OpenSolaris-based OSes a Threat to Linux?

sunBoy asks: "A number of OpenSolaris-based Operating Systems are popping up on the map. BeleniX (screenshots), SchilliX and Nexenta (screenshots) are a few OSes which have hit the headlines in the past couple of weeks. Some say OpenSolaris has a leg up on Linux - 'For Linux, we're trying to push many distributions through to compress them into a standard. With OpenSolaris, we are already at the small end of standardization. What will follow is more OpenSolaris distributions spreading out from that core.' Is OpenSolaris really a threat to Linux?" Less of a threat and more of an alternative. Would more Unix-based alternatives on the market really be a bad thing?

Comments

Competion is good for you

[davecb]

Linux and BSD should compete head to head with Solaris for stuff that matters to nerds, like quality, scalability, performance and so on.

Companies can do the competing over money.

--dave (who works for a conpany and definitely likes money (:-)) c-b

Re:Competion is good for you

[mikehoskins]

Agreed, and I, for one welcome the new competition.

And no, OpenSolaris will never de-throne Linux -- at least not without a lot of money and time....

Linux has several things on its side:
    More users in general
    More fanatical users, that do things "for the love of the code"
    More software (OK, most things are ported/portable, but much more software is "first tier" on Linux)
    More momentum
    More installations
    More documentation
    More press (Is OpenSolaris a household word? Has grandma ever heard about OpenSolaris?)
    More freedom
    More flexibility
    More user-friendliness
    More drivers and hardware support, in general
    More platforms
    More open sourcecode

With a wave of a wand, you can't expect OpenSolaris to take off overnight, this year, next year, nor the following year. In fact, it's been open for months now, and what do we have? A very small amount of momentum and three new distros, much of which is a cobbling together of OpenSolaris and pieces of Linux distributions....

OpenSolaris is late to the party, but I probably will try it out, especially to re-sharpen my Solaris skills. But I probably won't host anything, unless it is a specific customer requirement.

What is scary is that many PHB's don't understand opensource. Many think OpenSolaris is not somehow bad, since Solaris went opensource. Go figure....

Re:Competion is good for you

[Anonymous Coward]

The freedom point is debatable to the end of time (BSD vs. GPL, blah blah blah), but one thing isn't: documentation. There are excellent books like Solaris Internals, plus docs.sun.com, plus, now, Sun's source code, that just put the scrap pile of Linux FAQs and man/info pages to shame. The reason: Sun for years has written their docs to an audience of $15/hour intern sysadmins, who need a checklist or list of steps immediately to get tasks done. Their set of manuals for Solaris is _enormous_. And it all applies to OpenSolaris, too.

If anything, just the ability of people to learn about Solaris from Sun's documentation could un-seat Linux in the long term. Hell, I passed Sun's own certification exams (sysadmin, networking) just by using their documentation, a couple prep books (for the sample tests and topic lists), and a small network of workstations to play with.

Re:Competion is good for you

[htd2]

I rather doubt that Linux has more commercial software available for it than Solaris. The reality is that much more commercial software is 1st tier on Solaris than Linux.

heh

[elmegil]

Would more Unix-based alternatives on the market really be a bad thing?

You must be new here. Sun is evil, don'tcha know.

BTW, for any CLUELESS MODS yes, I know who Cliff is.

Unix variants

[Jonnty]

I think it could well become not jsut Linux and linux distros, but Unix variants and the distros of the variants.

Surely, you jest. Whatevah.

[ubiquitin]

Look SunBoy, even Solaris incorporated GNOME so that Sun wouldn't have to build out their own desktop software. You can't be serious about the GNU-Free-World all of a sudden capitulating after more than a decade to just decide that for a few minor improvements that they would rather work on top of an OS by Sun, open or not.

Totally appropriate that the fortune cookie that came up on the bottom of that story's page is:
"I am not sure what this is, but an `F' would only dignify it."

OT - OS Dir screenshits

[twilight30]

Is there anyone who honestly believes that OSDir.com provides any service of any use whatsoever? Christ, it's the same set of 60-80 screenshots of the same window managers and office apps, just using different themes.

They could just make up the names of the themes and distros used and no one would notice the frigging difference...

Re:OT - OS Dir screenshits

[Mendenhall]

Gotta love the title!

Re:OT - OS Dir screenshits

[Anonymous Coward]

I for one like osdir screen shots. Saves me the time of installing a distro to check it out.

Re:OT - OS Dir screenshits

[Gaima]

I don't think the service is totally useless, just pointlessly overkill with a stupid navigation system.

All we need are a few screenies of the install system, couple pointing out the features of any sort of package management, some more pointing out administrative features of note, and perhaps half a dozen detailing the main interface (which is likely to be an X desktop of some kind).
20 in total would be more than enough, with a navigation system that works.

Re:OT - OS Dir screenshits

[jZnat]

Something tells me it was intentional. :P

evolution has already spoken

[aminorex]

The fitness race has already picked a winner, IMHO: Solaris kernel with Debian user-space
and a mixed KDE/Gnome desktop. It's just a question of how long it will take for the market
to find this global optimum, via stochastic walk.

Re:evolution has already spoken

[DrSkwid]

In evolution there is no winner.

Re:evolution has already spoken

[Murphy Murph]

In evolution there is no winner.

Tell that to the dinosaurs.

Phrase it however you want, but any game with losers has winners.

Re:evolution has already spoken

[Anonymous Coward]

Linux is evolution. Solaris is intelligent design.

Re:evolution has already spoken

[AuMatar]

Play the slots or blackjack for a while. You can win short term, but long term your money is going away.

Re:evolution has already spoken

[Elwood P Dowd]

In evolution there is no winner.

Tell that to the dinosaurs. Phrase it however you want, but any game with losers has winners.

Tell that to all of your dead ancestors. No one wins at life.

Re:evolution has already spoken

[maxume]

Sharks?

Re:evolution has already spoken

[mnmn]

Sure there is.
The one still standing is the winner.

Do I see Tru64, System V, Ultrix, Xenix, IRIX around?

Re:evolution has already spoken

[CaptainPinko]

The fact that there are losers doesn't mean there are winners.

Re:evolution has already spoken

[mnmn]

The presence of losers define the winners. The presence of the night defines the day. The presence of the 0 defines the 1. The presence of poor define the rich. The presence of Microsoft defines the free software community. The presence of redundant posts define the insightful mods.

I'm tired.

Re:evolution has already spoken

[drsmithy]

Play the slots or blackjack for a while. You can win short term, but long term your money is going away.

That's because the winner in that game is the casino.

Re:evolution has already spoken

[andrewski]

Tell that to the trichordates.

Re:evolution has already spoken

[richie2000]

Tell that to all of your dead ancestors. No one wins at life.

The ones who get to be ancestors are the winners, as are the ancestors themselves.

Life does not end with death. Life ends with the inability to create new life. The winners get to spread their genes, the losers' die off.

Therefore, Slashdot really is full of losers. QED.

Re:evolution has already spoken

[DrSkwid]

If that person is the only living creature left - no bacteria nothing then what is the prize ?

If there is so much as a bacteria with him then it's business as usual in the evolution machine.

Re:evolution has already spoken

[LizardKing]

The fitness race has already picked a winner, IMHO: Solaris kernel with Debian user-space

Debian user-space wins the fitness race?!? You obvisouly haven't taken a look at the code for the bloated pig called glibc.

Re:evolution has already spoken

[mnmn]

That bacteria and that person have won. Now they compete with each other.
You also compete with your own species, so that person has won while everyone that doesnt exist, have lost out now havent they?

Say youre hungry, in a war zone, and your girl/wife is being kidnapped by some soldiers. What do you do and why?

You look for food.
You try to get back your girl/wife
You try to get out of a war zone.

Notice youre trying to survive/reproduce. All organisms work very hard to do that. Why? To (1) survive (2) beat the others.

Granted if only one person has survived, hes still lost since theres no reproductive partner. In that case the bacteria has won.

Re:evolution has already spoken

[DrSkwid]

Now they compete with each other.

ergo : neither has won

Re:evolution has already spoken

[mnmn]

Now theres a difference between our concepts of 'won'. You expect the winner to win absolutely. Win all conflicts. But you can win some conflicts and lose others. That still means youre the winner in the conflict that you won.

Now the human and the bacteria's conflicts have started, but the conflict between that human and everything else has ended. He has won those.

Say the bacteria wins and theres no human. Just one bacteria in the world.

Still has to compete against the elements. But this is a new conflict.

Re:evolution has already spoken

[aminorex]

Well put. Moreover, coexistence is not competition. If the human and the bacterium do not have conflicting survival interests, they do not compete.

Re:evolution has already spoken

[CaptainPinko]

You are assuming there are only two states... rarely thats the case. There are prime numbers and there are composite numbers... and yet there are numbers that are neither (namely one and zero).

Not really.

[keesh]

OpenSolaris is an occasionally amusing diversion, like Hurd or Darwin. Sure, it has a small band of fanatics who're pushing it like crazy, but it'll never reach the serious computing crowd.

Re:Not really.

[DAldredge]

That is because the Serious coumputing crowd is already using Solaris :)

Re:Not really.

[FidelCatsro]

Well Hurd has RMS(and a fair few others) and he is a one man Army of fanatics in and of himself , will have to wait and see how HURD comes along (eventually) .
As for Darwin well I am sure Apple and OS X users may disagree there

Re:Not really.

[DAldredge]

HURD will be production ready about the same time we have working AI, flying cars, and Nuclear Fusion.

Re:Not really.

[Yonder Way]

That is because the Serious coumputing crowd is already using Solaris :)

No, we dropped the Solaris boxes for Linux awhile ago (and haven't looked back).

Well, yeah...

[wbren]

Less of a threat and more of an alternative. Would more Unix-based alternatives on the market really be a bad thing?

Well yeah, more alternatives might be a bad thing, depending on how you look at it. One of my biggest gripes with Linux in particular is the abundance of distributions. While it's sometimes handy to have a distribution tailored to different groups of users, it is a pain to support and use.

I know every distribution wants to be unique and innovative, but why do we need so many different package managers, for example. I don't want 20 different text editors in my accessories menu; I want one that does the job really well. The same thing goes for distributions. I want one that does everything well. It would make users' lives easier and much less confusing. Hell, there isn't even a standard windows manager in use today. Come on people, if we ever hope to make Linux popular, it has to be standard, in every way possible. It needs a standard look and feel. It needs standard applications and protocols for installing programs. The way things look now, it won't be standard anytime soon. I know this article is about UNIX, but I think the same idea applies.

Re:Well, yeah...

[yamla]

Windows has succeeded without a standard way to install programs, a standard way to uninstall programs (though progress has been made on this front), and without a standard text editor.

Now, this doesn't necessarily dispute your claim. What Linux needs may be substantially different to what Windows needs. But still.

Re:Well, yeah...

[Haeleth]

Windows has succeeded without a standard way to install programs,

Windows Installer. Some third-parties use other systems, sure, but it's still the standard. I suppose you also think the SI units aren't standard, because the US public insists on using pounds and inches?

a standard way to uninstall programs (though progress has been made on this front),

Um, Add/Remove Programs is the standard way to uninstall programs and has been since Windows 98 or so. Again, the fact that a few badly-behaved programs don't respect the standard doesn't mean that the standard doesn't exist.

and without a standard text editor.

Notepad.

(Okay, Notepad was a joke. The other two weren't, though.)

Re:Well, yeah...

[yamla]

Well, if you say Windows installer, couldn't you also say RPM is the standard because it is part of the linux filesystem standard? Over 99% of the applications I install (on Ubuntu), I just grab from the central repository. Compare this to Windows. Some have a nice installer, though not necessarily the Windows installer. Others require unpacking and _then_ running. etc. etc.

The uninstall procedures are indeed more standardised, any counterarguments would simply be the occasional program that refuses to uninstall that way.

Notepad may be standard but wordpad is also installed standard. And many computers also ship with word processing software in Windows.

Re:Well, yeah...

[Jellybob]

I don't want 20 different text editors in my accessories menu; I want one that does the job really well.

So... Emacs or VIm?

Which is it gonna be?

Re:Well, yeah...

[wbren]

Well, if you say Windows installer, couldn't you also say RPM is the standard because it is part of the linux filesystem standard?

No, it isn't standard. Simple. It's popular, but it isn't a standard by any stretch of the word. It's just one of many package management systems. Also, I really dislike the idea of a "central repository". That's just my preference, I know, but I thought I'd throw it in :-)

Re:Well, yeah...

[yamla]

I'm not sure I can agree with you here. LSB specifies RPM as the standard.

In the end, I find a central repository incredibly convenient when I am an end-user. When I am a developer (that is, developing my own software), the central repository is a pain in the behind.

Re:Well, yeah...

[i.r.id10t]

joe

Re:Well, yeah...

[lambu]

I think there is a limit on how much standardisation I'd like to see. Text editors, for instance, should be a matter of choice - just think of the vi vs emacs debate/war. Likewise I have no problem with desktop managers and other software remaining a matter of choice.

However at some lower level system aspects, I think more standardisation would definitely improve the appeal of the open source operating systems. The two aspects that instantly spring to mind are file system layout and installation packaging.

I've run linux off and on over the last 7 or 8 years (including at least one lengthy stint where it was my primary OS) and working out what went where in the filesystem used to sh!t me to tears. That particular frustration was mainly to do with not knowing where to get good reading material about such things, but even once you've mastered that for a particular distro, the realisation that another distro may well put common things in slightly different locations is enough to drive you back to windows (ok maybe not, but you know what I'm saying).

This leads on to packaging. Firstly you've got competing methods (rpm vs deb etc). Leave that aside and check rpmfind for any well-known piece of software. There isn't much we can do about separate hardware platforms, but the need to re-package for each of the major rpm-compatible distros is a particularly poor situation to find ourselves in.

A standard file system structure and installation vector would ease the burden on software writers and distro maintainers, make it simpler for end-users to get hold of and install software they want, which in turn would increase the appeal of open-source operating systems. I'm sure there are other aspects of the OS's that could be standardised with similar benefit. I think that what we really need is for the base platform (with which end users rarely interact directly) to become common, leaving distro creators and software writers free to expend their energy on design and a single implementation (per hardware platform, I know, I know...) rather than repeating themselves over and over in the implementation phase.

It's easy to make such statements as an end user without appreciating the difficulties involved. These annoying aspects are side-effects of the bazaar approach which has yielded such amazing progress in the last 10 to 15 years, yet I think that without some standardisation, the open source OS's are going to find it difficult to broaden their appeal.

Re:Well, yeah...

[mnmn]

Youre right.

Think of Solaris in this regard. One window system with one window manager (2 recently).

You make an app, you know it'll work on all the workstations and servers out there. The paths are the same, the tools are the same, API, interfaces you name it. The kernel is the same one everywhere and nothing needs recompiling because of the uniformity.

Now think of Linux. "Linux" has lost its meaning apart from the kernel because the distros are so disparate. Will it run on slackware? Will it show in a window manager in knoppix? Will it just take the nice fonts from ubuntu? And what about uClinux on ARM, Linux on MIPS, Yellowdog linux, or whatevers installed on linksys routers.

New developers have this reason to choose DOS/Windows as their base platform for projects with low risk tolerance.

Re:Well, yeah...

[Billly Gates]

Just like the flamewars a few days ago about consistant abi's and api's for drivers, the problem is mute with solaris.

Solaris has a consistant stable abi and api's for drivers. Things like Linux which is why unix ISV's prefer solaris. Oracle infact has a script that refuses to run itself if RHEA3 is modified in anyway. Its a serious problem but Sun cares about software being reliable and consistant for closed source vendors. Something Linus and RMS do not care about as they view Linux as ideology only.

Lets wait...

[OneFix at Work]

Lets just wait till there's a major distributor (besides Sun) that offers support for OpenSolaris...

What you say? Sun won't even offer commercial support for OpenSolaris?

So, you can't get support for any OpenSolaris distro from a major vendor...yea, I can see how it may have the upper hand here...

Till I can find at least 2 major distributors that offer commercial support for OpenSolaris, I wouldn't count on it being anymore than an interesting project.

You can bet that Sun is gonna make sure that any commecrial support from other companies comes through Solaris and not OpenSolaris...

Re:Lets wait...

[Anonymous Coward]

"You can bet that Sun is gonna make sure that any commecrial support from other companies comes through Solaris and not OpenSolaris..."

There's good reason for this. If I'm going to set up a database for an important business...do I run it on a tried-and-true quarterly release of Solaris with regular patch updates and documentation...or do I just compile the latest nightly OpenSolaris commits and pray for the best?

Even for Linux, I'd run older kernels, perhaps even a 2.4 one for servers that have to stay up.

More like BSD

[duffbeer703]

OpenSolaris will be more like the *BSDs, since the core is controlled by one organization and will dictate architectural things. (And avoid the bickering and bullshit that often hinders Linux development)

Competition is a good thing. If OpenSolaris takes marketshare from Linux, the end result will be a better Solaris and a better Linux.

Re:More like BSD

[Homology]

OpenSolaris will be more like the *BSDs, since the core is controlled by one organization and will dictate architectural things. (And avoid the bickering and bullshit that often hinders Linux development)

There is one crucial difference: the restrictive Solaris license that is designed to keep Sun entirely in control. With a *BSD you can fork, and that has happened on occasion.

Re:More like BSD

[euthyphro]

The CDDL license does not prevent you from forking; it's a derivative of MPL, with the addition of choice of venue and some "patent peace" provisions.

(Of course, and like the *BSDs, if you choose to fork, you need to build your own core team that shares the principles that caused the fork...)

Re:Surely, you jest. Whatevah.

[Anonymous Coward]

"....for a few minor improvements that they would rather work on top of an OS by Sun, open or not."

If you think the internals of Solaris are just a few minor improvements then you're misinformed on this particular issue.

Less puffing, more development please

[andyross]

As much as I think a Solaris-based free unix would be a good thing, OpenSolaris just isn't there yet. Linux has long since cross the threshold where a typical user (an enthusiast perhaps, but not a hacker) can drop a CD into a typical desktop machine and get a working, internet-connected workstation. OpenSolaris isn't even close yet.

The immediate problem is a sad lack of drivers for very common hardware that Sun has never shipped, like wireless networking (an Atheros driver just went into their tree a few weeks back; I believe that's the only one so far), ACPI power management, etc... Solaris has always been an OS for servers and managed workstations, so there are big holes in the coverage for "consumer" devices and laptop hardware.

Note that Sun itself has no "OpenSolaris" distribution you can download, only a source tree. The void has been filled heretofore by hand-cooked distros like SchilliX and BeleniX, which are roughly analagous to early linux distros like SLS and Slackware -- no (or minimal) package management, no exhaustive software selection, etc... Just a bare machine with a userspace into which you can compile your own stuff.

Nexenta looks promising, being an attempt to port the Debian (i.e. GNU, not Solaris) userspace onto the OpenSolaris kernel. I haven't tried it so I'll withhold judgement. But honestly, it's got a long way to go. Note that the existing linux desktops tend to rely on the hotplug/udev/hal/dbus architecture for much of their hardware interface, and none of this exists on Solaris to my knowlege. Someone will have to port it.

Honestly, at the moment OpenSolaris advocates would be better advised to spend time writing drivers and packaging a distro than submitting flame wars to slashdot. The world has lots of space for another free unix, but it needs to catch up before puffing about itself as "Linux killer".

Re:Less puffing, more development please

[Billly Gates]

Go read the previous story a few days ago about Linux lacking a api for kernel drivers and a consistant abi?

Solaris just works. The drivers just work. Things have models and solaris is well tested. Linux used to be stable but its ideology about having to be opensource is hurting it and preventing many stable third party drivers.

Solarisx86 works and you can install and compile packages off the net and they will magically work too! In linux they are all incompatible with each other and you need to use apt_get, portage, or rpm's from the distributor. They are so different that compiling one app will usually introduce bugs and is not tested.

Solarisxx86 is solarisx86. The software you download like perl cpan's will just run because its consistant.

Linux is going to fork soon in my opinion unless it can get its act together about being compatible with itself and offering standards.

Re:Less puffing, more development please

[elmegil]

Two words: dependency hell. Four more: Solaris doesn't have it.

Really at the end of the day, what's the point?

[bernywork]

Why would I go for OpenSolaris? What benefits do I get as an end user? (Lets just run with this as a theory for a second)

I don't have the packages that are developed for Linux, there isn't any major "killer app" out there to make me want to switch. Really at the end of the day, what's the REALLY big bonus to running OpenSolaris now?

This is the question that I pose to you all here, this is the same question that a lot of IT Managers ask about Linux when comparing it to Microsoft Windows, but we have a few answers to that question.

Admittedly if you are a 100% Solaris shop (Solaris SysAdmin for example who wants to run Solaris on his 3Ghz P4 that sits under his desk) then you might consider it. There isn't a community around this to support it yet either.

One thing that could turn up would be application support from vendors that currently don't support Linux. If that turns up, then things could heat up.

I know it's early days, and choices are great, but Linux I think has filled that void. There are how many Linux distributions out there now? Do we really need 400 Solaris kerneled distributions out there?

I know this sounds like a FUD session, but I don't want it to be. Just trying to encourage some comments.

Anyway enough ranting, what do you guys think?

Re:Really at the end of the day, what's the point?

[Orrin Bloquy]

I tend to agree with most of your questions -- and I spend my days building interactive websites on Solaris boxen. Solaris' selling factor is being a Unix platform consistent enough that commercial developers can sell binaries for and thereby keep their source closed.

If OpenSolaris is compatible, it kills the commercial version. If it isn't, your questions come back: Why?

There was a time when it was fashionable to say that Linux users hated Microsoft but BSD users love Unix. The rhetoric I hear from OpenSolaris is that they hate Linux but aren't interested in investing the intellectual capital to produce an equivalent desktop product.

The number of people who are impressed with a non-mini distro that dumps you into a text prompt is diminishing.

Re:Really at the end of the day, what's the point?

[hackstraw]

I don't have the packages that are developed for Linux, there isn't any major "killer app" out there to make me want to switch.

The killer app is the kernel itself and its supported utilities. Solaris is very mature and robust, and being able to learn more about its internals and contribute back to the code for future releases by Sun is a good thing. Look at Linux, and to a lesser degree, BSD. Linux is more interesting IMHO because of its development process and the fact that big commercial players have joined in and contributed significant amounts of code. IBM, HP, SGI, and Cray have embraced linux. Linux is a big player in the embedded market. Does anybody even have a ballpark count off the top of their head how many architectures Linux has been ported to? Its nuts, everything from 16bit to 64bit CPUs, SMP and NUMA boxes, all the way to some of the most powerful computers in the world.

Re:Really at the end of the day, what's the point?

[bernywork]

Its nuts, everything from 16bit to 64bit CPUs, SMP and NUMA boxes, all the way to some of the most powerful computers in the world.

Solaris / Old School Unix (Lets face it, Solaris is still pretty true to it's system V roots) hasn't been ported, admittedly it's not the "hobbiest" OS that Linux is, and therefore hasn't been ported (I was tempted to put in a YET there). But at the same time, I think Linux has filled this void very very well, the hobbiest style people are now getting supplemented by the large corporations, a number of whom are taking on the people that have been kernel hacking for a while now.

By the very nature of the supplementation by these corporations, the supported utilities are slowly coming to Linux. People don't use an OS for the kernel (Unless you are a kernel hacker) they use it for the rest of the system, the applications and everything else that sits on top.

OpenSolaris, I think although will pick up some support, I don't think will ever reach the market share that Linux has today.

Fine with me

[Intangion]

the more people moving to any Open OS should be helpful to any other open OS. more people, more developers, more applications, more open source that can usually build/run on many other platforms (sometimes with a little effort) i really dont think the guy talking about standards gets it ;) one of the main reasons people goto linux is for alternatives i do think it is pretty cheesy how so many distributions lately just repackage debian and slap their name on it and start charging for it claiming it looks like windows cause they use KDE...

Hey clubie, RTFA

[ikewillis]

Linux has long since cross the threshold where a typical user (an enthusiast perhaps, but not a hacker) can drop a CD into a typical desktop machine and get a working, internet-connected workstation. OpenSolaris isn't even close yet.

There's the BeleniX LiveCD which includes a Gnome desktop. Drop it into a typical desktop machine and get a working, Internet-connected workstation.

It Pays To Read The Article! Yay!

Re:Hey clubie, RTFA

[andyross]

It Pays To Read The Article! Yay!

Yes, but alas: apparently it doesn't pay as well as responding to the second sentence of a post out of context and ignoring all the expository text that followed it. Yes, very lucrative. Yay.

Flame on, I guess. Rah rah. Slowlaris is teh sux0rs. Thbbt.

Re:Hey clubie, RTFA

[schon]

There's the BeleniX LiveCD which includes a Gnome desktop. Drop it into a typical desktop machine and get a working, Internet-connected workstation.

I might try it if I could download it. Their HTTP link gives me 1.2KB/s, and the torrents are fscked (they tell me 'my client is spoofing' - which it's not.)

Solaris bloat

[countach]

I'd like to hear from the people who've seen the code, but I suspect the solaris kernel is a bloated somewhat krufty collection of code that hackers won't want to touch or improve. That means it won't have the rapid development, driver support, general hackability of the Linux kernel. The Linux code is now big, but it is well sorted and continually being refined.

Re:Solaris bloat

[DAldredge]

http://cvs.opensolaris.org/source/ for most of the code with a link to download most of it.
http://cvs.opensolaris.org/source/xref/on/usr/src/ uts/ for the Solaris Kernel

This is a Good Thing

[Quattro Vezina]

I'm a Linux zealot, but more OSS competition is good.

Is OpenSolaris ready for the desktop?

[OneDeeTenTee]

I just had to ask it.

Re:no (or minimal) package management

[kensai]

Um, you might want to check out www.blastwave.org

Not with that license, it ain't

[Just Some Guy]

Sun's CDDL isn't GPL-compatible. I'm not a diehard GPL fan (I think BSD is also pretty spiffy), but a lot of the most talented developers are. In short, I doubt that they'll ever get much of a mindshare among would-be helpers - certainly not enough to unseat Linux.

Competition is great and I welcome them to the game, but I'd be quite surprised to see this get much momentum.

Re:Not with that license, it ain't

[Anonymous Coward]

No one cares anymore about GPL compatibility. That whole debate was a big political to-do that didn't accomplish much (with Groklaw pouring gasoline on the fire the whole time). The end result: there really isn't much that could be cut-n-pasted, anyway, and the important stuff (like GNOME) are applications so get shared regardless of license.

Solaris does have a leg up on Linux

[mnmn]

Its hard to argue Linux is better than Solaris at everything. We now have the Linux, Solaris and BSD kernels to choose from. Solaris is probably the highest scaling free OS available.

Solaris was built for the enterprise. Most Oracle installations are run on Solaris. It is also the most popular UNIX as such (so we can now run a free UNIX rather than unix-clone (not that it matters)).

It has been well designed and has features that are years ahead, while it lacks some of Linux's features.

I cant wait for dtrace/zones for Linux, or the Linuxization of Solaris. In this regard, I'd prefer Solaris' libraries and tools over Linux's for the most part. I wonder when will they release their compilers as free, to compete with gcc on sparc and i386.

Hopefully this will not confuse the market, and developers will release Solaris, Linux and BSD versions of their applications for each major architecture (i386, x64, PPC)...

Yup. Too many people in the boat.

Re:Solaris does have a leg up on Linux

[Xross_Ied]

I have yet to see benchmarks comparing Linux2.6, Solaris10 with BSD variants. Anyone have any links?

Most Oracle installations aren't run on Solaris..
a) Granted larger Oracle installations (8Processors or more) are on SPARC/Solaris.
b) Many small to medium sized installations are run on x86/Linux. Has been this way for a few years now, ever since Oracle started supporting Linux really aggressively.

Solaris's major advantage is standardized kernel, kernel APIs and system libraries.
It allows application developers to better target the platform they want to develop on and support and for how long. In the commercial space this is a big advantage for Solaris.

Where Solaris fails compared to say Redhat (note I am talking about commercial version) is how easy it is to manage the system.
Want to apply the latest patches that have been approved by Redhat?
up2date
For Solaris?
go to sun's site hunt for the right page that will list the latest patch cluster.
verify this patch cluster doesn't break any of your Sun applications (e.g. SunOne messaging 5.2 sp2 has problems with the latest patch cluster for solaris 9).

Why should a user have to hunt for this information?
Why should I have to phone support and have them hunt for it?
Why isn't this information on the patch cluster's download page?
Why doesn't Solaris have a patch management system that covers all Sun products installed on a server?

0.02c

Re:Solaris does have a leg up on Linux

[Anonymous Coward]

smpatch(1M).

Re:Solaris does have a leg up on Linux

[bolthole]

Want to apply the latest patches that have been approved by Redhat? up2date For Solaris?

you run sun's "patchmanager". autograbs "the latest" patches and installs if thats what you want.

verify this patch cluster doesn't break any of your Sun applications

As opposed to "verifying the latest linux kernel patch doesnt break app [foo]". no win for linux there either.

Re:Solaris does have a leg up on Linux

[Xross_Ied]

As opposed to "verifying the latest linux kernel patch doesnt break app [foo]". no win for linux there either.

Please note I meant, Redhat patches don't break Redhat Applications vs Sun patches breaking SunOne Messaging server.

Thanks for pointing out sun's patchmanager.

Re:Solaris does have a leg up on Linux

[Anonymous Coward]

Well, Sun has just released Sun Studio 11, and it is free if you download the software.

bah with screenshots

[tod_miller]

You can make all (latest) OS's look pixelly identical with a screenshot.

I hope OpenSolaris gains some ground, I want to play with it. I think I should, since I use solaris on many companies servers all around. It seems to have had some real innovative thought at all kinda of dark and dank levels.

w00t

You know what?

[aliquis]

Just run Solaris and you'll get support, simple as that, I'm told it's cheaper than RedHat to.

Re:You know what?

[OneFix at Work]

You know that's true, as a matter of fact, Sun changed their pricing a while back to be more competitive with RedHat ...but you notice who is changing their strategy here? Sun...

Yea, Solaris might be cheaper on paper, but there's gonna be more problems running it on just any hardware...not so with RedHat...And you can bet that if you are running Solaris on a competitors machine, the fingers are gonna point to hardware whenever there's a problem...

Besides, Sun support sucks...about half the time when I call for something like a system down problem, I get transfered to some guy in India who is reading from a script and doesn't know what he's talking about...and after like 2 hours of him reading from some script I'm finally transfered to a developer that actually knows what he's talking about and get a comment like "well, I don't know what the problem is"...oh, that's great...waste 3 hours just to tell me that you can't help me...

I'm just guessing here because most of them are extremely hard to understand...and I used to work in a company that had employees mostly from India and Pakistan...now just imagine how someone feels that doesn't have the background that I have...Sun made the mistake of offshoring a sizable portion of their US tech support and most of em suck...as far as I'm concerned, it has ruined their reputation for support...

Re:You know what?

[Anonymous Coward]

.Sun made the mistake of offshoring a sizable portion of their US tech support and most of em suck.

So did RedHat, it's just that they eventually figured out it was causing problems and brought a lot (maybe all, I'm not quite clear on that point) of their support back to the US.

OpenSolaris

[metamatic]

It has poor driver support.

It has System V intellectual property in it, meaning it's legitimately at risk from SCO.

Its license isn't GPL-compatible.

There's no commercial support available for it.

I'll give it the benefit of the doubt and assume the bugginess has improved drastically since Solaris 2.6 days. Still, it doesn't seem compelling to me.

Re:OpenSolaris

[Anonymous Coward]

It has System V intellectual property in it, meaning it's legitimately at risk from SCO.

Huh?

Re:OpenSolaris

[Art Deco]

I hear people complaining about driver support but to tell the truth with server class machines I never really had a problem with not having drivers. If you are trying to run it on every machine that comes down the bend your results may be different. As far as SVr4 IP in Solaris, remember that Sun has a very liberal UNIX license from when they were cozy with AT&T (remember the stock swap that sent ripples through the industry leading to everyone else forming OSF?). If anyone can get away with this Sun can. True Solaris code can't be GPL'ed. Nothting is preventing anyone from using FSF tools under Solaris though. 2.6 buggy? In my experience every version of Solaris since 2.5.1 has been more robust than any version of Linux I've ever tried. We currently have a lot of Solaris on SPARC and a lot of Linux on Intel. Our Lintel customers have no interest in trying Solaris of any flavor so in the case of our company we still don't see growth in Solaris from OpenSolaris

Re:OpenSolaris

[metamatic]

http://en.wikipedia.org/wiki/System_V

Re:OpenSolaris

[htd2]

It has good enough driver support

It is not at any risk from SCO, Sun bought out their UNIX license from Unix Software Labs AKA Novell before SCO were in the picture.

No one really cares about the CDDL vs GPL issue except people who like the idea of flaming Sun. The reality is that there is very limitted scope for cutting and pasting Solaris kernel code into Linux and the other way arround and because of that CDDL vs GPL incompatibility is close to irrelevant. You can of course mix the user space stuff.

You may not be able to get OpenSolaris support but you can get Solaris x86 support roughly equivalent to the relationship between RHEL and Fedora.

I run Solaris 10 x86 on a couple of x86 boxes the normal things that trouble a Linux distribution troubled x86 such as Graphics adaptor, Wifi and Centrino drivers, apart from that it seems very stable and has some rather interesting features such as dtrace.

Re:OpenSolaris

[Anonymous Coward]

SCO has even given their blessings to OpenSolaris.
Here-s the article: http://www.eweek.com/article2/0,1759,1785664,00.as p

Re:Surely, you jest. Whatevah.

[aliquis]

The reason I didn't run Solaris earlier was that it costed money, but that one is gone. I don't mind that much if it's free or not, I won't look in the code anyway, but even if I did it (more or less) will be free anyway so that's not a reason either. Only intresting question if it's better for my usage or not.

Re:Surely, you jest. Whatevah.

[htd2]

Hum, so you discount the rather large contribution that Sun make and has made to Gnome. Or for that matter the huge investment that Sun has put into OpenOffice/StarOffice one of the prime candidates to run in that desktop environment.

Re: you're right, localization is a big deal

[ubiquitin]

Actually, the internationalization code that Sun donated to GNOME was a pretty big deal. Also, you're right about the [Star|Open]Office. Point well taken.

Linux already has zones...

[Anonymous Coward]

It's called Xen, User-space Linux, [other alternatives] ...