Illumos sporks OpenSolaris

If you were hoping that someone would fork the OpenSolaris operating system, you are going to have to settle for a spork. You know, half spoon and half fork. That, in essence, is what the Illumos, an alternative open source project to continue development on the core bits of OpenSolaris, is all about. The disgruntled …

[Anonymous Coward]

The author hasn't done his homework - mistakes, half-thruths, etc.

For example - development on Open Solaris hasn't stopped - check number of commits on daily basis and it is progressing as usual. Also there have been community contributions recently which have been accepted and integrated, etc.

[Paul Johnston]

But what use is it if..

The posts from Alan Coopersmith about ongoing work do continue but as he says

[I know that for external users this isn't very useful while /dev is

frozen, though at least one person outside Oracle has built our 144

repo so it is possible to start testing these now if you build them

yourself.]

[Matt Bryant]

The Denial is strong wth this one!

OK, so if not enough people wanted it when it was being pushed by Sun, and not enough people want it when it's being half-pushed by Snoreacle, why on Earth does he thnk any more people will be interested if they manage to get it to the point where it is without any corporate backing behind it? A fork of OpenSlowaris will die just as surely as the current version is. Didn't anyone tell him about Linux?

[Oninoshiko]

Re: The Denial is strong wth this one!

WELL, if it isn't MATT, our favorite anti-solaris troll!

If you actually paid any attention, rather then just troll about things you know nothing about you would know that there are a surprisingly high number of interested people in this project, despite Oracle's mismanagement. We see a number of interested companies because no other system offers the same feature-set as well integrated.

FTR, those of us interested are aware of Linux, as well as it's inability to keep a stable set of interfaces and it's list of awful "stable" filesystems.

[Matt Bryant]

RE: Re: The Denial is strong wth this one!

".....If you actually paid any attention, rather then just troll about things you know nothing about....." Hmmmm, let's see, I know that Sun used to ship five x64 servers with Linux for every x64 server it shipped with Slowaris installed. I know that RedHat makes a profit and has for years, whereas Sun didn't make a profit from Slowaris on x64 or OpenSlowaris EVER. I also work with large numbers of x64 servers and desktops and - guess what! - I know that not a single one of them runs Slowaris, though we do use RedHat quite a lot. And, I know of at least a dozen techies that run Linux of one release or another at home, but I don't know one that has any interest in OpenSlowaris.

"....those of us interested are aware of Linux, as well as it's inability to keep a stable set of interfaces and it's list of awful "stable" filesystems." Yeah, just tell that to all those real businesses that are actually using RHEL or SLES as opposed to all your "interested companies" that are sniffing round your vapourware. Your FUD is the same junk that was thrown out by Sunshiners when Sun was trying to kill Linux, and it sounds just as laughable then as it did now. Why? Because those of us in real business roles KNOW that Linux is already here, established and works. What I also know from your response is you don't have a clue about commercial Linux, it's features or stability, so I suggest a place for you to start would be one of the many Reg articles on RedHat, such as http://www.theregister.co.uk/2009/01/20/redhat_rel_53/. However, I don't suggest you look at articles like http://www.reuters.com/article/idUSN2213065720091223?type=marketsNews as it will only upset you to know that RedHat know a lot more about opensource and business than Sun, Nexanta or any of the other left-over Sunshiner hold-outs can ever hope to realise.

Enjoy!

/SP&L

[Jad]

Submit post: Illumos sporks OpenSolaris

"I know that RedHat makes a profit and has for years, whereas Sun didn't make a profit from Slowaris on x64 or OpenSlowaris EVER."

I don't know where you got your figures from but:

http://www.theregister.co.uk/2009/04/10/sun_software_biz/

"... Sun said that based on results in the first and second quarters of fiscal 2009 (from July through December 2008), its software business was humming along at a $600m annual run rate and growing at a 21 per cent rate ... "

and "... But that is not really the Sun software story. Sun had $1.91bn in services revenues - which includes hardware and software support not included in the above numbers - in those six months ..."

The article talks about how they're not making money like they used to, but It's still well ahead of RedHat http://www.redhat.com/about/news/prarchive/2008/fourthquarter.html for instance.

I know you're a troll, I know that facts don't stop you, I know you hate Solaris for existing, and that somehow it's existence causes you to foam at the mouth, but give it a rest already

[Matt Bryant]

RE: Submit post: Illumos sporks OpenSolaris

Please break your figures out into OpenSlowaris, Slowaris on x64 and SPARC-Slowaris, and then I think you'll find those "profits" were all on SPARC-Slowaris and depended on the sale of SPARC servers and the existing SPARC base. I know you Sunshiners like to try and hide the differences, but SPARC-Slowaris and any Slowaris on x86 or x64 are different and cannot run the same binaries, do not have the same level of application availability, and are not interchangeable. This conversation is around OpenSlowaris, and Nexenta don't have SPARC hardware sales to tie Slowaris sales to. They won't have an x64 server or desktop offering either, which puts them massively behind competitors like hp or Dell (and possibly IBM) that will also offer Slowaris support on their hardware, let alone direct software support from Oracle.

Comparing to RedHat is even funnier given that Wall Street had a much higher value for smaller RH than Sun during the Sunset. Indeed, it was Sun's declining marketcap that led to the Sunset and Oracle purchase. The Reuter's article I linked to shows why Wall Street loves RedHat, whilst it was happy to dump Sun stock.

You are also only looking at software sales, not the costs to the company (Sun). Sun couldn't make an overall profit because the profits from the areas where it did make sales were overwhelmed by the overall costs, and losses of selling the servers needed to make those software sales. Don't let those facts stop you. Nexenta will have costs if only the cost of building support facilities such as call centers, sales teams, consulting teams for services, and even just download servers. Their competitors already have all of those in spades, along with massive marketing budgets and real services capabilities. Those are facts.

So, you quote software profit figures that have nothing to do with OpenSlowaris, and services revenue that definately didn't, and probably can't supply anything other than wishful thinking about how Nexenta will be able to provide a product and services to match existing and much more capable competitiors even just inside the Slowaris market, let alone when we consider the competitor's advantage of being able to offer options around Windows, Linux, commercial UNIX, and then pull-through products such as storage, monitoring software, etc, etc.

I may be a troll, but I'm not willfully blind to reality. I got here from looking at the "facts" over many years, the same ones that saw Sun's decline through egotistical denial of market facts. I don't "hate Solaris", I just have little time for Sunshiners like you trying to beat down anyone that dissents with you, especially when you have already been proven wrong by the market. I suggest you don't take a rest, I suggest you try looking at the reality of the situation and then gearing up to learning an OS with a real future, and that's not OpenSlowaris or any fork of OpenSlowaris.

/SP&L

[AndyS]

(untitled)

Snoreacle? OpenSlowaris? Do grow up.

[Matt Bryant]

RE: (untitled)

".....Do grow up." Apologies if it offended your delicate sensibilities, but I find it hard not resort to childish humour when dealing with that level of willful denial. I'll try something more "grown-up". Solaris, in any form, has one chance - hp and Dell (and possibly IBM if they re-negotiate a support deal with Oracle) selling it on their x64 servers, in replacement for all those ancient UltraSPARC servers out there. It's chance in the desktop market are in the region of hand-formed projectiles, made from frozen precipitation, in a bonfire. Care to reply with exactly how you think Nexenta or anyone will be able to make an OpenSolaris fork somehow take marketshare and make a profit from already-established, trusted and profitable Linux releases like RedHat? Since you get upset by "childish" references, please try and reply with some "grown up" facts and theorems, and not just more Sunshine and wishful thinking.

[A J Stiles]

Way to miss the point

The existence of another Open Source operating system will benefit everyone.

Too much of OpenSolaris consists of binary blobs, which is hampering porting efforts and discouraging adoption. (Some of us actually have procurement policies requiring i-tal code.)

The beauty of Linux is that Source Code written for Debian will build absolutely fine on Slackware, CentOS or any other distro. It *generally* builds *almost* fine on other Unix-like systems, too, apart from a few schoolboyish assumptions -- mainly that the userland utilities will be GNU (ever used `killall` in a script?) and that devices will be named Linux-wise.

Generally, unfortunately, isn't good enough for everybody. Installing a piece of software from Source Code *should* be as simple as typing

$ ./configure

$ make

$ sudo make install

..... or clicking on an icon, having the file type recognised as a source package, and the automated package installer doing the above steps on your behalf.

More Source-compatible Operating Systems out there, and greater willingness on the part of programmers to make their packages operating system agnostic, will mean that it doesn't matter *which* OS you are using underneath.

After all, you don't specify a particular manufacturer when ordering fasteners; you specify a thread gauge, such as M10, and know that your purchasing department know that they're all interchangeable with one another.

[Captain Thyratron]

Of course he knows about Linux.

You know, if you ease up on the silly epithets, people will be more inclined to listen to your posts. I've been reading them from a while, and that's the one thing that sticks out. You usually do pretty good fact-checking, and you usually make a reasonable argument. The trouble is that, in the course of it, you go out of your way to make an enemy of anybody who's read that far. The guy who stands in the corner doing nothing but pointing and laughing at people is going to have that corner all to himself. I guess that's fine if you need the space.

Now, I can get your take on it if I pretend that the following are true:

* Everybody uses nothing but Windows, Mac OS X, or Linux, unless they're at a business.

* Nobody has lifted the hood on Linux and frowned at what passes for software engineering in the Linux community.

* The sole merit of an OS (besides the aforementioned three operating systems) is that using it looks good on a balance sheet to some pointy-haired boss.

But it ain't quite so. Sometimes it works as an approximation businessmen can use, I suppose; most of the people you meet will probably fit it. But a lot of other people use Solaris not as a business choice, but because some of us just want a good, solid, reliable Unix to hack on, don't care for the mediocre duct-tape antics of Linux, can't afford to pay more than our computers are worth for AIX or HP-UX licenses (or for Power or Itanium hardware that can run them), are pretty well tired of Microsoft, think Mac OS X is an eyesore, and prefer a System V userland. If you take the turns I just listed, you'll end up at this weird little cul de sac and the only house on it is Solaris. Well, now there's another house being built next to it, but the construction crew seems to be debating with themselves on whether to commit ritual suicide or to nick all the tools and run. Also, somebody seems to have nailed a big "Free for 90 days" sign over the front door of the first one. Inconvenient, really.

If, displeased with this, you go back a street (before the turn in the direction of System V userlands), at least there are the BSDs. It's not a System V userland, but at least they're good Unixes, even if it'll be necessary to learn a new set of administration commands. That in mind, FreeBSD is likely the only upgrade path for a lot of Solaris users--especially if they've got a lot of older SPARC gear that OpenSolaris proper won't support.

Will a fork die? Yeah, probably, but getting it away from Oracle, if that can be done without killing OpenSolaris, is the only way to keep it alive. The trouble with OpenSolaris is that it was never really open. It was bad enough with Sun, back when you could at least get things like patches without a service contract. As things are, the community around it is too small, too young, and still too dependent on Sun^H^H^HOracle as yet to get by on its own /or/ to get much bigger. Besides, it takes a while for the smell of an old corporate Unix to wear off, and that may as well be DEET for open-source enthusiasts. To step away from Oracle /might/ kill OpenSolaris, but to stay there /will/ kill it.

[PushF12]

Solaris refugees should escape to FreeBSD

Let's carve a tombstone for Solaris. When it stops twitching, we will bury it beside Irix, Tru64, and all of the other gloriously dead unix platforms.

FreeBSD is adopting worthwhile technology from Solaris, and it already has critical mass independent of any single corporation.

The parenting habits of Apple and Oracle are remarkably similar. Illuminos can meet Darwin on the short bus to the slow school.

[Captain Thyratron]

It's the only upgrade path on SPARC.

That's what I'm having to do these days. Even of OpenSolaris lives, it doesn't support a lot of my hardware. Despite all their fanfare about open systems and such, Sun was pretty reluctant to provide hardware information to people and, in fact, a lot of it only came out because the BSD people hounded them about it until they dug the stuff up: Things like specifications for processors, memory controllers, disk controllers, graphics devices, all that. My only upgrade path is FreeBSD.

Meanwhile, OpenSolaris refuses to run on anything that isn't newer than what I've got. Well, it works damn fine and I'm not shelling out for new hardware.

As it stands, if I want to run something on, say, my Sunblade 2000 or my Enterprise 450, I can:

* Install one of the older, pre-Oracle versions of Solaris 10 that both supports the hardware and doesn't require me to get a service contract; however, much of the OS is old and nasty because things like the porting of Xorg to Solaris on SPARC have been terminally frustrated by ideological battles and Sun's general unhelpfulness regarding its SPARC graphics devices. Good luck if you want to write a driver for an XVR-1000! (Maybe if you come to Oracle with a pound of cocaine you can bribe your way into getting a manual for the MAJC instruction set.)

* Install FreeBSD and get ZFS, dtrace, and an OS that supports most of my hardware (except for my cthulhoid graphics card; time to take a zero off its name and buy that instead), but also things like Xorg, better Unicode support, and libraries that aren't all dusty from years of people who seem to think that Unix should act like it's 1994. The only thing I lose is probably my current graphics card; I guess I can either buy an XVR-100 or switch in my old Elite3D-M6 card.

[Matt Bryant]

RE: It's the only upgrade path on SPARC.

".......My only upgrade path is FreeBSD....." Last year I slopey-shouldered a project for trying Linux on SPARC to another team. Management know we still have some SPARC kit hidden away in the bowels of the company and want any and every option for possible re-use rather than throwing away the old hardware. When I last looked, there were Fedora-based and Debian-based releases for UltraSPARC in the works. Personally, I'm not sure of the value of continuing to use old hardware commercially as the hardware support costs will ramp up as the kit gets older, and both IBM and hp will give extra discount for converting, especially if you pretend you're looking at a Fujitsu solution. But, for hobbyists or small businesses that have the knowledge, using secondhand SPARC with spares off ebay and loading Linux is an option.

[Jad]

Since I'm running OpenSolaris build 134 ...

I kinda really hope that this works.

[Kebabbert]

PushF12

You are making the mistake of confusing OpenSolaris and Solaris. Solaris have shipped something like 15 million licenses in total. OpenSolaris has shipped another millions, in total something like 20 million licenses for both OS. For a while, statistics showed that Solaris (not Opensolaris) was downloaded more than several major Linux distros together. For every IBM server, there are multiple Solaris servers shipped. According to a study recently, by HP, it says that Solaris is the most wide spread Unix in companies today and Solaris has "the brightest future". In addition to Solaris spread in the server halls at companies, OpenSolaris will spread to desktop users.

On the other hand, IBM has officially said that they will phase out AIX in favour of Linux. IBM will kill off AIX. This is outspoken and official. Development pace of AIX has slowed down. More and more IBM resources are shifted to Linux. Development pace of POWER will never catch up x86. Intel and AMD are far surpassing POWER development. In just a few years x86 will be faster, to a much much lower price. And you can run high end Unix on x86 as well: Solaris. You can get a high end Unix on cheap x86. I dont see how Solaris is going to be killed? But AIX will be killed, it is outspoken and official by IBM. Solaris will live much longer than AIX.

[Matt Bryant]

RE: PushF12

Kebby, you are quoting downloads which didn't make any money for Sun, even Sun was forced to admit their "millions of downloads" had resulted in less than 0.1% return in licences. OpenSlowaris was makig a loss just in the systems requireed for the free downloads, let alone the development costs. And the exisitng Slowaris base, that is all SPARC Slowaris, the majority of which is on old and over-due-for-replacement UltraSPARC kit. When hp talk about the massive Slowaris base they are talking about a target market that is ripe for porting to hp-ux and Linux. The measures to judge Slowaris on x64 and OpenSlowaris, as with any OS, are marketshare and profit, and Slowaris on x64 has zero penetration in the market and hasn't made a profit for anyone. If Sun or Snoreacle or Nexenta had managed to build up a single percent of server OS marketshare by giving it away for free then they could claim the possibility of future profits through support sales, but they haven't even done that.

And before Nexenta can do anything they have to reverse engineer all the features that are closed source in Open-only-in-name-Slowaris. That's not five minutes of work, and if they annoy Larry at any point they will find themselves on the end of Larry's army of lawyers, and that would keep Nexenta tied up in courtcases until they went bankrupt. Remember, Larry doesn't have to prove they have infringing IP in their rewritten code, he just has to draw out the courtcases and bleed them dry to win.

But even suppose Nexenta do manage to avoid Larry's wrath, and manage to re-write all those Slowaris features, they would then find themelves going into a market dominated by Windows and Linux. Servers? Linux and M$. Desktop? Windows by a massive margin, followed by Apple and then Linux. Emebedded? Linux, fullstop. Indeed, Slowaris in any form on x64 only had one possible claim of advantage and that's for storage appliances, using ZFS, and that is stymied by the NetApp suit for now. Other opensource offerings are chosing BTRFS instead. So, Nexenta would be coming from massively behind, with an unpopular product, and trying to push it without the marketting grunt of a major corporation, after the market had seen two uber corporations fail with it. Good luck with that!

/FOMCL&P

[Anonymous Coward]

@Matt -- Can't spell, does not know what dead is, and needs an education

You spelled OpenSolaris incorrectly.

-- why on Earth does he thnk any more people will be interested if they manage to get it to the point where it is without any corporate backing behind it?

1) Because huge Oracle is not as responsive as some community members desire

2) Because code which Oracle is not interested in accepting into the OpenSolaris will be able to be accepted into Illumos.

3) Because Illumos distributions can be built with special embedded features that Oracle is uninterested in (i.e. 128Meg memory, old devices drivers, etc.)

4) Because people like Solaris features (i.e. Xen, DTrace, Containers, ZFS, future Lutre integration, run older Linux & Solaris apps in Branded Zones, etc.) which huge Oracle may not find as necessary to advance in their business model

5) Because multiple other businesses are dependent upon Solaris and they don't want their businesses to be placed at risk due to the whims of a single person (i.e. linus) or company (i.e. oracle)

6) Because some people caught the gnu religion but still love real operating systems (i.e. gnu has not delivered decent kernel goods)

-- A fork... will die just as surely as the current version is.

OpenSolaris ARC's are being checked in daily - it is more alive than the fictitious Matt Bryant (or whatever he calls himself when the time/place suits him)

Illumos is not a fork but can offer other communities the possibility of a fork

-- Didn't anyone tell him about Linux?

Yea - doesn't run thousands of lines of POSIX shell scripts, lame 16 Gig file system limitations on proven file systems, lame production debugging instrumentation, lame backward compatibility, lame non-linux compatibility, lame licensing that makes other open source commercial contributions difficult to leverage

[Matt Bryant]

RE: @Matt -- Can't spell, does not know what dead is, and needs an education

Yeah, and I say sticks and stones, chum. Or maybe that should be "chump".

"....1) Because huge Oracle is not as responsive as some community members desire...." That would only explain existing Sunshiners, your answer does not explain how you think that would attarct NEW interest, especially commercial interest willing to pay for support contracts.

"....2) Because code which Oracle is not interested in accepting into the OpenSolaris will be able to be accepted into Illumos....." Welcome to opensource! Not every submision gets integrated into the main Linux tree. Could it actually be that Oracle don't think much of the submissions? In Linux, you fork it and off you go, and if your work is good then your fork flourishes and may be pulled back into the main core release. Oh, wait! I see your problem - OpenSlowrais is not open, it's in reality very closed and controlled by Oracle, and Larry wants profits rather than myopic and wandering codefests. Can I suggest you might get a better response from Oracle if you tried porting the "features" to Oracle Linux or Linux in general?

".....3) Because Illumos distributions can be built with special embedded features that Oracle is uninterested in (i.e. 128Meg memory, old devices drivers, etc.)...." <Yawn> There's already this stuff called Linux, it can be found in flavours from tiny embedded solutions, through phone OSs, desktops and HPC to full-blown, mission-cirtical-ready bundles. Oh, and it has mindshare and marketshare, and has been eating real Slowaris, despite Sun's marketting clout, for breakfast for years.

".....4) Because people like Solaris features (i.e. Xen, DTrace, Containers, ZFS, future Lutre integration, run older Linux & Solaris apps in Branded Zones, etc.)...." Yeah, so interested that Slowaris has zero % of the desktop market, and virtually zero in the x64 server market. If people were choosing Linux, Windows or Apple over OpenSlowaris when it was being pushed by a big corporation like Sun, why on Earth do you think they will be interested when they hear about it from an unknown like Nexenta? Especially when you consider none of the features you mentioned are truly unique, not even DTrace, which is a very late response to hp's Glance toolset (which has been around on Linux for years). Actually, I'm told that hp's Glance used to be the tool used by Slowaris admins for real OS performance monitoring for years, so I'm not surprised Sun had to give DTrace away for free given their own customers' preference for an hp product. So, please explain how you think tiny Nexenta can do "free" better than much bigger Sun could, with something other than bravado and blind optimism.

"....5) Because multiple other businesses are dependent upon Solaris and they don't want their businesses to be placed at risk due to the whims of a single person (i.e. linus) or company (i.e. oracle)...." Assuming you mean the massive commercial base of SPARC-Slowaris, you have failed to show why they would take Nexenta's offering over anything else. After all, x64 and SPARC Slowaris binaries are completely different, and if they are going to port and recompile they might as well do it on hp-ux, AIX, Linux or Windows, all of which come with much better support offerings and presence than Nexenta can offer. That's if their SPARC-Slowaris app stack isn't made up of commercial apps already offered, supported and proven on hp-ux, AIX, Linux or Windows, all of which have a much bigger commercial app portfolio than OpenSlowaris.

When commercial companies look at change they look at cost and risk to the business, and at the moment I can guarantee you they would see Nexenta's Illumos as a massive risk compared to commercial offerings such as RHEL, Windows or another UNIX, or even to "free" offerings such as CentOS. Trying to pretend that the commercial Slowaris base are gagging for a "truly open Slowaris offering" is wishful thinking and goes against the historical fact of their lack of acceptance of Slowaris on x64 and OpenSlowaris.

".....6) Because some people caught the gnu religion but still love real operating systems (i.e. gnu has not delivered decent kernel goods)...." Yeah, and that's called the Linux community, and the BSD communtiy. Please try and pretend either is not a "real operating system" because all you will be doing is decieving yourself or just flat-out lying. Remember, us customers are not stupid, we use Linux (and BSD), so we know when you're talking male bovine manure. We're not likely to just take your word for it when we see examples to the contrary every day. Your attempt to label OpenSlowaris as the only "real operating system" option just smacks of the same, old Sunshiner schpiel. Please go back and look at the recent history of Sun and maybe you'l realise why Sunshine just won't work anymore.

/SP&L

[Tiddles the cat]

huh?

"Especially when you consider none of the features you mentioned are truly unique, not even DTrace, which is a very late response to hp's Glance toolset"

The last clause is truly surreal.

[Matt Bryant]

RE: huh?

I know you Sunshiners have a very limited view fo the World, but there were management, monitoring and performance tools long before Dtrace. Wander into your local coding shop and ask them about hp's Glance, I can guarantee you nine-out-of-ten will talk positivley about it if they have done real, enterprise coding. Especially since Glance has been multi-OS for years, including Slowaris.

But, if you want a free option already covered by the GPL (which Dtrace is not) then you can look at Systemtap. Oh, but that might mean you have to leave your little bubble of Sunshiner alternate reality....

[Willy Messerschmitt]

Solaris, Linux, x86, GNU tools, Croos-Polination

I have several years of experience developing in C++ (and minor Java) on Solaris, Linux and Windows. Recently I worked on a C++&Solaris-based system which is a rather critical part of German Finance.

In general, Solaris is considered a very good platform by everyone involved. The institution currently operates their systems on Solaris SPARC, Solaris x86 and VMS. They are very happy with that, except that the VMS solution does not scale and the next generation system will be completely based on C++&Linux. No Java or .Net, except for demo clients.

My personal experience is that the Solaris dbx debuggr is excellent, but the Java-based GUI is utter crap. I simply used dbx only. Also, the Sun tools like grep and find are vastly inferior to GNU grep, find.

Also, x86 is very fast and SPARC is a slug.

I suggest the OpenSolaris people use as many GNU userland tools as possible, as the Sun equivalents are simply ten years behind. Maybe other stuff like NFS and file systems could be adapted, too.

Certainly the Java-based crap should be removed completely, as it is not just slow and memory-devouring, but also extremely buggy.

In general, my feeling as a developer is that Solaris is a very good OS and that was also the consensus with my colleagues there. It just has to be rescued from the pointy-haired at Oracle/SUN.

[Numpty]

What Java-based crap?

Exactly what "Java Based Crap" is present in OpenSolaris to be removed?

[Willy Messerschmitt]

Well

At least the SUNWspro debugger has an alpha-quality GUI made in Java. dbx itself is great, but the GUI is just killing a developer's time and increasing the cleaner's workload because of all the torn-out hair that must be wiped away (around the places where SUNWspro is used).

I have also heard about something called "Java Desktop System" and from the name alone I imagine it is used by Merkin Intel in Gitmo to torture inmates who have stated their profession as being programmers.

According to

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

it actually does not contain Java in anything critical, rather it seems to be a kind of Gnome Desktop. So that's all perfect, but if they have slipped in the Java cancer - don't touch it if you care about the cleaner !

[asdf]

I agree

Also as a fellow developer I totally agree with you points that SPARC sucks bad and is a good thing it is dead in the water. I also agree the Solaris userland tools are very 1991 (no xargs for find, no gzip option in tar, wtf?). Finally I am not quite as positive on Solaris as an OS as I was ten years ago but it is stable at least. Funny though how any box I work on outside of production ends up looking like and acting like my Linux box at home (Gnome, GNU userland, scripts etc).

[asdf]

omg Matt B ftw

I can't believe I am saying it but I actually largely agree with Matts points. As a lowly very impartial developer who does not have his job on the line based on his fanboi purchasing decisions or market sales I can honestly say Slowaris is going the way of many extinct UNIXs of days past (btw Matt this include HPUX, its just dying a bit slower due to the vendor actually having a business plan). Ten years ago Solaris was a respectable choice but in 2010 any shop that has not been actively looking at transitioning away from SPARC and even Solaris for several years now is not serving its shareholders very well. I have actually done a fair bit of work moving applications off both SPARC and Solaris for a very large company that had to be considered one of Suns most loyal customers. I can say from a developer standpoint also good riddance. Very sad that code I ported to Linux will run faster in a Virtual Box on my x86 netbook than it will on a 6 year old fairly top end SPARC/Solaris machine. Also kudos to Matt for realizing Linux/BSD is the way forward and not another fail UNIX like HPUX.

[Matt Bryant]

RE: omg Matt B ftw

"......btw Matt this include HPUX, its just dying a bit slower....." Shush! I just need to stretch it out for a few more years until retirement! ;) After that point, I will be just playing with OSs to avoid going loopy with boredom. What happens to hp-ux after I sail off into the sunset is of no consequence to me, I really do just see it as a business tool rather than a religion. Whilst I do respect the passion which seems to afflict most Sunshiners, I do often think it would be better if they got into the habit of stepping back, taking a deep breath and saying to themselves; "It's just software."

PS: A certain well-known contributor to UNIX who worked with the Bell Labs team can be found spending his retirement on WoW (yes, it's not all spotty nerds, there is the odd geriatric nerd as well!). He told me a funny story about how one of the original UNIX team, Ken Thompson, was driven not by "the greater good" or even business, but because he wanted to get a game he had written working on a PDP, and to do so they ended up writing what became UNICS. UNICS gestated into UNIX, and everything that follows (BSD, hp-ux, AIX, Slowaris) are basically the developments of a gamer's OS!

[David Halko]

asdf: yea, any 6 year old machine is slower than a modern desktop

The speed of virtual box on a desktop running the same architecture is no big deal.

Comparing a 6 year old computer to anything current is a poor comparison regarding performance.

The fact that the year old machine can run any modern operating system and modern software reasonably well is a testimony to decent engineering.

Solaris 10 runs OK on a 10 year old desktop. Let's see if Vista runs OK on a 6 year old desktop.

[asdf]

SPARC sucks, Solaris ok

I am sorry but if spend the big bucks on semi big Iron and in 5 or 6 years apps run faster on a $250 laptop running in a virtual machine then your architecture sucks. I have a x64 AMD home computer from the same era and even it runs the apps I have ported faster. I have not heard any developer use the word fast and SPARC in the same sentence without irony in years. SPARC blows and the marketplace is rightfully killing it off.

[Anonymous Coward]

@ everyone

Thank you all for making me feel less of a nerd. I remember reading similar fanboism, fud, rants et al in the letters section of zzap64 and other such computing press.

Please, stop. Have a beer. Get laid - if you can.

I sincerely I doubt Solaris is being retired any time soon and best of luck to this spork project. More competition can only be a good thing.

[Anonymous Coward]

Remember the Palestine Liberation Front in the Life of Brian?

Rather than fight blindly in support of our prejudices would it not be better to do the opposite of forking.

Why not converge? Bringing the strengths of Linux and Solaris together. have more brains working on one thing rather than dissipating all that effort. Oracle could even slip off the burden without loss of face (and potentially even gain some some credit).

Everyone will get a better system - all the better to fight the real evil.........

Paris - cos she is happy to come together or fork separately.