Submitted by A Computerworld blog speculates that the open-source Unix distribution may live on, but Oracle won’t be supporting it. At this point, “OpenSolaris’ only real future is as a fork, which would not be easy to pull off. Still, with enough interest from developers it could be done. OpenSolaris is licensed under the
Yeah, I think we all saw that one coming… at least currently there’s other versions to choose from.
Should be something more along the lines of, “F*** You Oracle.”
Or something more along the lines of, “Thank You Oracle!”
Did Solaris rape you or steal your lunch or something?
OpenSolaris is licensed under the CDDL, not GPLv3
http://practical-tech.com/operating-system/sun-to-release-opensolar…
Still, gplv3 is delightfully linux-kernel-incompatible, so it doesn’t make any difference.
Do you have another source? The one you link to are based on an article from the same guy as this ‘RIP Open Solaris’ article.
And if they went for something that was compatible with the Linux kernel it wouldn’t have made a lick of difference. The two operating system are so different in nature any cross pollination would be non existent. People here have this deluded idea that some how that if OpenSolaris kernel was at least GPL then there would be this great code sharing fest between the two projects – you find that people who make such declarations have never written a line of code in their life any more complicated than a ‘hello world’.
If Solaris was GPL, you would see code sharing. How do you explain porting ZFS to BSD? They are so vastly different how is that possible?
God f-cking knows why you were moderated up – I had ignored your post hoping it would die the death of a thousand people ignoring it.
The gap between OpenSolaris and FreeBSD is a lot smaller than the gap between Linux and FreeBSD. With that being said, the sharing of code still required massive amounts of work, it wasn’t simply a matter of grabbing the code, rejigging some headers and compilation. The question is whether it is worth all the effort put into porting code to and fro between platforms of any benefit given the work required to do so. Also, not all kernel modules are created the same, some can be easy ported whilst others can’t – some simply require a basic abstraction layer such as Haiku-OS and the FreeBSD Wireless stack and others are more complex such as audio drivers that dig deep into the kernel for maximum performance and low latency.
Edited 2010-07-20 01:50 UTC
I have a feeling this is being announced prematurely. I don’t OpenSolaris being totally abandon by Oracle. I will wait till i hear it from Oracle… or another 3 months, what ever comes first.
“… OpenSolaris is licensed under the GPLv3, so the base code is available. …”
Are you sure?
OpenSolaris is licensed under CDDL as far as I know and only small fragments of it (mostly userland) are GPLed. The juicy parts like the kernel and ZFS are all licensed under CDDL and it seems there are even proprietary components.
I know at least ZFS is CDDL for a fact.. this was why the kernel couldn’t use ZFS code and btrfs was created.
Edited 2010-07-16 20:29 UTC
ZFS most likely wouldn’t have ever made it into the Linux kernel as an official patch regardless of licensing because of the way it’s built (“rampant layering violations”) and politics.
Beyond that BTRFS existing owes a lot more to ReiserFS/Reiser4’s groundwork than it does to ZFS coming along.
“SUN MicrosystemsThe open-source Unix distribution may live on, but Oracle won’t be supporting it.”
This sounds like a fact. When did Larry Ellison say this officially? Where is the link? Can somebody link to it?
I mean, OS News would never disguise a guess, as a fact because that wouldnt be appropriate. Some would even call it FUD, and OS News dont FUD. Where is the link to Larry Ellisons or Oracle Executives claiming this?
Yes I seen the link in the article, but that is just a guy writing down his guesses (wishes). It is not facts. So where are the facts and the true link?
Mod this guy up.
Sure, the governing board might be in a huff, but that doesn’t mean OpenSolaris is dead. The linked piece is just a damn opinion piece with conjecture and no facts.
As for forking? Why the hell not? It’s been done before. There are plenty of OpenSolaris forks (Nexenta, SchillIX, Belenix etc.)
I think it’s a bit premature to write OpenSolaris off yet.
Did they really fork it, or just aggregate it differently?
Opensolaris can only really be forked by former sun/oracle employees. One could guess that they could as well use their time on Linux, instead of a company that kicked them out / they themselves chose to leave at some stage?
SchiliX was created by cdrecord developer J~APrg Schilling who actually replaced many closed source parts with FOSS ones.
He upstreamed his modifications and those are part of OpenSolaris since quite some time.
He is also a member of the OpenSolaris Governing Board and a total jackass (infamous for his anti-Linux/anti-Debian/anti-Anything flamewars and “only my way is right and you’re stupid” kind of comments — very similar to glibc’s Ulrich Drepper in this respect).
Why? It’s not like Solaris is some totally weird OS. It’s just another Unix with some nifty features, but overall not special.
Any somewhat experienced Linux, FreeBSD, … developer could just as well also work on OpenSolaris or a potential fork.
The problem is that OpenSolaris hasn’t garnered the attention of outside developers.
You guys are right. I should have marked this link as speculation. I’ve made an edit to the article to clarify.
You show that OS News prefer a two way communication with the readers, and not a one way communication. Thank you for rewording the article. :o)
I still don’t see a reason for linking to an article, which actually is just some guys blog with random thoughts, wild fantasies and uninformed guesses with zero information value. Why should this blog post be worth reading?
The author was the Director of Open Source at Sun and has been involved with Open Solaris board since it was created.
SVJN worked at Sun? You got it wrong.
You are probably talking about Simon Phipps, but linked article is written by SVJN. SVJN is resident troll on CW and he doesn’t know to tell the difference between Java and JavaScript. It is funny that he thinks that OpenSolaris is GPLv3.
SJVN once again shows how little he reads about other operating systems. But that won’t stop him from giving his analysis.
I caught him before making up stuff about Windows security. Could he not spend 10 minutes reading about the subject?
Yeah, he was lying about Windows, whatever he said. Windows is a even less secure than he says it is.
On the other hand, even a broken clock is right two times a day. Maybe he was right about that.
Edited 2010-07-17 15:52 UTC
He claimed that Vista had no security improvements over XP.
I think his entire opinion of Vista was based on screenshots and hearsay.
Not that I ever got that upset with him. He’s just another tech tabloid writer that cares more about pageviews than anything else.
You are serious? Support contracts are not available from Oracle anymore and last official release of OpenSolaris (2009.6) is EOLed. Burden of proof is on you, you claim that support is available.
Tell ya what. Go on oracle.com and find a support contract for OpenSolaris, and I will buy one for you.
As for Larry Ellison saying… again: are you drugged? He’s not saying anything. Oracle only supply FAQs and PDFs, and in those they say they will continue MySQL, OpenOffice, Virtual Box, but no word of OpenSolaris. Like it doesn’t exist. Plan is to kill it quietly, that would be most effective. Rising the issue will bring bad press to Oracle, they hope people to give up by themselves. Besides, Larry is too busy mourning the loss of Golden State Warriors to care about something he already bought. It is not every day he get outbid.
FUD and Trollery – this article
I agree.
Also, does it really matter what happens to OpenSolaris since all the interesting bits (zfs/dtrace) are already available in FreeBSD? If you want an OSS OS, would you really want to even consider looking at Oracle? I’d surely would not.
Instead of pure guesswork, how about looking at the actual source code repo instead?
Here: http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/
If you browse through the directories, you’ll see that many files were last modified within the last few days (even if the date of a folder itself says May).
Maybe Oracle is not interested in OpenSolaris, but we can’t know for a fact. All we know is that Oracle is just silent, the source repo is still active, and OpenSolaris never was GPLv3-licensed.
Or even the ONNV mailing list:
http://mail.opensolaris.org/pipermail/onnv-notify/2010-July/thread….
I guess the problem is that for many there is a transition from the culture at Sun to the culture at Oracle with Oracle having to take a good stock take of what they’ve bought and turning each into a revenue generating machine. Personally I’d be more concerned if we are having the same discussion in 2011 at this time though but from what I have heard when Oracle has their big development expo later this year apparently the future will be made a lot clearer.
Indeed you can also see the activity on the ARC cases page and the flagdays pages:
http://static.opensolaris.org/on/flagdays/141-145.html
http://arc.opensolaris.org/caselog/year/2010.html
I really tend to doubt that OpenSolaris will die, just because Oracle now owns it and has decided to basically abandon collaboration with the open effort. People have been talking about the untimely “deaths” of Amiga/AmigaOS, IBM’s OS/2, BeOS and most recently, Windows XP. Yet, all of these OSs enjoy a faithful, thriving following. OpenSolaris may stumble, go dormant for awhile; maybe years, but rest assured it will live on in some form or another. Maybe ten years from now we’ll all be on here reminiscing about a day when Linux was once this popular little OS people tinkered with until OpenSolaris arose from the ashes and took the world by storm! Just one possibility among many.
OpenSolaris is dead meat as I see it. Oracle has already gotten out of the desktop hardware/software business. I think Oracle will pull support from OpenSolaris and shut it down. The reason, I think, Oracle bought Sun, was to get Solaris as a platform for Oracle’s database. I think Oracle will dump Red Hat linux as the biggest complaint by their customers was that they didn’t control the OS. So now Oracle owns an enterprise OS, the enterprise database solution, and Java. One nice neat stack from a single company. Convenient huh?
My opinion is that Oracle doesn’t actually care about Solaris either. I think Oracle wants to be a one-stop shop and doesn’t care if the customers buy Linux as long as it moves them closer to buying OracleDB.
The server world is converging around Linux and Windows Server on x86. I really don’t see Ellison trying to stop that convergence.
I see that Ellison stops that convergence.
Larry has said that Linux is for lowend, and Solaris is for highend. He also said that OracleDB is run on Solaris, more than Linux or any other OS. Also, OracleDB is first developed on Solaris and then ported to other OSes, maybe because of DTrace (several Firefox developers have switched from Linux to Solaris just because of DTrace):
http://blog.mozilla.com/rob-sayre/2007/09/06/land-of-dtrace/
“I deleted the Ubuntu partition that I had. Now I^aEURTMm using Solaris. It^aEURTMs much nicer to use, except for the fonts. I don^aEURTMt think I can work on Mozilla without DTrace ever again. Too useful.”
http://www.computerworld.com/s/article/9131829/Oracle_s_Sun_buy_Ell…
“The Solaris operating system is by far the best Unix technology available in the market,” Ellison said.
http://www.serverwatch.com/news/article.php/3861376/Whats-the-Futur…
“We think it will be a long time before Linux ever catches up,” Ellison said.
…
“I love Linux. We’re big supporters of Linux, [but] Solaris is an older and more capable operating system,” Ellison said.
So the conclusion is clear. Ellison thinks Solaris is superior to Linux and targets it as a highend OS (which it is). I do think he cares about Solaris, especially because he said in the second link:
“In explaining his decision to make the acquisition today, Oracle CEO Larry Ellison boiled down Sun into a company of two key assets: Java and the Solaris operating system.”
So, yes, he cares about Solaris and he is betting on it. He is increasing research and development, more than Sun ever had – he has said this officially.
But regarding OpenSolaris, it is a different thing. We will see what happens when we see an official statement from Oracle. But we can see in the source code that OpenSolaris is rapid developed, maybe to be released as Solaris 11? There are frenetic activity going on in the source code today. Development has not stopped. You can check it out and compile if you want the latest build. There are no ISO files, but you can create one.
Elison said all that before acquisition closed. EU was watching, he had to say something. And, as you see, he didn’t mentioned OpenSolaris at all. Just proprietary Solaris. Even if Oracle puts huge amounts of money in Solaris, it will get completely overtaken by Linux at some point in future. Solaris’ future is to be Oracle DB appliance and I don’t see Oracle developing any general purpose features which might help Solaris rise again. Sun fought one man war against entire industry which support Linux. Larry isn’t that dumb.
As for OpenSolaris, there are no ISOs, no support, OpenSolaris will be like Darwin at best. If that.
PS: As for DTrace. That article is old. Linux now has SystemTap.
Edited 2010-07-18 14:12 UTC
Well, I dont share your viewpoint. Oracle has invested more money in Solaris than ever Linux. Oracle’s own Linux version is just a repackage of RedHat. And that does not cost billions of USD to do. On the other hand, Larry has invested 7.4 billion in Sun to get the two key assets: Solaris and Java. Larry would not invest billions if he didnt really want Solaris. So when you claim that Larry praised Solaris as the best highend OS out there, just because EU was watching – I dont think so. He meant what he said. Otherwise it would be dumb to invest billions in something that is inferior. You invest billions because you believe in the tech.
OracleDB plays best together with Solaris, but that is not unique. Because Solaris is so good, it also plays best together with SAP also, for instance. Even though Solaris uses slower RAM and slower CPUs than Linux, Solaris still wins official SAP benchmarks. That is because Solaris scales better, Solaris had 99% cpu utilization and Linux had 87%! It doesnt matter how fast hw Linux will run on, Solaris will use the hw better and win.
So, I am saying that Solaris plays best together with lots of software. Not only OracleDB. Therefore Solaris is not doomed to be an OracleDB appliance only. If you are going to build a large SAP installation for instance, then Solaris clearly is a better choice than Linux. Or OracleDB. etc.
Of course, any high end Unix is better than Linux in the Enterprise server halls. For instance, HP says the same thing about Linux and HP-UX: HP-UX is for highend and Linux is for lowend:
http://www.internetnews.com/dev-news/article.php/3809181/HP+Unix+Vi…
I disagree. Clearly, OpenSolaris will live in the future in some way or the other, because Solaris 11 is built upon OpenSolaris. Without OpenSolaris, there will be no Solaris 11.
Please, dont mention Systemtap. It is just a DTrace wannabe. And it is a bad copy. Here we see that even Linux developers think that Systemtap is “crap”:
http://blogs.sun.com/ahl/entry/dtrace_knockoffs
It is as bad as comparing BTRFS with ZFS. In the BTRFS mail lists, the developers have discussed the bad code in BTRFS. Havent you seen it?
You fail to grasp that Oracle is only one investing in Solaris, while entire industry invests in Linux. No matter how much money Larry puts into Solaris, Linux will get more investment. Also Linux will get more users and more testing because it’s free while Solaris is now trial-ware with huge pricetag.
I wanna to see that test. I bet it is as credible as other links you posted.
You just linked HP advertising and you’re acting like its credible source. Also, you obviously don’t know what “High End” means to those companies. It means very expensive product. It doesn’t mean the best in the world. HP is normally advertising their stuff as high end, but PHUX is medioce OS at best, in terms of features. And it runs on crappy hardware.
Without OpenAIX, there will be no AIX 7… oh wait… LMAO…
LMAO… again. You linked Sun employee blog, where Sunshiner is bashing SystemTap. How’s that helping your argument?
Truth is that SystemTap has some things which Sunshiners are envious about.
You know, in open source world people talk about software they are developing. They talk about it’s flaws freely. Reason why you don’t hear about that in ZFS is because ZFS is developed in proprietary manner. Behind closed doors. They just dump the code when it’s done, and tell you it’s perfect. So you repeat that here in attempt to get everyone drink your kool-aid. If you really think Solaris is so perfect, then I overestimated your IQ.
But just for fun, I would see the link where “BTRFS has bad code”. It is probably another Sunshiner. BTRFS blows the doors of ZFS. Get over it.
Yes, but there are some companies where the business are more worth than a few million USD. For some companies, that handle billions of USD, reliablity and such are far more important. If the business go down, they loose millions each hour. Then you need the most stable OS out there and price is secondary. Solaris.
What do mean with “credible as other links”? Which of my links do you think are false or propaganda? Regarding, “Solaris achieves 99% cpu utilization on slower hardware and Linux achieves 87% cpu utilization”, here are the official white papers from SAP:
http://download.sap.com/download.epd?context=40E2D9D5E00EEF7CCDB058…
http://download.sap.com/download.epd?context=B1FEF26EB0CC34664FC7E8…
Yes? I linked to HP advertising that confirms what I say: HP claims that Linux is for lowend and HP-UX for highend. Just read it yourself in the link. HP says so. I do not make it up. I am correct on this, too.
This one made me smile a bit, actually. Look. Enterprise means not that you must have much features. It means long term stability and support. No downtime. And HP-UX is regarded as a very stable OS, in the Enterprise server halls. Jesus, dont you know that? It is like you claim “Mainframes are mediocre because they dont have new features such as compiz, iTunes, etc”. But you can not take your desktop preference into the Enterprise server halls. There are other rules there. For instance, you need hardware that have much redundancy and fail checks. Which what you call “crappy” hardware have. For instance, if a cpu instruction was errorneous, SPARC cpus and Mainframe CPUs can rollback and replay that instruction.
Yes, I know that desktop users focus on high cpu performance and many FPS in Crysis, but Enterprise is a different thing. Most of the time, you dont need extreme cpu performance, but you need reliability. If your server handle billions of USD, you need to do it reliable, not fast and unreliable. Most modern cpus are powerful enough to do all business work load, you dont need highly overclocked cpus at 7GHz which reaches many many FPS in crysis. Jesus. You have absolutely no clue about Enterprise server halls?
Hm? I didnt understand this one. OpenSolaris is like Fedora and Solaris 10 is like RedHat. OpenSolaris gets all new tech first, and then it is backported to Solaris. For instance, ZFS came first to OpenSolaris: just read wikipedia article on ZFS. So OpenSolaris code is the basis for Solaris 11. OpenSolaris contains all new tech. If it is scrapped, then all development on Solaris 11 has to start anew, from scratch. Highly unlikely.
Regarding OpenAIX, do you mean that IBM puts all new tech into OpenAIX first, and then backports it to AIX and even bases new AIX versions on OpenAIX? Or what do you mean?
Maybe you should read my link. On the link there is a photo of some text that Linux kernel devs have written. One of them writes that Systemtap is crap in comparison with DTrace. Do you mean the Sun employee have made up the photo of the text?
Wow! I didnt knew that. What things would that be? Tell us. Or are you just making this up?
Even IBM has copied DTrace, and it is called ProbeVue in AIX. Everyone copies Solaris tech.
Ehrm. No. Wrong again. ZFS has also lots of bugs and issues, havent you seen the ZFS mail lists? If you want, you can examine the ZFS code yourself, it is available. But there are no discussions of “broken design” as BTRFS is subject to. See below.
But the thing is, ZFS exists today, it is run in the Enterprise server halls. BTRFS is a prototype of a ZFS wannabe. A lesser copy. When a filesystem is released, it takes at least 5 years before you can start to trust it. When BTRFS is released as 1.0, it will take another five years before it is let into the server halls. But you know, ZFS development will not stop. When BTRFS is released, ZFS will have tons of new features that BTRFS does not have.
The another “Sunshiner” is actually a RedHat developer that was asked to evaluate BTRFS for Enterprise use by a RedHat customer. Here is his verdict: (do you also claim I made this link up, too? It is also false Sun propaganda?)
http://lkml.org/lkml/2010/6/18/144
“In the meanwhile I confirm that Btrfs design is completely broken: records stored in the B-tree differ greatly from each other (it is unacceptable!), and the balancing algorithms have been modified in insane manner. All these factors has led to loss of *all* boundaries holding internal fragmentation and to exhaustive waste of disk space (and memory!) in spite of the property “scaling in their ability to address large storage”.
…
It seems that nobody have reviewed Btrfs before its inclusion to the mainline. I have only found a pair of recommendations with a common idea that Btrfs maintainer is “not a crazy man”. Plus a number of papers which admire with the “Btrfs phenomena”. Sigh.
…
The first obvious point here is that we *can not* put such file system to production.”
Ok, if you say so, then it must be true. I will try to get over it. But apparently some Linux devs disagrees.
ZFS was integrated into Solaris first but released publicly in OpenSolaris.
Solaris doesn’t need OpenSolaris to progress, especially when all the work of OpenSolaris was performed by Sun employees. How ever did Solaris progress before the advent of OpenSolaris? Oh that’s right, with paychecks to engineers.
OpenSolaris/Solaris on x86 is a dead end. Big iron is a shrinking market thanks to cpus like the Xeon 7500.
Sun waited too long to move against Linux. The technical advantages of Solaris aren’t significant enough to counter the inertia behind Linux. Sun acted arrogantly by being dismissive of Linux during the late 90’s and has paid for that arrogance with the Oracle takeover.
Jesus. Not this one again. I have explained to you several times that ZFS came not to Solaris first. It is all written on wikipedia article, the same one you linked to. Yes, the same one.
Let us straighten this one out, so you stop spreading FUD and lies, ok? Read the wikipedia article and then come back and stop lie about ZFS and Solaris, ok? It is getting tiresome. You are just repeating the same old stuff all the time. Ok to repeat if you where true, but it is not. It is false. And I have told you are wrong, so either you are deliberately lying, or you just dont understand text. Or, just FUDing.
Oh, no. Not this one also. Look. All new tech goes first into OpenSolaris, and then backported to Solaris. Solaris is the stable enterprise OS. You can not include new tech into a stable OS. First you introduce the tech into the beta OpenSolaris. And then backport it when most of the bugs are ironed out. Jesus.
So, you say that Solaris on SPARC is the future? I dont agree with you again. x86 is developing faster than x86, so I think that x86 is the future. Not a dead end.
This one, I agree with you. But x86 is far behind yet. But it is catching up. The question is if it ever will catch up? x86 is too buggy and bloated so will it ever be as reliable as SPARC or POWER? Some says no. x86 is broken by design. But truth is that x86 is developing faster than any other CPU.
You have lot of opinions. Anyone is free to have opinions. I would like it better if you backed up your opinions with somewhat credible links.
And please, stop lie. Ok?
You tell someone to stop lying?
Lets see:
First you lied that OpenSolaris is commercially supported by Oracle. When I told you to find support contract and I will buy it for you, then you stopped talking about that and started lying about other things.
Like for example you linked HP advertising blog and acted like it is some credible source. Then you admired that is advertisement, but you still claim it to be a credible source. I really don’t get.
Then like for example you linked Sunshiner blog where Sunshiner dig out some old statement from Linux dev, statement which is from time whem SystemTap was in it’s infancy.
As for your SAP tests, I didn’t expected to be any more credible than the rest. For those who won’t bother to analyze it, here is what is wrong with it:
Linux box run MaxDB 7.8 which is pretty crappy DB. Solaris box run Oracle 10g which is very fast DB.
Linux box has 128GB of RAM while Solaris box has 256GB. That is probably the main reason why Linux utilized the CPU less – it lacks memory.
Also, why is SuSE used over Red Hat? I tell you why. Because it sux. Solaris probably can’t outperform nicely set up RHEL box even if it has more memory and if test is not done with same DB. So they slip SuSE which Microsoft also use every time when they want to prove that Windows is faster.
What’s that sound…. oh that’s last shred of your credibility crashing down and falling on the head.
Edited 2010-07-19 15:08 UTC
No I did not. What are you talking about? I do not lie. I have links and white papers to all my claims.
Look, I claimed that HP says Linux is for lowend and HP-UX is for highend, and then I linked to a blog where HP says so. Am I lying on this? Did not HP say so? Or did I made it up? Are you drunk?
But was that link a lie, or is it the truth? If you are going to call me a liar, then you better find posts where I lie. If you cannot, then it is actually YOU that lie.
And if you read about Systemtap, it really seems inferior to DTrace. But you said that Solaris people are envious on the DTrace wannabe, so let me ask you again: are you lying now? Which functions are we envious about? Tell us. Or, stop lie.
No, the reason Linux used 128GB was because the HP machine can use faster RAM sticks. If Linux uses 256GB, then Linux must use as slow RAM sticks as Solaris machine.
The SAP benchmark is not Database bound, it is a SAP test, not a database test. All important data is cached. And also, SAP benchmark does not require more than 32(?)GB RAM or so. It does not matter if you throw in 1024GB or more. It will still only use some 10s of GB RAM. It is only a speed benchmark.
Look, it was HP Enterprise benchmarking team that did the benchmark. They have access to all sorts of hardware and know what they are doing. These benchmarks are very prestigous and the benchmark team gets all resources they want. For instance, if the HP machine runs faster on another 128GB RAM, they get it. HP tries out lots of different configurations, and then trim it.
Havent you heard about IBMs benchmarking team? For TPC-C they used a Unix server that costed 35 million USD list price. They used several THOUSAND short stroked hard drives. Short stroke means that you only use ~10GB of the entire drive: you make a small partition on the outer edge of the drive. Because the reading head can read faster if the data is near the edge, near the FAT table. The head doesnt have to move to much from the FAT table which lies in the outer edge of the disc. IBM benchmarking team goes to great lengths and have all resources they want. HP Enterprise benchmarking team does the same thing. If you think that another 128GB RAM would have made the benchmark faster, or another Database, then HP would have got it. A full SAP installation can cost more than 100 million USD, it is VERY important to get high benchmarking scores on such a big business. HP will do anything. They try out all different configurations and all different hardware, etc.
Didnt you know this? Maybe you dont work with Enterprise stuff?
Ah, damn! I lost all my credibility. Now I am a broken man.
Look, if you claim that I lie or FUD, then you have to prove it. Quote a post where I lie. If you can not then it is you that lie. FYI, mathematicians hates lies. I never lie. Good luck on finding a post where I lie. I can always back up my claims with links and white papers. You, on the other hand, can not. You do lie. What are the functions in systemtap Solaris people are so envious about? There exist none. You, Sir, are a liar. And FUDer.
You linked the test which is Sun server vs HP server showdown. It is comparison of two servers with different hardware. And they happen to run different operating systems. Test is done to prove that cheaper HP server is comparable with Sun’s, no to prove it’s faster. And you paint it as it was Solaris vs Linux test. It isn’t. When you are testing 2 OSes, you want all other variables to be equal. Find test with same hardware and then we can talk, any link you shown so far is either misleading or straight out lies.
It is also very funny that you think that faster memory is better than more memory. That shows how much you are clueless.
Also, what about OpenSolaris support contract? You now claim that you didn’t say it is available? Or it is? Give me the link so I can buy it. Or you admired you were lying.
Anything you said so far was just hot air, and sometimes you link something and claim that it backs your story. While link in question doesn’t have anything to do with your story.
BEEP. Wrong.
No, it is not a server vs server test. It is SAP benchmark test. Look, if it where a server vs server test, then it would have lots of different benchmarks, CPU vs CPU, RAM speed vs RAM speed, power usage vs power usage, etc. Have you not read benchmark tests? But here it is one test. SAP vs SAP. HP gets the fastest SAP machine they can offer, and Sun gets the fastest SAP machine they can offer. And who wins the SAP score?
The HP machine will certainly win other benchmarks, for instance, RAM speed. The HP machine have higher RAM bandwidth.
Yes it is. The HP linux machine used 2.8GHz cpu and PC2-6400 RAM sticks. And the Solaris machine used 2.6GHz cpu and PC2-5300 RAM sticks.
So, Linux used faster hardware, and Solaris used slower hardware. The reason Solaris won, is because it scales better. It uses the 48 cores much better than Linux. So, even if Linux uses faster hardware, it scales bad. Solaris will always use the cores much better, and therefore get higher score.
When you have more than 32GB RAM, you dont need more RAM for the SAP benchmark. Then the speed of the RAM is more important.
If you have 8 cpus in the chassi, and you can not put in more cpus, then the speed of the cpus is more important. It is “very funny” you dont understand this.
Again, link to when I said this. I did not say this. You are now lying about me.
Ok, so point out which links had nothing to what I claimed. You sounds like Kraftman, he is really really an original. But he admitted he Trolls, he confessed it. He also lied about me. And you lie about me. Prove your claims about me. Quote the posts.
And, where are those nifty functions in Systemtap that all Solaris people wants? Did you just lie about that, also? Do you always FUD and lie? Because you have not supplied any proofs, even though I asked you. Just like the Troll Kraftman.
Yes and you never explained this sentence in the article:
Source code for ZFS was integrated into the main trunk of Solaris development on October 31, 2005[4] and released as part of build 27 of OpenSolaris on November 16, 2005.
Here’s the original blog post from the head developer:
http://blogs.sun.com/bonwick/entry/zfs_the_last_word_in
ZFS went into Solaris first. It was released publicly first in OpenSolaris. Unless you want to claim that the creator of ZFS is also lying.
It’s called a development branch. Does OSX have an OpenOSX? As I pointed out Solaris progressed before without OpenSolaris and can do so again. OpenSolaris has only been around a few years.
Jesus. If you read his blog again, you will see
“And today, 10/31/2005, we integrated into Solaris. ZFS will hit the street in a couple of weeks via Solaris Express.”
Do you know what Solaris EXPRESS is? No? But I do know. A hint, it is not Solaris 10. It is a different distro. Solaris Express was developed for some years, the ISO file was a DVD on 4GB size. Then Sun basically scrapped some of the bundled programs, made some changes and compressed the OS down to CD on 700MB and called it… OpenSolaris. Marketing thing. OPENsolaris sounds good, it sounds as it if is open, like Linux.
So ZFS was not introduced into Solaris 10 first. According to your wikipedia article, ZFS was one year later backported into Solaris 10.
Ok? Can you stop saying that now? I have told you that you are wrong on this too, and I told you to read your own link. Must I explain your own links to you? Jesus.
Jesus. Look, I dont know how many times I have to say it, but all new tech gets introduced first to OpenSolaris, the beta version of Solaris next, and then backported to Solaris 10. Jesus. I feel like talking to a madman.
Look at this quote from the same article:
“I’m a Linux fan and if you want Linux we have the best Linux in the world. If you want UNIX, we have the best UNIX in the world.”
That’s a guy that will sell whichever Nix the market wants, and the market has been moving towards Linux on X86. He makes his money by selling a DB that runs just fine on Linux x86. Why should he make a significant investment in Solaris when the market is moving towards Linux? He isn’t interested in changing that market direction and that is what counts.
Yes he is investing in Solaris but it is at the mainframe level. It isn’t like with Sun where the plan was to displace Linux with Unix. Ellison doesn’t have an emotional attachment to Solaris.
You don’t need to wait for a statement, their position is clear by their actions.
The fact that OpenSolaris isn’t on the main Oracle.com website speaks volumes. They’re clearly interested in pushing other Sun acquisitions like Java and MySQL but not OpenSolaris. They would have killed it immediately if it weren’t for the bad press that such a move would elicit.
I dont understand your question: “Why should he make a significant investment in Solaris?”. He IS already doing that: betting much more money on Solaris than he ever has betted on Linux. It is happening as we speak, have you missed it? Dont you read my posts? I just wrote in the post above: He has invested many billions in Solaris and Java. And he will continue to invest even more than that. If Larry invests many billions then he is committing and he (almost) goes all in. It is almost as 95% of Larrys billions goes to Solaris, and 5% goes to Linux.
Regarding why he prefers Solaris, I told you recently in the post above, that OracleDB runs better on Solaris than Linux. Solaris is the prefered OracleDB platform – this is outspoken and official. The official reference platform is Solaris – this was official long before Oracle bought Sun. And also, OracleDB runs on Solaris more than any OS. And also, OracleDB is developed first on Solaris, and then ported to other OS, maybe because of DTrace.
Almost all of this, I wrote just above. It is getting tiresome to repeat everything all the time. It would be much easier if you just read my posts?
Ok, so now we agree that he is investing in Solaris? But above you seemed to imply he would stop to invest more? I dont get it? Is he investing or not?
What “actions”? The problem is that Oracle has taken no actions at all yet. That is what this OS news article is about. There are no actions from Oracle. Or do you mean that the lack of actions is a bad sign? But below you write that they should have killed OpenSolaris? It is far worse to kill OpenSolaris, than to link to it from Oracle web page. So no actions proves nothing. It could be that Oracle is busy with all the income generating stuff first, and then turn attention to OpenSolaris. So no actions, does not prove anything.
When Oracle talks about Solaris, it is because they earn money on Solaris 10. Oracle is interested in pushing Solaris which costs money. But OpenSolaris is on their Solaris web page. So I am not as convinced as you.
I dont think that Oracle cares about what others think. If they want to fire people, they do it. If they want to kill stuff, they do it. But the truth is that OpenSolaris development is continuing in a rapid pace which you can check yourself.
And how is Oracle supposed to release overdue Solaris 11, if all OpenSolaris code is scrapped? All new research and development goes into OpenSolaris first – so OpenSolaris will soon be forked off, polished and debugged – and then released as Solaris 11. There are lots of research and development into OpenSolaris, why throw away and kill it?
Your posts just dont make sense. It is like when we talked about Enterprise server halls, and you showed that Windows server sales are increasing. So what? What has Windows to do with Enterprise server halls? Totally off from reality.
Edited 2010-07-19 09:09 UTC
He isn’t investing significantly in the OS with the goal of displacing Linux. He’ll just be investing in Solaris stacks that are built around the DB. Unlike Sun he has no interest in fighting a war against Linux and that is what matters. As for investing in Linux he doesn’t need to when he can just take RHEL source and sell it with an Oracle stamp.
No one knows the exact amount that is going to Solaris.
Consider the following actions since the takeover:
1. Making positive statements about acquisitions like Java and MySql while saying nothing about OpenSolaris
2. Ignoring the open letter from Ben Rockwood
3. Providing no response to the recent ultimatum by the OpenSolaris board
4. Providing no link to OpenSolaris from the main website
They have not forgotten about OpenSolaris. They have simply chosen to ignore it.
Where are the performance advantages over Solaris? Even if OpenSolaris had some special tech it could be moved over to Solaris.
They probably don’t from your perspective because you are deluding yourself when it comes to the fate of OpenSolaris.
OpenSolaris will either be forked or ignored by Oracle. Either way it will stagnate due to a lack of developer interest. Building around OpenSolaris or even Solaris x86 at this point is a mistake.
The purpose IS to move over the special tech to Solaris 10 and to form Solaris 11. I have told you this umpteen times; OpenSolaris is like Fedora, and Solaris 10 is like RedHat.
Or maybe you are really confused and know nothing? At least it is very hard to try to explain things to you. And your own links are supporting ME, not you. But you dont understand that. I must read and interpret your own links and explain them to you. Do you have a hard time understanding written text? It seems so. I can explain again and again, to no avail.
Great! At last you understand! OpenSolaris WILL be forked into Solaris 11. That IS the purpose. That is also the reason OpenSolaris will not die. Jesus.