Sun failed miserably in embracing Linux and cloud computing (as a service). They hung on to their old business that was getting commoditized rapidly, plus having huge dependencies on Wall St. customers didn't help.
I agree...but, now what about Schwartz time at the helm? The past couple of years, I think, have set a pretty interesting new course for Sun. The acquisition of MySQL is merely the most obvious example, but Sun has open sourced everything in their software stable. Literally all of Sun is Open Source now (hell, they even have an open CPU design or two). This has beaten down their revenues a bit (OK, maybe a lot) in the short term...but I wonder if it's the beginning of a dramatic turnaround. It feels kinda like they've put the nerds back in charge. I don't know if that's always a good thing. Would Apple have become what it is today if Woz had been at the helm? It's almost comically obvious that it wouldn't be...but still, Sun was founded by nerds for nerds, and was explosively successful because of that.
In other words, I have no idea what's going to happen to Sun. But, I like them better as a company under Schwartz than at any time in the past ten of fifteen years.
'Literally all of Sun is Open Source now...this has beaten down their revenues a bit (OK, maybe a lot) in the short term.'
Why do you think that? All the Solaris (most of Sun's revenue) Sun's former customers actually use is Solaris, not OpenSolaris. No Sun customers I know would go without a support plan if they use Sun's OSS work. Right now they already pay Red Hat for their support plans.
Because Schwartz as much as said so at a conference last year. Maybe I misinterpret the remark, but it was something along the lines of, "I've caught some flak for our Open Source strategy from Wall Street and some of our institutional investors, but I believe that long-term this is the only viable option for Sun." I took this to mean he believes, and that some investors believe, that in the short term, going all Open Source is a net negative for their revenue.
I am pretty sure you could rent time on a Sun cloud long before there was EC2 or AppEngine. They were doing $1/CPU/hour back in '04. Except they called it a "grid", not a cloud.
Yes, but then they basically never improved it. Hardware got dramatically cheaper, but Sun didn't drop prices AFAIK. People started wanting to run Web servers and databases in the cloud, but Sun did nothing. Now they are racing to catch up by productizing Caroline and Q-Layer.
It's true they are focussed on compute, but it's also true that EC2 and AppEngine are not. I'm not entirely sure how you'd, for example run an FEA or CFD job on AppEngine.
I've heard that MPI works (granted, at 1Gbps speed) on EC2 and it's only $0.20/core/hour. When EC2 is cheaper and strictly more flexible than Sun Grid, it's hard to argue why Sun Grid exists at all. Maybe that's why Sun shut it down.
I remember when Sun announced this. I don't know if it was ever open to walk-up credit-card customers, like EC2.
The price was absurd even then. But further, you had to purchase exact blocks of time in advance. And if you bought 12 hours, but then only needed 2, tough cookies.
