Slashdot Mirror
Will Open Source Solaris Kill Linux?
Scott_Blayney writes "This guy contends that now that Sun is releasing Solaris 10 under an open source license, Linux will begin to wane in popularity. To quote, "Linux will probably not grow much beyond its current market share of about 10 % leaving Red Hat and especially Novell with a big problem."
"
Frankly, an Open Source Solaris stands a better chance at killing Solaris than it does Linux.
The author starts the article by declaring "Let me start by saying that I'm not a technical expert. Although this might be a handicap it could very well be a positive in looking at the issue of Sun versus Linux in the server business from a decision maker's point of view." He's basically lost most of his credibility there, because a good decision maker needs to bring in technical people to explain this.
Though I understand the PHB philosophy of needing a brand name on their *nix product, I don't understand what's wrong with Red Hat or Novell now?
Moreover, it's setting itself up for patent lawsuits once the usual suspects start going through the newly open sourced code. Not that it's an agreeable method, but it's a fact of life.
The points the author makes are weak:
o The Solaris software is of proven quality and at least equal or better then Linux and the open source model will assure that it stays up there.
An open source model assures it of nothing, unless they get good contributors....but:
o Because Sun by default is the only designated party managing the open source software, there will be no risk of a version bonanza with multiple incompatible versions.
No risk of a version bonanza with multiple incompatible versions? Does this individual not recall that this is the company who had put out Java 2 (1.2) and Java 5 (1.5) before a real version 2?
o Sun has a proven reputation in terms of quality of support. This should be at least as good or better then that of the Linux supporters.
Sun's support isn't bad no, but IBM's been an awesome proponent of Linux - including Red Hat's 3.0 offering. Big names do know Linux, and work with you on it. Not to mention that Google is an awesome knowledge base for Linux users. And how much does it cost to search google vs. a support incident with Sun?
I don't buy that this will affect Linux's growth in the server market any more than the stray bullet SCO lawsuits. Open Source Solaris will definitely invite more hobbyist interest, but I don't understand why any of this makes it a more viable server product.
If Sun makes inroads to the desktop, then I would agree. But if Linus finally picks a GUI and starts up a desktop linux fork, I will disagree. But neither are going anywhere until they innovate, simplify and give desktop users a reason to use it.
/.
When I say "desktop users", I'm not talking about anyone here on
Life is the leading cause of death in America.
I think this will only be true for shops that would have choosen Solaris anyway. The draw of open source is more than just cheap software. It has to do with familiarity and the availability of expert administratiors, developers, etc. Solaris will still have the same number of developers and, especially, administrators for awhile.
Agreed. It's more than just OS capabilities. It is also an emotional thing. And I believe on that issue Linux has the mind-share.
This article is a pointless exercise, but since my alternative to commenting on it is to go do actual work, let's consider:
In terms of Sun's conversion to open source: Too little, way too late. If Sun had pulled this seven or eight years ago they might have had a chance at stopping Linux before it got rolling, but frankly I don't see any way a system developed as proprietary in-house software is going to be able to come out and defeat the reigning champion of the OSS movement. Everything else aside, consider the simple fact that you can't go open source overnight. Look at how long it took to turn Mozilla into something useful -- by the time the open source version finally hit the mainstream it was hugely reorganized and largely rewritten. This process took years to complete, and that was just for a web browser. So unless Sun's programmers had a *lot* more discipline than Netscape's (doubtful) and a lot fewer tangled licenses (impossible), forget about it right there.
Aside from the huge initial development issue, consider the business side: What possible reason is there to think that Solaris is going to be able to come out and impact Linux's market share? The author himself uses language that doesn't support the idea that more people are going to switch to Solaris as a result of this. In fact, the most logical outcome of the points made in the article is that Sun *may* slow the erosion of their install base over to Linux. Not exactly killing Linux there.
Then we get some more of the same 'ol. The reasons he states in the article for the impending demise of Linux could have been (and were repeatedly) written ten years a go and are just as wrong now as they were then -- sure, any of the fears he listed *could* happen, but they haven't.
To sum up, this article makes incorrect assumptions, starts from a place of poor understanding and leaps to unsupported conclusions. The greatest revelation is creates is probably that YEALD needs to find themselves a new COO.
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
If Sun opens Solaris, and if my mind serves me correctly they are switching to using AMD CPUs instead of their SPARCs, what exactly will Sun be producing?
Solaris can't be "open-sourced" because of the confidentiality parts of the Unix license agreements. They'd be in trouble for the same things that got IBM a lawsuit and started Groklaw, leaking trade secrets.
main(O){10<putchar((O--,102-((O&4)*16| (31&60>>5*(O&3)))))&&main(2+ O);}
LN2 is cool!
Exactly.
Never forget that there will be differences between Solaris code and LINUX code (well, we hope). Even if every bit of software were successfully ported, there are enough people that know LINUX inside out that don't have time to learn the Solaris code to the same level.
The hardcore purists on either side will argue until they're blue in the face about how theirs is better (hey, isn't that why we're all here?). That fight alone will be enough for some people to keep it alive.
Let's not forget that even Sun has and supports LINUX. Their Java Desktop is a bunch of Java tools with a Sun setup on SUSE. I'm sure someday they'd drop that support, but for now, even they don't think that Solaris is enough of a LINUX killer to release the product on Solaris.
Additionally, even if Solaris were ported hard enough to get the same or more hardware compatibility of LINUX, they dropped all LINUX support in favor of Solaris, a guy from Sun came out to your place to install it for free, and made sure you were comfortable before he left, some people love the penguin too much...
Now, before you start the flaming, let me state I'm not an idiot. I'm not particularly partial to Solaris or LINUX. I use Solaris on my Sun Blade Sparc workstation and an x86 as a server. I have SUSE and RedHat running, and a lab machine on which I've also installed Fedora, Debian, Slacker and a few others. And I have an old machine I use as a thin client to both Solaris and LINUX using Knoppix and DSL. Yes, I have some Windows machines, too.
My wife does think it's too much, but what does she know? (Really, honey, I meant it as a rib, not a slam...)
End the FUD
RTFA. The torroise did beat the hare in the end. :D I would say that linux is the tortoise anyhow, slow and steady. Just gaining ground. What do I know though. I am a BSD guy.
m l
http://www.childrenstory.com/tales/indexhare.ht
Later on he goes further and says he is "not a technical person". WTF. Why would anyone choose an operating system based on the advice of a non-technical person? He even seems to think that being non-technical makes him more qualified for giving advice on choosing an operating system.
Should I let a vegetarian tell me how to cook a steak? Should a let a blind person choose what color to paint my house?
His logic seems to say that the only people who are well qualifed to give advice in a field are the ones that aren't experts in that field. Maybe he is looking to get a job with SCO.
We don't see the world as it is, we see it as we are.
-- Anais Nin
Yep Solaris only runs well on Sparc and Opteron processors, with very limited other hardware.
Linux runs on anything to one degree or another. One can force linux to even run Windows Drivers to load hardware that Linux drivers don't exsit. it's standard at Nvidia. Nvidia uses a closed Binary driver, and a source code wrapper. It's how they get around the little details of GPL.
Solaris, well is there even a nvidia/ 3D hardware driver for it?
i thought once I was found, but it was only a dream.
The 'do exactly as I say and nobody will get hurt' attitude is why many people hate microsoft.
Only in slashdot circles. Look outside of the headspace around here, and you'll see that most of the world is, at best, indifferent to Microsoft.
They feel comfortable with Windows because that's what they learned on in college/high school. That's the system that their nephew showed them how to play solitair on, etc.
Options are great, but some people don't need them. Sometimes they just confuse the issue.
Imagine a car that shipped with manual, automatic and slap-stick transmissions, and how confused you'd feel when you sat down the first time to see a gear shifter marked "P R N D L", one marked "R N 1 2 3 4", and a clutch.
I don't need no instructions to know how to rock!!!!
I did... 5 years as a sysadmin.
And when I see how many Sun Servers were replaced by Linux boxes, I guess it's Linux that's killing Solaris by forcing Sun to Open Source it.
Trolling using another account since 2005.
I know those are his sentiments, not necessarily yours, but the reason new rollouts will go with linux is that linux is the hot sport for new app development. You know, that little advantage that has allowed Microsoft to dominate the computing arena the last 10 years or so.
It will make it easier to get management to switch from Windows to Solaris than Linux. You bring up Linux to managers in an enterprise and they instantly start pointing out lack support contracts. Management wants someone whose feet they can hold to the fire when something goes wrong. Sun is a well known name so it will be an easy sell.
Also Solaris is more of a real enterprise class operating system. Linux is working on it, but not there yet. Solaris has decades of use and refinement, it is a solid OS that scales great as the load increases.
OK, imagine with me...
/atleast/ on par if not far ahead feature-for-feature with their competition. Yet, they arn't adopted more than their niche(but rabid) markets.
Sun has, in Solaris 10, arguably the most powerful core OS of any on the market for general/server/production use(not counting embedded, rt, or mainframe systems and I know not enough about them). They have some killer stats, powerful new features, and amazing development coming out. This all sits under xorg+gnome for a gui...
ok, transition time... please don't kill me, just hear me out...
Apple(*ducks*) has arguably, the most powerful gui on the market. Feature wise, simplicity, elegance and consistant... Very powerful, though not perfect, ofcourse. Darwin is a nice enough system but doesn't have much more the 'average' features. Nothing makes it have a selling point other than it exists and it is freebsd-like.
What both companies have is an amazing half OS that by all accounts is
I propose a merger. Sun is marginally(25-50%, i think) larger than Apple income wise and whanot. Merging the two systems, the Aqua and solid gui on top of Solaris 10 core system. This would take time and would coincide with the System 11 release(ironically) for both companies.
To make this really work though, they need to migrate. Procide highend workstations and servers(as Sun and Apple always have) but offer a lowend as well... Sun offers x86 systems for lowend servers, and has talked about a ppc port of solaris. Personally, I think ppc is superior, with IBM's work they have made it very powerful. But an AMD64 system would be fine as well.
The companies need to flood the markets with their OS, even if it is on amd64, only, at a less-supported(community support for the free version)... I mean, send free boxed copies of their software to every university with a cs department, free of charge. They need to literally hand this software out to everyone they can. Free download for personal or non-profit use (minimum) and have a good resource/community site for support.
The more open the system the better, but both companies use open source where it makes sense in their systems, not because they are sucking up, but because it is the best choice for the product.
PS. yes, I know Apple is a hardware company. I wrote this on my 12" pbook. Sun is a hardware company too, which makes demand for an operating system that(for full use/support like sparc or ppc's cooler features) require a more highend(expensive) machine.
IBM does this with Linux, they are a hardware company using a generic operating system on highend machines.
The way I put it is: "When all you know is shit, garbage tastes like heaven".
Most people have only known Windows, and they think that that's the way that computers are supposed to work. I've thrown Linux at just about every roommate that I've had, and they've generally been very happy with it -- one of them turned into a downright Linux crusader (much to my surprise).
For most of the people that I offer Linux to, their only real question is "will I still be able to do the things important for me". If the answer to that question is yes, then they're pretty happy to move over from the 'dark side' of computing.
Free Software: Like love, it grows best when given away.
As an employee for a very large ISP (the one that's dispised), I can say that Linux is going to win this one. Sun has been the primary platform for the last 5 years, and now many of those old 400mhz Sun boxes are coming off lease. Their replacments are all Linux. Why? Simple - It's disposable. The hardware, the software, everything. Once we use it up, we throw it away like a bent paperclip. Sun simply can't compete. For one thing, their prices are still insane (not Apple Hardware insane, but close). Their support, while excellent, is way too expensive. In fact, it's so expensive, that third parties offer discount Sun support. HP does. And third, their hardware underperforms.
Compare that with run of the mill dual proc 3Ghz Intel boxes, which are super fast, cheaper than a used Camry, and abundant. And Linux runs great on them.
Linux has won this contest, and this is Suns attempt at making amends to all those customers they raped for the last 10 years by overcharging for support and hardware.
Depends on what you mean. Others have already commented on hardware support, so I'll talk about speed. While developing Solaris 10, Sun's standard was, "If it's faster on Linux, it's a bug." They explicitly set themselves the goal of beating Linux in terms of performance.
Now, I haven't seen a head-to-head performance shootout on identical hardware, but I wouldn't automatically assume that Linux is ahead...
"Open Source" is a trademark... the OSI should be sending letters to all the journalists and to Sun saying that they must not use the "Open Source" mark until the license is finalised, available to the public and OSI approved.
It doesn't even make sense to call this Open Source because it is air-ware... nothing is real, Sun can talk for years but they should not be allowed to claim Open Source until they deliver some goods.
I am, sadly, going to have to agree with this statement. My group is currently going through a very large problem becuase our default distro (SuSE Enterprise) and the distro of one of our primary software providers (RHEL) are not binary compatible. This has led to quite a lot of frustration since our software provider cannot be convinced to release source, and must now QA releases for 2 disparate Linux Distros.
What does etc stand for? Not et cetera?
"Will Open Source Solaris Kill Linux?"
No.
First off, I'd like to state that this isn't exactly a convincing argument.
Secondly, I'd like to point out that Solaris is one helluvagood operating system. I used to work at a place where we moved our production database from an Ultra 1/170 to an Ultra 2/2200 with no glitches. We didn't recompile anything, we didn't have to reconfigure anything. We moved the boot drive and RAM from one chassis to another, moved the SCSI cable for the external RAID chassis to the new box, and powered up.
Needless to say, it was easier than moving a Windows installation by a huge margin. It was easier than any machine migration I've ever witnessed. Everything automatically recognized the fact that we'd gone from a single processor box to a multiprocessor box.
Now I realize that Solaris on SPARC had the advantage of going from one Sun-engineered box to another, making it likely that the underlying chipsets and such were identical or at least compatible, but the point is that the OS was rock-solid on a single-processor box and it was rock-solid after being migrated to a dual-processor box with no configuration changes.
That said, in spite of all of Solaris's goodness (and there's plenty of it), I seriously doubt even an open source Solaris will kill off Linux. Why? Freedom (as in speech).
Let's face it: Linux isn't the only open source-licensed Unix or Unix-like OS available. So why the hell is it so popular? Obviously, Linus is a huge reason why Linux is popular, but the GPL sure doesn't hurt. Contributors to Linux know that their contributions are being used by others who are required by the terms of the license to contribute any improvements they make back to the kernel. Nobody can take the source and close it off from the rest of us. Nobody can build proprietary extensions to the kernel.
There's no way in Hell that Sun's license will be anywhere near as free as the GPL. This is why open source Solaris can't kill Linux.
sh and ksh are separate for a reason. Solaris sh (more so than other sh's) is *very* stripped down
/lib and /usr/lib are separate for similar reasons. /lib holds system libraries, while /usr/lib holds user-installed libraries. It makes threat containment easier.
That's like saying that an egg is a "very stripped down" chicken.
My, what will those brilliant military-grade security wizards at Sun invent next? Passwords? Access protection? The mind boggles.
This is the sort of stuff Schwartz is talking about when he mentions military-grade security.
Schwartz talks a lot, but he seems to actually know very little.
While I'm a GNU/Linux user myself, I can't understand all those people who insist on pitting Linux vs. BSD, Debian vs. Redhat, or Gnome vs. KDE. If opensource Solaris is as free as Java is now, I doubt it will make much headway. But if Sun licenses Solaris the way it licensed OpenOffice, then I guess Linux is in danger. But who would care then, but the few Linux überzealots? Redhat is already packaging OpenOffice into their OS offerings. I fancy in the future, they might come up with a co-branded Redhat Solaris. Maybe we could even hear of Linus contributing patches to make sure free Solaris interoperates well with legacy Linux.
Solaris in it's current and past incarnations on x86 hardware is slow and stupid. The Solaris x86 project at Sun never had any backing before Solaris 10. Now, Sun plans on selling Solaris (support licences) on their hardware which, guess what, isn't going to be just SPARCs, but also x86 servers. So, the Solaris engineers (who are still some of the brightest people left at Sun), have done a lot with a lot of support to make Solaris for x86 a viable platform.
Yes, I do work for Sun. No I don't work in the Solaris group. No, I don't particularly like Sun, especially the upper management. But Solaris is a top notch OS, and if Solaris 10 works as well as I've heard it does (from engineers at Sun... not marketing), then I for one am happy that they are going to release it under some sort of "open" license. I don't expect it to beat Linux, but I think if it's done properly, the curiousity of all those geeks out there will get the better of them, and they'll be pouring over the source as soon as possible.
*yawn*
Do you really think the hardware vendors who now have linux in their lines will let Sun eat their market share? Even with x86 Solaris from Sun as a token of collaboration in the IA platfrom, you can be sure the hardware guys will be pushing back. Not to mention Intel.
Correct me if I'm wrong but isn't Solaris full of copyrighted, licensed SysV code? Solaris is also a SCO License holder. How on earth could Sun even attempt to OpenSource it?
- Will Solaris suit for desktops?
Historically no. Sun has never made their own KDE build, leaving those who favor kde to rely on dedicated (non-sun) volunteers.
As for their GNOME buld, last I checked it has been lagging about 3 revisions behind... Our office dumped Solaris sun blades for Macs last year and we're quite happy. And less you think I'm a mac zealot, I run linux at home--even on my sparc...
Acquiescence leads to obliteration
Solaris is more for the enterprise, Linux is more at a server or workstation for personal or small business.
Really? Our enterprise is dumping Sun and SGI boxes in favor of Linux. Over 60 new Linux servers in the past year (and that doesn't include the personal desktop machines). The payoff is in commodity hardware instead of high-priced boxes and reduction of the very expensive support costs. There are tradeoffs in performance, sometimes worse but generally better. Sun would have to come up with a really good reason to get us to go back.
This has happened before, hasn't it? The creator of Minix shot his mouth off then, just like the Sun COO is shooting his off now. There isn't any grounds for this. Linux has a strong base because it -is- a "Hacker's Project". It means you can probably get something 'fixed' faster in the kernel than the development cycles of other projects. So, whereever Linux might be deficient now, it gets to the point where it'll be better. A year ago, 2.4.x was stable and there were massive problems. A year later, we have a stable 2.6.x tree with a supreme amount of functionality and performance, and patches which can leverage even more of that performance if we feel we should be on the bleeding edge. That's one of the big reasons which keeps me away from FreeBSD on my Desktop, let alone Windows. I wouldn't be surprised if Linus has a similar response to this as he did to the "Linux is Obsolete" dig. Though who would blame him?
"A Goddess rarely smiles for she is forced by others to be an island unto herself." - Zephiris
There is no mind share to keep. What will happen is what has always happened. UNIX religious wars, which became *NIX religious wars. Solaris will start to look like a very advanced stable and robust Linux distro to big business users and developers. SuSE will become Solaris's kid sister and fast tract to merge with Solaris x86.
Eventually Novell and SUN will merge and make an honest woman out of SuSE. 64 bit development will flood the market and AMD 64 and Ultrasparc cpus will start flooding the market. Intel and Redhat will start flooding the market with Intel 64 related stuff. And poor microserf will have to kick XP into the 64 bit world alot sooner the Ballmer's timetable plans for.
Suddenly all those over 35 years old coders will admit they used Solaris and can now still claim they are always Linux programmers. Solaris will at last get a decent user interface and CDE will be dropkicked back to what ever icy part of hell it came from.
Second Bubble here we come.
Sorry about the writing. Robot fingers, you know? Cliff Steele in DOOM PATROL #23