Slashdot Mirror
UserBSD vs. UserLinux - Is It Feasible?
A not-so-anonymous Anonymous Coward asks: "Someone has suggested to make a UserBSD instead of a UserLinux. From what Bruce Perens' anonymous 1-million-$ backers seem to want (no GPL-/Commercial dual-licensed development toolkit like Qt in any library, but only gratis LPGL stuff), this seems to make a lot of sense. After all, only the kernel would be different, the rest of the stuff (including the KDE or GNOME desktops) runs pretty much the same on BSD as it does on Linux. Is it possible to get the legal problems solved with licenses and still create a usable enterprise Unix desktop system on *BSD?" The idea, in and of itself, sounds fine, but does the choice of kernel really matter? What advantages would BSD have over Linux in such a project, and vice-versa?
MaxOS X, no?
Seriously, I've been using FreeBSD as my desktop in various forms for about 4 years now. It's nice, although my experience became a lot more agreeable once I partitioned my work into two bits--the "hardcore" stuff, like scripting, testing, compiling, sniffing networks, etc. and the "soft" stuff like doing presentations, writing docs, etc.
I do not like Star/OpenOffice, and the lack of something like Crossover Office (from CodeWeavers) running nicely on FreeBSD, despite Linux binary compatibility, has made me keep an XP box around. For some things, FreeBSD is just faster and better, and for others, XP involves a lot less knob-dicking around to get application xyz working.
So in short, yeah it's doable, yeah, it'd be nice (I've had _no_ stability issues with FreeBSD at all, and the whole thing is organized nicely) but it would need a lot of work to get it all prettied up for the masses.
That's honestly why I'm considering buying a Powerbook...
Cole's Law: Thinly sliced cabbage
Mac OSX is NOT free software. Also, you can run BSD on reasonably priced standard hardware. To run Mac OSX you need Apple hardware, which is double-priced, quirky, hard to find (relatively) and non-standard. You can get a PC notebook that is faster for half the price of a Powerbook.
Bruce Perens says:
This looks like someone who want to throw a few principles over board. O, yeah. Please let us not confront our proprietary overlords with something as basic as Free Software. Lets not call it GNU. Copyleft protects us and makes us strong, so lets throw it out as much as we can. Jeez. Cowards.Sure it would be nice to get popular really quickly. And if it was easy to change some laws all at once. But we have not done so bad. Several countries are now demanding free software when it impacts their relation with the public or in schools. America isn't one of the fastest to get those changes, but even in America some states are now seriously investigating free software. But changing laws and views takes time.
Please Bruce Perens come back to your roots. I was so proud when I read your "It's Time to Talk About Free Software Again" essay. Please don't let a million dollars put you back into the open source and proprietary camp again.
Now I can rest in peace knowing it was a poor attempt at trolling the front-page (something that doesn't even take skill these days, as slashdot editors are never reading entries anyway).
"Try using the FreeBSD kernel with OpenBSD"
True, although FreeBSD has had a lot more development time thrown at it recently with the upgrade to the 5.x kernel.
"It has very little commerical support for commodity hardware."
I suspect you mean that there aren't a lot of closed source drivers for hardware. The question then comes, do you really want a whole load of closed source drivers hanging around? To date, I think the most that Linux has got is Nvidia drivers and possibly some ADSL stuff. I *believe* that the Nvidia drivers run under Linux compat, but I've not checked, mainly because I don't really want my freebsd box to do that kind of thing.
"BSD has always been a geek os and will never reach the same popularity as Linux. Linux is everywhere."
Replace 'BSD' with Linux and 'Linux' with windows to see how fallacious this argument actually is. For shame.
"It is harder to install. You have to compile everything manually from a huge CVS respitory. It dosen't havge advanced Linux Packageing Systems such as APT, RPM, Portage, friendly GUI based autodectecting installers such as YaST, DrakX, Anaconda, etc."
The ports collection is not a CVS repository, and it's actually extremely easy to install FreeBSD. Open and NetBSD are completely different, however, and not for beginners.
Having said that, I've gotten on extremely well with the ports collection and CVS updates, although I don't class myself as anything approaching a hardcore geek.
"BSD is easy to make propreitery. You can take the code and make it propreitery. I can make Microsoft WinBSD XP and you can't do anything about it hahaha!"
Except ignore it. I fail to see your point here. Are you against freely available source code or the BSD license?
"It is not secure. People do not trust bsd to run it. Even OpenBSD, which claims to be "the securest OS ever made" runs on Solaris. They can make up excuses, but if you don't eat your own dog food, why should my Dog eat it?"
Nice, you picked out a good point there. Did you check the Freebsd or NetBSD servers? Did you not thing that not checking those out would drop you to the level of a troll? FWIW, Linux and FreeBSD tend to share the same security advisories.
"It has no support."
Au contraire. You can buy support, or avail yourself of the many newsgroups (ick), books or mailing lists without those dumbass kids trying to make you feel stupid. It's one reason why I'm really happy that FreeBSD stays under the radar compared with the religious wars over which 'distro is best'. I just can't abide this constant comparison with other OS's based on the 'whizzy' features that made Windows such a bloaty piece of crap.
FreeBSD does it's job, and it does it's job well according to my five odd years of using it.
"BSD are ex-linux geeks who coudn't take the fact that KDE and GNome made linux useable for the masses, so he ran and cried to his command line mistress!"
Ah, you're suggesting that the graphical user interface is the way forward? Have you considered windows 2003 server for your needs?
Oddly Draconis
Too cynical to live, too stubborn to die.
FreeBSD is for people who like Unix
Linux is for people who hate Windows
OpenBSD != NetBSD != FreeBSD
Loser
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
This seem a lot like an uneducated rant then anything helpful.
I have used freeBSD as my mail computer for years and it has been rock solid.
1. You do not need to compile everything. You can add packages easily, but compiling is as easy as typing "make install"
2. FreeBSD is far more secure then Linux and windows. There is a reason that major ISPs use an OS like FreeBSD, Not nearly the amount of bugs that are found in Linux and Windows. And it is far more stable. (Try reading about it)
3. FreeBSD is a Complete operating system, Unlike Linux which is a kernel that is packaged with someones distribution i.e. Red Hat. Free BSD is the whole system and it is so damn good that most Linux distros take large chunks of the code.
4. Every time I have personally compared the speed of applications running on FreeBSD to Linux (on the same machine) FreeBSD has been the winner.
5. BSD is the best operating system to run a network server like apache, or exim. Believe it. It's true. There is a reason yahoo runs on it. I have been a part of ten ISPs now and ALL of them have used *BSD for critical servers.
The fact is that the BSDs are geared towards being servers. So people who can actually run a server usually set them up. This means a cute GUI installer that holds your hand has not really been needed.
I think many of the Linux users would really like FreeBSD and I for one would love to see more desktop hand-holding in FreeBSD
is that someone could take UserBSD, and for a small fortune, make something better but closed source and patented. Then the work that everyone else put into it would be used against them.
Debian is a big project, and includes a GNU/NetBSD distro.
-Peter
nVidia drivers run *natively* under FreeBSD.
You can also run Linux OpenGL software via FreeBSD's Linux support without a recompile.
nVidia FreeBSD Native Drivers
What? I'm not a BSD user (Linux - not really a lot to do with it) but do you know what you're talking about? Abysmal sales? Err, FreeBSD doesn't sell software retard. Any BSD does make a cracking good server, and if you want to do anything seriously network related (large traffic, firewalling, bandwidth shaping) use a BSD.
This is a great idea. There are a huge number of technical advantages to building atop a BSD, rather than a Linux base for this sort of project, entirely aside from licensing issues. I've used both extensively, and even the best Linux distros are dramatically less stable and robust than the BSDs. (I've also found that those arguing otherwise have usually never really tried the BSDs - just managing to have installed them doesn't count - use in actual production does.)
But it's also quite clear, given Bruce's requirements for the UserLinux project, that BSD would be a better fit, since it is not a commercially hostile license. It's great that even a guy like Bruce now realizes that GPL-only licensing is the kiss of death for the kinds of large-scale commercial support such a venture needs.
I personally would be in favor of a modified BSD license that would add only one stipulation: that the code can never be placed under another more restrictive license, preventing the modified-BSD-licensed code from being relicensed (and thus effectively "stolen" from the community) under the GPL or similar viral licenses. In this way, it can be assured that truly free software remains that way and cannot be co-opted and limited to Stallman's twisted idea of "free".
"The future's good and the present is nothing to sneeze at." - Roblimo's last
But seriously, the idea does make some sense. I switched from Mandrake to FreeBSD, and actually find it easier to use. Yes, seriously. The biggest problem for a user of OSS is installing software, and FreeBSD's ports collection does a very nice job of handling dependencies automatically, finding the server that has the software you want, etc. It's similar to Debian's apt. I also like the fact that FreeBSD is a single OS, rather than being fragmented into a lot of different distributions; you can walk into Barnes and Noble and buy a book on FreeBSD, and it actually describes the system you're using.
As I see it, FreeBSD has a couple of advantages over Debian.
One is that Debian has set itself a very ambitious goal of supporting lots and lots of architectures. While I'm sure this is great for the tiny number of people who want to run OSS on obscure machines, I think it's also had the effect of making it impossible to create a good installer. Good doesn't necessarily mean GUI-based -- FreeBSD's is text-based, too -- but my own experiences with the Debian installer have been really really painful. Reading Debian's documentation for its installer is like reading code: "If you're installing on a Cray with a punched-card reader attached, do this, unless it's the first Tuesday after the new moon, in which case..."
The other big disadvantage of Debian, as I see it, is that they simply have a history of being elitist. Now there's nothing inherently wrong with elitism -- the alternative to elitism is sitcoms and top-40 music -- but when it comes to usability for ordinary end-users, they just don't seem to get it. (I say this after several attempts to get a working Debian system going. The third one resulted in a running system, but way too much stuff wasn't going to work right without many weekends of pain and suffering...)
Find free books.
How many times are you going to post the text of this same rant to Slashdot? Every time BSD is mentioned? Don't take it to the poo-poo and get over it!
I belive the success of Linux(and GNU) over BSD was because of the license also. Most of the opensource/free software developers want their work to REMAIN free. Face it, most of the people dont like their work to be taken away by corporations and made proprietory. If we look in to repositrories like freshmeat and sourceforge this trend is clearer. A good majority of the programs are released under GPL/LGPL. Even many of those who use *BSD as their platform also tend to follow this pattern more or less. Actually if BSDs were licensed GPL, then I think they would have continued to enjoy the developer base they had in the past. If UserLinux is to be dropped in favour of UserBSD, I think its going to enounter the same fate of BSDs.
http://www.nasirudheen.blogspot/
It's interesting to note the GNU project's essay about the X Consortium and Open Group's plan for developing X and an essay on GNU/Linux naming. Both essays cite some calls for chasing popularity at the expense of software freedom. I think we're better off as a community letting popularity take a back seat to freedom.
Your point is well-taken, but I'm not sure what roots you're talking about here. Bruce Perens is part of the group that started the Open Source movement--the movement that wants to "sell" Free Software by never mentioning the software freedom and only talking about the practical advantages of a software development methodology. They do this because they believe this will make the software easier for other people (chiefly businesses) to adopt. The advantages that the Open Souce movement touts come from having software freedom, so it's quite ironic that the Open Source movement champions the practical benefits of software freedom while distancing themselves from freedom talk. This leads to a built-in flaw for the Open Source movement: people can't justify choosing the software the Open Source movement wants you to run and develop if there's a proprietary program that better meets your practical needs.
The Free Software Foundation wrote an interesting essay on the differences between the Open Source and Free Software movements that also discusses shortcomings of focusing on the Open Source movement's goals.
Digital Citizen
I've often wondered lately, why Linux? Why didn't FreeBSD or OpenBSD take off the way Linux did and get the kind of corporate interest and deployment that Linux enjoys? The answer is interesting.
Althought it is true that a "UserBSD" or any BSD could be a very good product, having the same features in every way as Linux, BSD will never have the same success as Linux is having. There never will be any large traditional corporate sponsers of BSD. Why? The reason is the GPL. Although many proclaim the BSD license as a more liberal license, and one that is business friendly, companies that contribute to Linux, such as IBM or Novell will not touch it.
From a business standpoint, contributing to a BSD-licensed project makes no economic sense. To do so would be tantamount to subsidizing your competitors, such as Microsoft. Linux, on the other hand, is licensed under the GPL. IBM and Novell can contribute greatly to it, in order to build themselves a better platform to support their business. Any improvements they submit to the community benefit everyone, but no one can take their contribution and use it for their own proprietary projects. So the GPL makes it so that what's good for IBM is good for all, but that any benefit that others get from using IBM's code also comes back to benefit IBM again. Thus the playing field is leveled and fair. IBM actually gets the spirit of the GPL whereas SCO just does not.
Anyway, regardless of the technical merit of using BSD as a platform for a User distribution, because of the BSD license, there will never be the corporate backing needed to really ensure the rapid development needed to support an initiative such as UserBSD.
I am always wary of corporate dependence, but I feel that companies that work within the letter and spirit of the open source licenses are a great benefit to all, while they themselves can benefit. And this benefits projects like UserLinux.
Of course, the classic flawed argument people make is to assume that if Mac OS X ran on PC hardware, that the Windows lemmings would jump ship.
BeOS proved that generally speaking, Windows users aren't interested in switching to another OS, even if it's far superior to what they have. Furthermore, the adoption rates of recent versions of Windows demonstrate that in general, users don't upgrade their OS. There are way more people running Windows 95/98/Me than there are people running XP, and that's including all those corporate machines. The only way Joe Sixpack is going to upgrade to a newer version of Windows is if he gets it when he buys his next computer, and at that point he could just as easily buy a Mac.
This is one of the things Apple has got right. I imagine they did the market research and drew the obvious conclusion. The Apple stores even have people who will help you transfer all your data.
Yes, a few people, probably including anyone reading this, are prepared to upgrade OSs, or maybe even switch OSs. We're special. We're the 1%.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
The OpenBSD main site runs on Solaris because it's hosted for FREE at a sunsite.
The policy of the United States is worse than bad---it is insane. -- Ludwig von Mises, Economic Policy(1959)
This is hardly a funny comment. It is completely serious. Where are the metamoderators when you need them.
I just finished setting up a freeBSD web server and it was very easy. How hard is typing:
/usr/ports/www/apache13
cd
make install
apachectl start
And that was remotly so a GUI wouldn't have been any help anyway.
Linux is also for people who hate Unix. Unix is a nasty, primitive, backwards OS. All the worst parts of Linux are the things that come from Unix (the filesystem, for example).
Linux succeeds because of a willingness among its developers to not create another Unix. The original was bad enough.
Isn't it reasonable to think that the success of BSD as a server OS is due to its complete and utter lack of concessions to the desktop-user?
Frankly, I'd be much inclined to have BSD focus on rock-solid stability and security for server applications, and let Linux focus on GUIs and other user-comforts. That way each has a better chance of excelling at what it does.
I read through all that, expecting humour to be mixed in. As it drew on & on without anything funny, I then expected a punch line.
Maybe the moderators were making fun of you.
testing out my trending skills
Closed source drivers are the kiss of death
for security-conscious installation
environments. If I wanted to buy a pig in a
poke, I'd get Windows 2003 Server.
-I like my women like I like my tea: green-
Well, if you were deploying a few thousand appliance boxes of some sort, say file and print servers or firewalls or what have you, then you might be able to avoid the "does this expose us to a SCO lawsuit ?" dance with management.
( the thought of dealing with "managment" makes me glad I don't have a job just now )
So why is Bruce settling for the GPL'd MySQL?
Postgres seems to be the better one for a commercial friendly distro.
Promoting the BSD license is certainly a bad idea or any license which doesn't carry the requirements of the GPL. If the million dollar contributor doesn't like it, can him and his million with him.
Besides which the linux kernel surpassed the BSD kernel long ago. I certainly has more in terms of feature and capability, and gives nothing to BSD's kernel in terms of stability.
I'd almost think it's Microsoft or Apple making that Donation since they obviously have the intention of stealing the blood and sweat of open source developers and putting it in their proprietary code.
What We Can Learn From BSD
By Chinese Karma Whore, Version 1.0
Everyone knows about BSD's failure and imminent demise. As we pore over the history of BSD, we'll uncover a story of fatal mistakes, poor priorities, and personal rivalry, and we'll learn what mistakes to avoid so as to save Linux from a similarly grisly fate.
Let's not be overly morbid and give BSD credit for its early successes. In the 1970s, Ken Thompson and Bill Joy both made significant contributions to the computing world on the BSD platform. In the 80s, DARPA saw BSD as the premiere open platform, and, after initial successes with the 4.1BSD product, gave the BSD company a 2 year contract.
These early triumphs would soon be forgotten in a series of internal conflicts that would mar BSD's progress. In 1992, AT&T filed suit against Berkeley Software, claiming that proprietary code agreements had been haphazardly violated. In the same year, BSD filed countersuit, reciprocating bad intentions and fueling internal rivalry. While AT&T and Berkeley Software lawyers battled in court, lead developers of various BSD distributions quarreled on Usenet. In 1995, Theo de Raadt, one of the founders of the NetBSD project, formed his own rival distribution, OpenBSD, as the result of a quarrel that he documents on his website. Mr. de Raadt's stubborn arrogance was later seen in his clash with Darren Reed, which resulted in the expulsion of IPF from the OpenBSD distribution.
As personal rivalries took precedence over a quality product, BSD's codebase became worse and worse. As we all know, incompatibilities between each BSD distribution make code sharing an arduous task. Research conducted at MIT found BSD's filesystem implementation to be "very poorly performing." Even BSD's acclaimed TCP/IP stack has lagged behind, according to this study.
Problems with BSD's codebase were compounded by fundamental flaws in the BSD design approach. As argued by Eric Raymond in his watershed essay, The Cathedral and the Bazaar, rapid, decentralized development models are inherently superior to slow, centralized ones in software development. BSD developers never heeded Mr. Raymond's lesson and insisted that centralized models lead to 'cleaner code.' Don't believe their hype - BSD's development model has significantly impaired its progress. Any achievements that BSD managed to make were nullified by the BSD license, which allows corporations and coders alike to reap profits without reciprocating the goodwill of open-source. Fortunately, Linux is not prone to this exploitation, as it is licensed under the GPL.
The failure of BSD culminated in the resignation of Jordan Hubbard and Michael Smith from the FreeBSD core team. They both believed that FreeBSD had long lost its earlier vitality. Like an empire in decline, BSD had become bureaucratic and stagnant. As Linux gains market share and as BSD sinks deeper into the mire of decay, their parting addresses will resound as fitting eulogies to BSD's demise.
Simply put GPL licensed software is a better platform to build any kind of *free* enterprise. It just keeps everyone honest, period.
BSD - even being a very nice OS- is hurt by its license.
> 2. FreeBSD is far more secure then Linux and windows. There is a reason that major ISPs use an OS like FreeBSD, Not nearly the amount of bugs that are found in Linux and Windows. And it is far more stable. (Try reading about it)
Actually as you state in 3, Linux is just a kernel. There have been a few bugs here and there, but because Linux is just a kernel there have been no remote root exploits! Those would be in the daemons, which also can run on FreeBSD
> 4. Every time I have personally compared the speed of applications running on FreeBSD to Linux (on the same machine) FreeBSD has been the winner.
Are you interested in getting work done or do you constantly redo your work on the same machine in both OS's? Where are the numbers buddy? Come back with numbers.
> 5. BSD is the best operating system to run a network server like apache, or exim. Believe it. It's true. There is a reason yahoo runs on it. I have been a part of ten ISPs now and ALL of them have used *BSD for critical servers.
Ya-Who? Who the fuck are they? Let's not forget google is run on Linux!
Can I get an eye poke?
Dog House Forum
Nobody runs pure Mach. There just aren't
any apps for it, except the X server.
To solve that problem, the Mach developers
got BSD. They ripped out the virtual-memory
part of BSD, along with the low-level parts
of process/thread context handling. The
remaining disemboweled BSD then got grafted
onto the Mach microkernel.
The result can run both native-Mach and BSD
apps. I suppose MacOS Carbon apps are now
a 3rd type of personality.
I don't make it a habit to look up where someone previously associated with the Open Source Initiative might have said that it was "probably the biggest mistake I've ever made" just before I post. But I'm glad to know that he said this, thanks for pointing it out to me. By the way, the quote you cite is nowhere on that Slashdot page; it's from the grandparent article (see http://ask.slashdot.org/comments.pl?sid=37241&cid= 4000207).
Digital Citizen
Umm, I think you have Microsoft confused with Apple. Apple licensed Streams from Mentat, Inc, and created Open Transport from it. Actually it can be argued that OT is a better TCP/IP stack than the BSD-Sockets based stack now in MacOS X, but Classic MacOS has the stability of a Jello mold so the rest of the upside of MacOS X beats any advantage OT gives Classic MacOS.
As far as I know, if you run strings on WinSock you can see the UC Regents copyright string there. I could be wrong, but I understand this to be the case.
I may have missed someone else posting this, but it doesn't seem like anyone has brought it up. One of the key advantages with starting with the BSDL is that because it is weaker than any other open source licences you can then mould it to be whatever you want.
Someone has already pointed out that the LGPL works favorably for companies, they can be sure competitors won't take their code and use it a proprietry system, but at the same time it can mingle easily with their proprietry programs.
Now if you start with GPL programs you can't backpedal and then release modifications under the LGPL, but you can take BSDL code and then release new versions under the LGPL. So if the intention is to move away from a GPL style licence, they will either have to start from scratch or start with a weaker licence like BSDL.
The average user, sitting in front of a KDE desktop, would neither know nor care which kernel was actually in use, however the choice ought to be available. Nor would he or she care which licence applied, it is likely that most people have some GPL software on a BSD system and vice-versa. That is OK for end users, it is only if you are coding or distributing that you even need to read either licence. Most people will also have stuff licensed under Apache, or Netscape, or Sun........
It is fundamentally important to maintain choice, for those who wish to exercise it.
AFAIK (and I do run Linux, OpenBSD and FreeBSD, but don't usually look at the internals) the system call mechanisms of Linux and BSD are orthogonal, so you can put both into one kernel without them colliding. So most versions of Linux have the BSD kernel capabilities compiled in. I think the reverse is probably true also. The small amount of code which should be needed to do this can be written from scratch to avoid needing to put Linux GPL code into BSD or vice-versa. The only difficulties are the few system calls which do not have an exact correspondence between one OS and the other, these few will need most of the work. Any code that uses a standard subset of calls can be easily compiled for either OS without change, it is usual for applications to have conditional code to cover the difficult areas. All of this is much like the days when you had Unix from AT&T, Microport, SCO (first incarnation, part owned by Bill), HP, Plexus.... (some now long gone). The similarites far outweighed the differences, as they still do.
Sticking to a Unix-like OS is the way forward, since apart from the bug-ridden and insecure products of the Convicted Monopolist, all other current OSs are in a tiny minority, numerically at least. There is at the same time a high degree of commonality in the API set, and usually the user interface, and a great deal of diversity in the code base. The commonality benefits users and programmers, the diversity improves resistance to hackers and virii, (OpenBSD on my firewall, Linux on my server for example) and also avoids being tied to one hardware monopoly (although that one is being constantly challenged by AMD and the positions might reverse in the 64-bit world). There is far more commonality between Linux and BSD than between Win 9x and NT/2000. You really need to work very hard to make an application work on all versions ow Wincrash, it is a fairly predictable set of ifdefs etc for Linux/BSD or even the hated SCO.
If BSD were to become as popular as Linux, it would actually strengthen the position of both against the Monopolist. One does not detract from the other, rather they complement each other.
It is purely as a result of the well-known stupid court case between AT&T, later Unixware IIRC, and BSD (UCB), that Linux came into existence. That happened about the same time. I for one was wanting a Unix I could afford, and was impatiently waiting fore the court case to finish, so BSD would become available. I looked at the awful thing Linus was playing with (forgotten the name now, by Tannenbaum, not very efficient and no memory protection) and decided it was a waste of time, having wasted my money on the book. I waited a bit longer, and first to go public was Linus, but variants of BSD were not very far behind IIRC. So it may be just by accident that Linux is the leader (no bad thing, by the way), the positions could have been reversed if the legal process had been quicker. I sometimes wonder if Linus would have bothered, if he could have obtained BSD at the time, but am glad that he did. We need both. It is sad that the present owners of BeOS killed it off. I know that an open-source clone is in development, but BeOS woud have also been a worthy addition to the small group of free operating stystems which are worth using.
i mentioned userlinux going lgpl only on the no kde in userlinux thread before i found this.
why would anyone be suprised to learn kde was denied entry given that it is actually designed to be lgpl?
in terms of the topic, the hardware support for linux is most excellent. more importantly, linux is a corporate name.
although being gpl does add some annoyances, its not like a software app where you likely have to expose your whole apps source code. if your really in a pinch on some hardware, abstract kernel interfaces with userland drivers always work. (tivo w/ nfs...)