For some time now, Sun has been trying to push its way into non-proprietary Unix markets, and Solaris 10 is its crowning achievement. An abundance of innovative new features, mostly aimed at administrators rather than users, contributes further to Solaris 10’s value proposition. An interesting balance of administrative features and support for new hardware implies Sun is trying hard to maintain its current niche, as well as move into server and workstation territories where Linux has encroached. Read more here.
Solaris 10 Gets Ready to Roll
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
If they granted FreeBSD it’s Java license. Another aspect of fitting into the FOSS world is to share and help others.
and it’s too little, too late. This won’t help save solaris from RIP.
As Gates told a few weeks ago, “In ten years it will be just windows vs. linux”. I completely agree. And the pattern is right out there.
1- It’s too late. The situation would be different if Sun make this 2 or 3 years ago.
2- Sun would be smarter if Solaris could use more linux-standards like rpm (or deb), GRUB as boot manager, Fedora’s Anaconda as installer, apt/yum/urpmi for install/update/upgrade, etc. They could transform Solaris into a complete GNU/Solaris system, changing linux kernel by Solaris kernel. The Solaris kernel ans some related utilities/filesystems would be an atractive to people who need a enterprise-supported unix-like OS.
3- While Solaris was not free software, people will not see reasons to contribute for it.
> If they granted FreeBSD it’s Java license. Another aspect of fitting into the FOSS world is to share and help others.
The only reason the FreeBSD license was revoked is to award another one with revised clauses, there are absolutely no concerns from FreeBSD accroding ot its own people. Sun’s intententions are absolutely benevolent in respect to FreeBSD. In other words stop spreading uncertainty about things you do not fully understand.
> It’s too late. The situation would be different if Sun make this 2 or 3 years ago.
It is not much different than 2, 3 or 10 years ago — price of licensing and support and TCO longterm are the deciding factors in choosing software. Solaris 10 addresses these concerns in spades and has a pretty significant edge over Linux as Solaris is much cheaper than Linux in respect to both licensing and support. Solaris is a win-win situation compared with Linux. The only edge Linux has over Solaris is the hype being spread by the rather clueless Linux fanboys.
> Sun would be smarter if Solaris could use more linux-standards like rpm (or deb), GRUB as boot manager, Fedora’s Anaconda as installer, apt/yum/urpmi for install/update/upgrade, etc.
What is the poing of switching to inferior technology when you already have something that is both superior and proven by time? Solaris already has better packaging and patching technology than Linux. Why fix something that isn’t broken in the first place and is better than the alternatives?
Sorry for my ignorance, but since it’s been already sometime that people are talking about Solaris 10 as an open source OS, can somebody tell me under what license has Solaris 10 been released? Or hasn’t been open-sourced yet?
If the latter is the case, then it’s not very honest to talk about it as an open source OS.
Again, this is not a rhetoric question; I really don’t know about it. So please, somebody clarify the situation.
Klaus
“It is not much different than 2, 3 or 10 years ago — price of licensing and support and TCO longterm are the deciding factors in choosing software. Solaris 10 addresses these concerns in spades and has a pretty significant edge over Linux as Solaris is much cheaper than Linux in respect to both licensing and support. Solaris is a win-win situation compared with Linux. The only edge Linux has over Solaris is the hype being spread by the rather clueless Linux fanboys.”
Solaris can be cheaper than RED HAT linux, Suse Enterprise or other enterprise-level linux distributions. Sun can explore this market but there is no advantage to switch to Solaris for little and medium companies that already use linux.
Linux development is cheap and it is not smart to be different of linux at administrative and program views. If Solaris use the linux style of administration/programming there would be more developers and users interested in switch.
Sorry, but Solaris is late and now linux is the STANDARD in unix development/administration/instalation.
“What is the poing of switching to inferior technology when you already have something that is both superior and proven by time? Solaris already has better packaging and patching technology than Linux. Why fix something that isn’t broken in the first place and is better than the alternatives?”
Can you point these “inferior technologies” ? Install free software on linux is much more easy because you have much more ready packages on the major linux distributions.
Solaris has advantage only on BIG computers but linux is much more friendly and flexible.
…The only edge Linux has over Solaris is the hype being spread by the rather clueless Linux fanboys…
What is the poing of switching to inferior technology when you already have something that is both superior and proven by time? Solaris already has better packaging and patching technology than Linux. Why fix something that isn’t broken in the first place and is better than the alternatives?
Would you please contact the companies who switch from Solaris to Linux and let them know immediately that they make a very BIG mistake? Also, would you mind telling IBM that it’s really stupid to support Linux and they really, really need to look at the superiority of Solaris?<br)
In fact, I would go as far as suggesting that you immediately start a campaign to inform the clueless IT around the world that it would be a costly mistake for them switch from Solaris to Linux or to even consider using Linux in their business places, especially those with mission criticals.
It would be a disservice to mankind if you do not make it your responsibility to spread the gospel of the superior Solaris to uneducated mass. Do it now and do it fast before it’s too late to stop the hype from spreading further. Please step forward and do the the world that big favor.
Solaris could reign on the server market and be big on the desktop if they’d ported to x86 10/12 years ago and done the work Gnome or KDE have done to date.
At the time, Windows was a nightmare without solution (NT4 would come only a few years later) and Solaris ran happily 4 users environment and Sybase on a Sparc 20 with 128MB or RAM. Most large banks had Sparc stations by the hundred, let alone engineering departments in most companies. And it was already fantastically stable.
But they were blinded by the astronomical margins they made (and still do) on their hardware and thought they were very smart, shunning the humble PC. Now it’s going to cost more bilions than they have to come out of the hole they are in.
Good luck.
I agree completelly. Apple is following the same wrong way on its MacOS X strategy. Apple only look for the proprietary and expensive hardware who sells and its market share is decreasing.
It is fun that the technically worst american company (Microsoft) succeed while (technically) good companies like Sun, Apple, IBM and many others failed.
The only hope for Sun is to open Solaris kernel and embrace (and not extend) linux technologies.
CDE is a bad joke as GUI, Solaris installer sucks and there are better packaging and updating formats like apt/portage/*BSD ports. I repeat: the only (temporary) advantage os Solaris are some kernel features for scalability. For typical 1 or 2 CPUs machines linux is easier and faster.
IBM is a services company. So it makes sense for them to use GNU/Linux. Despite the fact that the code is licensed under the GNU General Proprietary License. The sell services, so it doesn’t matter to them. They fight Sun on one front and Microsoft on the other. That’s why they throw cash at Java. Linux is getting there on big iron because IBM and SGI are throwing money and engineers at the problem. Solaris is still a very robust system on high end servers, though.
As far as installation, packaging and patching goes, Linux is
way beyond Solaris. And, on x86, Linux’s hardware support is light-years beyond anything except Windows.
I’d be tempted to say that’s it’s easier to admin a Linux box as well but Webmin, a contribution of the OSS world, has made them equally easy ( or difficult ).
Yes, we’ve come a long way from the days when linux was all about “fighting a monoculture”. Now the linux fanboys adore monoculture – they want a linux monoculture and even though they may be almost entirely ignorant in the most breathtaking way on topics like solaris, IBM, Sun, and the IT business in general, this will not stop them spreading FUD and lies in order to see linux installed in a monopoly position. Linux zealots: haters of diversity (and pretty ignorant too).
As far as I can see linux and solaris aren’t even competing. linux is for low margin use on cheap x86 cloned boxes to run a mail server or php+apache and the likes on. Solaris is not in that market at all. Its at an altogether higher margin level where linux barely plays at all (at this level IBM tends to use AIX). Comparing linux and solaris is like comapring apples and oranges. If you have a mission critical and expensive piece of software to run, you don’t run it on free software. You don’t run a a multi-multi-thousands database like oracle on free software, for god’s sake – this is why the linux version of oracle accounts for about 1% of oracle’s revenue.
I suppose nobody will ever convince the monoculture-loving linux zealots that different users have different needs, and that linux will never, ever be able to serve all of them. In fact, it will only ever be able to serve a relatively small amount. And hooray for that. Variety is preserved.
The linux bubble will burst when it runs into this reality. But I fear linux zealots will never join the reality-based community..
“IBM is a services company”.
IBM is an everything company in the IT world. They pretty much do it all although they sold their disk manufacturing to Hitachi and, very recently, their PC desktop unit to China’s Lenovo Group.
“GNU General Proprietary License” – PUBLIC, not Proprietary
“They sell services, so it doesn’t matter to them” –
How many services companies do you know spend a billion dollars enhancing, marketing ( the only TV spots for Linux I’ve ever seen were from IBM ) and porting someone else’s OS.
The only place where I see IBM’s commitment to Linux coming up short is on their PCs and notebooks – I’m not aware of a Linux factory-install option.
> It’s too late. The situation would be different if Sun make this 2 or 3 years ago.
It is not much different than 2, 3 or 10 years ago — price of licensing and support and TCO longterm are the deciding factors in choosing software. Solaris 10 addresses these concerns in spades and has a pretty significant edge over Linux as Solaris is much cheaper than Linux in respect to both licensing and support. Solaris is a win-win situation compared with Linux. The only edge Linux has over Solaris is the hype being spread by the rather clueless Linux fanboys.
Solaris cheaper than “Linux”. You gotta be kidding. Various GNU/Linux distrobutions are freely downloadable. Support may be another matter, however as the Linux market-share increases so will the number of companies offering support for it and thus making support for it cheaper through increased competition.
Its not too late for Sun. They still make an excellent product , have great support and have a huge number of customers. Solaris 10 has many features that are not available in linux making it very attractive and I for one will run it and most admins I know plan on doing the same.
The one BIG mistake Sun is making is not GPLing it. This is the fatal flaw in my opinion and this is why it will never be as successful as linux in the long term. Few people are going to want to contribute to Solaris under Sun’s terms and this is where they will lose the war with linux. At this point I consider Solaris superior to linux but because of Sun’s stupid decision for licensing Solaris it will never command the popularity that linux has across a wide range of areas and will eventually die off.
> Solaris cheaper than “Linux”. You gotta be kidding. Various GNU/Linux distrobutions are freely downloadable.
Incidentally Solaris is also freely downloadable. On the same note RedHat and SuSE for that matter are going to charge you a lot more for support and the license than Sun will for Solaris. So yeah, Solaris is cheaper than Linux.
> Solaris installer sucks and there are better packaging and updating formats like apt/portage/*BSD ports.
If you judge the Solaris installer by the flashiness of the icons than may be Linux has an advantage, but on a more applicable note JumpStart install is light years ahead of anything there is under Linux — you can roll out Solaris on hundreds and thousands of machines in your infrastructure with minimum effort (RedHat’s kickstart is pretty crippled in this respect). Solaris packaging and patching (especially patching) facilities are also superior to anything under Linux.
> Few people are going to want to contribute to Solaris under Sun’s terms and this is where they will lose the war with linux.
I think the whole point a lot of people make that Linux will outpace the competition because of the community contributions is utterly misleading. Sun was quite successful outpacing Linux and actually widening the feature/technology gap, which according to Linux zealots should have narrowed — even if the whole community doesn’t just jump onboard with contributing code to Solaris, it is not going to stop Sun from continuing to beat Linux on the technology front. Linux kernel development community is not all that large and not all that productive — according to Linus’s own words there are only about 200 of more or less active contributors to Linux kernel worldwide and only a few dozen major contributors, the rest of the community are just “dipping the toes” but doesn’t actually contribute anything useful. Sun can easily match this workforce internally on the Solaris front with more efficient developers and thus continue to deliver a more superior product to Linux. Solaris can be very successfull being open source and not relying on external contributions for development, this is the model MySQL is using and yet it is the #1 open source database on the market.
As far as I can see linux and solaris aren’t even competing. linux is for low margin use on cheap x86 cloned boxes to run a mail server or php+apache and the likes on. Solaris is not in that market at all.
Then why is Sun obsessed with it then ? And for your info, the mercantile exchanged has switched from unix to linux to gain in both performance and cost. And we are not talking printing servers for workgroups.
Had SUN (or IBM or HP) done the right thing 10 years ago (a cheap PC version of solaris or Aix or HPUX and partnership with Lotus and other Corel and Netscape), there would probably never have been any need for linux as a kernel because there would have been a reliable unix kernel out there for those not happy with Windows. And then there wouldn’t be a monoculture.
SUN (and IBM and other HP) were all too happy to see Intel and Microsoft take on the mass market whilst they made massive margins on every bit of kit they sold. They fostered the Windows monoculture because it was beneath them to touch low margin hardware.
I remember the sysadmin where I worked, not happy about deploying windows servers but the x86 hardware was a tenth of the price demanded then by the big iron companies of the time. had they been able to run their favorite server OS (and it was then either vax or unix) on this hardware, there would never have been any windows servers (or very few). They were not stupid but there was no choice.
I can’t be sorry for SUN now. IBM has mutated into service and HP into mass market hardware firm. SUN was arrogant and kept their stance. The internet bubble has propped them up temporarily but Linux and Windows and now MacOSX all show that SUN is badly overpriced. They should have got the hint when WallStreet banks started dumping thousand of SUN workstations for multiple per trading desk PCs setup.
Linux and the free BSDs have just allowed a lot of people to do what the big vendors didn’t want to let them (and smaller vendors) do : cheap clusters, web maintainable boxes, mix and match distros, freely choose their support provider.
And it’s not about their product, it’s just because there is less and less of a market for their products at that price point.
“”In ten years it will be just windows vs. linux”. I completely agree.”
You believe Bill Gates but not Scott McNealy? You should, by that logic.
Solaris will still be around. There are lots of people who are not satisfied with the Linux development process (even if they really like other OSS like Firefox), and there are people who really don’t like Red Hat (the biggest Linux player right now). Solaris really does have a solid place in the market, Solaris 10 clinches it, Opteron Niagara and UltraSPARC IV+ clinches it a little further, and Sun’s new pricing model closes the deal.
“Solaris will still be around. There are lots of people who are not satisfied with the Linux development process (even if they really like other OSS like Firefox)”
and a whole lot of people dont like sun people blatantly lying repeatedly about redhat being a proprietary distro. ask them to name any SINGLE proprietary software in that distro and they will just shy off.
dont pull pathetic stuff like this for an argument
http://www.sun.com/software/solaris/10/index.jsp
So after visiting this link, can somebody tell me how those with slow download connections can get a copy of Solaris 10?
Now the linux fanboys adore monoculture – they want a linux monoculture and even though they may be almost entirely ignorant in the most breathtaking way on topics like solaris, IBM, Sun, and the IT business in general, this will not stop them spreading FUD and lies in order to see linux installed in a monopoly position. Linux zealots: haters of diversity (and pretty ignorant too).
So how much is Mr.Schwartz or Mr.McNealy paying you to troll on tech sites again? Where do you get this nonsense from? Please inform the masses of how you could possibly know what the whole Linux community is thinking and what their goals are. Talk about displaying ignorance…
The linux bubble will burst when it runs into this reality. But I fear linux zealots will never join the reality-based community..
Is this the same reality where Sun trolls and groupies like yourself dwell in? No thanks. By the way you seem like you have been hurt by Linux in the past. Was it a misconfigue of your .config, that led to a kernel panic, that makes you so angry and out to insult every Linux user you see? Did your stuffed animal Tux doll abuse you?
OS News is the official “Solaris vs Linux” flamewar site 🙂
I can even recycle a post from yesterday’s news…
Solaris is superior to Linux for many kinds of problems. I think anybody who would argue that point is being foolish. But then again, you could argue that z/OS on an IBM Mainframe is an even superior solution as well as VMS on a cluster of Alpha’s. The point I’m trying to make is that Linux has never been, and may never be the “Best of Breed” technical solution. That is not its strength.
I have always been fond of a paraphrase of Winston Churchill. It goes along the lines of “Democracy stinks, but it’s better than everything else!” Linux is a lot like Democracy, for a couple reasons.
First, the technology either solves the problem or it doesn’t. When Solaris, z/OS, VMS and Linux are all adequate for a certain task, Linux really has a leg up because it’s cheap and starting to become comfortable. Windows succeeded mainly because it was “cheap and comfortable” to all the suits. All us geeks like jeans and t-shirts because they’re cheap and comfortable. Don’t underestimate Cheap and Comfortable. And don’t underestimate Comfortable, it will sometimes trump Cheap!
Secondly, Linux fulfills one of the original promises of Unix – openness. In the early days of computing vendors could keep their customers hostage, and thusly computing sucked. Unix was this little toy operating system (The CEO of Digital really did call it that) but it ran on practically everything and the cocky kids told everybody it would take over the world. The Mainframe and minicomputer vendors said “Bring it on!” and now they’re all gone. But Unix never really fulfilled that promise of openness. You may argue with me on what the real value of openness is, but even the Penguin Suit-wearing Scott McNealy thinks enough of openness that Solaris 10 will be open sourced in some fashion.
Solaris 10 will indeed be a “Kick Butt” operating system, but nevertheless I will probably not invest too much time on it. But if I had a multi-terabyte data warehouse to look after, or a large e-commerce site, I probably would. Personally, I cut my teeth on Dynix/ptx. That was an operating system that did a couple of things well, and you may not see its influence on AIX, but I assure you it’s there. It Kicked Butt, but I don’t really mourn its loss. I’m not saying Sun is going to disappear, but apparently its attitude of the day is “Hey Linux! BRING IT ON!”…where have we heard that before? Not all the Mainframe companies are extinct, heck IBM is going like gangbusters. But the IBM of today looks nothing like the IBM of the 70’s and 80’s and it was a hell of a bumpy ride to get here. If it is to survive, the Sun Microsystems of 2010 and 2020 will look nothing like the Sun of today.
In summary, what I’m trying to say is that Solaris 10, for all its many technical merits, may be irrelevant to what’s going on at the Chicago Mercantile Exchange, and in many other organizations. Is this something most people can agree on?
I think many of the Linux people are subconciously offended when Solaris tries to open-source and impede upon Linux ground on low-end and middle-end servers. I also think Solaris people may be subconciously offended when Linux impedes on big-iron high-end servers.
Linux really has a leg up because it’s cheap and starting to become comfortable.
And now Solaris is as cheap or cheaper, and for many shops it is already comfortable. For Solaris shops, they’re down right intimate. For Linux shops, it’s not glaringly different. For Windows shops Linux is as painful as Solaris. Remember that almost all of the new benefits in Sol 10 are aimed at Admins.
Secondly, Linux fulfills one of the original promises of Unix – openness.
All of the Unixen are as proprietary as they are open. Openess is in regards to portability of source code across implementations, and all of the Unixes have their sins. Having the source code to the OS is a benefit in porting, but not as critical as you’d think. And if/when Open Solaris comes out, that benefit will vanish as well.
In summary, what I’m trying to say is that Solaris 10, for all its many technical merits, may be irrelevant to what’s going on at the Chicago Mercantile Exchange, and in many other organizations.
Companies are moving to commercial Linux distros because they leverage the cheaper hardware, and the cheaper hardware is simply getting better and better.
With Solaris being “as cheap or cheaper”, and in some ways offering a better mousetrap, companies that are in the early enough stages of considering Linux will take a hard look at Solaris. No company that’s hip deep in an implementation or transition will bother, of course, there’s no need. Any cost savings would be eaten by transition costs.
But those companies that have yet to make this switch, and who are looking in this space now have more choice. They’ve always had the BSD’s as an option, but there’s no marketing behing BSD. Red Hat, Novell, and Sun are going to go to the stops on marketing and start entering into a price war. What has been left to smash the prices in the commodity hardware base is going to hit the commodity software base.
And here I think Solaris has an edge. They have a more stable software base, and a large, experienced international support team. Solaris will be easier for them to support, it will be easier for vendors to support their software on Solaris, and that will mean lower costs for both vendors and Sun.
”
All of the Unixen are as proprietary as they are open”
wrong. linux, freebsd.netbsd, openbsd, dragonflybsd are so on are all unix systems unless you want to redefine unix by openunix certifications
“Openess is in regards to portability of source code across implementations, and all of the Unixes have their sins”
wrong again. it depends on the license and the amount of freedom it allows to end users and developers
“if they’d ported to x86 10/12 years ago”
They *did* port Solaris to x86 10/12 years ago. I still have a copy of Solaris 2.1 for Intel from 1992 or so kicking around somewhere. However the first decent release of Solaris on x86 was probably 2.4 which was released back in ’95 or so.
Hell, I still have a machine quite happily running 2.5.1 on an ancient dual P90 which dates from ’96 acting as a mail/news server which has not been upgraded as, well, it just works.
>
Linux kernel development community is not all that large and not all that productive
<
Don’t forget IBM is strongly behind Linux. As I understand it, IBM has already contributed about $1 billion to linux developement.
In fact, this is what msft and sunw are so upset about. This why both msft and sunw have been supporting scox.
In order to really trump Linux, Sun need to release Solaris under the GPL for x86 AND Sun hardware, because as I understand it, Solaris on x86 is really nothing more than a curiosity for people who want to try out a new OS. The real value of Solaris is on Sun hw, where most Solaris applications are developed. Solaris on x86 just doesn’t seem to be a worthwhile proposition compared to Linux on x86, (among other reasons) because Solaris x86 supports only a small subset of the x86 hardware Linux supports, and that won’t change as long as Sun is in control of the software-writing process. In contrast, Sun can write Solaris drivers for all the hardware on their SPARC platform, (or supply the info necessary for other people to write it) because they control it.
Another facet is openness of the (leadership of) the community. There really is only one person in the Linux kernel community whose presence is a fixture for the foreseeable future, and sooner or later even he, Linus, will need to be replaced. But if Sun retains total control of Solaris, then even if they change their programming team, to anyone outside of Sun they look exactly the same with the new as with the old – the same thing that happened with Java and the same thing that is prompting the Eclipse people to write an open-source implementation of it. If Sun run the risk of having a GPL’d or BSD-licensed version of Solaris out there, they had better make sure they’re involved.
wrong. linux, freebsd.netbsd, openbsd, dragonflybsd are so on are all unix systems unless you want to redefine unix by openunix certifications.
And I include them in my statement of “All of the Unixen are as proprietary as they are open”.
And I misused the term “proprietary”. Rather I should have used “exclusive” or “unique”. Proprietary is in regards to something owned by a company and has legal protection, whereas exclusive means something with a unique attribute.
A better phrasing would be “All of the Unixen have exclusive features that make them as unique as they are open”. Some of the Unix systems happen to have exclusive features that are also proprietary features.
Simple example. The entire point of DragonFlyBSD is to do something different than FreeBSD. That something different is going to be exposed to applications through some API. Applications that rely on that API are then bound to the DragonFlyBSD implementation and by definition will not port to either FreeBSD, Linux, Solaris, HP-UX, etc. If you choose to write code against the exclusive APIs of an OS, then you are pretty much locked in to that OS. Using exclusive APIs is not conducive to open, portable software.
It is irrelevant if the source code is available, it is irrelevant if the license is free. If the platform you wish to port the code to does not support the API, then you’re not going to get that functionality from the application. And considering that the Dragonfly team decided that their changes were dramatic enough to necesitate a fork in the BSD kernel, it’s not like I have to simply port a library to enable the application to work. For all practical purposes, while the unique and exclusive features that DragonFly provides are not proprietary, they lock in an application to that system anyway, just as if you relied on an exclusive, yet proprietary feature of an OS. So, DragonFlyBSD is just as open as any other Unix, yet can still force “vendor lock in”. It depends on the application.
“Openess is in regards to portability of source code across implementations, and all of the Unixes have their sins”
wrong again. it depends on the license and the amount of freedom it allows to end users and developers
With regards to proprietary vs exclusive, correct, license matters. But, it’s the APIs and compatability across systems that opens them up. Writing software for a VMS system is a completely different experience from writing software for a UNIX based system. The similarity of the UNIX based system, along with the fact that UNIX itself was reasonably portable is what made UNIX systems viable at all and what let them become dominant in the market.
In the late 80’s and early 90’s, everybody and their brother had come out with a UNIX based system. This was because it let them leverage their skills in hardware and system design without having to reinvent everything from scratch in terms of crafting a new OS. It also let sofware vendors use a vast amount of common code to more easily port their software from platform to platform.
Back in the day, we were able to deliver our Informix based applications, and did, on PCs running SCO, DG’s, Suns, IBM, Sequents, HP’s, Sequoia’s and AT&T/NCR machines. Open Systems let customers choose not just their software, but their hardware as well.
Before you had to pick the hardware with the software. If you wanted to pick VMS apps, Prime apps, Pick apps, Wang apps or IBM apps, you bought a VAX, Prime, Pick, Wang or IBM machine. Today, if I want to pick Windows, I have to pick x86 (despite earlier efforts by Microsoft). But if I want to run SPARC, or PowerPC I can run a wide assortment of OS’s and software on those platforms.
We had one customer upgrade from their AT&T/NCR machine running x86 to a Data General running 88K RISC processors, with our code needing little more than a recompile on their new system. Informix enabled us to do that, and the openess and standards (many de facto) of the Unix platform made that portablility viable for Informix. That’s a great power of flexibility that was rare in the day.
This portability enabled systems like Perl and such to actually become dominant members in this growing community. Ever see the Perl 4 configuration script? Ever see that line “Congratulations! You’re not running Eunice!”? That was back in the day where we had LOTS of UNIX based systems, and not the few that we have today.
Today, if you want the features of SQL Server, you’re stuck to running a Windows based machine on an x86. If I want the features of Oracle, I can run on a much wider range of OS’s and hardware. If I want to write an application leveraging ReiserFS, I’m stuck to a Linux machine, but if I write to Berkley DB, I can run it on pretty much anything I want.
Most folks do not want to port software, it is easier to find similar software that is imlpemented widely if they want to have a portable application. They would rather write to APIs that help ensure portablity.
So, all of the UNIXen: Linux, *BSD, Solaris, AIX, HP-UX, IRIX, etc. are as exclusive and vendor locking as the next, but all are also capable of participating in the world of open systems.
Finally, here’s the last example of system lock in, and where all of the systems fail. You can’t take a Linux administrator and plop him into a Solaris shop, or a BSD shop, or whatever. The systems still do things at the lower level in different ways. In this sense they are all unique systems.
Sure, they all do the same thing, it’s just all done differently. Sys admin scripts written for Linux simply won’t work on BSD. Heck, scripts written for one distro of Linux won’t work on another distro. Certainly a good administrator can make due and figure it all out, and there are similarities, but in this sense with most everything else the same, the Administrative switch over will be one factor that will deter a Linux shop from becoming a Solaris shop.
So, that’s what an “open” system means to the market at large, the flexibility to move your applications across different vendors. This is how a proprietary system such as Solaris can still also be an open system. Why something as locked down as Java is really an open system because it has so many different implementations, or how a single implementation system like Perl is open because of its wide availability.
Open source can make it easier for motivated individuals to move an exlusive feature from one OS to another, especially compared to a proprietary feature. But that doesn’t mean it is practical to do so, and can hinder application portability just as easily as a proprietary feature.
Actually, either porting to another system is easy, or it’s hard. Not both. (It may be easier to do some parts of the port, and hard to do others, but then those parts become hard or easy, not both.) The fact is, despite what your post might suggest (in some places; it contradicts in others), it is much easier to port from Unix to Unix than from Windows to Unix or from VMS to Unix. It may even be the case that, due to instability of APIs on the Windows side, it is easier to port from Unix to Unix than Windows version to Windows version.
But however hard it is to do either, people have done and will continue to do it.
It is also much easier to port from BSD to BSD than even from Unix to Unix, and easier still to turn an rpm into a deb. Even porting shell scripts can be made easier by writing them in csh or even sh, and by using the PATH variable to determine the location of commands rather than hard-coded paths. Linux and the BSDs change the Unix porting game because they cross-pollinate without having to invest in industrial espionage, or the fear of being taken to court over trade secrets. The point of open source is that you can see how it is done, instead of being locked out of some parts of the system (the source code and/or some of the APIs, as some have alleged happens with Windows).
“Simple example. The entire point of DragonFlyBSD is to do something different than FreeBSD. That something different is going to be exposed to applications through some API. Applications that rely on that API are then bound to the DragonFlyBSD implementation and by definition will not port to either FreeBSD, Linux, Solaris, HP-UX, etc. If you choose to write code against the exclusive APIs of an OS, then you are pretty much locked in to that OS. Using exclusive APIs is not conducive to open, portable software.”
It’s really not clear at this point how much of an API change the differences between DragonFly- and FreeBSD will require, but my guess is, “not much”, in the same way that the change from a two-level to a four-level paging system required “not much” change in applications written to the Linux kernel. The differences in philosophy seem to be much more to do with how the system is implemented, not with the interface it presents to users – the same way that Linus was able to create a complete Unix-compatible kernel without using any SystemV code. As such, any headaches are more likely to be reserved for the core team than for J Random DBSD developer.
Solaris cheaper than “Linux”. You gotta be kidding. Various GNU/Linux distrobutions are freely downloadable.
The initial acquisition cost of software is generally the *cheapest* part of IT infrastructure.