Hewlett-Packard has formally demanded that Sun Microsystems and its president, Jonathan Schwartz, stop publishing what it calls “misleading and factually incorrect statements” about HP’s commitment to its version of Unix–but Sun is standing firm.
HP tells Sun to get its facts straight
2004-10-05 Unix 25 Comments
…but this is ridiculous. HP is so afraid to admit Itanic is a disaster that when they see Schwartz critizicing their long term strategy, they threaten to sue for libel. Sorry HP, but you knew that it would take a long, long time for your EPIC architecture to mature, and forcing a preemptive switch before the architecture is truly ready for prime time was really a stupid move on your part. Sun has made its share of them, and Schwartz has a few dozen horrible PR blunders attributed to him, but this is not one of them!
I am a fan of Sun and thier products but Schwartz
is a wind-up merchant one day he says Linux is cool, next
he says it’s flawed. He should get a haircut and get back in
his pram. Besides Sun has done so much backpedaling with
thier x86-64 machines and Solaris on Intel that Schwartz
is in no position to shout the odd’s
because of linux being supported by IBM and the likes, more and more “big iron features” will find their way into linux. As long as the UNIXes have a technical advantage over linux thei won’t fade, but keeping development on an operating system going if there is a cheap and equally capable system available for a fraction of the price is simply not going to happen. The companies will compete once again on hardware and support, not on operating system because all will be using and developing the same system.
i beg to differ your one-sided view, you left out other open sourced OS’s and you must understand that if linux is the only OS in existance then it will really suck. You need other OS’s to encourage innovation.
“Sun has done so much backpedaling with their x86-64 machines and Solaris on Intel”
Well, I can see Sun wanting to drop Solaris for Intel. But hey, at least Sun listened to their customers who complained. Customer desire is actually why did not drop Solaris for Intel. Apparently a lot more people were using it then anyone thought.
That’s more than I can say for IBM, who really blew it big time with OS/2. Nevermind the fact that a lot of customer wanted OS/2 to keep being developed and sold… Probably more so than Solaris on Intel.
After 3 straight years of losses they really had no choice!
I think you need to go and price up a Linux solution and a
Solaris solution (with support). You may find that there
is not such a gulf in your “cheap and equally capable system available for a fraction of the price” as what you seem to
think. As people here have, correctly, pointed out Sun has
been doing some strange things for the last 3-4 years (
including not making a profit (they have always had a positive
cash flow from operations)). Things have changed and are
changing for the better. The new x86 systems are competitive,
the x86 systems in development (Andy Bechtolsheim) look to
be awesome, Solaris 10 is a great OS (but you can run linux
on these boxes if you want to).
Maybe Schwartz will realize before it’s too late what he says won’t go away. It’ll be remembered for a long time and I think it’ll come to bite them in the ass instead of helping Sun.
Schwartz just doesn’t come off as a proffesional to me. It’s crap like this is why I like the low-tone companies who compete on merit instead of hype.
Grow up Schwartz, this isn’t the playground anymore.
if Schwartz and Gassee are friends?
Sun and Schwartz really have no room to talk, but he’s right. Apart from their printer business, HP is a directionless disaster, let alone HP-UX. Quite clearly, HP-UX and even AIX and Solaris are heading for very marginal markets if not for extinction. If it’s HP-UX today, why not Solaris tomorrow?
The really sad thing is though, this might even be a play by Schwartz to grab hold of HP-UX’s miniscule market and customers (and it almost crtainly is from the article). An attack on HP-UX doesn’t really reflect well on Jonathan Schwartz, the state of Sun now or what its markets will be in the future.
Part of the campaign touts the fact that Sun is bringing its Solaris version of Unix to servers based on the widely used x86 chips such as Xeon and Opteron.
Now that really is funny.
was just one of a handful sent through a private “back channel” that Sun and HP use to discreetly point out factual slip-ups
I’d like to eavesdrop on that one.
You may find that there is not such a gulf in your “cheap and equally capable system available for a fraction of the price” as what you seem to think.
Probably true, especially with their Opteron machines, but companies like Red Hat are using free software to provide cost effective software stacks expanding way beyond what they do currently, so you don’t just get an OS. While Sun mucks about with stuff that is meaningless to end users like DTrace, Novell/Suse and Red Hat have immediately useful enterprise software that will run, fully supported as part of a whole package, on their platforms now (Novell as a single entity very soon). Ultimately, Red Hat and not Novell, have the flexibility to make all that work though.
“Probably true, especially with their Opteron machines, but companies like Red Hat are using free software to provide cost effective software stacks expanding way beyond what they do currently, so you don’t just get an OS. While Sun mucks about with stuff that is meaningless to end users like DTrace, Novell/Suse and Red Hat have immediately useful enterprise software that will run, fully supported as part of a whole package, on their platforms now (Novell as a single entity very soon). Ultimately, Red Hat and not Novell, have the flexibility to make all that work though.”
Hmm. I agree somewhat with a lot of your other points but I
have to take issue with what you have written above. I don’t
see how you could rationally arrive at your conclusions.
How does using RedHats stack differ from using a Solaris
stack? (apart from it currently costs you more and gives
you lower quality support). Have you seen how integrated
Solaris’ stack is now? you don’t just get an OS with
Solaris either (application server comes bundled). Solaris
10 will expand on this, shipping with more integrated
software than ever before. Also, given your remarks on
Dtrace, you certainly have never used it so I don’t see
how you are qualified to dismiss it as ‘meaningless’.
In the corporate sector, Linux=RedHat and has little to
differentiate it from Sun/HP/IBM (certainly not price
and definitely not tco)
Even if Schwartz is right, I’m tired of the attitude of Sun. Around ’98 I thought Sun and Java were great. I even because a Sun certified Java Programmer in ’00. But over the last few years I’ve grown really tired of the behaviors I’ve seen regarding Linux, Open Office, Competition and even regarding Java. Sun had so many chances to make a good impression and take a strong leadership role. They need to clean their image up if they want me back. I haven’t even be motivated to download Java 1.5 yet…not that I doubt it is good, but I’m not sure I even want anything to do with that platform.
Linux was only mentioned once in the article and only in passing. There are plenty of Linux threads to discuss it on, this is neither the time nor the place.
> but companies like Red Hat are using free software to provide cost effective software stacks
Saying that RedHat is cost effective is not only old but also untrue. Solaris is 40% cheaper to license and support than RedHat, plus Solaris is a much more superior product compared to RedHat. Solaris’s software (middleware) stack stands head and shoulders above the competition (including IBM and Microsoft) in both features and cost — I would say Sun’s application server (7/8), directory server, and portal server are the best products on the market in their respective categories.
HP is attempting to transition their Enterprise Level customers from PA-RISC to an essentially unproven platform (Itanium2) with sales pitches of being able to run multiple operating systems on the same box (Linux, HP-UX, OpenVMS, and Windows) through nPars similar to the way IBM runs Linux on a Z Series mainframe using an LPAR. See the following pdf for more details:
If you are an Enterprise Level customer who has spent millions of dollars on high-end equipment for your business, the last thing you want is your vendor to “blow sunshine up your ass” saying this new platform will be the “greatest thing since sliced bread” with not a whole lot to back it up.
Large businesses see big iron such as IBM pSeries, Sun Sparc, and HP 9000 Series machines as investments, not “disposable assets” like PC’s. And when you spend between $100,000 and $5,000,000 on a single server, you expect the vendor to not want to dissuade you to change your entire architecture, which is exactly what HP expects its Enterprise Level customers to do. This involves way more than just hardware and can become very, very expensive.
I see Sun’s move as being very aggressive, and in this environment you have to be to survive. And obviously it is taking its toll on HP. Never mind Carly trash talking Sun at every opportunity (if you RTFA). And Sun figures (and rightly so), if you have to change your infrastructure, change to Sun.
So, does anyone know how HP-UX software programmers will be able to continue programming HP-UX now that PA-RISC will be EOLed soon and HP has just abolished their Itanium workstations? Looks like you need to buy an expensive Itanium server just to program for HP-UX pretty soon…
I don’t see how you could rationally arrive at your conclusions. How does using RedHats stack differ from using a Solaris stack?
Red Hat uses GPL’d software to spread the cost of development with other companies, the community and individuals. It’s not just customer cost – it’s Red Hat’s costs and the costs of development and support. Of course, a competitor could conceivably use the software Red Hat has put into the public domain against them, but the advantages outweigh the disadvantages and Red Hat take that risk without thinking.
Also, given your remarks on Dtrace, you certainly have never used it so I don’t see how you are qualified to dismiss it as ‘meaningless’.
DTrace is just not a marketable feature – yet. Companies want groupware, applications, to use directory services, look into thin client architecture…. It’s a marvellous bit of development, but it’s just not what Sun need right now. If Linux distributors and customers can live without such a feature then it isn’t required. Do administrators really want to know the status of the whole system in detail? Probably not. If it goes down that’s all many people want to know. It either works or it doesn’t, in much the same way as the air conditioning unit. Sounds ludicrous I know, but that’s the mass market low cost world Sun has to get into.
Have you seen how integrated Solaris’ stack is now? you don’t just get an OS with Solaris either (application server comes bundled).
Tell people about it. Look at how it all fits together and go and do it. Look at what people actually need in organisations at the coal face, not brilliant technical solutions with no immediate use for the majority.
I have a vague idea of what’s in Sun’s software stack (haven’t looked into it recently), but that never seems to take focus when you talk to Sun. Look at groupworking, mail, directory services, clients and bring it together with your Java Desktop system. Give your software stack catchy names, give them pretty logos and make sure they get equal billing with Solaris and everything else. Describe exactly how you would deploy Sun Ray in an organisational setting, and make sure people actually use it. Having a strategy with Linux would give incredible publicity, but that will never happen.
The fact that I have a vague idea of what Sun has, and a pretty good idea of what Microsoft, Novell and Red Hat have isn’t a good sign. The problem is Sun have been about being protectionist towards Solaris, which isn’t what actually matters.
In the corporate sector, Linux=RedHat and has little to differentiate it from Sun/HP/IBM (certainly not price and definitely not tco)
I’m sure that’s true in many ways, but the difference is that Sun are just not bringing it together.
Please don’t use the letters TCO. They mean nothing or very different things in different circumstances. Microsoft has screamed TCO until it is blue in the face, but no one cares because they don’t understand it. What they do understand is cheaper and cutting stuff out that they have found they don’t need and paying for it when they do. That’s going to be a bitter pill for many of Sun’s excellent technical engineers to swallow.
Several years ago I bought a Cobalt Qube for an organisation, and it was probably the best thing I’ve bought. It sat there, did it’s job and it was cheap. As an added bonus, it saved so much on electricity compared to a Pentium 4 server it was unreal (TCO?!). Pentium 4/Itanium servers, Linux installed or otherwise, I think are a joke personally. A Sun server (got that?) – cheap – in every sense, apart from the quality. In its wisdom, Sun then withdrew support for it forcing me to look into other alternatives I didn’t really want to. There are a lot of former Sun customers who have had such experiences, and they are going to be nigh on impossible to get back.
However, given that this article is about HP’s high-end customers (spending hundreds of thousands on high-end stuff and moving to Itanic?!) this is a bit OT. It’s just a bit of a concern that Sun are seeming to be going for high-end customers when that is not where the threat is.
DTrace is just not a marketable feature – yet. Companies want groupware, applications, to use directory services, look into thin client architecture…. It’s a marvellous bit of development, but it’s just not what Sun need right now. If Linux distributors and customers can live without such a feature then it isn’t required.
Every time I think I’ve seen the most outrageous possible act of arrogance from a Linux zealot, along comes another to prove to me that the arrogance actually knows no bounds. If something isn’t already in Linux, then “it isn’t required”? ! Are you kidding? I’ll ask the obvious question: so if Linux is the ne plus ultra of operating systems, why even continue Linux development? How can one possibly improve on the perfection that is Linux?
Suffice it to say, you’re wrong about the marketability of DTrace, you’re (very!) wrong about the degree to which customers want and need observability, and (more generally) you’re wrong about the willingness of Linux users to live without features found in other operating systems. But please, continue to spread your misconceptions: the arrogance of Linux zealots like yourself is so obviously unfounded, it only serves to help those of us offering an alternative…
If it’s good enough for Jonathan Schwartz to make misleading, factually incorrect statements and pass them off as facts on his blog, its good enough for us plebs here on OSNews.
If you want the FUD around Sun technologies to stop, how bout you start with the clowns at the top of your own organisation?
Well, I will try and address some of your points.
1) RedHat. OK, they are an open source company. What does
this mean exactly? From where I am sitting they are getting
fat off other peoples hard work. What exactly do RedHat
provide that differentiates them? (this is a genuine question,
I am not being facetious). Solaris IS being opensourced,
there is good community support for Solaris (blastwave).
2) Dtrace. Brian has responded to this already. I will just
reiterate that you clearly have not used, don’t understand
it or the benefits it can (does and has) provide. If you
don’t know anything about a technology it’s better not to
post as if you are an authority on it. You comments are
almost offensive in their ignorance and presumptuousness.
3) Software stack.
You start with:
“I have a vague idea of what’s in Sun’s software stack (haven’t looked into it recently),” Then proceed to comment
with authority on it.
TCO – to say that it doesn’t matter is just plain dumb. OK,
Solaris on x86 is cheaper (to buy and maintain) than a
supported RedHat solution. You can’t put down Sun for
being proprietory and expensive anymore because it’s not
true. And the article btw, was about HP trying to tell JS
to shut up.
Even though I’ve been a Sun critic (for various well deserved reasons), I hope they survive. It’s fun to have big-mouthed Schwartz and McNealy around to shoot their clueless mouths off with silly FUD.
Sun is in deep trouble. They keep losing money. They have some great technologies and products, but they are having a hard time competing with both MS, and cheap commodity OS (Linux) on cheap commodity hardware. They are trying to turn it around with some aggresive pricing and product strategies, but they have an uphill climb to get the market to go their way.
As for HP – they’re screwing up with HP-UX on PA-RISC migrating to Itanic. But HP-UX is not the big part of their overall strategy. Like IBM, HP are selling Linux on x86 (to the expanding small and medium size market, as well as large clusters), and getting big revenue on hardware and services. Both HP and IBM have recognized the market trend and have capitalized on it. Both HP and IBM are keeping their respective Unices around for their own biggest of big iron hardware, and to support existing customer base. But their big market strategy is Linux on x86 with bookoo $$$ in service revenue. Meanwhile, Sun is banking on Solaris and FUD.
IS this a lawsuit?
No, that’s simply HP telling Schwartz to grow up.
Executives that talk from their ass like Schwartz usually tend to go down in history as fools, i.e. DeLorean, McNealy. McBride, Schwartz and/or being forced to tone down.
You lives and you learns…
“…and cheap commodity OS (Linux) on cheap commodity hardware.”
Sorry Jeff but how is this true?
Solaris runs on the same commodity hardware.
Solaris is cheaper to license and support.
What’s on Sun’s mind ?
1) Fight with HP
2) Fight with Red Hat
3) Big friends with Microsoft, yet “supporting” Linux and rivalling them with Solaris and StarOffice
4) Looking for a fight with the Linux community (?) over OpenOffice (deal with MS) and Java
Last time a company took on so many people or companies at once, was when SCO sued IBM … Look what they got, heheh.