While I personally liked him, I couldn't figure out their strategy, leading me to believe that they didn't really know what they were doing.
You used to be able to buy low-end SPARC systems for $995 (v100, Blade 100) now the cheapest system is much more than that - the same pricing strategy that ruined SGI: as their market share declined they retreated to higher end customers, which led to more market share declines.
Companies that are getting eaten from below have very few options. Do you really think Sun could have made a profit selling equipment at PC margins? That requires an entirely different cost structure. If they had tried, you could have written a slightly different comment:
Sun used to make really high-quality equipment, but now they are making crappy PCs like everyone else - maybe they should have focused on the customers willing to pay for quality.
Schwartz's task was essentially impossible, and selling to IBM or Oracle was probably the best outcome he could have hoped for.
> Companies that are getting eaten from below have very few options.
Companies die if there isn't an entry point that makes sense for customers.
It doesn't have to be at the same price point as the folks eating them from below, but it does have to be some place where customers are willing to enter.
The mid/high-range sun systems are about as expensive as comparable PC based systems. However, the PC systems have a low cost option for development. Suns, not so much.
It doesn't have to cost $500, but it can't cost $10k either.
"...leaving the company in the hands of new owner Oracle (ORCL) and its very profit-minded leadership"
Maybe I'm reading too much into the above statement, but the point of a company is to make profits. There are these things called "charities" for cases when you don't want to do that.
That said, Sun's approach in recent years has certainly been a mess. They have failed to really make open source work for them as a strategy in the same way companies like IBM and Apple have.
Apple's open source strategy is well defined, but not, shall we say, integral to their company. They take what works for them, give a bit back here and there, and mostly do proprietary, closed software and "ecosystems". In some ways that's better than a mess (Nokia), but they're not really big open source boosters in the way IBM are.
Sun has given a lot to open source, but they don't seem to have gotten as much back as IBM. I think once again the issue is that there wasn't a particularly clear strategy.
Yep, it sucks being the company that owns Java and puts open source on a pedestal, but gets beat badly in creating the ecosystem for Java development... i.e. Eclipse. NetBeans was never even a contender.
The sad part is that it is far superior now. A friend of mine converted me 8 months ago and I am petrified that it will get lost in the shuffle now. The thought of going back to Eclipse is mortifying after having such a consistent experience with Netbeans.
I'll concur, as I've been getting quite comfy using Netbeans. If Netbeans gets hung out to dry, it does mean going back to Eclipse (I'm not really into the idea of buying an IDEA license).
Actually, the new IDEA 9 now has a free, open-source version, with most of what you'll want. There's still a commercial version with some more features enabled.
http://www.jetbrains.com/idea/features/editions_comparison_m...
NetBeans' profiler is great though. IDEA doesn't have any - I use YourKit, a 3rd-party commercial profiler.
That those are open source is coincidental to Apple. It was just a faster means to an end for Apple, not part of a defined open source strategy. WebKit, for all practical purposes, killed the OSS project on which it was based (KHTML).
Well, it works for them, certainly. I think the important thing for companies is to have a strategy that's reasonably well-defined, and to be pragmatic about it. "No open source" is more or less an idiotic strategy (at the very least you should be open to taking advantage of it where it makes sense), and "all open source" makes it difficult to turn a profit (but not impossible...).
Well, superior for Apple. But Apple is in a different business. Apple's biggest competitor is Microsoft, another closed-source company. So, for them to open up a little bit gave them an edge over Microsoft. They used what open source code they needed (BSD), and added what they could (WebKit). Apple only uses open source when there is an advantage. For example, they are using LLVM and OpenCL (GCD) to battle DirectX. The only way for others to support it is to make it open source. They also gained significant geek credit, which has been very good for them.
But Sun, on the other hand, had a whole other set of problems. Apple is a consumer facing company. Sun is a business facing company. Sun's biggest competitor was Linux on any-brand PC server. That's a big challenge. So, they fought open source the only way they knew how... join them. If your biggest competitor is free running on commodity boxes, you have to go to that space. Unfortunately, there wasn't enough of a difference between Solaris and Linux to matter to most people, so they went with Linux. There is some great open source stuff in Solaris (ZFS, DTrace), but ultimately, there just wasn't enough 'there' there.
And then once you've lost the OS war, all that was left was hardware. Which, while high-quality, just wasn't enough to keep _all_ of Sun afloat. Perhaps though, it would have been enough to support the hardware divisions of Sun.
I probably should have said Apple's biggest competitor is the Microsoft ecosystem. True, hardware wise it's Apple versus HP, Dell, Sony, etc... but in terms of mindshare, it's Apple versus Microsoft.
Similarly with Sun, it wasn't Sun verus the nebulous Linux. It was Sun versus the Linux ecosystem. You could argue too that one problem with Sun was that they sold hardware and software, and that you could get them separately.
Just for a point of clarity, I was merely making the argument that both Apple and IBM have figured out how to leverage open source in their respective business models in a coherent way. I'm not necessarily claiming that either of them are religious, RMS-style open source boosters. They each have figured out ways they can use open source to benefit their bottom line.
IBM is basically a services company for the most part. The more they can do to reduce the amount of money a customer is paying in software licensing (sometimes to other companies), the more they can charge for their own services on a project. It is to their benefit for there to be as much free (as in beer) software out there as possible so a higher percentage of a project's overall budget goes to IBM.
Apple is primarily a hardware company. It is in their interest to make use of whatever community software is out there so that they don't have to re-invent the wheel for everything. They look to open source software as a way to augment what they can do internally without having to acquire tons of software companies just to get basic software technology in-house. This frees up cash for hardware R&D and aquisitions (PA Semi, Fingerworks, etc...).
Sun has suffered from an identity crisis in recent years. Were they a software company? A hardware company? Both? They don't appear to have been able to answer that question themselves. It's hard to have an open source strategy (or any other kind of strategy for that matter) when you don't even know what business you're in.
24 comments
[ 2.8 ms ] story [ 63.1 ms ] threadYou used to be able to buy low-end SPARC systems for $995 (v100, Blade 100) now the cheapest system is much more than that - the same pricing strategy that ruined SGI: as their market share declined they retreated to higher end customers, which led to more market share declines.
Sun used to make really high-quality equipment, but now they are making crappy PCs like everyone else - maybe they should have focused on the customers willing to pay for quality.
Schwartz's task was essentially impossible, and selling to IBM or Oracle was probably the best outcome he could have hoped for.
Companies die if there isn't an entry point that makes sense for customers.
It doesn't have to be at the same price point as the folks eating them from below, but it does have to be some place where customers are willing to enter.
The mid/high-range sun systems are about as expensive as comparable PC based systems. However, the PC systems have a low cost option for development. Suns, not so much.
It doesn't have to cost $500, but it can't cost $10k either.
Which lead to decreased network effects that, in turn, hurt the viability of their whole stack (SPARC+Solaris+apps)
You have to sell something entry-level unless you want to be servicing legacy customers forever.
I believe IBM is making with POWER/AIX exactly the same mistake Sun made. Fortunately for them, their legacy customers are less loyal than Sun's were.
In order to sell servers a year in the future, you must captivate developers right now.
Maybe I'm reading too much into the above statement, but the point of a company is to make profits. There are these things called "charities" for cases when you don't want to do that.
That said, Sun's approach in recent years has certainly been a mess. They have failed to really make open source work for them as a strategy in the same way companies like IBM and Apple have.
Sun has given a lot to open source, but they don't seem to have gotten as much back as IBM. I think once again the issue is that there wasn't a particularly clear strategy.
The sad part is that it is far superior now. A friend of mine converted me 8 months ago and I am petrified that it will get lost in the shuffle now. The thought of going back to Eclipse is mortifying after having such a consistent experience with Netbeans.
I don't know why but I feel that NetBeans is more organized and less "rugged" compare to Eclipse.
And it looks like that turned out to be a superior strategy, in a profit-minded sort of way.
But Sun, on the other hand, had a whole other set of problems. Apple is a consumer facing company. Sun is a business facing company. Sun's biggest competitor was Linux on any-brand PC server. That's a big challenge. So, they fought open source the only way they knew how... join them. If your biggest competitor is free running on commodity boxes, you have to go to that space. Unfortunately, there wasn't enough of a difference between Solaris and Linux to matter to most people, so they went with Linux. There is some great open source stuff in Solaris (ZFS, DTrace), but ultimately, there just wasn't enough 'there' there.
And then once you've lost the OS war, all that was left was hardware. Which, while high-quality, just wasn't enough to keep _all_ of Sun afloat. Perhaps though, it would have been enough to support the hardware divisions of Sun.
Similarly with Sun, it wasn't Sun verus the nebulous Linux. It was Sun versus the Linux ecosystem. You could argue too that one problem with Sun was that they sold hardware and software, and that you could get them separately.
IBM is basically a services company for the most part. The more they can do to reduce the amount of money a customer is paying in software licensing (sometimes to other companies), the more they can charge for their own services on a project. It is to their benefit for there to be as much free (as in beer) software out there as possible so a higher percentage of a project's overall budget goes to IBM.
Apple is primarily a hardware company. It is in their interest to make use of whatever community software is out there so that they don't have to re-invent the wheel for everything. They look to open source software as a way to augment what they can do internally without having to acquire tons of software companies just to get basic software technology in-house. This frees up cash for hardware R&D and aquisitions (PA Semi, Fingerworks, etc...).
Sun has suffered from an identity crisis in recent years. Were they a software company? A hardware company? Both? They don't appear to have been able to answer that question themselves. It's hard to have an open source strategy (or any other kind of strategy for that matter) when you don't even know what business you're in.
I would like to point out that the leadership of Sun for the past good couple of years seemed completely averse to profit.
Only that would explain such a long series of blunders.