Slashdot Mirror
Sun-isms Debunked
Newman writes "We're all aware of the hole-ridden arguments that Sun executives Scott McNealy and Jonathan Schwartz use to attack Linux. This guy at NewsForge really grilled them at the Solaris launch party last Monday, and actually got some straight answers out of them. At the end of the article, both execs have some specific words for Slashdot readers."
Just because you hear something from Schwarz today doesn't mean he'll say the same thing tomorrow. Today: we're going to give the hardware away and charge for support! Tomorrow: we're going to "open source" the OS, give it away, and charge for hardware.
The vultures are circling, and I, for one, can't wait until *something* happens to Sun. Get bought, go bankrupt, develop a *real* open source strategy, just something. I'm tired of the bullshit.
Bruce Perens' basement is in the same league as Sun and RedHat. I bet many Fortune 500 corporations sleep soundly knowing that their financial systems are cool because they have Theo's home phone number.
Other GNU/Linux distros may not have military grade security like Trusted Solaris 8, but Security Enhanced Linux (SELinux) was developed by the National Security Agency -- surely that's good enough for government work.
Not really. Many people see different distributions of Linux as different OSes -- sure, under the hood it's the same, but dependencies, packaging systems, etc. etc. feed the perception that different Linux systems are just that -- where as Trusted Solaris 8 sounds to me like it's just a very secure version of Solaris 8. Keep in mind I don't know a whole lot about this stuff, but then, neither will the Execs who ultimately control the money towards paying for these systems.
Before I could thank them for their time, I was interrupted by a Sun PR flack, who informed me that I was not supposed to be there and that she was going to escort me to the door. It turns out that the press was supposed to leave a half hour before that, and that the end of the party was for Sun employees only. Somehow my colleague Chris Preimesberger and I were overlooked during the press and analyst roundup. So like the cops arresting the Monty Python cast at the end of "The Holy Grail," my colleague and I were ever so gently forced to leave the building. If only we'd been developers instead.
Then they would have given you a cheeseburger and a coke on your way out?
Sun does make some decent hardware. My friend bought one of their opteron 150 boxes off of ebay, and he loves it. I believe he has gentoo running on it iirc.
Viral software licensing is not freedom, it is in fact GNU/Socialism.
Sun is in it to make profits to please the shareholders. If the job doesn't requite Solaris, go ahead and use Linux, be it SuSE or Red Hat. That's what McNealy sees anyway... dollar signs.
Sun has done some pretty cool things with Solaris 10. It might actually end up making Sun a lil' bit of spending cash..
Discuss.
Vanya's Law: "In any culture without irony, fart jokes will be the highest form of humor."
"Fuck you're a bunch of loser nerds who could never get laid if your life depended on it. Go home and leave our marketing FUD alone. We like our lifestyles and need money to fund them."
Using software that infringes a patent violates system and probably method claims. Unless you have a contract agreement with the software company that says they'll protect you against patent infringement lawsuits then you're screwed. And if you know the software you're using infringes a patent then you're screwed x3.
The reason you rarely see companies going after users is because they tend not to have as much money as the company making the software.
Nerd: Derogatory term typically directed at anybody with a lower Slashdot ID than you.
At least that's the way I understand the masses.
This issue is a bit more complicated than you think.
I think everybody ought to give these Sun boys more of a chance, they're attempting to sass their way to the top rather than butt kiss their way there.
So let's see if they put their money where their mouths are.
Nescience Redemption
Right now the only thing that differentiates Sun from the rest of their market place is their expensive high end hardware. They need to squeese as much out of it as possible till it caves into the x86 - 64 commodity CPU market. Then their ability to gain high profit margins will be gone, as well as their position to compete in the computer space. Part of that differentation is solaris, that's way they need to squeese as much out of it as they can even if Linux is the one taking over the server-space industry.
I used to like Sun, but now-a-days... why bother?
This issue is a bit more complicated than you think.
You say: "And we want developers back on our side. If there's more for us to do, we'll go do it,"
maybe then you can show some good faith and put some action behind those words? don't want to open up java any more? fine, then at least remove the ridiculous redistribtion limitations
since you're giving away the software, you're left to making $$ from hardware and services. try being a little more friendly with those seeking to buy your hardware, and maybe you'll see a resurgence in hardware sales
because of the tactics you've employed in the past, i chose to not buy an opteron system from you and instead opted for another vendor. change your tactics and perhaps you'll start seeing a resurgance of your customer base
vodka, straight up, thank you!
It's a bit more complicated than that. If you read the SELinux FAQ:
The NSA itself says that it's NOT one, so on its own SELinux is not good enough for secure US government work, despite its being developed by the NSA.
Qu'on me donne six lignes écrites de la main du plus honnête homme, j'y trouverai de quoi le faire pendre.
Sun still has to address the issue that their old market seems to have gone away to a large extent. While it made sense 5 years ago to drop ~$500k on an E10k box to get reliability and support, nowadays you can get that same reliability for much less by using piles of clustered Intel hardware and a fairly-competent Linux or BSD admin.
...), so I just can't see why people would go with Solaris these days.
Outside of academia, the only reason people bought smaller Sun boxes is so they could develop for Sun's big iron with minimal migration issues at deployment or "scaling up" time. With the disappearance of Sun's big iron market, their low level market disappeared as well.
Open sourcing Solaris 10 is fine and dandy, but I think it's too little too late. There's brands of Linux and BSD (e.g. RHES, Debian, SE Linux, OpenBSD) that cover every one of Sun's old sweet spots (e.g. uptime, security, Oracle support,
Bottom line: Where is their sweet spot for selling their product? Why would I buy Sun these days?
It's a pity - Sun had a terrific product line that no-one else could match, but they didn't see the tide turning.
At the end of the launch event Jonathan Schwartz made an impromptu speech; I didn't hear most of it, as I was too far away, but he did end his comments with something about Slashdotters. I ambled over to Schwartz and said, "If anyone here is going to get an article onto Slashdot, it's probably going to be me (since NewsForge and Slashdot are both part of OSTG). Tell me what you'd like Slashdot readers to know."
"Tell them that we're returning to our roots," Schwartz said, referring to the company's renewed focus on the Solaris operating environment.
"And we want developers back on our side. If there's more for us to do, we'll go do it," McNealy added. It was the first time all day that I felt that the two had broken character and simply told me what was on their minds.
a shame because Sun can be a champion.
Question: if you were leading Sun,
what would you choose to do and why?
Wow! And you Linux folks talk about the Windows users being gullible.
Does that mean they're going back to BSD? Oh, I see: Motorola hardware.
How about solving the chick problem! We really could use some help on that one.
sigs, as if you care.
I hope Sun reads this. The interesting tension mentioned is that between a server and a client OS. The most interesting promise of Linux is that it might eventually get to the desktop. Why is that important? Because that is how MS is getting to the server. They have a great GUI environment. It's not just the ability to configure everything from a GUI vs. just storing things in files, although that's a big part.
I personally think it's in the whole GUI environment. Whoever has the most advanced GUI technologies will win people's hearts. Look at Apple. They did something different, and thus they have a chance to have their server technologies paid attention to. I write client-side Java for a living and like it a lot, but it just pales in comparison to Microsoft client-side tech. It looks ugly. Even Eclipse doesn't have the polish and jazzy look that something like VS.NET has. Sure, VS may be much less technologically advanced under the covers, but for God's sake, can't we just realize and admit that PEOPLE ARE SUPERFICIAL AND WILL BE ATTRACTED TO WHAT LOOKS BEST. Even when they finally figure out that it was only skin deep, it's just the way it is. When you are the beautiful hot chick you will always get attention. Whoever comes up with whatever dethrones MS will have to be a beautiful hot chick with four PhD's from MIT, whose also a star athelete.
If Sun really went after the desktop computing market beyond just making another window manager, they might have a chance. Looking Glass is a step in that direction, but I think they need to get much, much more serious about it. Turn Solaris/Linux/JDS into the absolute best GUI/gaming/graphics platform there is. *Only* if they do that will they have a chance really to get back into it.
That said, I would prefer to work with Solaris over Linux, assuming it supports my particular hardware. I used to manage a bunch of Sun servers and the stuff was just a cut above, especially with the superior consistency in things like design and documentation. The stuff just worked as documented and promised. There wasn't as much guesswork and fiddling.
That doesn't mean Linux doesn't have tremendous advantages, not the least of which is hardware support.
look at linux but have a hard time switching from solaris because once you realize the cost of, for example, RHE3 support and it's about the same and solaris is a much more mature OS. The linux kernel has some things to iron out still. Plus with solaris you have one vendor to deal with for hardware and software. There are very valid reasons why solaris is still alive.
http://tinyurl.com/3t236
"Sun does not have the operating system market on Wall Street -- they're not even close"
Yeah, like the author really knows what he's talking about. Wall Street is Sun's to lose. Everyone likes solaris, it's just slow and the hardware is expensive. Now that Sun's moving downmarket, it's faster and the hardware is cheaper.
Last I was in the space (over a year ago) Sun was losing share in the lower middle market, but the high-end was sticking in a wait-and-see mode. Their share on WS might have collapsed dramatically, but the numbers from IDC (unit shipment) don't bear that out.
So I guess the reporter was exaggerating to make a point? Does he actually have data to back this up?
Red Hat does a lot of work with SELinux, thank you very much, including employing its author Russel Coker. EL4 will be, AFAIK, the first distro to ship with it turned on by default.
ReiserFS isn't supported (although it is there, start the installer with the 'reiserfs' parameter)
as it requires reinstalls, has had stability issues (in particular, it used to have arguments with the NFS driver, even when Suse were shipping it). But Red Hat have GFS for use as a SAN filesystem and underlying on disk filesystem.
So, I was there at the launch (being a Sun employee tasked with providing some technical support for one of the kiosks... I don't want to go into too much detail because I'd like to retain at least a little anonymity on /.).
I haven't yet understood any message that the Cult of Personality(tm) has been putting forth, but one has to realize that Sun is a big company that has many competing interests vying for control within it. JDS sucks, and everyone there knows it, because we have to use it (that or Solaris, which in my group would be next to impossible... at least the current version). But JDS had a groundswell of support and when policies are made, they are often tough to kill, even in the obvious face of failure. Red Hat is the name that is used to fight against because they are the market leader, even though the Solaris people know damn well that Red Hat != GNU/Linux. Red Hat had a banner plane flying over the Tech Museum in a marketing gimick meant to draw attention away from the launch. Of course they will be the target of the CoP(tm) attacks.
I can't say I like Sun corporate, and I think that the infighting there is ridiculous. There is some really cool technology that is being developed however, and some people with some good ideas. I just hope (for my stock's sake) that those people and projects manage to get the attention and funding from the talk-boxes who make the decisions.
On a positive techie note, one of the cooler things I saw was the dtrace support in Solaris 10 for doing kernel tracing. As an engineer, I find that very fun.
*yawn*
In consequence, Netscape's SSL is considered acceptable for Government use (and DES has only just had its permission revoked), but the DoD's own implementation of IPSec and the NSA's work on SELinux are not. Rijndael-128 is OK, but Rijndael-256 is not. Even though all the evidence so far is that both versions of Rijndael are perfectly good.
A version of SuSE Linux (with help and funding from IBM) has been certified by the NSA as secure under the "Common Criteria" at about the same sort of level as Windows NT. This was on a PC I believe. No other platform for Linux, and no other distribution of Linux, has been certified.
So, you CAN run that specific version of SuSE on the specific PC platform it was tested on on military unclassified or confidential networks. Because so few OS' have been certified (only a tiny number of Unix manufacturers have the money for the approval process, never mind the development!!!) it's common practice to run any "approved" OS on Secret and Top Secret networks, even though they're not supposed to.
(Having worked as a contractor for the DoD, I can tell you that it is also not uncommon for software companies to request and receive waivers exempting them from NSA security auditing. The main appeal of COTS solutions, such as Microsoft, is that it's a lot cheaper than most GOTS solutions and the quality is about the same.)
For real "military grade" security (the stuff the military would like, if they weren't spending all their money in strip clubs) you'd need to take one of the existing security patches and add the following:
All that would give Linux a clearance comparable to the old B2 or B1 levels, which would be more than adequate for most classified networks. Relative to the work already put into Linux, it's really not that much. If IBM and SGI wanted to pool resources to make a B2/B1 version of Linux, I see absolutely no reason why they couldn't.
Now comes the fun part! What if you were to do all the above, and then do a line-by-line full coding audit with formal validation? IBM has something like 10,000 Linux coders. There are 50,000,000 lines of code. Assuming you could do the audit at no more than 10 lines a day, it would take 100 days to audit the kernel to this degree. For a real bare-bones box, it would probably take about the same to do the user-space stuff.
What would this give you? Well, the ONLY COTS Operating System to be A1-certifiable. There simply aren't any other. Nobody makes software to the A1 standard. At least, not that
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
" ... we want developers back on our side. If there's more for us to do, we'll go do it," McNealy added.
... followed by ...
... my colleague and I were ever so gently forced to leave the building.
Sun is just that friendly! Hey, we want you on our side, now get out of here!
[signature]
Indeed, but the thing Sun is worried about is the question: How long? (until linux is mature enough at the speed it is going forward)
We're all aware of the hole-ridden arguments that Sun executives Scott McNealy and Jonathan Schwartz use to attack Linux.
Sun is not anti-Linux, Sun sells Linux, Sun will even sell you a full rack of x86 servers all running Linux. Get over it, Slashdot!
-- "Makes Little Debbie look like a pile of puke!" - Moe Szyslak
UML has substantially low performance compared to N1 Grid Containers. If you're going to compare a server virtualization feature, compare to something like the Xen Virtual Machine, in this performance comparison, you can see the performance of UML is rather appalling, especially compared to Xen.
The performance of Solaris Grid Containers is more akin to Xen or FreeBSD jails. However, the advantage N1 Grid Containers have over Xen is that they are portable to every platform Solaris runs on (SPARC, IA32, AMD64) whereas Xen only emulates one platform (IA32). Also, other Solaris features to which there are currently no Linux counterparts such as the Fair Share Scheduler, which allows a N1 Grid Container to be bound to certain processors, and given a dedicated percentage (or share) of available processor resources. This provides an advantage over Xen and UML which can't even use multiple CPUs. It has an advantage over FreeBSD jails where monopolization of system resources by a single jail cannot be easily avoided.
While Linux may have counterparts to various Solaris features, in terms of maturity, feature set, and performance of these features Solaris has Linux trumped.
http://www.sun.com/bigadmin/features/articles/sola ris_zones.html
It's interesting that FreeBSD influence is getting
recognition at SUN... Maybe now they will be persuaded
to support some of their products on FreeBSD.(aka Java, and yes, i know about the FreeBSD java group
and their agreement on the 1.3.X jdk with sun)
Top 10 reasons Trusted Solaris 8 is so great.
10) Written by Sun
9) Used extensively by the government in classified situations
8) Containers
7) Roles
6) It costs quite a lot of money.
5) It takes 2 people to do one simple job for 'securities sake' at least where i work.
4) Its proprietary
3) Its unique
2) SELinux doesn't have the same feature set
1) Oh yea, its got that Sexy Swing look (Java)
Why is sun bothering to position themselves against Linux, when all the market share they're after is running either a Microsoft product, or MVS?
It's like Ford saying that they're going to target Vespa.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Solaris in a WiFi access point? No.
Solaris in a watch? No.
Solaris in a TiVo clone? No.
Solaris in an auto's ECM? No.
Solaris on a laptop? No.
Solaris on a mobile? No.
Sun doesn't get populism. They're too used to their own history to reinvent themselves, or understand community OSS efforts. I watched Scott slam IBM's pension funding, and Schwartz look fawningly on in the press/analyst conference. I watched the little plane with the Red Hat "Business as Usual" banner flying behind. Sun wants Linux love a lot. And they're clueless on how to get it. Not McNealy's bruskness or Schwartz's blogged ponytail are going to get it. It's a shame, because Solaris 10 looks droolingly good, if proprietary (look closely at ZFS, dtrace, and so on). Sigh.
---- Teach Peace. It's Cheaper Than War.
At the end of the launch event Jonathan Schwartz made an impromptu speech; I didn't hear most of it, as I was too far away, but he did end his comments with something about Slashdotters. I ambled over to Schwartz and said, "If anyone here is going to get an article onto Slashdot, it's probably going to be me (since NewsForge and Slashdot are both part of OSTG). Tell me what you'd like Slashdot readers to know."
"Tell them that we're returning to our roots," Schwartz said, referring to the company's renewed focus on the Solaris operating environment.
"And we want developers back on our side. If there's more for us to do, we'll go do it," McNealy added. It was the first time all day that I felt that the two had broken character and simply told me what was on their minds.
As a long time Slashdotter who has had to use and deploy Solaris on occasion, let me tell Mr. McNealy and Mr. Scwartz what's on my mind about Sun. I know they'll be reading, so here goes:
First, cut the marketing BS. No press wars with Redhat, IBM or HP. No trumped up, spin laden press releases about Solaris 10. I don't even want to see a comaprison paper. Give me a technical white paper about what the OS can do and STFU - I then can see for myself whether Solaris 10 is a good or great OS. I can also then decide for myself if it's a good fit in my architecture. Most on Slashdot are technically adept - that's why we can run and support Linux or *BSD without Redhat's help. It's the PHBs who require that kind of hand holding, not us. (Hey, I just invented a new comic book villian - Spin Laden, the Marketing Terrarist!)
Open your dev process, as well as your code. I don't (necessarily) mean provide CVS access, I mean accept and credit quality patches to the code base. Open code would mean we can fix our own damned stuff when things in Solaris break and get our jobs done, while benefiting anyone else who has the same bug - we tend to like to share the fact we're smart enough to repair someone else's broken code. For large contributions, pay the contributor and pay him well.
Stay away from the rest of my systems unless I ask you in. No embedded Java in the OS, no Sun only core stuff (think Microsoft and Kerberos 5), just a big box of properly impelmented tools that I can use to make systems work, work well and work reliably. Your products will be sharing my network with other vendors, so play nice whenever you can. If that means re-writing some Solaris code to put into linux so it interoperates properly and GPLing it, so be it. That way I know that you're concerned about me and not just "maximizing value".
Contriubute to the industry. Some of us think RMS is a real looney, but we have the utmost respect him and his contributions. Mr. Gates, IMHO, does not contribute to the general cause or making my life easier unless there's a price tag, be it in dollars or having to shut out one of his colleagues - he calls them compeditors - from my architecture. Real contributions move the whole industry forward, and provide new opportunities for everyone to make a little $_CURRENCY, not just a select few.
Censure that person who 'escorted' out the interviewer. We like plain talk. We know you have fiduciary responsibilities, and most of us try to take those into account, but trying to hide what you really want to say doesn't wash. If you hate linux or love it, say so, and say why - with no spin on the matter. Speaking of plain talk, you'll get some from us. We know you're the head of a big, powerful Corp., but you should be willing to learn from us. When it comes to putting the tech on the floor, we are your betters, not your underlings.
Lastly, put your engineering department off limits to marketing personnel. OFF-LIMITS. Spin Laden should be shot on sight (by a Nerf gun, of course) if he dares tread where something cool is being made. No "That's a killer system, and we can leverage it to sell..." baloney please. I'm still loathe to implement AD because it's actually proprietary technology, even though it would make administrating my network a little easier.
Thanks for tuning in to my little rant. HAND.
Soko
"Depression is merely anger without enthusiasm." - Anonymous
Unless I'm mistaken, Sun used to have a great reputation in the CAD world. Am I imagining things again?
SUPPORT.
BLAM. That's IT.
You pay half a million for your box breaking to be SOMEBODY ELSE'S PROBLEM.
That money makes your box Sun's priority. Period. They'll FIX IT. Software or hardware. They'll roll you custom Solaris patches, because you're paying for it.
You're paying for a COMPANY to give you some LOVE. Not some snotnosed Admin whose first-line defense is an O'Reilly bookshelf.
There's a definite market for this kind of service. Just because you're not in it doesn't mean it isn't there.
There are several tools I consider to be very important to me when I'm using a remote environment. These include gnu/screen, less and (most importantly) vim.
I've done development work and sysadmin work on Sun boxes and on linux boxes. I have consistently found that linux boxes come set up in a way that is well-fitted to my usage, whereas operating under remote Solaris environments is a struggle (this comes from having worked on it as developer or user in three separate and very different organisations). In all cases I have had a struggle to get these three key tools installed in environments (and at times lost) where it's been perfectly obvious to anyone with a clue that they are important. (picture several-megabyte log files where the most advanced pager is more and the only editor is vi, which breaks in ugly ways with even moderately sized files and which can't read more than a certain number of characters per line)
The admins at these companies have even at times given me excuses against these tools such as "that's not compatible with our security policy", yet the same environments they have perl installed!! While I realise that that's not Sun's fault directly, linux admins are more open to the idea of using these tools. Thus, when I've been a Sun user I've been unhappy with the experience, and when I'm a linux user I tend to like the experience.
Just in case anyone's tempted to write my opinions off as those only of an utterly naive linux user who couldn't get by in a slightly-unusual world: I do know a bit about the Solaris environment - I'm familiar with ksh, use set -o vi, and am fluent with vi.
I'm under the impression that once upon a time Sun was at the cutting edge of trying to improve the environment - competing against other unix providers to put killer tools in that made it stand out as excellent. Now I realise there were downsides to this (unix wars mentality, etc) - but there was something in that. These were the days when somebody would write/find a great new tool and just ship it.
This unix geek see his linux experiences as the bar against which everything else is measured and I suspect many people are in my shoes because that's what we grow up with. It's possible my experiences have just been an unlucky coincidence, but unlikely.
Guys - you need to win me back by doing things to ensure the Sun user environments I get exposed to are up to scratch.
Here's what I'd recommend to Sun by way of improving the situation:
- move to the version of 'more' derived from the 'less' codebase (if you don't already use that) and ship both by standard in the operating system (if you don't already - can't justify looking this up right atm)
- terminal definitions need to standardise so that vim and less work 'out of the box', *including syntax highlighting in vim!!*
- I'd recommend that people who go to Solaris courses, education programs or read the official books get exposure to the tools that developers love to use so they don't get paranoid at the prospect of exposing them to users
- make sure your evangelists match the target audience. In the two experiences I've had of Sun marketing people, they have not been from a cultural background to allow them to appreciate the difficulties I'm describing here. They'd rather talk about Sun One, or Java or current buzzwords, and they look a lot more comfortable talking about them with managers than developers. If you're serious about evangelising to developers you're going to have to do it properly
- distribute some sort of security policy for high-level secure environments that validates versions of significant tools that are important to users. eg: maybe you could have a program of forking vim every now and then and having a 'Sun-endorsed' version.
- get ahead of the rest by distributing a pager that's specifically designed to make it easy to bounce around huge log files without loading the whole thing into RAM (there may be something like this already - I don't know of it)
Believe with me, my saplings.
I am my 24x7 support source. I built and developed my high performance server clusters on slackware and would not have thought to use solaris because meeting my needs would have been considerably more cumbersome to accomplish with a boxed solaris product. It would have required dozens of third party (GPLed apps) addons before I could begin development for real world. Perhaps more importantly than this I was able to package my solutions for replication across hundreds of servers in a few hours and deployment takes 15 minutes per node. Every machine is running a high performance journalled filesystem and is smart enough to email me if it's having a hardware problem. Eat your heart out SUN. Linux and free BSDs are for people that love to hack this stuff out and have some idea or someone that knows what they are doing. Redhat's solutions much like SCO and SUN solutions are for people that don't want to know what's happening in their own environment.
Don't get me wrong I love solaris but unless it's running on a 32 way e series machine it has no place trying to compete with ANY Opensource OS for real world usability out of the box or in the closet.
there is no legal basis for Kodak to sue end users over their use of the JRE or JDK. End users did not infringe upon Kodak's patents -- they downloaded the Java software in good faith that it was perfectly legal, and they presumably abided by the license terms. Kodak would have absolutely no right to try to recover any damages from an end user or anyone else who was not a party to adding the allegedly infringing code to the Java source code.
We probably all wish for it, but that is not how patent law infact works.
Using something in good faith is no defence against a patent-lawsuit. Neither does it save you that the patented algorithm was added to the software you use by someone else, without your knowledge.
If this was a valid defence, then most Linux-users would also be equally safe, afterall they *also* tend to use Linux in good faith, abiding by its license terms, and they *also* had the hypotetical patented technique added by someone else without their knowledge.
Sadly, that's not how patent-law works. There are basically only 3 relevant questions in a patent-infringement-lawsuit:
If the answers to those are yes, yes and no, then you are guilty. Even if you didn't *know* the patent existed. Even if you had absolutely no idea that your software was doing this. Even if the software infringing on the patent was written by someone else. Hell, even if the software is closed-source and you thus reasonably *couldn't* know that it was doing this. Those are all irrelevant.
And surely the government knows what's best for us... NOT!
Without getting into the Linux/Sun arguement (which could go on forever), don't forget this relation:
Newsforge::C|Net
Fox News::BBC World News
(And I'm not too fond of C|Net either)
NewsForge is news about Linux, for Linux, and related to Linux. It is so utterly biased and laughable (and amateur) that any 'expose'' it does is almost entirely suspect. This article certainly was.
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
Even when I could get it for free, I will not. I'm totally uninterested in it. To me Sun equates to cool looking hardware boxes with shitty software.
A few years ago, I attened an SGI conference in Las Vegas. Didn't lose too much that time either... Anyway, the point of this conference was to communicate the SGI vision going forward. This was right about the time SGI got done getting their ass kicked in the win32 / Intel space.
Sgi said they needed to return to their roots too, just like Sun is saying now. For SGI this meant, taking their best tech forward while cutting costs on everything else. Good message, seemed the right thing to do.
Well, how are they going to cut costs? Enter the chief scientist, an Asian GUY Goh, I believe. Very personable, very smart, very excited about --- Linux and OSS.
The SGI plan was very simple. Keep IRIX doing what it does best. At the same time, begin working on Linux. SGI learned they had to accept the community as a partner. This means if they submit something and it gets rejected, they either don't do it that way, or submit again, or maintain it as an add on, until the community catches up with them in that particular area. The idea being that either their solution would be accepted, or the community would evolve one that SGI could use.
(This does have to do with SUN, bear with me!)
So, SGI did go back to their roots, worked with the OSS community, and ended up once again able to do what they do best; namely, low latency, NUMA supercomputing. They are 2nd on the top 500 again, for now, and their flagship machine runs Linux!
At the time, I thought: "uh Oh, there goes SGI..." You can say what you want about IRIX, but it does what it does very very well. Linux looked impossible at the time. But it worked, and worked very well for them. SGI lost a lot of smart people, but obviously kept the ones that mattered. There was one other significant thing: After the banquet, I got a chance to talk with Bishop. Very interesting fellow in that he is totally geeky, but has solid business sense, and a direct line to NASA... He told me SGI was going to commit to this new course no matter what. Half way was not going to cut it. SGI makes the lions share of its money making powerful systems that do things that are near-impossible to do. Anything else would only prolong the death spiral. That meant getting rid of the baggage in measured steps, then build again lean 'n mean.
So, now we look at Sun.
All of SGI was committed to doing one thing, well actually two: Building their Linux / Itanium platform while doing everything they can for IRIX / Mips. To this day, they have not deviated from this vision at all and it is now paying off, just like Bishop said it would.
Sun? Lots of infighting, no core vision to drive forward. Until they fix that, they are doomed to fail because nobody is going to pay for 'almost the greatest' solutions, which is what Sun is selling right now.**
**Please don't flame for that. Sun makes good stuff, but they don't have clear niches where they are the absolute best and where there are few to no alternative solutions.) Massive SGI NUMA, mixed with graphics, insane I/O, and big low latency memory machines solve a class of problems that nothing else solves. There are only a few players, none as mature as SGI is. Ok, back to my points...
Sun needs to cut the baggage. Carrying Solaris forward is not going to be the answer. The cool hardware features, redundancy, hot swap, etc... can be solved in other ways. That means Solaris really does not have anything the market must have and that's the key to this whole thing.
SGI realized this with IRIX. However, some bits were needed on the Linux side, such as their XFS filesystem. The few bits we are clamoring for, Sun wants to keep tight hold of and this is a mistake. The market is not going to rebuild onto Solaris, all the work done with Linux, just to get Java, or redundancy, for example. Instead, they are going to just figure out how to do it with Linux, just as they have everything else.
The SGI approach at least got their technology in wide us
Blogging because I can...
I think you'll find most of us don't have anything against Solaris as such, but a lot of us have had enough of Sun's bullshit and FUD.
At least SCO, since you mentioned it, is consistent. It keeps saying the same thing. "All your base are belong to us." Err... I mean, "We own Unix, Linux is evil pink commie stuff, everyone copied our code." I can deal with that.
Sun's upper management is plain old multiple-personality schizophrenic, taken as a whole. You never know with which personality you'll deal today. Or even at different hours in the same day. Will it be McNealy 1 who loves Linux and OSS more than he loves his mother? Or maybe McNealy 2 who doesn't even have any strategy, and just foams at the mouth worse than any zealot? Or McNealy 3 who's as convinced that Linux sucks as Darl from SCO is, and spreads FUD about Linux? Or who?
I wish Sun just made up its mind already. These fits and hourly changes of mind are getting tiresome.
A polar bear is a cartesian bear after a coordinate transform.
That is all great, but it may be too expensive to pay for all waht comes around.
Sun sells solutions (OS, Machines, Services, application software). But perhaps many clients do not want that any more for the premium price Sun charges.
Right now we have two bugs with Solaris 8, one has not been solved for long time (and we know they don't want to solve it) and the other we were told they are not going to fix it, that we should upgrade.
Now tell me, what use is all the bells and whistles of a superior OS if when basic functionality breaks you don't obtain an adequate reponse for the gazillions you are paying for support?
All those things are on the todo for Xen, and are expected within a month or two.
I work for a large software company and we only support Redhat for linux platforms. We are now moving our customers to Enterprise 3. Apparently it is really difficult to take an application that was developed for Redhat and then use it on other Linux distributions. We don't support them if they try. If Redhat played nice, perhaps this would be less of an issue.
I'm probably going to get my first negative karma from this one but here goes.
I'm a linux developer... and a FreeBSD, Solaris developer. The only 'major' OS I don't code for is Windows (intentionally).
Anyhow, to my point now, I must say the "Elitism" of Linux supporters is sometimes overwhelmingly sick. I've encountered people who refuse to believe that there's anything at all better than Linux, "Linux is great". I'm personally sick of it.
Sun has some great tools and some great developer networks. I don't use them much myself as I simply don't write programs which get down to the levels where OS differences become a major factor.
I for one am greatly looking forward to ZFS and DTRACE. For years I've been asking around in the linux community for something like dtrace, best responses I got was "Why would you want that? Use top" or "Profile your application" (like wtf??). I'm willing to bet that now that Sun as released DTRACE there's going to be a clambering to release a similar thing in linux.
To be fair, linux has also some lovely tools, valgrind is the one major tool which keeps me holding onto linux.... for now.
PLD.
You may be working for large banks and telcos, I am working for the largest bank and worwed for the largest oil field company.
What you are saying standing here is bunk.
There is no freaking bloody way that we would outsource our Solaris support to other companies.
The reason is glarongly obvious: consultancies do not have access to the source code.
That is the wasy with closed software, if you want the best support you want to talk to the guy that actually wrote the damn thing, not to one guy to know how to configure it or fine tunie it.
>> While Linux may have counterparts to various Solaris features, in terms of maturity, feature set, and performance of these features Solaris has Linux trumped.
:-( :-)
I just spent 10 mins ranting in my own post - and you summarise in one line
Incidently, the N1 grid containers are going to be very useful indeed. I can imagine ISP's picking this up as a nice way to offer 'dedicated OS' hosting solutions.
And the machines will fine tune and configure themselves.
You obviously have no idea what it costs to provide true contingency, even with "cheap" hardware.
IANAL but write like a drunk one.
Find it kinda amusing that the article basically is a criticism against Sun for making broad statements which are either half thruths or not explained anywhere, especially against Red Hat. Then the author himself goes onto criticizing Red Hat and calling it inferior to Suse without explaining why. Maybe he should be a Solaris 10 user, seems he has the same mindset at Jonathan Schwartz :)
From TFA:
``I asked Scott McNealy if he ever considered Java's closed licensing from a user's perspective, and I gave him the example of FreeBSD/AMD64, which has no native 64-bit JRE because Sun has not yet provided one.''
How about Sun Community Source Licensing? Sure, you cannot distribute modified versions, but the typical operation of BSD ports is to download original + patches anyway (so the modified version is created locally). I don't see how this can't be used to make a native port for FreeBSD/AMD64, or any system at all.
Please correct me if I got my facts wrong.
How many of us need military grade security (MGS)? Sure MGS gets Sun bonus points with intelligence and defense but the rest of the market looks at it very differently.
:)
I'm about to finish a report that compares a Linux i686 and a Solaris SPARC III/IV solution. CPU power is the issue here (not I/O or FS quality) and the costs for the Linux solution (in my specific case) are almost negligible compared to the Solaris solution.
Don't get me wrong, I have a long history with Solaris (started with SunOS 4.1.3 and even developed for and administered SunOS 4.0.x) and I think they were great to universities in late 80s and begin 90s. But they lost the commodity hardware race. Or did they actually compete there?
So intelligence and defense will support Sun the current and next decade. Does that mean Jeb Bush should be the next president? Lucky for me I'm not a US citizen
I hadn't the slightest objection to his spending his time planning massacres for the bourgeoisie... (P.G. Wodehouse)
Red Hat is dead.
There you are, staring at me again.
In the corporate world there is no such thing as "natural allies". Especially not with competing products. But generally, when a business man/woman shakes your hand, you can bet his/her other hand is behind his/her back, holding a dagger.
People, this is not Tolkien, where the elves are your natural allies for eternity, and the orcs are your natural enemies. In Tolkien's world you know where you stand. It's a simplified world. That's why we like to escape to phantasy worlds: they're a refuge from the madness of the real world.
Real world is nowhere near that simple.
In the corporate world, there are no heroes in shiny spandex, and no villain cackling over death ray blueprints. There's only a bunch of greedy people trying to make a buck. Your buck.
Throughout the history of computing, as little of it as we have, one thing stayed a constant: whoever is in the lead wants proprietary stuff and tries to lock you into their incompatible formats. Whoever is losing badly wants open standards and generally a fair chance to have a go at the big guys' locked-in customers.
Then the wheel turns, companies go from top to bottom and viceversa, and they switch the tune without missing a beat. And things stay the same. The ones who are now winning, try to lock you in, the ones at the bottom suddenly become open-standards evangelists.
That's why IBM and the rest are supporting Linux nowadays, for example. That's why Sun would even give away OpenOffice, even with sources, to try to break MS's file format lock-in.
There are a lot of has-beens in this industry. People who once owned the market, but were too stupid to keep it.
E.g., PCs once had to be "IBM Compatible", then it was "Intel Compatible", while nowadays it's "MS Windows Compatible". Intel doesn't single-handedly decide new architectures any more, but has to beg MS for support in Windows. (And just got refused recently!) IBM had its ass handed to it a longer time ago, when the PS/2 microchannel architecture was basically rejected by everyone else. The company that created the PC was no longer in control of its architecture. Novell once owned the network server market, but thought it could ignore NT and stick to charging outrageous prices. Prices for which you could buy not only 2 NT server licenses, but also 2 high end PCs to run them on. Etc.
And when they still were at the top, neither of them has acted any better than MS does. E.g., although nowadays "FUD" is synonimous with Microsoft, once it was synonimous with IBM: In fact, it was _invented_ by IBM.
Now all those has-beens are suddenly pro-Linux and pro-open-standards, to get their righteous vengeance against MS. But if either got back on top, they'd start doing the same shit all over again.
And Sun is the prime example. Sun is somewhere in the middle, and can't decide if it's losing, or still has a chance of being king. As soon as it thinks it's losing, it starts being a Linux zealot. As soon as it thinks "hey, maybe everyone will convert to Solaris if we port it to the Opteron", it starts openly trying to kill Linux.
And as management perceptions and sales figures fluctuate, pushing them a little up or a little down from that middle position, Sun flip-flops between the two extremes several times a month. Or sometimes even within the same day.
Sad.
A polar bear is a cartesian bear after a coordinate transform.
RedHat is the most visible enterprise Linux vendor. Most software and hardware vendors that I have dealt with automatically mention RHEL when I ask them about Linux support. Most of my sysadmin friends who use Linux in enterprise seem to be using RHEL too. At the same time, RHEL users are exactly the type of users Sun wants to be their customers (e.g. who care about using an OS that has received a huge number of software and hardware certifications and has vendor's support for piece of mind). I don't think that there are too many other Linux vendors who can claim this level of acceptance in the enterprise market. SuSE might have a good enterprise product but they probably aren't on Sun's radar yet due to their small market share. So, I think it makes a perfect economic sense for Sun marketing to target (and bash) RedHat. RedHat has made it specially easy for them to do that with RHEL pricing. But RedHat is not Linux you might say? Yes. But Sun doesn't care. RedHat customer base is what their target is. So, I don't cosider it to be FUD when Sun implies that RedHat == Linux. RedHat IS Linux as far as enterprise customers are concerned. Sun doesn't care about others much.
We're using Sun One App server for a project and it's a fairly dreadful application server. Slow, clunky and just on the whole inferior to Apache Tomcat (even though Sun One is based on Tomcat)
Solaris might be supported and have some enterprise friendly funtionality, but it's also rather archaic. Some of it's daemons are rather old and needlessly installed.
Linux has JFS, XFS and Reiserfs, so don't tell me it doesn't have any decent filesystems. The performance of 2.6 is great, beats all the BSDs and i'm sure it would beat Solaris.
As for security, this is a problem definately, but it's up to the vendors to harden their products, this (and support) is what you pay your license fees for when you buy Red Hat. But since Sun now ships with Gnome and has all the GNU software I can't see how Solaris on the whole is much more secure. It all comes down to the kernel.
You don't need to use a processor set in order to
use the fair share scheduler. You can indeed just
bind the zone to a processor pool, but it's
cleaner to just put it under FSS and give it
a fixed number of shares. It makes far more sense
to have two zones, each with one share, than to
bind each one to half of the processors.
ian
It's specifically a reply to some verbal manure from HP, but it cuts through a lot of the ignorance and hostility here, including your own troll which has been modded Interesting due to the current Sun-basshing fashion here.
Their expensive high-end hardware? Why is that a problem? High-end hardware is reliable, degrades gracefully under load, and detects incipient failures.
a) x86 servers are getting more powerful
b) x86 server-class machines can be pretty damn stable too, given the right hardware
c) clusters are eating away at high-end segments
Yes, their high-end hardware gets more powerful too, just as x86. Moore's law and all that. But the users' needs don't grow according to Moore's law.
5 years ago, 10 years ago I could have told you of lots of companies that could never survive on x86 hardware. Today, I can name very few. It is an evil circle, where a smaller market increases your R&D costs per machine, increasing cost, providing an even smaller market still.
What happens when the premium is so high? It is cheaper to throw more boxes at it (serve less people per box, dedicated cache/replication/load balancing). Unless you abso-f*cking-lutely need it. But that is a damn small niche to be left with.
Kjella
Live today, because you never know what tomorrow brings
This is the same guy who wrote a pathetic review of a Sun Blade that the eds for some unknown reason thought was worthy of mention here some months ago. The guy is ill-informed and is not a Solaris user. Both his Sun-related articles mention his inability to get Solaris to perform and yet he blames it on Solaris; of course it would have nothing to do with his inexperience -- after all he is the author of the world-famous The Jem Report , The Internet's Best Computer Review Site (!!!), so he's obviously a guru.
And the ad for the "critically acclaimed" novel was cringe-worthy -- for some fun, read some of the comments from those that bought the novel.
I just hope the eds remember not to run stories from this site again.
If open source zealots delivered even a tenth of what they promised even a tenth as fast, we'd all be running Linux and OpenOffice and have perfect binary compatibility with Win32 and Microsoft would be out of business. I don't see anyone selling Microsoft stock just yet...
While Linux may have counterparts to various Solaris features, in terms of maturity, feature set, and performance of these features Solaris has Linux trumped.
Linux has those features that its users want and need because those are the features its users invest in. Virtualization and "Fair Share Scheduling" make sense for Sun with its overpriced server hardware. For x86-based systems, it doesn't usually make sense to virtualize--you just deploy on multiple, individually cheaper systems.
Red Hat is not the only product out there, though. There is a far superior server operating system called SUSE Linux Enterprise Server 9, and it buries Red Hat Enterprise Server 3 in every way.
.. most of the time.
Bullshit.
While RH's updating system is utter crap and then some, SuSE's response is desktop-only version of linux, with from non-existant to useless CLI tools. And this on server products from SuSE, such as "openexchange server" which I personally have had the agonising pain to deal with.
Even Mandrake succeeds over these two mammoths from the past, with it's fairly well working urpmi update tool. While it's not quite apt-get or emerge, it does it's job in correctly locating dependencies and merging then into the system
Not particularly surprising, when you consider that once upon a time SunOS was BSD. Long time ago, admittedly.
True. They told me when I went for an interview the other day....
"And who will support that? Red Hat won't support Debian,"
The context of Sun's words, essentially a press party, dictated that they speak in simple, repetitive statements designed to convey a message that the press would NOT screw up. I've done a half dozen or so press interviews, and believe me, 'the press' can distort a clear, direct, statement. So, consider the context. Also, consider tha Scott and party did not know the interviewer.
As for business. Who can know Sun's real intentions. In business the only intentions that are worth a damn are those that: are in the contract and not open to interpretation; lie in the cards you hold close.
Maybe Sun is headed toward a more (than is is so far) open OS. It's not something that can be turned on overnight and it's certainly something to be done slowly and carefully, as long as you've enough $ in the bank to be a lawyer target.
I don't understand why developers might eschew Solaris. It's 'another channel'.
Now I'm the grandest Tiger in the Jungle!
Thats bullshit and will be even more evident when 8 and 16 way opteron boxes are out in CY05. Most ORGANIZATIONS consider real estate, power consumption and cooling when specifying an installation. Having application virtualization built in, for free, can't be anything other than a good thing. Especially when the OS is free and the support cheaper than the competition.
Could you please elaborate? I haven't had much luck about finding howw Grid containers are actually implemented. A FreeBSD jail still uses the systems main kernel. UML doesn't. Do Containers?
Despite all that sun's done for the community, most linux-obsessed slash dotters (who see themselves as the archetypes) fall for the HP/IBM rhetoric against sun. sun started from bsd (tell me that's not "open"). You guys just don't get it, do you?Show me half a line of code that HP's contributed to the community. Ask ibm to open up it's crown jewel- DB2 for starts. And yes, open source solaris might draw some linux developers towards it, but, they(the developers) don't turn into hideous,thieving villians;they're still contributing to open source, still fuelling innovation, still strengthening the original bsd tribe (is that anti-linux now?) and still workin against M$. The 2 (solaris & linux) are under the same umbrella.They're siblings in the *nix family. Yes, running a uname -s on the 2 systems will yield a different name- if that's something you want to hold against solaris (you can laugh, but slashdotters are getting that gullible these days)- but that doesn't mean "Solaris's gonna kill Linux". What crap. The 2 are now part of the same moment. If you still want to "boycott" the new kid in town & shoot yourself in the foot (by selling your soul to HP,IBM & hence M$), no one's stopping you.
They tried that once. It worked about as well as a "field cannon". Everybody jokes about the Mac Portable. This was worse. It was actually passed off as a laptop.
The computer itself was little more than a regular pizzabox sparc painted black, a typewriter case handle, with a bad LCD screen and keyboard...and a car battery glued onto the back.
It had a power brick that needed its own high-speed fan, and was the size of half an unabridged dictionary.
One of two times I've ever seen one in person was when a NetApp sales engineer (sorry for the oxymoron) toted one to us to demonstrate solaris compatibility or something.
PS: were Sun's words "zOMG all your linux box are belong to us!", and "thrust vector calculations rule the OS!"?
Please help metamoderate.
How did this get modded up so high?
Ideas such as jails, chroot(), nested virtual machines / environments, running applications with least privilege, etc. didn't originate with either FreeBSD or Solaris.
The article uses jails for comparison.
Please give up the GNU/Elitism!!! Linux is finally
seen as a threat to one of the big boys. Sun is
about to release a product which for the application
I'm running my testing has shown is superior to ANY
linux release out there. If you cut through all the
bullcrap that Sun is spewing and you cut through all
the bullcrap the linux side is spewing it comes down
to which OS makes management feel more comfortable.
Solaris 10 x86 is just as cheap as Linux if not
cheaper if you choose support and for many
applications it is faster. Which would you choose?
Never trust a Jew in power. He will always lie, cheat, and steal to get what he wants.
McNealy equates "proprietary" with "interoperable only with the same brand." While that may be true from a narrow frame of reference, the free software world tends to use a different definition; when we say "proprietary," we mean that all of the rights to that software are locked away from us.
Back in the old days, before RMS and ESR got into a fight over what free meant, and we just gave away our code because we thought it was cool what other people did with it, proprietary meant "you buy this, you're stuck with it". Open systems, whatever the status of their code base, were a response to that.
Write your code to an open API and it'll run, with some effort, anywhere that API was implemented. If you used a proprietary API, you had to either rewrite a lot of your code when you wanted to transport it, or create your own transportable API and port it to each platform. One of the reasons UNIX was so popular is that the API was abstract, distant from the implementation, so it served BOTH purposes well enough that everyone, Microsoft included, ended up with UNIX emulation of some kind or another.
But benefiting from an open system requires remaining aware of the open API and what's not open. And this gives a back door for proprietary interfaces to sneak in again. You can get yourself locked in to an API without intending to. It takes effort to fight that, and a lot of the open source community doesn't seem interested in spending that effort. Apart from the unnecessarily complex X11 toolkit situation, there's just too much code that depends on proprietary GCC features, or on specific extensions to open-source versions of open-systems tools.
So McNealy is quite justified in using proprietary in terms of interfaces and protocols, and there's a lot of open-source developers out there who ought to pay attention. The source isn't enough. If we have to pull things like "a ?: b" out of your code to get it running on other implementations of open systems, then your software isn't as "open" as you think it is.
Whether Solaris is actually as open, in this older sense, as Scott would like you to think it is... possibly not. Sun's played the 'stealth extensions' game themselves in the past. But that's a different matter. I'm only talking about the meaning of the word here.
I think you'll have to get used to Linux fans looking down on every other OS. Because the others are running out of advantages. Compare your first Linux installation that you mentioned with any recent Linux version to see what I mean.
With large companies agreeing (for Linux, etc.) with the GPL and contributing to OSS, the rate of improvement should only increase. IBM may only contribute things of interest to IBM but this is OK; these things get done (by IBM) and the community is free to make improvements where "it" wishes. We do not require individuals to work on specific projects; each person finds the project he/she enjoys and works on it. Why not let companies like HP and IBM do the same? If they want to work on SMP support and you want to work on the TCP/IP stack, great. In both cases (if Linus agrees), Linux improves.
This is funny because I work for the military and I see RedHat being used extensively. Talk to a fellow named Brian Mikkelsen the federal sales director (DoD too) for RedHat and tell him that he has no military clients.
0 00072
H at_Cert ificate.jpg
Here's just one example with a quick search on google: http://afmsrr.afams.af.mil/index.cfm?RID=MDL_AF_1
The US Defense Information Systems Agency
Here's a link to RedHat's DII COE compliance cert:
http://diicoe.disa.mil/coe/kpc/RHLinux/Red
Now Sun does have a mucher larger market share as does AIX but thats only because they have been players in the DoD world longer and run the OS on their own hardware. As time goes on SuSE and RedHat will get a stronger foothold into this market. All that said about RedHat, SuSE rocks!
The real reason that Sun paid the $9 million over is to buy rights over code included in Solaris, which will shortly be GPLed. This is what they said at the time and this is what everybody with half a brain knows now, except for the headcase linux zealot types that read groklaw and think coincidence + their own personal prejudice = an open and shut case.
I'm so fed up of linux zealots. Why can't they just all go away and get a sense of perspective and some semblance of reason about how the world works. Please! Why do they all reason that "what is good for linux" is somehow the absolute moral good of the unvierse and must be followed by everyone, regardless of what is in their interests? They need to understand that when someone criticises Red Hat or SuSE they are not criticising linux and they are not eating babies and they are not saying you have a small tiny penis.
Christ.
From the article:
"We had to pay SCO more money so we could open the code -- I couldn't say anything about that at the time, but now I can tell you that we paid them that license fee to expand our rights to the code," he said, referring to the February 2003 multi-million-dollar purchase of expanded Unix SVR4 license rights from the SCO Group."
I would love to know if SCO even has the right to allow this, considering that they do not even own the copyrights. Even if they do, I think that it will be interesting to see what license will be used for an open sourced version of Solaris.
Check out this entertaining response to HP (thanks James Gosling).
So why does it have to be "Linux VS everything else" (like Windows, Solaris, xBSD, etc)? Why can't it be "Linux AND everything else"? I think there's room for everything...right tool for the job and all that.
Just a thought.
Never by hatred has hatred been appeased, only by kindness - the Buddha
Most ORGANIZATIONS consider real estate, power consumption and cooling when specifying an installation
Fortunately, they don't have to settle for inefficient 8 or 16 way machines, they can have 8 or 16 blades in about the same space.
Of course, I'm not saying that virtualization is useless. Quite to the contrary: it does have some important specialized uses. But Linux supports a wider spectrum of virtualization and partitioning options than Solaris.
The problem is that Sun is overselling virtualization and selling virtualization to people who don't need it, and they are overhyping what they have. When a Sun salesperson tries to sell you a big Solaris server plus virtualization, hold on to your wallet.
There is plenty of information at BigAdmin, including technical whitepapers and tutorials. I can't imagine you looked very hard if you couldn't find this information.
I'll give you the 60 second summary. Containers are zones plus resource managers. Zones are very similar to BSD jails. A single kernel is shared by all zones. There are potentially 1000s of zones per server. Each zone has its own copy of Solaris userspace including applications. All zones sit inside the "global zone"; the Solaris running on the hardware. Upgrading the global zone (eg, with patches) will automatically upgrade all the other zones. Resource managers can limit the resources used by a zone; think CPU and memory quotas for zones. Zones can use multiple CPUs, or part of a single CPU, or whatever.
UML isn't nearly as good; UML runs a whole new kernel per instance. Completely unlike VMware; VMware runs a whole new virtual PC per instance! Jails are the closest equivalent but still not exactly.
"Tell them that we're returning to our roots," Schwartz said, referring to the company's renewed focus on the Solaris operating environment.
That's not their roots.
Solaris is not SunOS.
If you're smart, you date 10's, and you marry 7-8's with great personalities and intellect. ..
in other words your right palm is a 10 and your left is an 8.\
This, of course, is assuming that you yourself are able to carry on an intelligent conversation, but given the audience here, the vast majority of us are.
You're new here, aren't you?
"Danke daß Du mich gemolken hast" said the German cow.
So they've open sourced it...but how "open"?
Open enough that I can take the source and create my own version of Solaris? Are we going to see "White Box Solaris", "Tao Solaris", "Debian Solaris", etc, soon?
It's pretty simple. Most ISVs (Oracle, for example), will only certify their software under RHEL. Most companies will only run software under platforms that are fully certified and supported by the vendors. This makes them locked in to RHEL, and to paying the licensing costs for that system. From that perspective, Redhat and Solaris cost the same, and are direct competitors.
If you saw the video of Jon Schwartz at the Solaris 10 launch, he directly addressed this. He mentions Debian, Gentoo, and Yellowdog by name, and acknowleged that they were not what Sun is targeting.
The writer of this report has no clue what the straw man fallacy is. He means that Sun committed the fallacy of a biased sample. Straw Man is when you distort someone's position and attack the distorted view. Biased sample is when you pick X out of population P that best supports your cause and draw conclusions on P.
24x7 guaranteed on Sun?
I'd have thought OpenVMS or Tandem systems would be what you go for if you _needed_ 24x7 _guaranteed_.
Sun hardware/software isn't that much more reliable/available as decent PC hardware/software, (and depends on model - it's like comparing models of Seagate vs Maxtor vs et all - you have the usual duds and lemons).
An OpenVMS site migrated their cluster to a different _physical_ location miles away and during the migration (when the parts of the cluster were on different sites) the users complained that things were a bit slower. That's about it.. No downtime.
Try that on Sun or Linux. You'd need scheduled downtime. Same for IBM systems.
An E10K isn't going to save you if a CPU goes poof - you lose the processes running on that CPU. Whereas if you are using Tandem Nonstop, no problem.
Ironically Sun SPARCs didn't even have hardware instruction retry, whereas _Fujitsu_ SPARCS had.
The sad thing is given the environment - these systems are probably more reliable than the companies selling and "supporting" them through the decades. I suspect at least a few have clusters/systems that have been up since the days when Compaq bought Tandem+Digital and now Compaq "isn't around" but these clusters are still chugging along...
Sure doesn't look like HP is making the best of these systems either.
Solving the real HA software issues on Solaris will be very similar to doing the same on Linux. Unless Solaris is going to come up with a few tricks using their Zone stuff, or some other new Solaris 10 feature.
With the mid-high end Suns, if you are lucky you can probably detect soft-errors - e.g. oops RAM/CPU is flaky (ECC failing), then move stuff off the relevant board(s), switch em off and take em out.
In the x86 environment isn't really that different - detect ECC problems etc, move stuff off the relevant server(s) and take it offline. But how much are those boards compared to a single equivalent performance Dell Server or x86 blade server?
Sure you don't get the single huge memory space you get with the E10K etc. But ( correct me if I'm wrong ) that single memory space system isn't HA - it's more for performance. You'd have to cluster for HA anyway using similar cluster architectures as you would for x86 servers.
The last I checked x86 address space and performance is getting rather better, as is I doubt it's a problem for 95% of the companies out there.
That's the problem Sun has. Their stuff is not really that high-end. And x86 servers are killing em.
Will be interesting to see what they do with their next range of opteron servers.
"...but it's the same price as Fedora Core and White Box Linux, which use a more cutting-edge codebase than RHEL3..."
Uh, White Box *IS* RHEL3, without redhat's artwork, etc, recompiled from the freely-available SRPMS that RedHat releases.
No sig for you.
I work at a company that ships product for Solaris/HP/AIX/Linux/Windows. I have my choice of developing code on whatever platform I want. Before checkin I build on all platforms. Solaris is by far the slowest compiler. Even linux on a single CPU machine can compile faster than solaris on a Quad! Make the compiler faster. I also can debug on any platform, compared to the windows debugger ever single UNIX debugger is far inferior. I have used debugger on all the platforms. The Solaris debugger is the most robust but even still it is slower than Windows and has far fewer valuable features. Make the debugger faster. I think those are very important issues to address if they want to get developers to write code for their platform.
I paraphrase:
"He says Linux doesn't have X,Y and Z."
"But it does... this distro has X."
"This distro has Y."
"This distro has Z."
I bet he voted for Kerry too.....
It would violate the GPL for Red Hat to require a binary license fee of end users. Red Hat charges for its products of course, but has neither the right nor the motivation to stop people from using it for free. The Red Hat license is the GPL, but the services connected with that software -- Red Hat Network -- are only available to paying customers.
Hah. Where are the RHEL isos ?
That the money was paid to SCO by Sun BEFORE the lawsuits started.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Save if for the Oprah crowd buddy.
Yep, when I read that, the credibility of the article fell by 5 points...
But Linux does have a feature comparable to Zones (is that the same thing as grid containers?), it's called Linux VServer, it'd be nice if it became part of the stock kernel some day.
Another thing that Linux desparately needs IMO is something similar to FreeBSD union mount or the translucent filesystem. There are a few projects out there, but they seem abandoned, and this is something that should be part of ext3. LVM could use improvement also - AFAIK snapshots are broken in 2.6.9. Last, but not the least, umount needs an -f option which forces unmounting no matter what.
These are a few little things off the top of my head, I could probably come up with a ton more if I had the time to think about it....
very true, this was probably giving /. crowd too much credit. g0atsex and FP -- nuff' said.
A year spent in artificial intelligence is enough to make one believe in God.
This next bit was extra good :-
Sun does not have the operating system market on Wall Street -- they're not even close. So by McNealy's own reasoning, Sun is as insignificant as Novell is.
Sun's attacks on Red Hat amount to straw man tactics
uh-huh
(As an aside, what about Eclipse and swt, open source from *IBM*...)
-- andrew international ? consonants : http://grkvlt.blogspot.com/
In any case if you think 60% is good, wait until it is a penny stock, then you will be able to watch it flail around for +/-80% daily swings like United AirLines and other DeadFish(tm).
There are REASONS people use massively redundant arrays of rack PCs. Do you think we have these setups just because the first PC is cheap?
"click click" is not what makes a debugger good.
Can your Windows debugger even load core dumps, so you can see what went wrong with the application in production?
As for compilation speed - that is absolutely, totally, and completely irrelevant (unless you are a Gentoo user). What matters is how fast and correct the output is.
> Make the debugger faster.
The only thing that needs to be faster is watched expressions. But then again, they're not fast on ANY platform if you're watching at the global scope.
My debugger runs as fast as I can type. Any faster would be counterproductive, because it would have to make up what to do.
Do daemons dream of electric sleep()?
Just like DEC back in the day: the brain is dead, but the body lives on.
Dumping MIPS and hitching a ride on the Itanium was the dumbest thing SGI did. SGI is basically dead. HP did that with its PA-RISC, and the effects of that will be obvious a few years down the road.
What this poster doesn't understand is that you can survive in your niche, if you're good enough. By moving to commodity hardware and software, well, what is your competitive advantage? Nothing. How can SGI survive on service when there's nothing to service? How can SGI keep selling machines when they're exactly the same as other machines?
Answer: they can't. That's why their stock is in the toilet. They have no future, and probably will be bought out for their customer base and maintenance stream.
The SGI core vision is technical computing. In this area they are continuing to innovate, and quite nicely I might add, and make money. The really big players consider SGI for new projects.
Does this make them the money they used to make in their workstation heyday? No. They still make very powerful workstations, but only a small niche needs them and continues to buy them.
They are not dead however, just gone back to the world of big contracts, near impossible problems and non-disclosure. That's where they started. Most of us don't see that transpire for obvious reasons, but that means little about how SGI is doing.
Blogging because I can...
Sgi is doing exactly what you said. They do not make commidity hardware. Their stuff is so far from that...
Their Itanium machines are full, low latency, modular NUMA machines. Thousands of CPU's running on ONE OS IMAGE. Yep, run top and cpu(s) > 1000! When you purchase one of these, you get a small army of SGI folks on site to literally build your supercomputer, just the way you want it, tuned just for your problems.
You cannot get that any other way and that is the brilliance of Bishop's business vision. What he did was make SGI able to survive on that, because there will always be a market for that.
The modular part is cool too. If you want, you can break your computer into several computers. This is very important as the supercomputers age. They will be replaced with other supercomputers. However if you happened to buy an SGI supercomputer, the massive machine can be broken down into little computers to be used in lots of places, or sold to others as smaller units, or repurposed for other problems. Nobody offers that in the way SGI does.
I miss them in the lower tier markets though. SGI computers are a lot of fun to work with. This is why most people say SGI is dead, and for most of them, sadly, they are.
Anyway, your point is totally valid, but it just does not apply to SGI, that's all.
Blogging because I can...
Some of the Sun folks I know, have told me about some pretty cool stuff. (It's on Sparc, so I worry a little.) And I have the highest respect for Sun people. They love what they do and support their stuff like no other.
I suppose my post was a bit too much in the gloom and doom department.
Somewhere inside Sun there probably is some niche they can just nail to the wall. Probably a couple. Their success will depend on if and when they can come to consensus on whatever it is.
Blogging because I can...
And it's a damn cool envoronment to work in. I like multi-cpu systems feeding groups of users. All the admin complexity required for higher-end cad can be centralized in one place, right along with the data. For version control on large models, this works very well.
Users can run whatever they want on the desktop so long as they have a decent X server running.
These kinds of solutions are still competetive, per user, when second hand hardware is used, but support from the MCAD vendors, for this mode of computing is beginning to seriously wane. Just try calling support for an issue on a box like this --they have no clue.
I'm just rambling, I suppose.
If these guys would get off their arse and produce a Linux port, they might just find the market will still consider many advantages of multi-user computing, particularly when data management issues are factored into the equation.
Each one of them says the same thing however. Too many Linuxes, how do we support them all.
What they don't get is that they don't have to. They could pick one Linux, such as SuSE, for example. All they need to do is make sure their stuff runs solid on that particular Linux, then let their customers do what they want. Open OSes mean lots of choices. The fact that they are UNIXey means, multiple Linuxes can all work together to provide solutions to the end user, without them even knowing.
Maybe some day, once they have all been bitten by the whole win32 deal enough times, we might see this begin to happen...
Blogging because I can...
Show me half a line of code that HP's contributed to the community.
h elds.org/
... yup HP, along with a few other of the top SAMBA guys. You think maybe they've contributed code?
A half a line of code would be pointless (unless you're SCO).
So how about some full projects?
http://opensource.hp.com/ is a start for HP and Compaq contribtions. Such as Vesta (GPL), DSpace (BSD), Open Source DB Benchmark, OpenSSI (clustering) (GPL). http://hpinkjet.sourceforge.net/
http://www.hand
My group has submitted patches to Postfix, with some Mailman patches coming soon (contributed by yours truly). In a previous HP incarnation we submitted kernel patches in the SCSI subsystem.
We've got people working on things like distcc, Mondo (came -again- from HP), elilo, various math libraries, and more. Such as the SiS suite (go to sourceforge), perfmon on IA64 Linux, LDAPWeb and others.
The maintainer of the IA64 Linux 2.6 kernel maintainer is an HP guy, as is the 2.4 of the same. I dare say they've added at least half a line of code each. Jeremy Allison of SAMBA fame works at
How about IBM?
SponsoringOf/sourceOf/patchesFor STAF, Eclipse, OpenHPi, Kprobes, udev, JFS, EVMS, Many kernel patches, HotPlug, SAMBA, CIFS Channel bonding for ethernet, ATM, drivers, SiS, and many, many others.
The list goes on, and I could karma-whore for a really long comment showing you dozens more. But I won't hold my breath that you'd correct your statements.
Try going to http://www.hp.com/linux http://www.ibm.com/linux and
http://www.sun.com/linux
Compare the intent of those pages.
Now, please, someone build a similar list for Sun. Of course starting with OpenOffice.org. http://www.sun.com/linux is not as useful as one would think when looking at the others. Let's see if we can compile what Sun has done for the opensource community. HAS, not SAYS it has.
My Suburban burns less gasoline than your Prius.
Nuff said.
As quoted from Andrew Morton:
Update: Linux keeper, SpikeSource CEO talk up open source
quote: "Top contributors to the Linux kernel have been Red Hat Software and SuSE, he said. Also contributing have been IBM, SGI, Hewlett-Packard.and Intel" For a company that has done so much to promote and --- BUILD ---- Linux, Sun insulting them is no different then them going after Debian or Gentoo, yeah so what they charge a premium for support, have you considered they must offer something of value to their customers for them to be willing to pay a premium, such as their expertise in Linux kernel development... We are not communists at the end of the day, we want all the Linux players to compete and succeed... Even the article quotes Mortan's view on that: "Leading-edge projects are the exception in the open source world," he said. If anyone is developing leading-edge technology, "they should get their act together and form a company and take a shot at getting rich with it," said Morton. It used to be funny when Sun's McNeally insulted Microsoft, since they are much smaller than Microsoft, so MS just probably laughed it off, but when they pick on RedHat which is an order of magnitude smaller company, they just appear like an arrogant bully...
You could have at least given a link to those comments. I missed the ad for the novel, but found these reviews on amazon.
My Journal.
(t)csh is garbage and should die a painful death.
ksh in emacs mode rocks for most tasks, with the exception that vi mode rocks when I want to recall a single loop as a single command. Other Bourne shell inspired shells are also cool (zsh, bash). Once again (t)csh sucks.
However, vim for text editting is required by me since emacs is physically painful to use; it's cntl-whatever editing commands aggrivate my Guyon's Canal Syndrome.
Thanks. The Sol. 10 marketing stuff should perhaps link to bigadmin more. All I get is 'containers! woo!'. And neither my read HW nor VMWare seem capable of installing S10 (which, despite being downloaded two days ago, is weirdly marked as a beta), so I haven't been able to try for myself.
There's different advantages from different approaches. Containers seem really nice from an admin point of view, but more abstraction means more layers of protection. A kernel compromise in a UML mode session is much less likely to lead to an intrusion on the host.
I don't whether UML is multithreaded - I should find out...
The problem with Sun is not products. They have some real jewels: to mention Solaris (cute tricky bitch - I love you!), Java, Opterons. The problem is with their closed mind flip flop management. Sometimes there PR seems to be follow the mess principle from SCO.
P.S: It's time to free Sol and Java under the GPL - when will you get this at Sun?
I think you'll find most of us don't have anything against Solaris as such [...]
:-)
:-)
I think you'll find that many of the old punters that read this thread hate both Sun and Solaris2 with a vengeance
Solaris2, for those tuning in late, is the Windows NT 3.51 of *NIX: unfinished at release time, overly hyped, and released in that broken state for political reasons only. Oh, and for years after the initial release, the picture didn't improve appreciably (which OS, four releases later, still couldn't manage to give more or less accurate output for who(1) without patching? Yup, that'd be Solaris 2.4)
Sun was hellbent on destroying SunOS 4. Which is fine by me; by the time it became apparent that simple things like the DNS resolver were not going to be fixed in SunOS 4, NetBSD was ready to fill the void and I never looked back.
I'm exceedingly happy to note that by trying to kill off BSD, Sun succeeded in alienating enough people to sustain BSD development long after the Regents gave up on it. And then, when Linux became a marketing success, Sun forgot to include BSD in its list of initiatives to marginalize.
I for one would not lose sleep if McNealy, Gates and Stallman managed to keep each other off the street by trying to slash their respective counterparts throats whenever the opponent isn't looking. Heck, I'm not losing any sleep watching them do just that as we speak
Bert Driehuis -- All I asked was a friggin' rotatin' chair. Throw me a bone here, people.
Sorry I didn't, as Mr Sensitive has removed the banner for his book as well as blocked access to the reviews on Amazon (which is where the banner linked to in any case).
Suffice it to say that the reviews were more critical than critical acclaim.
I didn't even know authors could do that. That's a bigger indictment of a novel than just about anything else I could imagine. Not even Anne Rice turned off comments on her book after that big ol' flamewar a while back.
Linux from Day 1 has been out to get Sun. My proof? Why label the Solaris partition as swap under linux?