Man Survives Steve Ballmer’s Flying Chair To Build ’21st Century Linux’

As I was reading that story I was imagining the pitch email that was sent by VMWare's PR team to this journalist.

Sometimes it is a little too artificial. You are right though, if you give most writers an interesting hook, they will use it and then recite the remainder of your pitch deck almost verbatim (as in this case)

If you want to give it another try, email me at jeremy@appfog.com and I'll set you up with a private beta account. We're committed to making it as easy as possible to get started.

I highly recommend this 2005 PG essay on PR: http://www.paulgraham.com/submarine.html Very enlightening.

That wasn't the entire team. The original team was only 3! It grew to larger than 6 before Cloud Foundry launched in April (I'm talking specifically about Engineers). If you add other people the team is far, far larger.

In some senses we have this already: with some clouds you just have to boot a node from a tiny PXE image or a USB key, and it'll discover the cloud, join in, and VMs will migrate onto it.

Unfortunately physics simply won't permit transparent scaling of workloads across nodes. For any application involving communication between parts (anything not "embarrassingly parallel") messaging between the parts gets slower and harder as the parts get further away and there are more of them communicating. 1000 nodes have a million possible connections between them, so you can't directly wire them. You need a switching fabric, so your messages have to deal with multiple hops and contention -- slower and harder.

so this is essentially an open source/VMware backed competitor to Heroku?

and Strobe for sproutcore is now facebooked i suppose

The article wasn't vague. It explicitly pointed out that there's no business model at all yet:

Maritz says VMware doesn’t have a concentrate plan for making money from Cloud Foundry. “A leap of faith,” he calls it.

I imagine the obvious business model though is to position themselves as the world experts on this platform and provide hosting and support to enterprise customers that would rather pay a premium for peace of mind as opposed to startups that would shop around for the cheapest option.

I wouldn't assume that VMware have to make money off this project directly. It's enough that they keep customers locked into their eye-wateringly expensive hypervisor (which Cloud Foundry is built on).

It's like asking why is Internet Explorer free ... because it traps businesses into requiring Windows on their desktops to run all those pesky IE6-only webapps.

We're dogfooding at AppFog. We self-host our web management console (see http://blog.appfog.com/creating-your-first-app-in-appfog-in-...).

When we needed a staging environment, it was just an `af push` away.

As the article says the cloud foundry project already has a life of its own outside of VMWare with both ActiveState's Stackato and AppFog already offering products and services built on Cloud Foundry with a growing community of contributors to the project, the best is yet to come!

added eric at ericschrijver.nl , curious!

added thepumpkin1979 at gmail. thanks!

Very wise. VMWare has established a precedent with ESXi V4 starting out free, then drastically changing terms, conditions, and pricing with V5. Their prerogative of course, but there is no way I would build a business around free beer software unless it was also free speech so I can fork it if I have to. They have open sourced this code, so you can take a chance on it, but you have to be certain there is enough of a critical mass of code commits from the outside that it could fly on its own were VMWare to close it back up.

He's a sales guy, and it's hard to overestimate just how much of Microsoft's revenue depends on sales relationships with big, big companies.

He was Bill Gates' closest friend, according to an interview with Gates.