS. I run Debian Linux BTW, after running slack since '95. I like debian because I can just use Dselect to get new packages. I just have to figure out apt.
Lowmag.net
After I wrote the above comment (and coined "DebianBSD"), it occured to me that this effort may very well go forward...
I say this without malice toward anybody.
Besides the various "why"s such as "because it's there" and "to improve the portability of our software" and the like, there is a political reason to do this.
If Debian makes a BSD distribution, it would almost certainly be called "Debian GNU/BSD". This would reinforce the idea that, according to the FSF, et.al., the correct name of the Linux distributions is SoAndSo GNU/Linux. By creating a GNU/BSD, the media attention would make people pay attention to the "GNU/" prefix and would also result in comparisons to GNU/Linux.
Christopher A. Bohn
cb
Oooh! What does this button do!?
Your comments on the size of the Linux kernel are, um, bullshit. Unless you are referring to the size of the SOURCE tree, which is irrelevant because Linux has tons more drivers than FreeBSD. I personally have compiled BSD kernels and Linux kernels. Recently, I compiled a NetBSD kernel and the resulting binary exceeded 1.5M. My Linux kernel is 1/3rd the size.
...to unite *BSD and Linux,maybe ? That will be great! Just think of the combination of Linux's flexibility and ,may be, the OpenBSD's security.... Or sth. alike..May be the BSD license isn't good enough in these times,since GPL and linux took over the u**x people .....
All this kernel independence surely would help a company like Sun if it wanted to base a Solaris release on the Debian system. The techie in me drools. But there would be no question of "official" support by the Debian organization, which seems to be the real sticking point with the FreeBSD effort.
Debian GNU/NT, anyone? ;-)
A significant merit to adding FreeBSD to the mix is that this makes Debian less and less kernel-dependent. In the long run, that makes it more and more possible for Debian to support more "UNIX variants."
Interesting, in the longer run, would be:
Thus, support for FreeBSD tomorrow may help there to be support for more unusual OS selections a couple years from now.
That seems to me to be a Good Thing.
If you're not part of the solution, you're part of the precipitate.
Well Linux kernels are usually compressed by
default, comparing image sizes isn't really
fair.
There is a mailing list for such a project...
(For the curious, I am a FreeBSD user)
:).
There are several reasons I do not think DebianBSD is going to be advantageous:
1.) Most people using FreeBSD enjoy the less restrictive, more open BSD license.
2.) The FreeBSD Ports Collection is truly incredible; there little or no need or desire for a Debian package system.
3.) I personally love the fact that FreeBSD is *not* a distribution, it is a full featured, integrated Operating System. I think this allows FreeBSD to be more intelligently organized and to work more smoothly than the various Linux distributions forced to integrate an operating system beyond their control with software largely beyond their control. DebianBSD sacrifies this advantage.
4.) The licensing issues are difficult, if not impossible to resolve.
5.) Any variation on the name GNU/BSD is going to cause an uproar and permenantly doom any Debian based FreeBSD distribution.
6.) Finally, from quickly scanning the Debian mailing lists, it seems as if most of the Debian developers have no respect for FreeBSD. One called it "dying software" and others claimed it offered no advantages over Linux. While everyone is entitled to an opinion, however ill-informed and erroneous, I wonder how dedicated Debian could be to an operating system it does not like and does not respect; after all, part of the allure of working on open source software is being able to code for your own pleasure rather than someone else's.
In short, DebianBSD seems like an abortion from the start, althougth I am prepared to be surprised.
(Note, even if DebianBSD did come into existence I would never switch; Debian and the FSF irritate me with their holier-than-art-thou pronouncements on freedom. Sadly enough for DebianBSD, I have a feeling a large number of FreeBSD users will agree with me.)
I love Debian (been using it for the past year after switching from RedHat), but shouldn't they be concentrating on getting the next 'official' version out rather than adding more platforms? Slink was released back in the February/March timeframe, and it would sure be nice to an official version running the 2.2.x kernel series before 2.4.x comes out.
The unfortunately part of this is that there are packages that have been updated in unstable, but are unusable on an unmodified slink system (slink's dhcpd-beta / unstable's dhcpd comes to mind). While not updated for security risks, it would be nice if I didn't have to update half a dozen packages and move to the 2.2.x kernel just to fix confirmed memory leak problems.
Before the 'unstable is as stable as RedHat' flames start, there have been problems with unstable (the perl problems come to mind). And for those 'well pitch in and help' flamers, I just dropped off my United Way sign-up sheet to HR and made a nice contribution to the Debian project.
This is better than the opposite observation that I was going to point out, which is that there are components of Debian, such as Perl, Python, and XFree86 that already use non-GPL-like licenses.
If you're not part of the solution, you're part of the precipitate.
This would be a whole new userbase, not the crochety old men that run FreeBSD.
I have been trying to get freebsd to install for over 2 weeks at home. The sysinstall utility only works with the beta 3.3 kernel and the 3.2 kernel edition of sysinstall can not install booteasy properly on a second partition of a second drive.
The sysinstall program even screwed up my disk geomtry on my first partition on the second drive
(even thyough I didn't touch this partition during the installation).
The bsd kernel is a great kernel but alot of bsders love to attack linux as only a kernel so I have to judge freebsd on the merrits of the utilities as well.
I changed my bios to normal to LBA to see if thius fixed the problem and then the booteasy program trashed my MBR during another install and I could not boot windows. ARGGG!
IF you linux guys think suse or debian is tough you aint seen nothing yet.
If debian replaces the installation with there's but keeps everything else, I will be a happy supporter. Its not fair that the bsd guys get a bad rap for something silly like an installation program when the bsd kernel already has a multithreaded tcp/ip stack and usb support.
(Perhaps some bsd guys could help with my installation problem. I need all the help I can get.)
Hey, I was pretty interested in upgrading my server to debian a while ago, and have been waiting for potato for far too long now. It has really become a joke. Sure porting to HURD and *BSD are possibly a Good Thing, but maybe we should concentrate on one thing at a time here eh?
P.S. Don't reply about using apt-get to use the latest stuff, read: tested.
--- Stampede linux for me! I play with fire to break the ice..
Personally, I run RedHat, but that's because Debian took too long to upgrade and I was really concerned about the age of some of the packages. However, I love Debian's ease of package management. Just dselect, pick and choose, and it downloads and installs for you, w/o needing X, unlike a certain RPM-based distro... :) If they could get people to actually keep their packages up to date, then Debian/*, be it Linux, Hurd, or FreeBSD, would be cool, very cool, and it would make using FreeBSD alot easier to use, so long as it still includes the ports area (it *does* work nicely once you figure it out....)
.
This is typical of BSD folks over the years! It's nothing personal. But it is exactly why I have linux on my system instead of BSD. Just an unwashed linux luser... ;)
Interesting... I feel completely the opposite. I found the RH init scripts infinitely more logical and easier to admin than Debian's. In particular, network configuration was miles better in Red Hat with all the relevant details for each interface in a config file in /etc/sysconfig, rather than hardcoded into an rc.d script. Still, that's one of the great things about Linux (and indeed, free software in general). Everyone is free to use what they're most comfortable with.
Disclaimer: I haven't looked at Debian since 1.3, so things may have improved since then. Also, much as I like Red Hat, I freely admit it's not perfect.
"The invisible and the non-existent look very much alike." -- Delos B. McKown
> Debian GNU/NT, anyone? ;-)
There was such a project, basically it was to be a Debian format distribution of Cygwin. It would have been very useful, given that Cygnus doesn't put in much work on making free distributions, but nothing seems to have come from it. Nobody volunteered to do actual work.
Just drop a BSD kernel compiled to run only Linux binaries in a Linux distribution.
This puts the Linux kernel in direct competition with BSD.
I would buy such a distro.
I have wished for that for so long, I love the feel of freeBSD but the ports confuse hello out of me being a 100% dpkg/apt man so yeah. go go go :)
I would really like to see this. I very much like Debian's amount of packages and ease of installation and configuration while I think FreeBSD is more stable and faster than the Linux Kernel (at the moment), but the configuration is doing everything by foot.
A Debian/FreeBSD would combine the best things, IMHO.
Sebastian.
..
:-)
The cygwin stuff is not quite there yet. It's close. I would use it a *lot* more of if I could get xterm running. The windows command terminal really really blows, even with bash running. So I see GNU/NT held back by the lack of an xserver + xlibs mostly. Once that was running, I think the rest of it would fall into place pretty fast. Why is this important? Couple of reasons. 1. Some of us are stuck in NT for at least the next few years. 2. It is much easier to train potential Linux users if there were a way to ease the transition from pure gui to not-pure-gui. This way they can see that DOS is a hopeless POS, but a modern shell is extremely useful. And that the only point of similarity between the two is the CLI. I am sure I will get roundly flamed for this, especially over point 1. But, whatever. I got stuff to do, and I have already gotten 1 pure linux box (debian!) and a dual boot box (debian!) in the office. It is just going to take a few years to complete the transition. (Actually, having a couple of NT boxes around *is* useful.).
Easy does not mean bug free. I think an easier install method like corel's or even debian's would beneifit bsd tremendously. Regarding your comment about linux is not really unix.Keep in mind that all of At&t's source code has been removed in bsd so its not unix either. Infact I believe only SCO has some unix code in it and we all know how wonderfull sco is. :-) the unix tradmark is owned by the open group. By deleting all th unix code was how bsd got through the lawsuits. :-)
There is nothing better then debian when it comes to package management and keeping your system free. Its proably the only reason I don't use fbsd now, I hate ports!
-- four
SICK
There are ports of the X libraries for cygwin, and there are gratis X servers that run under NT.
I haven't bothered with them, as Cygwin 1.0 is close to a perfect fit for my needs. It gives me a nice Unix programming environment under NT for producing win32 executables, and since XEmacs runs natively under win32 (and is included on the CD from Cygnus), I cannot think of any X11 programs I miss.
I think the port to FreeBSD is a cool idea. I'd also like to see a port to cygwin.
HURD is a collection of server programs, that run of top of the GNUMach microkernel, AFAIK. However, in theory, isn't the HURD itself portable to other kernels? Might a stripped-down linux kernel work better as a "microkernel" than Mach? Relatively few people have taken the time to understand Mach. HURD-on-Linux sounds to me like an interesting idea.
Choice of masters is not freedom.
I'd be glad to help you, but you did not leave an e-mail...
Alas, if sysinstall touched a disk, is because you told it to. There is *no way* it could have touched it without you selecting it.
Anyway, e-mail me if you want. Or e-mail freebsd-questions@freebsd.org.
(8-DCS)
I know it's not that easy. But I'd remembered a comment from about 15-18 months ago by the CTO of some company (can't remember which) which had previously only released products for commerical Unices (IIRC: Solaris, HP-UX, SCO, AIX, and maybe IRIX). This was when "there are no apps for Linux" was pretty much true in the commercial sense, and this was one of the first commercial apps ported to Linux. One of the industry mags asked him how hard it was to port their app to Linux, and the CTO replied "I typed 'make'."
Christopher A. Bohn
cb
Oooh! What does this button do!?
I'm a FreeBSD user (former Slackware 1.0 switched Solaris switched FreeBSD), and for using FreeBSD for the last two years, I've had FAR less troubles than I've had in the past two weeks with f'cking RedHat. First of all, I would like to know how it is that the kernel itself got to the point where no install scripts work right at all? They all seem to end up either toasting /etc/lilo.conf, or some files in /boot, or just not installing the thing at all.
Blah... Anyways, I've managed to make this machine panic on boot 5 times in one week.
make world never did that to me...
Anyhow, it just seems to me that maybe the Debian guys might take some hints away with them as far as PORTS (yes, PORTS PORTS PORTS), and the ENTIRE CVS theory. make world is a wonderful thing.
I'm all for any type of progress. This is why I use both Linux and FreeBSD. I need to try SuSE or Debian I suppose. Redhat sucks to me so far.
This comment sucks...
Debian is one of the last Linux distros which
are not yet share holder driven (watch out for SuSE).
It is IMHO a sad fact that the Debian development
model seems slower than it would be good for Debian.
I think that Debian's receipe contains too much
democracy and not enough (benevolent) dictatorship.
Embracing additional kernels can be a nice thing
if your development model is fast and you are
among those who march in front.
Splitting up forces further when you seem to
be falling behind is IMHO the wrong thing to do.
Debian first should do some streamlining of their
development model before increasing entropy
further.
Don't get me wrong: it might be a big plus 'for
the whole thing', if different exchangable kernels
are competing against each other. But increasing
complexity might also be another nail in the
coffin if your development system is too slow to
handle it.
I don't see where there is an advantage for anyone, save some rabbid GPLers that want to run the kernel from BSD.
/.)
It *IS* market differentation for Debian, and that is what the 107+ GNU/Linux compaines are all about. So they are going to try a different thing.
1) Between GNU/Linux and *BSD, if one group comes up with a feature, or a better way of doing something, the code is either moved directly, or the idea is re-implemented under the correct license.
2) Most of the user-level code doesn't care about the kernel being BSD or GPL. So, unless it's commerical software using a group of tools, the tools on *BSD are on GNU/Linux, and GNU/Linux tools are on *BSD.
3) *BSD has the ability to run most GNU/Linux binaries. So, being *BSD lets you allready run GNU/Linux stuff.
Given these 3 things, I don't see a win for customers. (Unless, of course, the BSD kernel is more stable/better written than the GNU/Linux kernel. But such talk is considered a TROLL on
As for the argument that *bsd is dying...Sorry to say this, but there is a whole group of people who think OpenSource is doomed, that GNU/Linux is dead on the vine, etc. Ask Micro$oft. Judge Jackson thinks that OpenSource is not relavant. The people at BSD/OS mock 'the part time OpenSource' world. Such people claiming *BSD is dying are no better than the blind Amiga or Macintosh advocates. They want to push down other OSes to make thier own OS look better.
If it was said on slashdot, it MUST be true!
FreeBSD is definitely faster than Linux. You would know if you had *actually* used FreeBSD before. Based on my recent Linux installs, I would say it is much more than 5% more stabile too.
Sorry, but I am at a point where I think Linux needs more time in the oven, maybe a year or two. BSD has been around for a long time, and that is why it is faster and more stabile.
EverCode
want to install lynx?
/usr/ports/www/lynx
cd
make install
installing needed packages ######
installing lynx ######
lynx www.slashdot.org
now how hard is that?
If you want Linux, try Caldera OpenLinux. It is the easiest ass install, you will not believe it.
However, Caldera has been very unstable for me, and am going to remove it.
EverCode
Even though UCB removed *their* advertising clause, a lot of contributors in the past have followed suit, copying the UCB/BSD license for their own code that was added to *BSD's. Thus, even though the parts of the code that have an advertising clause for UCB can be "cleaned up", you would still have to go to all the other contributors who used the advertising clause on their code and get them to remove it.
While *most* of *BSD is UCB code, a *LOT* of work has been contributed by a *LOT* of other people...
it's a release version. 4 is the current beta.
Just from this thread, it sounds as if there is enough support for Debian-FreeBSD to make it not DOA.
The idea of it all is very interesting, no matter the pros and cons of it.
FreeBSD and Linux users needs to wake up and learn to respect each other. Most of us know that competition is good, and FreeBSD and Linux are in direct competition in many areas.
If the BSD community does not want GPL, I can respect that. GPL is not some Holy Grail of licensing as many of you make it out to be. I believe things can be TOO open. Why? It is because things become chaotic and fragmented, slowing progression and causing all sorts of problems. You are going to see a lot of this in the Linux community over the next couple of years. (It is going to happen because commercial distros are going to try to get ahead by breaking away from the others. The lag time for others to pick up any advantages to their ideas is also enough to cause problems.)
Besides, for the vast majority of us, the license does not matter anyways. We just install it and use it. We are not modifying the libraries, etc. etc.
Debian-FreeBSD is a wake up call to us all. It's time to start repecting our differences. We are all individuals, not members of cults.
E
EverCode
This would be VERY nice indeed. One of the best things that could happen to BSD would be to have a centralized package managment system, simular to RPM and apt.
.deb files would be limited to Linux only, so once we have the package manager working correctly, it's just a matter of letting the BSD based packages stream in.
And the best part is, I really don;t think it'd be all that hard. There's no reason why
-- I'm the root of all that's evil, but you can call me cookie..
(I am also a FreeBSD user.. I don't want Debian dragging FreeBSD into slow releases either)
Sure, but let's remember that all BSD's forked from a common code. 'Distrobutions' could eventually end up going the same way as BSD if Linux does eventually fork for any particular reason.
And programs under one BSD aren't promised to work on ALL BSD's..
-- I'm the root of all that's evil, but you can call me cookie..
Where I think there might be an issue is that the BSD world does not have this idea that the operating system is the kernel, but rather the operating system is the whole distribution. A debian distribution based on the FreeBSD kernel is not the Debian distribution of FreeBSD but a completely different branch, which one might call DebianBSD.
I think there are advantages to the BSD way of looking at things: security issues around OSs are not issues of the kernel, but issues about the whole system as deployed. I would rather the BSD vision of the operating system was not buried because it is not the same as the Linux vision.
Now you know why the 95% of desktop users who run some variant of Windows laugh at you when you try to tell them that bash and pine is better than Explorer and Outlook.
Actually, they won't affect our FreeBSD. It will be their own distribution of it and their own responsibility. Technically it would be a different operating system. So don't worry about the real FreeBSD. :)
First, This is not going to be Debian FreeBSD. FreeBSD (as the other BSDs) regards itself as an operating system, not a kernel. If you replace userland, it's going to be something else - and I'd be surprised if FreeBSD Inc. would allow them to call it FreeBSD. If they are serious, they will end up with a name like Debian GNU/BSD. Let's call it DebianBSD for now.
Second, this is going to be a very linuxified BSD. You inherit practically all of Linux' organizational/architectural problems while sacrificing a number of BSD's most impressive strengths. I don't think it will be superior to either pure Linux or FreeBSD.
Third, having said all this, I think it's a Good Thing (TM) nonetheless. Becauser, hopefully, the developers will have a lot of fun doing it, they will find a number of interesting and challenging problems that eventually will find their way back into the respective communities, and they will fall across countless quirks nobody before has seen or thought about. The mutual technical understanding between the communities could be enhanced, and both systems should improve in the effort.
So, more power to them.
FreeBSDebian ?
The FreeBSD Copyright
e .html
Copyright 1994-1999 FreeBSD, Inc. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE FREEBSD PROJECT ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
The views and conclusions contained in the software and documentation are those of the authors and should not be interpreted as representing official policies, either expressed or implied, of the FreeBSD Project or FreeBSD, Inc.
This was taken straight from
http://www.freebsd.org/copyright/freebsd-licens
I would guess the license contained on the ftp site would take precedence but why the two licenses. Did the recently decide to put the ad clause back in? That seems like a strange thing to do.
Why don't all programs built for one Linux-based operating system work for another one? For myriads of reasons, including hardware version, kernel versions, operating system versions, adminstrative strategies, and user environment. I have numerous such examples, as do we all.
This seems to be the wrong direction- I think the whole linux community would benefit if instead the freebsd ports collection was ported to linux. (a point i think most are missing- just about all the debian packages are already avaliable on fbsd as a ports- repackaging them as .deb's seems silly.)
Hello Nick,
whom are you actually citing? I can't find any "hilarius" posts about "Our kernel has less featurs". What is your problem? Putting in people's mouth what they (actually, noone but you) haven't said?
I use Debian GNU/Linux happily now, but a Debian GNU/BSD, and Debian GNU/Hurd, and Debian GNU/EROS would be welcome additions to my arsenal.
I refuse to believe corporations are people until Texas executes one. -- desert rain on http://www.dailykos.com/user/
Debian GNU/BSD or Debian GNU/FreeBSD seems to be the likley name of such a product.
There was alot of worry over FreeBSD branching, but overall there seemed to be a positive reaction.
A good number of people like the idea of the .deb package under FreeBSD. just as alot would like to see the FreeBSD ports tree in a Linux Distro
A quick search of the FreeBSD mailing lists on the topic reveals some .
A few choice articles can bee seen below.
The seemng start of discusion: FreeLinux
A nice answer: RE: FreeLinux (Debian/GNU BSD)
Two weeks huh? An install shouldn't be more than 20 minutes (if you have the CD) and no more than 2 hours (if you have a fast connection for an FTP install.)
What exactly is a "beta 3.3 kernel" anyway?
As for your claim regarding 3.2, here is the errata page for the 3.2-RELEASE. No mention made whatsoever regarding sysinstall. Please stop talking out of your ass.
From the sysinstall man page:
HISTORY
This version of sysinstall first appeared in FreeBSD 2.0.
Also from the man page:
BUGS
This utility is a prototype which lasted several years past its expiration date and is greatly in need of death.
It's far from perfect, but it's not the nightmare you say it is.
FreeBSD isn't to blame for your incompetence.
The only one that trashed your MBR was you.
Btw, The "Linux is only a kernel" line isn't an attack, it's a fact, one that some people forget, or don't realize from the start.
--
My comments and opinions completely reflect those of anyone and anything I am remotely associated with.
I have recently switched to FreeBSD, since I beleive it is better than Linux. And I personally think that GNU/FreeBSD would be the worst thing possible for FBSD. Here's why:
/usr/ports//
.deb packages would start appearing for fbsd and there wen't good old pkg_add for anything you need... In addition to that, I'm afraid debian would make a sort of linux distribution around the fbsd kernel (actually, I sould say GNU in this case). This means that even at the absolute minimum, the install would fill up your HD with every available piece of free software, no matter how bad it is. In BSD, you get a clean, structured system.
FBSD is a whole lot better than Linux mainly because the distribution is centeralized. Just like only good code makes it into the Linux kernel, only good code makes it into the fbsd distribution - all of it. In fbsd, there are no 5 different libraries for a single task, and you can be sure there is a single good library all programs use. You can be sure all programs would run on your distro, since there is only one distro. In linux, you have to install 43 and half different libs for the same tasks just to be sure everything runs on it.
There is also one single distribution for fbsd, so you can be sure things work on it in any case - you can safley download binaries, unlike linux where you have binaries for redhat6-glibc2 redhat5-glibc2 redhat6-libc5 redhat5-libc5 slackware-libc5 slackware-glibc2 suse6-libc5 suse6-glibc2 and so on, you get the point. In addition to that, sources might not be compatible between distros, since they use different libs and in differend locations. Simply put, you must tweak the program to get it working, unless an archive is available for your distro especially.
When I switched from linux to fbsd, it was like entering heaven. Things are organized, code is better. No more tweaking. If I want a program, it's simply a matter of typing:
cd
make && make install
and presto, a network connection is established, the sources are downloaded, patches are applied, the code is built, and then installed. Also, all the neccessary dependancies are also downloaded and built. Or even simpler, just mount the packages cd and do a
pkg_add
All the binaries, libs, etc. will be installed, everything will be in place and registered for later uninstall, and without an error. This is all because of a centeralized distribution.
Now, if Debian (whom I do respect for their linux distro) were to make a Debian GNU/FreeBSD distro, things wouldn't work this way. First of all, glibc would probably be ported so fbsd would start having seperate binaries for glibc and fbsd libc. And probably seperate bins for debian and the official distro. In addition to that, debian would start using it's own sources for linux programs, which means ports loses a lot of it's value. And
Now, a point that *must* be emphasize: freebsd is an *operating system*, not a kernel. Linux is a kernel, fbsd is an os. Read this sentance ten times: fbsd is an os, linux is a kernel. Making a freebsd debian distro would be like taking window's solitair and building a linux distro around it. fbsd would lose alot since people would start using it just for it's kernel, and not the beutiful system it is.
It's been mentioned that this way in an open unix system you would be able to choose your kernel, libc, etc. But this is a plain wrong view! Have you seen a new linux distro with just libc5 or glibc2 on it? NO! You would need to have ALL the kernels, and ALL the libcs this way. This would also make things worst for the linux community... if libcs become interoperable, linux binaries are opt to show up with the fbsd libc, which means 3 libc's for linux, which means even more mess.
One last point I want to make is that the support gnu, debian, et al get from linus in the linux kernel might not repeat itself in fbsd if debian won't make sure the fbsd community wants this step - and my guess is we don't.
Please try to consider all sides to this before taking such drastic steps. Thanks.
--
Oren Sarig
sarig@bezeqint.net
Since Debian is a hard-core volunteer organization, Debian GNU/FreeBSD will be worked on only by those who want to see it happen. Therefore, to them, it's worth the effort. If you want to redirect that effort, the way to do it is to come up with and advance a more attractive project. Telling others to stop it because it's worthless just won't cut it---your valuation is not mandatory for everyone.
That's one of the things that's impressed me most about Debian over my years of using their software and observing their efforts. If enough people want it, it gets done. Otherwise, it peters out with no harm to anyone.
Of course, the discussions about whether it's worthwhile can get a bit tedious.... :-)
I refuse to believe corporations are people until Texas executes one. -- desert rain on http://www.dailykos.com/user/
As for the licensing, I well and truly believe that the lion's share of the daily users of these systems could not possibly care one wee whit less about the distinctions in the licences.
Finally, just some would prefer an operating system with an FSF-derived user environment but a BSD kernel, others would prefer the world the other way around. You read about what progress they're making toward this goal in active threads the BSD newsgroups today.
But please, please consider those threads read-only. Don't flame. Just read. In particular, don't have a coronary when you hear about how like Microsoft's dirty tricks some people find GNU's "embrace-and-extend" and anti-POSIX strategies.
Just let people have what they want to have. Sure, a BSD kernel and FSF non-kernel would end up making one more free Unix operating system than we had before, but likewise would a Linux kernel (I don't believe the FSF owns it yet, right?) combined with a BSD non-kernel.
So what? The Linuxes are so splintered and disorganized now that people would never even notice another one. Sure, you'd be upping the number of BSD-based operating systems by a far bigger jump than you'd be upping the number of Linux-based operating systems.
(4+1)/4 is a a bigger number than (122+1)/122--or whatever--is. Strangely, the complaints about the smaller number involved here dramatically outnumber those about the larger one. Isn't that peculiar? Try to resist.
TRY TO RESIST!
Let's just let everybody have what they want to have, ok? Unix is Unix. It's not Microsoft. Isn't that enough?
Hoping this isn't too far out of context:
I suggest you read the Debian Free Software Guidelines (DFSG), or their discussion of `freeness'.
To a first approximation, `Debian-free' means you can
Note that Debian is not terribly restrictive. Anything that meets the DFSG can be packaged and go into the main distribution. They'll happily take GPL, BSD, X, Artistic, and other licenses.
Even then, if the software is at least free enough to let a non-profit organization distribute it, it can go into the `non-free' collection, which, while not Official Debian, is nonetheless kept up by Debian maintainers and available on the Debian site. (If you're a CD purveyor, though, you'd better go through the licenses in `non-free' pretty carefully to make sure you're not stepping on anyone's toes.)
As for ``doctrinaire insistance", they're merely insisting that they stay on the right side of the copyright law. If a program, or its constituents, are licensed in such a way as to make re-distribution illegal, they won't do it. Period. If the licensing changes, it gets re-evaluated. It makes for some interesting discussions, to say the least, but under the current conditions (no court decision defining how copyright's concept of `derivative work' interacts with program linking, static or dynamic) I certain can't argue with taking the conservative course: it's their necks on the block, not mine.
I refuse to believe corporations are people until Texas executes one. -- desert rain on http://www.dailykos.com/user/
I've tried BSD, yes. BSDI, OpenBSD and FreeBSD. You know what I did not like with them? Their interface. Not the GUI - just elementary ease of use. It is not that I cannot handle it. But it is not pleasant. I hated Windows for it's stupid interface - it just lets you do nothing. If BSD people want to attract more users and get respect from more than just geeky sysadmins, well, they should consider some change. Now you're arguing that all the BSD's are whole OS'es - this is (IMO) where Debian is most different from other distributions. The 'stable' branch is guaranteed to work, with centralized control and configuration, places to store data, etc. - thanks to the Debian Policy. That is why I will stick with potato (hmm.. they should really rename the 'unstable' branch to something else - say 'devel'?) And (again IMO) the best part of all BSD's is the kernel - all the rest has to be completely rethought if intended for mass use. Also as a programmer I would never license anything bigger than a 20-lines-in-C utility under the BSD license because I don't want someone to close it and make money from it; I'm not afraid of chaos/enthropy/anarchy because there will be always someone intelligent enough to lead the development. I think a DebianBSD would be just great once they push potato stable. Just my $0.02. Sorry for the strong words, but that's how I see things.
The BSD license as I've read it allows you to do anything you want with BSD code. Any particular reason Debian cannot GPL their port?
Other people I've seen talk about this have used the term "FSF taint" instead. The kernel that Linux-based operating systems use is, for the most part, pretty well respected by other Unix programmers. The same cannot be universally said for the non-kernel clutter that various vendors slap together and call an operating system. So really, I'd try to avoid the term "Linux taint". I don't think it's really what you meant, and it won't get you in as much trouble. Try using the specific operating system, like "Redhat taint" or "Corel taint", or the specific political group, like "FSF taint", because that way you won't seem to be slamming the fine work that Linux Torvalds has done.
"we are essentially giving first aid to software that is dying (and rightfully so) because of its license". (from the Debian site)
What kind of asshole makes an absurd statement like that? Even Microsoft, who Linuxers love to point to as the world's primary FUD manufacturer, doesn't make statements like that. I've been using FreeBSD for about 3-4 years now, and it's made vast improvements over that time - and it's thriving compared to when I first used it. I'd be very interested to see Debian running on the FreeBSD kernel, if for nothing else than curiosity about how well it would work compared to the Linux kernel.
I think this is a great effort by the Debian folks, but if this is going to cause such a childish licensing battle between Linux and BSD, then just leave it alone, and Linux and BSD users will be happy to segregate themselves further. I certainly don't want to feel that I'm being given 'first aid' by the vastly superior development team of Debian.
-lx
It was post #38 in this thread that I think he was referring to.
Considering how GNU-free BSD already is (modulo in most cases the compiler, but that's largely irrelevant to the user's experience, and doesn't produce infectious output), this always sounds strangely redundantly, so I always have to rescan a few times until the words jumble back into place. :-)
I think you are operating under the mistaken impression that BSDI is based on FreeBSD's kernel -- it is not. It is based on its own kernel, which is based on the 4.4BSDLite2 kernel.
There are actually 4 main distros for *BSD -- but only 1 per kernel.
BSDI, OpenBSD, NetBSD, and FreeBSD all have a complete userland -- the "distro" part that distinguishes Redhat from Debian from SuSE from TurboLinux from Slackware from Caldera.
However, unlike the above, they all have separate kernels, as well, instead of being based on Linux.
There are still variant "distros" -- they just are not very large or general-purpose.
As an example, there is PicoBSD, a router-floppy "distro" for FreeBSD. However, unlike many other "distros" for linux, PicoBSD is not a standalone "distro" but is part of FreeBSD.
I hope this clears things up.
The license doesn't matter, I don't think. Remember that a number of distros are commercial ones (e.g. Caldera's) even in Linux-world.
The reason for a lack of alternative distros in the FreeBSD space is because the kernel comes bundled with a perfectly good distro.
Imagine, for a moment, that since the earliest reaches of time, Linus had distributed the "LDistro" distro of Linux.
Would there be Redhat, Caldera, SuSE, et.al, if Linus had distributed his own full OS of Linux? I doubt it. People would have declared his whole distribution, not just his kernel, "good enough" and wouldn't have forked it, as it is the natural tendency not to fork code unless you MUST.
Since there was no official distro for Linux, the community HAD to fork to produce one.
What has debian to offer any of the BSD's? Everyone is asking or stating how the BSD-kernel can improve Debian. Why?
My question: how is a Debian-like BSD better? OpenBSD is cool because it has been security audited and can use crypto. FreeBSD rocks. The BSD distros have managed to provide very stable systems. Linux is touch and go and very spotty. Debian can't even throw together a current linux distributuon. What is so superior in Debian to the existing BSD distros?
Why not reinvent Debin in the image of FreeBSD and release one consistent OS?
What impresses me about the FreeBSD distribution the clear distinction between the core OS and extensions. The system is very clean.
Why do we need debian when we (FreeBSD) have the Ports system. Even if debian uses a BSD kernel, the problems with porting linux software to BSD will not disappear. There are portions of the FreeBSD kernel that are incompatable with some of the linux calls. If you want glibc, run it in linux emulation. Done.
Just fork the stupid freebsd kernel, and license it GPL!
What's wrong with just sticking the libs in /compat/linux and running whatever binaries?
Already works with Netscape, RealPlayer, Quake 3, etc.
Am I missing something here???
Debian Linux is not as nice as the default FreeBSD distribution. What do they think they can improve on?
I have been using FreeBSD for about 3 years now (I started with a 0.99 Linux back in the day), so I can tell you from my perspective what I like about FreeBSD above Linux (I still use Linux at work) and what I would see as good and bad.
One of the things I really enjoy about FreeBSD is the centralized control. This means that there are never 2 libc's floating around that I have to deal with, never multiple filesystem layouts to deal with (well, I guess you do have all the major branches of FreeBSD that do differ some), you don't have to worry about binary compatibilty as much. The centralize FreeBSD model has led me to be much more productive, as I do not have to worry about library miscatches and keeping my system up to date as much. Everybody repeat after me: "CVSup is the most wonderful tool, the most wonderful tool, the most wonderful tool in the world."
I like how I just go to ports and make from there and rarely have to worry about things breaking or finding the right library.
If Debian were to just take the FreeBSD kernel and put their own, redundant packge system ontop of FreeBSD and ports GLibc, then you destroy two of my favorite parts of FreeBSD.
If, however, Debian were to integrate the package system ontop of ports and remove pkg_* utilities (or make wrappers for backwards compatability), that would be nice. If Debian were to ports their install to FreeBSD that would be nice (but not of monumental impact). If Debian were to get Gnome to run better on FreeBSD or ports some of the new Linux-gadgets (like the filesystems and drivers) to FreeBSD that would be nice, too.
I am certainly not going to push the Debian developers away, but I think that they should realize that FreeBSD already have many nice things going for it and by complicating matters with multiple libraries they will not be nearly as much appreciated as if they were to learn how to work withing the FreeBSD methodology.
Just my pennies.
-jason
Time to back to my code...
Yes, the fork/exec interface isn't ideally suited to that.
Well, it could spawn it with CreateProcess() (IE 4.0, at least on Solaris, doesn't handle mail or news itself, but punts to whatever mail or news reader you tell it to - which, amusingly enough, makes it arguably more UNIXy than Netscape on UNIX, at least in that regard!), but that interface isn't, as far as I know, any more friendly towards that sort of pluggable text-editor functionality than is fork/exec. A COM/OLE editor interface might be better suited to that.
I have the impression that in some message about a mailer for GNOME that they were thinking of making a Bonobo interface for text-editor widgets, so that different widgets could be plugged in.
If they do so, I'm curious whether it'd be the first desktop environment/toolkit to do so; I have the impression something such as that could be done in Windows, or in KDE, but I don't know whether anybody's actually done so.
(The Andrew toolkit had, as its text widget, a fairly powerful editor, which had, I think, some amount of vi compatibility available atop its more EMACSish base, but if you're used to a particular editor, that still might not be all you'd want.)
Thank you: you have brought a ray of laughter into what was otherwise a fairly grim day for me. The Debian Project should concentrate on getting a up to date Linux distribution out the door before talking about a BSD based Debian. Unless they planned to chuck Linux alogether and build on FreeBSD's strengths. IMO if they would concentate on making their Linux distro a little more like FreeBSD particularly in the installation (profiles) and package adding (ports) department, embrace & perfect linuxconf in the stable branch they could drive commercial Linuxen from the field, except the explicitly desktop distros. Not that driving anybody to or from fields is the point.
You're totally taking that out of context. That sentence was a quote from some of the comments about this project, and was clearly presented as such.
Well it is :) Look it up... About Mach... its a dog, destroyed micro-kernel reputation. Take a look at L4 or Eros for "slightly" better performing ones.
--
/usr/games/fortune: Segmentation fault (core dumped)
43rd Law of computing:
Anything that can go wr
It's messy to have to reboot to install tetrinet under leenux, reboot to fbsd and be able to run a good version of X with truetype support...
---
-
ping -f 255.255.255.255 # if only
Which was my point to the response. BSD has no more assurances then Linux as far as compiling capabilities based on distros. Distros in BSD I would equate the the many 'flavors' of BSD, as opposed to distros just being a different packaging of things under Linux.
-- I'm the root of all that's evil, but you can call me cookie..
..except for the fact that BSD and Linux have incompatable licences.
...
Bitchslapped? Give Rob a bitchslap from bitchslapped.com.
This is the most Horrible thing i've ever heard in my entire life!!! BSD tainted by Linux blood??? NOO!!!!!!!
I realized that the various BSDs actually have somewhat different kernels. I don't believe it's an important distinction. Some of the different BSD kernels probably have more in common with each other than the various 2.x (x = 0,1,2,3) kernels, for example, but that's not important.
What I said was that BSDI was the only BSD commercial "distro" currently. I would compare it to RedHat, Suse, Caldera, Corel, Turbolinux, etc.
The other BSD "distros" are more similar to Debian. The group behind them all is not a commercial for-profit concern.
Since there was no official distro for Linux, the community HAD to fork to produce one.
Well, you'll have to explain how this theory accounts for the fact that "the community" has produced and continues to produce numerous commercial and non-commercial distros, have none of them been "good enough" for most users?
I think the existence of BSDi in the BSD space scares potential competitors out of the commercial BSD space. It would be difficult to pop-up and compete with them based on any of the BSDL source bases as BSDI can simply use most of the BSDL work in their own offering while the competitor would not benefit from the proprietary BSDi code.
Then, there's also the problem that anyone who started a new commercial BSD distro would have to resist the powerful temptation to close off important parts of their distro to give them a competitive advantage, especially when your major competition, BSDI, has done exactly this.
This is not the case with fully GPL'd OSs. A new competitor can pop up and add value to an old competitor (witness Mandrake and RedHat).
BSD advocates like to make all sorts of excuses for the relative success of Linux vs. BSD. I think the accretive nature of the GPL has something to do with the rapid growth and acceptance of Linux.
A better question would be:
What does Debian need that FreeBSD has????
I use Debian GNU/Linux, and I am left wondering what is point Debian GNU/FreeBSD? If the FreeBSD kernel is so wonderful, then why not simply use FreeBSD? Someone might point to the Hurd, but the development of the Hurd is a Good Thing for Debian because the Hurd is a very different kernel (i.e. something new (and truely FREE)). If someone wants FreeBSD + the features of Debian, then fine, have at it, but I see no reason for this to become a subgroup of Debian, and take up space and bandwidth at ftp.debian.org.
The current Slashdot moderation system is made by gay communists!
Debian GNU/command.com
Debian GNU/Multics
Debian GNU/Babbage's Analytical Engine
The current Slashdot moderation system is made by gay communists!
Let us not forget this! What could it be? A penguin with horns and a pitchfork? A daemon in a tux? Or just both swirled together?
OK, so it was flamebait, but I'm replying anyway :-)
So I'm a dense RH user/sheep, am I? I was using Linux when it was just a root/boot disk combo (kernel 0.12), and my first real distro was MCC (kernel 0.99pl8+). What was yours? At various times, I've run most of the major distributions. I settled on Red Hat because I felt most comfortable with it -- purely techincal reasons, nothing to do with market share or anything else. After trying the available options, I picked the one I liked best. Can you say the same?
"The invisible and the non-existent look very much alike." -- Delos B. McKown
And it was a jackass comment to make. I assume it was made by someone close to or involved with the Debian project.
-lx
SGI has demonstrated the first Linux cluster based on Intel's IA-64 Itanium processor, using the Supercomputing 99 show as the venue for the event.
The demonstration achieved a number of public firsts including the first Itanium processor-based cluster and the first technical computing applications successfully run on the IA-64/Linux platform. The demo was based entirely on open source components including Cactus (an open-source simulation toolkit computing collisions of black holes), MM5 (a multidimensional weather simulation), Amira (an Advanced 3D Visualization and Volume Modeling System), C compiler, Fortran compiler, Linux kernel, and MPICH.
Macintosh developers occasionally laugh their butts off at the unix weenies that just do not understand why all the LINUX users in the world amount to only 0.24% (yes 0.24%) of the identifiable hits on innumerable tracked popular websites. LINUX zealots are usually so hotile to mice, scrolling, clicking, dragging, etc that their words are notming more than little pouting mewls of the ignorant. Apple's sales have increased in MARKET SHARE every quarter for the last 8 quarters or more and based on that a humorist could draw a graph showing that in the far far future all computer users will be running macintosh OS. Ignorant comments, such as "blind" that lump the only popular OS that runs a billion flops on one 1400 dolalr stock full blown desktop system are amusing at best. I run a mac, and write multimillion dollar selling mac programs because DOS and unix are stomach churning and wintel is annoying for code purists who hate compromise. There is a book I have read a few times called The UNIX Haters Book written by an accomplished developer and author. If I ever meet a UNIX zealot, I have them read that book FIRST and then I eagerly look to debate them. Funny thing is.... after reading "The UNIX Haters Book" the zealots lose all zeal to argue. I like OpenBSD because I totally despise the GPL pigs and always have ever since they stole public domain source code written by Darin Adler (the famous one) and then put that damned fascist GPL on his headers. The Mac in a cluster is a speed demon, and the Mac with multiple G4 processors in late january will be a supercomputer. And even my mom can run it, or old medical doctors or aged lawyers.
How is the HURD truly free and the Linux kernel is not? Both are under the GPL, both will have core teams that restrict what is inputted, etc. And I wont even get into the argument of whether the GPL is freer than the BSDL (it is all by definition, whether you mean freedom, or free software).
I'm a bit uneasy at the Debian project, but the more I think about it the more I like it. If the maintainer is not a spoiled GPL advocate (here I mean the type of person who rants about screwing over BSD projects, GPLing their code, etc), than its good. He will try to make it a benefit for the BSD community and hopefully give code back that they can use. My only fears is whether this is will try to fragment FreeBSD, though as there will always be FreeBSD, Inc, and if they are careful, there should be little harm. Still, chaos is more likely.
The Debian project is for free software, that is open source. They are also for freedom, so supporting many kernels isn't bad. If they do everything correctly, it will simply be the Debian enviroment with numerous kernels to choose from. FreeBSD is known for high performance and stability, but not as the greatest worksation/home system, where as Linux is. Take the kernel that does the best for your needs, keep everything else the same so little administration worries, and propigate. That is an interesting goal, and I hope I'm right as far as it being Debian's.
I do, however, wish to see a good reason for "GNU/FreeBSD". Whereas they might convert to everything but the kernel GPL, my understanding was RMS wanted GNU/Linux soley because GNU played a major role, and continues to, for Linux's development. Linux also requires GNU software to function, and always has. The BSD varients have not, and converting over so that they do doesn't mean GNU plays a significant role in development, as the kernel will still be maintained by FreeBSD, Inc.
"Open Source?" - Press any key to continue
FreeBSD does not want glibc.. They have been trying to stay away from GNU wherever possible, this is simply because they dont like relying on other efforts to support their own.. Glibc is GNU, they already have glibc compatibility in freebsd with the linux_base port.. The whole debian port is pointless and will turn BSD into a flea circus. Unless they can come up with a REAL point to doing this, then I dont see the point. All the other BSd's specialize in something, NetBSD is Portability, OpenBSD is security, and FreeBSD is speed and stability.. Can Debian think of something besides a linux freebsd hybrid which is pointless? If they can, go for it, if not, it's a waste of their time.
Wow, it's great to know that you dont know how to make a correct kernel config file, congrats!
My Background:
I use Debian GNU/Linux and FreeBSD in a networked environment.
Insights:
I think the Debian package mangement system, apt, is a good step ahead for the managing packages and dependencies with Linux. I like the Debian view on free software. However, Debian is for the most part the userland on a kernel, in this case Linux.
I love the ports collection offered in FreeBSD. I think its quite a bit easier, and farther along. I also like how it is source compiled so you can make any modification you need to instead of getting just stock software. Example: apt-get install irssi picks up irssi, and gnome, and all the related dependencies. What if you don't want gnome? A quick change in a Makefile will settle that on FreeBSD, whereas you'd have to make a new irssi in Linux from source.
Fact:
Linux is a kernel. Debian GNU/Linux is a distribution.
FreeBSD is an entire operating system. The kernel just makes up a nessecary core. It wouldn't be FreeBSD if the kernel was unlinked.
NetBSD and OpenBSD are not BSD distributions like most people tend to think. They are seperate operating systems, like Solaris, or Windows.
Question:
How would the Debian GNU/FreeBSD kernel get upgraded? In a lot of cases, any updates to the kernel have related, dependent modifications in userland. You really can't go ftp.freebsd.org and get just a kernel.
Cheers,
Sean