Picking up from Tim’s post, since he has comments disabled:

Solaris: I know that Sun is deeply emotionally invested in Solaris, and that they can point to technical superiority on several levels. Fantastic. Unfortunately that battle is over - no-one outside of Sun cares. Linux is the operating system of choice for servers in small, medium, and large enterprises. It will be a heart-wrenching decision, but Sun has to move on from Solaris, take its best pieces of technology, and bring them to Linux. They can even keep the Solaris moniker if they’re enamored with it, just have it be Solaris Linux. 

Picture a Sun Linux distribution geared for servers, with ZFS, DTrace, and a variety of other advances thrown in, running on reliable, power efficient Sun hardware. Wouldn’t that be nice?

Picture not having to invest so much in competing with Linux, with APT, with the vast river of advances pouring into the Linux train. Picture instead investing those resources into improving Linux, making high-end, enterprise Linux synonymous with Sun.

Put another way: what percentage of hardware buyers are looking at Solaris as an advantage versus a disadvantage in their hardware purchasing decisions?

My own experience: A few years ago I was trying to find a home for a very nice Sun server that I had essentially free access to. To my great surprise no-one wanted the free hardware: “it’s not worth it, we’ll just have to move everything over to Linux eventually anyway”. That was eye opening.

The Hardware: Sun’s strongest attribute, one it’s squandering away rapidly, is the reliability and performance of its hardware. If you talked to hardware buyers 5 few years ago you’d see a deep-seated loyalty to Sun, because the hardware just worked. Sun had nailed operational reliability, saving people from 3am “server down” phone calls. That wins you big points.

Again from personal experience I can tell you other hardware vendors have trouble meeting the same standards; heck, the failure rate from one well known vendor was so shockingly high it was mind blowing. 

Sun is putting too much focus on their other stories, ignoring this crucial advantage. Considering the cost of hardware purchase pales in comparison to the cost of operating that hardware over the following years, why not pouce on this, make it the focus of your story?

The Margins: Margins on servers have been eroding. That’s a fact of life, one that Sun cannot turn the page back on. The sweet spot is, more and more, distributed grids of commodity servers. Sun has to embrace this, realize it means their business will be a different size and a different shape, and move on.

If they have an interest in getting into the startup and small business world, Sun has to realize it can’t sell hardware in the same way anymore. Resellers aren’t going to be part of the equation. I agree with Tim on the sales organization.

Developer Tools: Netbeans may well be the best IDE on the planet. I wouldn’t know, because I haven’t used it. I’d ask my friends’ opinions, but I don’t know anybody that uses it. I’d try it, but I can’t figure out a good reason to pick it over Eclipse (which I use, but I certainly don’t love). They seem very similar, but Netbeans has a much smaller community.

Glassfish could be a fantastic piece of software. To be fair I know people who are interested in using it (if not already using it). But it’s competing with several other pretty decent options, and they’re all free.

Someone needs to explain to me the benefit of competing with free on so many fronts. I don’t buy the idea of winning developers over to your platform; it’s just not working out. 

Instead, find the most popular free tools and provide the best platform to run them on.

MySQL: This is a fantastic piece of technology to draw on, to leverage into selling your other wares. I’m picturing a Sun server (running Linux of course), geared specifically for high-performance MySQL. It may simply be a matter of marketing, but I know if Sun put its substantial technical expertise and reputation behind a MySQL server that was configured to give me great performance, I’d pay a few extra bucks for it. Make that a cluster configured for master-slave with reasonable BCP that I didn’t have to design myself and I’d pay even more for it.

MySQL itself is a very nice, growing technology, one of the bright spots for Sun. I’d invest in things like Drizzle and orienting it for cloud operation. 

Looking forward: I largely agree with Tim here, and hope Sun can pull it off.

Disclaimers: I don’t speak for my employer, my uncle, or your friend’s poodle. These are opinions held at a particular point in time, subject to change in the face of relevant and convincing arguments. Enlighten me with comments below.