Submitted by “There’s been a lot of talk on lists and blogs about an exchange between Linus Torvalds, Jonathan Schwartz and Theo de Raadt regarding licensing and documentation. It all started with a ‘cynical’ message from Linus about Sun’s motivation with regard to Open Source. Jonathan Schwartz responded by extending Linus a dinner invitation. What? The romance was briefly interrupted by a message from Theo pointing out the doublespeak.”
If you look at Schwartz’s original blog post, Theo was the second or third post. He raises some good points Sun and being open. But if you read the rest of the comments, most people pat Sun on the back for their stance on FOSS. What Theo has to say (very much valid IMHO) goes unnoticed. When I first heard about Theo, it was in articles and posts which referer to him as a cranky programmer with a chip on his shoulder. Most people seemed as though Theo should be ignored Given the lack of any reaction to what Theo had to say, it looks like many people still disregard Theo, his work, and what he has to say. It’s a shame because there is more to Theo and OpenBSD than the very strong opinions.
EDIT: Having looked at the blog post recently (after my initial post), many more comments have been added. Some of them support Theo and his comments. Still, the praise for Sun eclipses the criticism of Theo (one post even tries to refute Theo’s POV).
Edited 2007-06-16 22:07
yeah, i was noticing that too. I’ve talked to people that write Theo off as that… I like the guy though. He has strong opinions and isn’t afraid to tell them too you
Anyways, if what Theo talked about is true, which I’m sure it is, thats pretty pathetic.
If you look at Schwartz’s original blog post, Theo was the second or third post. He raises some good points Sun and being open.
Indeed. Both Linus and Theo raised some good points about Sun’s continually strange behaviour over these things.
I didn’t actually see anything particularly cynical or nasty about Linus’ post. He just told it the way he saw the situation, using examples such as the talk about porting ZFS to Linux and the difficulties involved (quite apart from any coding difficulties).
Linus then went on to say that Sun was merely looking out for its own self interests here, this wasn’t a bad thing per se, no one should be surprised and no one should buy too much into the ‘Sun open sourcing things’ hype.
Theo was spot on in pointing this out:
“There are two operating systems which surprisingly do not run on the Sun v215/v245 — Linux and OpenSolaris. OpenSolaris?? Yes — Sun isn’t even open enough to give the OpenSolaris community enough documentation to support their new machines.”
OpenSolaris not running on Sun’s own hardware? That shows you that Solaris simply isn’t open, if you can’t get from the source code to a running system.
Sadly, all we got in reply from Jonathan Schwartz was:
“But I disagree with a few of your points. Did the Linux community hurt Sun? No, not a bit. It was the companies that leveraged their work. I draw a very sharp distinction….”
Well, that wasn’t one of Linus’ points, but that’s a YES then. Linux did hurt you, even if you’re trying to split hairs by claiming that it was indirect.
“OpenSolaris scales on any hardware, has built in virtualization, great web service infrastucture, fault management, diagnosability”
Blah, blah, blah, blah blah. Let’s fill some space here so I can get to the end quicker.
“This has nothing to do with being afraid of the community (if it was, we wouldn’t be so interested in seeing ZFS everywhere, including Linux, with full patent indemnity).”
That doesn’t answer Linus’ direct point. The main obstacle to Linux having a ZFS implementation is because of the license incompatibility and the patents they feel the need to then hold on it. Patent indemnity simply isn’t required.
He didn’t deny that ZFS code would not be released under any GPL version.
“But most of all, from where I sit, we should put the swords down – you’re not the enemy for us, we’re not the enemy for you.”
Linus never said that. He just pointed out your strange behaviour.
We want to work together, we want to join hands and communities – we have no intention of holding anything back, or pulling patent nonsense.
Then do it then. It’s up to you. Make ZFS available under a compatible license, and certainly don’t hold patents on it that put question marks over a compatible implementation.
And to prove the sincerity of the offer, I invite you to my house for dinner. I’ll cook, you bring the wine. A mashup in the truest sense.
That’s nice, but I think the above might be a better idea first.
All we can ask as free software advocates is that software and hardware vendors establish and act upon plans to become more friendly to the free software community by opening specification and interfaces. In this respect, Sun is surpassing our expectations by also steadily moving to free software licensing for their own software products. They have not finished carrying out their plans, and I believe that this is part of the “duplicity” that Theo aptly notes.
What we see today is a large software vendor in the middle of a transition period that has already had a positive impact on the status of free software in the IT industry. Not all of their specifications are open yet, and therefore it is still hard to support and enhance their products with free software. Intel isn’t completely open either, but the free software community has been much more accepting of their analogous free software stump speeches.
I think that Linus’ “cynical” email was a mistake. It forced Schwartz to smooth things over without having anything to announce publicly. The challenge of rallying the troops as a leader of a free software project is that your enemies and “uneasy partners” can read the mailing list. For Linus to assert in the same email that license compatibility with OpenSolaris is mutually beneficial while also questioning Sun’s intentions regarding ZFS was a step forward and then a step back.
If the intention was to goad Sun into making a statement about their plans for ZFS, then it failed. All we got was a vague assertion that they would like ZFS to run on Linux without any explanation as to how they would make this possible. Obviously Sun has nothing to say about ZFS that we don’t already know. In the process, Linus has shown that his instinct is to respond to progress with cynicism.
Both Linus and Schwartz are technologists that have risen to prominence and share the burden of having far too many people analyzing their public comments. In this respect, Linus could learn a thing or two from Schwartz. When your counterpart announces progress, praise it as such. If he fails to capitalize on an opportunity to make progress, express regret. If he says nothing at all and you express distrust or discontent, expect him to try to take the matter offline.
As silly as it sounds, the dinner invite makes sense. Schwartz can’t have Linus publicly badmouthing Sun’s progress toward openness, and Linus can’t be so dismissive of Sun’s efforts. Schwartz needs to tell Linus that they are both on the same page as far as embracing the practical advantages of free software–it’s just that Linus had the advantage of starting from scratch, whereas Schwartz has to steer a ship.
I’m sure that Schwartz would appreciate a vote of confidence from Linus. But failing that, Linus should have just noted that license compatibility with OpenSolaris would be a good thing and left it at that. The rest of the the “cynical” email had nothing to do with the GPLv3 relicensing issue, and it simply should not have been said. I’m not sure that an apology is completely necessary, but I wouldn’t take it off the table–the dinner table, that is.
I agree with you that linux do not have a “corporate level” in his communications. But i would not be surprised if he simply don’t want to.
I agree that it does not seem to be professional, but “professional”, here, since everything is analyzed, means “with a strategy”. I don’t think that linux wants to play the industry game with its machinations. I think that he wants to keep his role of “leader of amateur project” and the naivety that goes with.
Why ? because linux hackers wont like the duplicity and unclear positions that any strategy in communication implies. Hackers are hard enough to manage, and since linux project do not have any mean to keep people working with them, you must not corrupt the soul of the project. And i’am pretty confident that Linus himself hates to calculate the impact of what he says to business.
The other point is that linus do not represent any corporation. He does not represent any business. So then, what are the reason to be happy of sun openness other than technical ? none ! simply none … And since technical (in computer and legal terms) are not really satisfying, linus says “i’am not satisfied”. This is the nature of open source, this is why open source is what it is. Business and open source mix well together under the condition that you do not try to keep your open source project under the business rules. You can run an open source project and try to do business with it, but the opposite will always fail.
This is the “here is my code and do business with it if you want” behavior. The developer might be happy to be paid for his job, but whatever happens, he don’t want his work to be conditioned by your business problems, especially if they are political. This is the main reason of open source success, and this is why i want linus to continue to be rude with business. Playing corporations games won’t pay, because it is not the how open source works.
[edit: typos]
Edited 2007-06-17 10:22
Linus said that there is no reason to move Linux to the GPLv3 unless Sun does the same with OpenSolaris and ZFS in particular. He implied that Sun will not use the GPLv3 if Linux goes GPLv3. Therefore, he claims, the ball is in Sun’s court. If Linux does not go GPLv3, it will be because Sun is not interested in license compatibility. Sun is the enemy.
This is Linus playing defense. He started by restating the fact that he doesn’t agree with the philosophy of the FSF. But many of the Linux kernel developers do, and they represent an important bloc that could leave if OpenSolaris goes GPLv3 and Linux does not. This is why he’s dialing up the rhetoric on Sun’s nefarious intentions concerning the GPLv3. Linus had the choice of whether to take the lead on GPLv3 or attack its proponents, and he made his choice.
Linus is wrong about Schwartz. They are so close in terms of their positions on various aspects of free software. I really do think they should have dinner and discover just how much they agree upon. Putting Linux and OpenSolaris under the same license (or mutually compatible licenses) would be a landmark moment in the history of free software. It’s bigger than any issue Linus may have with the de-fanged TiVo language.
License proliferation is perhaps a greater threat to free software than software patents. While we can only do so much about software patents, we have an opportunity to make a powerful statement against license proliferation. The first step is for Linus to realize that license proliferation isn’t his pet issue that he can use to attack Sun. What the free software community doesn’t need right now is for Linus’ cynicism to get in the way of a potential breakthrough.
What’s the basis for that statement? There hasn’t exactly been an outpouring of support from the kernel community to date for v3. On the contrary, the core developers (an inarguably important bloc) have taken a public unified stance against v3. Aside from that, contributing to openSolaris will require assigning copyright and permitting code use under CDDL as well as giving Sun leeway to license your code in the future as they see fit.
There always seems to be this mildly arrogant undertone from the FSF community that Linus is too obtuse to accept the “spirit” of the GPL, so they fail to accept that developers (and those commercial interests paying the developers’ salaries) may be contributing specifically because of his attitude, not despite it.
I’m really not convinced that if Sun goes ahead and dual-licenses openSolaris under v3 that there will be any mass or even miniscule exodus from the linux kernel camp. A significant amount of linux developers are paid, and those that contribute voluntarily are just as likely to be attracted to the open nature of the development model as they are by the four freedoms. Even allowing that some developers may be contributing in the “spirit” of the four freedoms, I can’t see them jumping at the opportunity to sign their code over to Sun with no guarantees that it will remain free and open.
If unwilling to submit patches/work upstream, then the alternative is that the v3 proponents wind up forking a v3-only openSolaris and keeping it unavailable to Sun or CDDL developers. I think that would potentially be more damaging and frankly, pointless, than the existing license clash.
Hell, I think the apparent inability of the linux kernel to shift licenses due to self-ownership of code underscores the strength and resilience of the license that Linus chose, and the reason he chose it. Some people will be drawn to that, as much as some will be drawn to v3 for philosophical reasons.
Theo’s main gripe against Sun is that it hasn’t released specs for an old UltraSparc chip so OpenBSD can’t support it.
Theo is spot on with his criticisns: Sun has poor harware documentation for a chip (OpenSparc) that is supposed to be a free (as in freedom) platform.
Another poster commented that ZFS has patents and Sun won’t give patent indemity for it.
Actually, the OpenSPARC chip is incredibly well documented. What Theo is actually banging on about is the documentation for the Schizo pci bridge chip. Totally different issue.
My beef with Theo is that he seems to be completely unable to try to approach the Schizo bridge chip issue from a different route – all he ever seems to do is yell that Sun isn’t open because of that chip… rather than trying to build a constructive working relationship with people from Sun who could actually help him. His preferences in this matter do him no credit.
Regarding ZFS and patent indemnity, ZFS is licensed under CDDL, which most definitely does have a patent indemnity grant as part of accepting the license. The “problem” is that if you cannot accept CDDL then you don’t get access to the IP that’s licensed under it.
Don’t spread fud, get the facts and spread those instead.
Forgot to add this url – OpenSPARC is available at http://www.opensparc.net and not only can you get the verilog code for the RTL, you can get the higher level documentation too.
My beef with Theo is that he seems to be completely unable to try to approach the Schizo bridge chip issue from a different route – all he ever seems to do is yell that Sun isn’t open because of that chip… rather than trying to build a constructive working relationship with people from Sun who could actually help him. His preferences in this matter do him no credit.
I don’t know, Theo has been pretty consistent about having an agenda to get hardware companies to open up their specs. It isn’t so much that HE wants/needs the specs for that chip, he is trying to get Sun to open them up for _anyone_.
Building constructive relationships is all well and good, and it might have gotten Sun to give him the specs (probably under NDA or something). But it won’t convince them to change _policy_, that requires a different kind of tactic.
Being that Sun is a very image aware company that is trying to put on a good face for the OSS community, openly berating them is probably as good a tactic as any.
If these people are going to complain about patents, why don’t we (CDDL/OpenSolaris) people complain about how the GPL deliberately inhibits cross pollination between projects.
There is a code orgy going on between the various *BSD’s, and Linux is the odd one out because Linus chose to use a licence that deliberately separates itself from the ‘mainstream open source movement’ – remember kiddies, the FSF and GPL do not equate to open source.
GPL’s projects don’t seem to be limited in number by that fact, nor do they seem to be lacking in developers.
If projects licensed under non-GPL licenses are having such a code orgy, then why do they need (or even care about) code which is licensed under the GPL?
Edited 2007-06-17 08:09
If they don’t seem to he limited, then why do we have people whining, whinging and complaining over ZFS and other CDDL and how evil OpenSolaris is because GPL projects can’t pinch things from it. Cry me a river.
We don’t need code from GPL projects – its your camp, the GPL camp who want code from our projects! please, our camp is more than happy to share, if we could do like wise from your code. You want a one way transfer, we want cross pollination, where code can be shared amongst all projects.
All I’m hearing from Linux is a giant sucking sound of code but giving very little in return to the parent projects whom which they leach code from. Its the equivalent of a proprietary vendor embracing BSD code and refusing to give back the changes.
“Its the equivalent of a proprietary vendor embracing BSD code and refusing to give back the changes.”
Which is what proprietary vendors actually do. Maybe that’s why so many FOSS developers prefer the GPL, to keep proprietary leechers away
GPL in effect IS a proprietary licence; GPL can take what ever it wants, but don’t expect the ability to take from a GPL project – again, that isn’t sharing or cross project pollination, its a giant sucking sound coming from one project.
CDDL proves that you don’t have to have that sort of proprietary/exclusivity licence to protect your IP from being raided – and the simple fact is, time and time again, those who DO embrace *BSD for their proprietary code do not give back because they have no relevance to the main project.
They give back fixes, but if there is a feature which is exclusive to their piece of hardware, what do the larger community benefit from something that won’t benefit them at all? How does open sourcing code which has no relevance to the project going to benefit the project itself beyond just adding *MORE* code to something.
Like I said, I’m confused; we have GPL advocates on one hand scream about sharing and open source and yet their licence does the exact opposit! it doesn’t promote it, infact, Stallman himself said not to call GPL licenced software as ‘open source’. There seems to be a great disconnect between those who advocate GPL and FSF software, and those who are actually the core of the GPL and FSF movement – the ones who actually created it.
Which is why Theo was complaining that Sun was thinking GPL and not BSD for OpenSolaris?
Are you retarded? Theo did not mention licensing the code of Solaris at all.
So when he was talking about how Sun should strive to make ZFS the defacto unix filesystem and talking about how the licensing issues with that he wasnt talking about licensing?
As for the ‘retarded’ bit… nice to see you picked up your social skills from Theo. I’m bothering to reply to you this time but language such as that will just get you a -1 in the future
So, you are retarded then? Theo did not mention ZFS.
[i]If projects licensed under non-GPL licenses are having such a code orgy, then why do they need (or even care about) code which is licensed under the GPL?<i/>
What we are talking about is open source, why spend unnecessary effort and reinvent the wheel? Apparently that viewpoint is fine as far as lifting BSD code into Linux. The other way around does not work without major license issues (a compiled FreeBSD kernel that includes a GPL licensed subsystem will thereby be GPL licensed).
There are lots of opensource projects that have stalled and died because of reinventing the wheel and lacking developers, that includes a lot of GPL software.
There is a code orgy going on between the various *BSD’s, and Linux is the odd one out because Linus chose to use a licence that deliberately separates itself from the ‘mainstream open source movement’ – remember kiddies, the FSF and GPL do not equate to open source.
Yeah, that sums up the GPL attitude nicely. I you use a license that’s incompatible with the GPL is your fault for choosing a crappy license and you deserve to be publicly ridiculed and hounded for your lack of freedom.
It’s stupid – how can you defend openness with such a closed mind ?
http://blogs.sun.com/jonathan/entry/one_plus_one_is_fifty#comment-1…
Edited 2007-06-16 22:30
It’s interesting that someone have answered Theo.
http://blogs.sun.com/jonathan/entry/one_plus_one_is_fifty#comment-1…
People should stop thinking about what people did wrong in the past and start seeing what people are doing right in present. History must not be forgotten, but we should not live in the past. By the way, Sun has been selected as one of the most ethical companies in this world. How many companies can say that??
http://www.sun.com/aboutsun/media/features/ethical/
Presently, Sun is using very dramatic language about how friendly they are with the FLOSS community. Theo is simply pointing out that this is not totally true. Sun is doing great things, and they should be thanked. But this does not give them reason to overstate what they are currently doing.
PS – Ethical when describing a corporation takes on a totally different meaning than an ethical person.
“People should stop thinking about what people did wrong in the past and start seeing what people are doing right in present. History must not be forgotten, but we should not live in the past.”
In general, I do agree. But people need to learn from the past in order not to repeat the mistakes done there. This essential knowledge will make people doing things right in present. It gives them the additional ability to judge educatedly about present developments (advertised “new technologies” that we had in the 80s).
People should stop thinking about what people did wrong in the past and start seeing what people are doing right in present.
They’ll only do the right thing so long as it serves their bottom line. Not that you shouldn’t support them when they are doing the right thing, but always watch your back, because as soon as it serves their bottom line to screw the open source community, you’ll quickly find a knife in it.
Note: This is not an attack on Sun specifically, but pretty much any for-profit company will do the same. Even Microsoft will be your friend if it has the potential to make them more money than being your enemy.
Just follow the money
if they were so they’d keep their word and release the sun lighthouse design applications for humanity to use.
What Sun is currently undergoing – as Schwartz points out – is a very difficult move. And I think Linus was spot-on with his criticism; one just needs to read the whole message. Apparently Linus is saying that Sun is acting like any company would do, which results in both good and bad stuff – and then Mr. Schwartz points out how great his company is in the follow-up…. like any company would do.
Sun won’t be GPLv2 or v3 overnight. It’ll take time and evaluation. Let’s say they just threw away all patents and gave away the source. How would that effect competition? Who would win or lose? Sun? Community? The economy? End-users? Heck, who knows.. and that’s why they’re so careful about this. If ZFS is one of their greatest assets, then they’ll probably be extra careful about that.
Just like my comment here, this discussion (at Schwartz’ blog) really lacks relevant content, but the interesting part surely lies in the facts of how Sun opens up – the order of their license migrations, and I think it’s a clear indication of how they value their assets…. word up Linus.
Dinner Invitation ?
Who cares about OpenSolaris and OpenSuse and Fedora ?
#That’s just a dirty trick to get people — and to engage
# free labour (exploitation) for the commercial distribution(s). Do not buy it.
#
Please, all readers, engage with Debian/Slackware or FreeBSD instead.
Why should Linus have a dinner with the people that are trying to make money with the Open Source Community ?! stealing developers for their own commercial purposes ?
I tried OpenSolaris and it’s a piece of bloated garbage, no real software difference from Linux / FreeBSD !
Do not buy it.
If anyone has a good point, it is David Miller. David is a very active Linux Kernel developer.
http://vger.kernel.org/~davem/cgi-bin/blog.cgi/2007/06/13#sun_theo_…
” Sun isn’t even open enough to give the OpenSolaris community enough documentation to support their new machines”
I’m not surprised corporations like SUN are in control, they will do whatever they want and some things that they do will not make sense. so with that in mind i can’t cry for everything corporations do that seems unfair.
Edited 2007-06-17 08:26
Linux acts like it has the MORAL HighGround here.
That’s a Laugh.
You’ve allowed Linux to be POLLUTED with the MONO project.
NOW company’s are PAYING Microsoft to install Linux.
Linux keeps his Big Fat Mouth Shut Tight about those Developments.
Interesting Times Indeed.
It’s NOT LInux’s job to ASSURE ONLY IBM get paid for Linux.
The Linux franchise is now nearly DEAD with Microsoft getting REVENUE from Linux Installs.
But, you guys KEEP BITCHING about SUN.
You’re truly living in an alternative universe, aren’t ya?
keep on trolling
Microsoft’s involvement is a chance for Linux to become an OS that everybody would want to use, without ideological restrictions that are understood by 1% percent of the population. Microsoft is boosting Linux, in fact.
I am still waiting for decent UNIX that I could run on AMD/Intel hardware without having to have C compiler installed. Havenig access to source code is no priviledge to me because I realy don’t care staring at compilers output. I don’t mind paying for software because my time is more expensive.
I am pretty sure that 99% of computer users in the world would agree with me, at least because they have no idea what C/C++ compiler is. That makes all those ?PL and BSD licenses much less relevant. In fact, it makes me tired.
Give me UNIX that just works and I’ll pay for it.
Have you tried calling SCO? They’ll happily sell you a fully supported closed source Unix for AMD/Intel hardware. By all accounts it’s a fairly reasonable unix, perhaps not as feature rich as Linux, but it should cover your basic Unix needs.
I was working as SCO admin couple years ago. It is very much a server OS. It is too expensive. I could by MAC with OS X and have some money left. MAC is OK, but it is a problem with hardware upgrades.
“I am still waiting for decent UNIX that I could run on AMD/Intel hardware without having to have C compiler installed.”
Having a cc installed does not imply you have to use it. Many UNIXes provide cc as a means to help yourself (as long as you’re able to, this requires a certain education).
“Havenig access to source code is no priviledge to me because I realy don’t care staring at compilers output. I don’t mind paying for software because my time is more expensive.”
May I buy some time from you?
To use the source (the force) is interesting to the ones whose time is worth doing this, such as developers, tweaking enthusiasts and people who want to gain knowledge and to educate theirselves. In fact, education is not a waste of time.
“I am pretty sure that 99% of computer users in the world would agree with me, at least because they have no idea what C/C++ compiler is.”
Users don’t need to know. Furthermore, they don’t want to know (anything).
“That makes all those ?PL and BSD licenses much less relevant. In fact, it makes me tired.”
In fact, licensing is not irrelevant because of the people abusing licenses. To the average user, of course, licenses are irrelevant. Just think about the huge pile of pirated software around.
“Give me UNIX that just works and I’ll pay for it.”
Consider the commercal UNIXes, such as Solaris, IRIX, or HP-UX.
For now, you may try PC-BSD, which is derived from FreeBSD, a UNIX. No need to compile anything. Just click on the pretty pictures (and maybe you see the dancing elephants). I think this is what you’re searching for. Maybe you want to give PC-BSD a try.
May I buy some time from you?
In fact you can, if you need a custom business application, and if you don’t mind Java or PHP.
Yes, I am about to try PC-BSD. I was running an older release, but switched to FreeBSD. PC-BSD was a little crude then. I will give it a try as soon as I get some time to play with it. Elephanst are nice, inteligent animals, although I preferr pigs, because there so many good things to eat made from them :-). Pigs are intelligent, too, although not when roasted.
IRIX and HP-UX are forgotten, more or less. I have a Solaris 10 on other partition, but I still need compile a few things, like DVD player. And I like neither Gnome
nor CDE.
I am putting a lot of hope in the latest PC-BSD.
In fact, PC-BSD is particulary suitable, because it has MSIE prepackaged, and I always need to test my applications on Explorer, they never run without adjustments. I am still going to need clumsy MS OS to run SQL server.
“May I buy some time from you?
In fact you can, if you need a custom business application, and if you don’t mind Java or PHP.”
No, I just want the time because I need it to relax. Put the time into an SAE or send it via e-mail.
“Yes, I am about to try PC-BSD. I was running an older release, but switched to FreeBSD.”
I tried PC-BSD earlier, DesktopBSD too, and I found it appealing to “newbies”. My neighbor is using it for a longer time now, he’s completely happy with it. Personally, I prefer a “real” FreeBSD because PC-BSD definitely is not designed for my needs. I would have to deinstall nearly everything first… and then install the stuff I need (which is not in the PBIs)…
“PC-BSD was a little crude then. I will give it a try as soon as I get some time to play with it. “
Since the 1.3+ versions, PC-BSD is a quite good solution. Sadly, you need to install it in order to try it. But it’s worth the time.
“Elephanst are nice, inteligent animals, although I preferr pigs, because there so many good things to eat made from them :-). Pigs are intelligent, too, although not when roasted.”
I prefer rats, because they’re intelligent, clean, and they smile. Of course not if they’re roasted via xratroast. =^_^=
“IRIX and HP-UX are forgotten, more or less.”
HP-UX is still in use in the commercial sector or in large scale data storage facilities, but in principle you’re right, they’re not used very often today. IRIX was very nice at its time, I still like my SGI Octane. It did stuff like video editing when the x86 PC was still beeping in its cradle.
“I have a Solaris 10 on other partition, but I still need compile a few things, like DVD player. And I like neither Gnome nor CDE.”
Maybe Solaris’ GUIs are not everyones case, but its power (under the hood) is awesome and still straight forward in OS technology (zfs, dtrace etc.).
“I am putting a lot of hope in the latest PC-BSD.”
Enjoy it! Especially its growing amount of PBI packages available is a good reason to try. You may – if you want – still use precompiled FreeBSD packages (pkg_add) and the FreeBSD ports collection (make et al.), if you need to tweak applications (such as mplayer’s various options via Makefile.local).
I’ve seen that there is a IceWM package for Solaris, I am going to try it. I know that Solaris is the most advanced OS today. I deploy applications on customer’s servers all the time, but if I ever deploy my own server, it will have Solaris installed.
My latest PC-BSD was 1.2-something.
What I expect of my PC is:
– Postgres and MySQL
– Apache and PHP
– Java and Tomcat
– Mozilla (Firefox)
– MS formats compatibile office
– Wireless support
– Bluefish / Quanta / Netbeans
– DVD and MP3 player
– Sylpheed or Thunderbird
That’s what I use 99% of the time.
“What I expect of my PC is:”
Before I comment on your expectations, I’d like to say: Most of them are not related to the OS itself, they are additional software. While OSes like the “real” BSDs do come as an “OS only”, PC-BSD comes with many preinstalled software that is targetted at the home user. You surely are not a usual home user, so some of your needs would require you to install precompiled packages (pkg_add) because no PBIs are available.
“- Postgres and MySQL
– Apache and PHP
– Java and Tomcat
– Mozilla (Firefox)”
All available as PBI, except tomcat which is available via ports / packages.
“- MS formats compatibile office”
I’d prefer standard compatible formats. It’s MICROS~1’s turn to be compatible, not the other way round. Refer to OpenOffice as a good solution (with the tendency for migration towards standard formats (ODF)), if you really need to stick with this old fashioned stuff.
“- Wireless support”
Except the other points mentioned, this is a concern of the OS. Wireless drivers are availabe for FreeBSD and that’s why they’re available for PC-BSD, too. Be sure to buy hardware that is compatible. Driver support is something the hardware vendor is responsible for, maybe by publishing documentations and specifications of by offering drivers for the particular OS. Refer to FreeBSD’s hardware support list in order to find out more.
“- Bluefish / Quanta / Netbeans”
Quanta and Netbeans via PBI, Bluefish via ports / packages.
“- DVD and MP3 player”
Choose what you like: KDE’s Kaffeine, [kg]mplayer, xine, xmms, ogle, Amarok… there is no “the media player”. The choice is yours here.
“- Sylpheed or Thunderbird”
Sylpheed? Wow, I thought I’m the only one using it! Of course available as PBI. If you use the same storage method (mbox, MH), you can use both of them together.
“That’s what I use 99% of the time.”
If these are your requirements, PC-BSD definitely is for you. As long as you like KDE, but there’s no need to stick with KDE if you don’t want, just install WindowMaker, Gnome, Metacity, IceWM etc. as you like.
And if you’re familiar with how things work in “real” FreeBSD, you’ll find PC-BSD a very good solution, because you can do the “low level stuff” (CLI and config files) and the “high level stuff” (GUI and clickityclick) without any problems.
PS. Looks like both of us are using similar stuff 99% of the time.
I don’t know abot PC-BSD, but all platforms I used required DVD player compilation (Mplayer, Xine, VLC, Ogle). They never worked good as a binary package. Well,
I guess that I will have to live with it. I’ve heard that Media Player on Windows, which is precompiled, has a lot of problems itself. Never tried.
I’ve seen icewm package for Solaris, I am going to try it. What source of precompiled Solaris packages do you recommend ?
“I don’t know abot PC-BSD, but all platforms I used required DVD player compilation (Mplayer, Xine, VLC, Ogle).”
I don’t know what platforms you have used but I have never had to compile (on OpenBSd or Ubuntu) MPlayer, VLC or Ogle to get DVD support.
If that is all you require, then Windows will handle all that with ease. And stuff like wireless support and MS office support will be handled with much less hassle than any Unix. You have already said you would prefer a closed source OS so is there any particular reason you won’t use Windows?
Yes, true, Windows handle it all. But, I am a UNIX person, if you know what I mean. My first serious job was on UNIX platform, and I need consistent logic of UNIX system to be able to use system. Faced with Windows I am like albatross on the ground, I can barely move.
Windows is the product that is easy to learn, and, by no means, easy to use.
Let me explain how I work. I have IDE, like Bluefish open in one workspace. Squirrel SQL is open in other, so I can monitor the impact on database caused by the code I write and test. I have a browser window with documentation in a third workspace. Application is tested in a fourth workspace. A window manager is IceWM, with taskbar, so I can switch between windows. In fact, OS desktop is a kind of big IDE, which is more flexible than conventional IDE’s.
During testing I utilize UNIX command line to customize the testing environment, to search the code with grep, or even modify it with awk.
I can’t work that way in Windows. If I used it, I would be forced to run some monolithic IDE, which can be customized in a limitted way.
I am just used to work and think in UNIX way, and I don’t want to change that.
Technically windows supports both bash and awk and multiple workspaces.
I still don’t get what your problem with GPL and BSD licensed operating systems are though? Why do you want a closed source OS. What do think that will gain you over an open source OS?
I have no problem with those licenses. Just trying to say that I have no benefit from them, since I avoid compiling things. If I don’t compile, what is the use of the source, which is available ?
And, most people don’t compile, too. They have no use of the source, as well.
Everybody is talking about the source, and nobody uses it, except 1% of users, that’s the point.
No offence but your view of this subject is a narrow-minded and extremely myopic. On a planet which has more than a billion PCs in use, and 10 times that number in embedded appliances licenses do matter.
An every single being on face of this planet is within their reach and that includes you. The sole difference between a developer and other people lies in fact that developer stands on the first line of a license’s scope. Software is a value, directly on indirectly used by everyone. Every human suffers from software failures, and benefits from its goods.
Every license promotes some values. Some license will benefit the developer, some users, some community, some software itself. A good license will promote many of those values. Successful licenses usually have mass appeal on many levels and do promote sharing of some kind (binary or source code or knowledge). This in turn leads to better software, or cheaper, or other kinds of impact.
For example almost all major software companies do have special EULA-s for software used in academic education. This leads to better education. Or GNU license which excludes the developer side from chain-of-trust and this makes it ideal for security appliances. Exactly because of this side-effect is Linux used by NATO as standard platform for communication channels. Wouldn’t you agree that this kind of usage is of extreme importance to any citizen of any country? I think you would.
To state that you do not have value from license is similar to stating “I have no use for free press because I am not a journalist”.
Decades have passed since software left math universities or military labs and become one of central pillars of society comparable to electricity, communication or books. So what if no more than 1% of population is cappable of writing sofware!? No more than 1% of all people know how to build an electric engine, and less than 0.1% is capable of writing a good book. But everybode enjoys the values of their work. In this software is no different.
In modern world licenses do matter. They matter to everyone.
Edited 2007-06-18 18:13
I am developing business applications for living. For my consumers I am a developer, but for lower level things, like OS, drivers, I am just a consumer.
I don’t see anything wrong with it. Software is just a mean to get things done faster, better and with less effort, it is not a goal itself.
So, I am having a perspective of consumer, because that’s what I am. I need certain functionality, and I am willing to pay for it. If some license conflicts with my needs, then it is bad for me.
Ok. First of all I do not treat software as gift from God. In most cases indeed it is just “a mean to get things done”. But as it leads to higher productivity of work, has great sociological impact and keeps airplanes in the air it is very important part of modern society.
Now back on subject. You speak about developer-consumer rift in the lines “only developer sees the license; it does not matter for consumer”. Well it isn’t true.
Look at the Apache License. In contrary to vast amounts of end user software published under the GNU license, majority of projects under Apache license are aimed only at developers. Does an end consumer have benefits from developer-only open source projects as Jakarta, Eclipse or Apache server? Well of course it does! Without these tools, costs and development time would be far larger. It’s obvious fact. And besides lower costs there are other aspects which matter to end user.
The next question is this: are these benefits properties of used tools or are they a property of license under which these tools are published?
It’s both!
Tools are used because some their qualities make them desirable. Low quality tool will not be used no matter how perfect its license is. However, a good tool does not come from thin air. It is made by a group of people during large timeframe. In order for tool to become a reality a good license is mandatory. Good license is the one which promotes desired aspects of that particular tool. World is full of good ideas which never became good tools exactly because of failure of its license to promote development of those tools. BSD for example (OS not the license itself).
Thus licenses do matter to an end user. And when the end user is your bank, your pension fond, your state, your health insurance or swimming club of your daughter license matters to you.
Edited 2007-06-19 20:49
What planet are you on? Sun sells Ultra 20 AMD64 workstations starting at the sub $1000 price tag, and SXCE/SXDE/Solaris 10 all supports it.
If you want a UNIX for your AMD/Intel machine, then purchase Solaris plus a support policy from Sun.
have to remember that Theo theRatt is an ass.
Generally, that does seem to be the case, but not this time. I braced myself when I clicked the link, but was pleasantly surprised. That was possibly the most mellow posting I have read from him. It appears he was wrong about the driver issue, but this was a great change of pace from his normal inflammatory attacks that he does. It seems like he has been lightening up lately.
Nice to see you linked to the flame feast creator that is Theo Raadt whilst ignoring the reply by David to the matte of PCIe support:
Interesting, apparently everyone else in the world knows about the PCIe chipset except little old Theo Raadt. Yes, there is a time and place for banging chests, but that is after all polite and reasonable avenues are exhausted. Simply doing a sermon on the mount expecting people to come to your beaconing call is arrogance in the highest possible way.
… SUN has to offer. And I fail to see why people dance around it like a golden calf. _Everything_ else SUN has to offer does not follow the high standards Linux has set in this respect (GUI, speed of kernel, GNU userland, drivers, support of major players etc.).
And yes, I think Theo raises some valid points. There is a lot of hype around SUN and zfs these days. It still remains to be seen how honest SUN is while working together with the community. Let’s face it: SUN is a business. And they have interests they hope to protect with patents.