Slashdot Mirror
Sun Plans Solaris Subscription Model
heliocentric writes "As reported in this CNet article. In an effort to make its version of Unix compare more favorably to Red Hat's Linux, Sun Microsystems plans in coming weeks to begin selling its Solaris operating system through a subscription model." On the down side, there was coverage of the announced layoffs, as well as the MSFT case being won. The article makes a good point, that Sun has reinvented itself before, and that no one should write Sun off.
What the article's author doesn't get is that maybe Redhat is not so much succesfull because they have a subscription model, but more because their is direct interaction with their userbase (fedora) and the source is Open. A subsciption fee based model tends to be *really* good for the vendor (guaranteed, known cashflow that you can put on the books as revenue) and not so good for the end-user (expensive, bad for your cashflow). I'm sure that this time, the users have wisened up, and are using Linux.
also, the MSFT case wasn't *won* it was settled...
People who think they know everything are a great annoyance to those of us who do.
Hasn't Sun Microsystems licensed Unix code from SCO? Wouldn't a Solaris subscription funnel even more money to SCO (even though that would only be an unintentional side effect)?
cpghost at Cordula's Web.
Saw an academic's analysis of Sun lately along the lines of "Sun is spending massively on R&D. Why is Sun doing this instead of copying the successful strategies of HP and IBM? Sun should pull the money out of R&D and immediately better its bottom line while lowering the costs of its products."
... but Sun's record is a bit better than theirs in this.
And I think, didn't Sun get to be where it once was by catching the front of the wave of network computing (become Internet)? Isn't their core skillset being able to ride such a wave well? Isn't their future in getting out ahead of the next one so they can apply that skill again?
R&D is always risky, but as a long-term investor, shouldn't you be buying R&D? As corporations put less and less into it (as most are) what's left becomes potentially even more valuable when and if it pays off. Of course, you don't want to go into the equivalent of the old Xerox or AT&T which never properly capitalized on their best inventions
"with their freedom lost all virtue lose" - Milton
Boycotting MSFT is not easy/practical
I've been doing it for years. It actually is quite easy if you are willing to pay for good software or find a good free replacement for the MSFT (or Windows only) software you think you need. For example, there are lots of office suites out there and more individual spreadsheets and word processors that are office compatible.
In any case, if you are running a Sun OS, what is the likelihood that you have any MS products on it? Didn't IE for Solaris get discontinued years ago?
They will slowly lose more and more to the PPC based systems and better OSes (Linux/MacOS X). The Apple offerings won't make a large dent, the expected IBM systems will.
I would sell any shares in Sun you still have (I left long ago).
Sun has a fundamental problem, one it shares with Microsoft. Both firms live by selling a premium product in a commodity market. Operating systems are no longer rare and valuable enough to pay for. Linux has demonstrated the feasability of a single, free, standard OS for all hardware, and despite ferocious resistance from many quarters, makes inexorable progress towards becoming the eventual standard.
Sun will die if they do not reinvent themselves away from selling proprietary OS products. Moves like the Sun Java Desktop are a sign of hope: Sun must adopt Linux and FOSS and become a services company selling value-added solutions. Then it can survive.
But changing a company like this is incredibly hard and there are few cases where it works. Most likely, Sun is doomed.
Ceci n'est pas une signature
Sun made its money selling expensive hardware that was rock-solid and fast -- and if you wanted to run Solaris, you bought a Sun box with a SPARC processor. Now Solaris is out for the x86 and hardware reliability and speed can be had at a fraction of the cost in x86-based servers from Dell, Compaq, IBM, and others. There is no longer a compelling argument to buy Sun boxes for most applications. Sure, there are some exotic, massively multi-CPU servers from Sun, but that's not bread and butter sales and there are nott enough of them sold to support a company the size of Sun (hence the layoffs).
We've seen this before with IBM. To the Slashdot youngsters who may not remember: It was IBM who created the x86 PC back in the early 80's. They were trouncing "hobbyist" venders like MITS (Altair), Imsai, and Cromemco. They owned the PC market as far as business was concerned. Then the smaller companies like Compaq hit the scenes aiming at the business market with equivalent, or better, performance at lower prices. When is the last time that you saw a new IBM brand PC for sale?
Next to take a hit: Cisco. Companies are seeing the cost advantage to going with simpler, less-expensive routers and firewalls wherever possible. Why pay for "Cisco-certified" personnel when they can buy an SMC router, some switches, and have their office up and running at a fraction of the cost? Again, there are niche areas where Cisco is still king, but that won't support a Cisco-sized company.
I never believed the official party line with regard to Sun. I saw great ideas devoured by Java -- as even most Java-lovers realize, it is absolutely unsuitable for desktop applications, yet it was marketed for them, and it was used for them. Remember JXTA? That was Sun's peer-to-peer initiative. I saw JXTA come and go and hundreds of peer-to-peer developers with it. Peer-to-peer and Java -- truly a winning combination.
Nor did I believe that this was merely a coincidence or gross incompetence. When the internal Sun memo the Java problem was released, where Sun engineers complained that Java was too slow even for internal use on Sun's operating system, it was clear to everyone what had always been clear to me: Java was never intended to be a fast, powerful programming language. It was intended to be a way to sell big hardware, and to tie people to a single company: Sun. They sucked up a lot of mindshare. Neither proprietary nor truly free, Java existed in that same state of justifiable coercion (by means of copyright) as, say, MacOS X. But both Apple and Sun have as their goal the same thing as Microsoft -- to become the only vendor that matters, to create a monopoly. Microsoft is just better at it.
If you want a powerful, truly free, cross-platform interpreted language, try Python or Perl. Just because your PHB hasn't heard about them because there are no glossy brochures doesn't mean they can't kick Java's ass any day, even (or especially) in "mission-critical" application. Both are modern, object-oriented languages, idiosyncratic to be sure, but scalable no less. This very website is probably a larger application than most of the stuff that runs in your company. When did you last lose a comment on Slashdot? And Slashdot's code is ugly and hackish.
Now it turns out that Steve Ballmer and Scott McNealy are on the same football team. Their common enemy: Linux. Well, you know what? Linux can kick Sun's ass, and Linux can kick Microsoft's ass as well. And that's not because "Micro$oft sux0rZ!" It's because Linux has behind it not just governments and corporations, but the power of thousands of unimaginably creative volunteers. It's because Linux is free and will always remain so. Technology is not just about gadgets. It's also about freedom, and in the long run, freedom will prevail. Say about RMS what you want, but sometimes being a little overzealous can be a good thing.
Are corporations like Sun and Microsoft evil? Of course not. They are amoral (that also goes for IBM, by the way). They will do anything if it's good for their bottom line. If corporate murder was legal, every corporation would immediately start murdering people, other than by exposing them to toxic chemicals and unsafe workplace conditions. That's because if the current CEO doesn't like murdering people, they will simply be replaced by someone who does - shareholder value.
Linux is a little bit of everything. It incorporates elements of socialism (sharing your work, writing software in your free time), capitalism (being paid to program), of dictatorships (Linus coordinating the development process), of democracies (various associations with elected representatives), of meritocracy (those who develop, lead), of plutocracy (those who have money can get stuff done). This is in many ways a model for society. There is no single way to run a complex world. You need to combine the elements in a smart fashion.
Who cares about Sun? Who cares about Microsof
Here's an example of an actual sequence of conversations I've had:
"We really care about our relationship with your school and will go out of our way to solidify our reputation with you."
Yeah, but what's this about you changing the support structure for our yearly support contract?
"Oh, well, we found that most schools weren't using the contract to the fullest, ending up with blank software entitlements instead of using them all, and so we changed it. We now no longer cover things like backup software and directory services as part of your contract. Oh, and it costs more now too."
Sun lost out to Redhat because Sun (and their authorized reseller) could never get their act together, and their treatment of us bordered on abuse. When you'd pay $15K for a Sun and end up with yearly fees approaching $1K for the warranty period (since we'd have to cover OS upgrades) and then $3K for continuing maintenance, yet you could get a more powerful Dell server with better maintenance coverage and Redhat, in the long run, it was cheaper.
In other words, the TCO bug didn't just hit Microsoft - it hit Sun pretty hard, too.
I talk about stuff.
A subscription, OTOH, implies a recurring payment for goods delivered on a predictable recurring basis. Such as Playboy, for example. My subscription obliges Playboy to deliver on-time or lose my business. This seems to work for magazines, but would be a horrible idea for Tom Clancy novels: either the quality would go way down, or I would get zero product for my money.
And this seems to be the deal software subscriptioners are getting: low quality or zero product (missed delivery dates).
They say the first thing to go is your penis. Well, it's either that or your brain. I forget which...