Friday, November 9, 2007

Looks like the The Gang, rounded up by Steve Gillmor, is back in the saddle

Jason Calacanis is blogging about the latest debut of The Gang, aka Gillmor Group, aka Bad Sinatra, aka Gillmor Gang. The first episode is on Facebook, in four parts. I was happy to be a part of this, nearly a year since the last real Gang recording.

It actually came out quite good, just like the olden days. And a critical mass of the original gang is on the call: Steve Gillmor, Nick Carr, Mike Arrington, Doc Searls, Robert Anderson, Jason Calacanis, Mike Vizard, and yours truly. Expect more.

At least this first weekly and lively discourse on the really important things in life is not in 18 revolting segmentations, as was the norm in some past iterations. I can only surmise that Steve is out hustling up some underwriters for the podcast. Why else break it up at all?

Anyone care to cut a check on this? Six figures? Jonathan? I'm sure Steve's voice-overs on your introductions will be inspiring. ... ("He'd never be in blogging if it weren't for me!") Actually, I'd probably not be in blogging if not for Steve either. Thanks, pal.

True to his attention-deficit marketing mentality, there is virtually no promotion of the new The Gang. Links are dead after all. It's all about negative gestures, don't ya know. Ya, and I buried Paul, you expert textpert.

I'm very glad to see that Steve is producing this independently. No more Pod.*. And Facebook will make a fascinating viral platform. It's good to experiment. Just open enough. He might even be able to measure the audience; might even be able to define the audience members, might even be able to invite the audience individually. Ah, the good old days of controlled circulation ... much better rates that way. And the list -- My God, he could sell the list! Elitism has its advantages.

And I'm glad it's not video either, leave that to the infomercials. See Gate, et al. Voice is plenty. Just repurpose it on iTunes and monetize on the Facebook picket-fence garden. Screw the rest of 'em.

And so, how do you post "music" to Facebook? Is that an application, or a feature?

Wednesday, November 7, 2007

Red Hat partners with Amazon for Enterprise Linux fabric in Elastic Compute Cloud

Seeking to make ISVs an offer they can't refuse, Red Hat has teamed with Amazon to provide Red Hat Enterprise Linux (RHEL) 5.x instances in the Elastic Compute Cloud (EC2) infrastructure as a service offering.

The beta Amazon-hosted monthly payment subscription service will allow developers -- be they in ISVs or enterprises -- to develop to RHEL and then deploy the applications as a service. The applications owners will then only pay for the hosting of those applications and services based on their use and support infrastructure demand -- known as pay as you drink.

Red Hat also announced Wednesday the release of RHEL 5.1, which follows from RHEL 5.0's arrival last March. The newer point release provides an even deeper and aggressive reach into virtualization benefits, including significant performance boosts.

Additionally, Red Hat plans in mid-2008 to debut a series of software appliances, beginning with Red Hat Appliance OS and an associated development kit. This will allow entire packages of platform, middleware and applications to be crafted into an easily deployed functionally targeted and optimized appliance.

The goal is to allow developers and ISVs myriad choices on how to deliver their applications and services, as long as they deploy to the RHEL stack of open source infrastructure. Red Hat then charges for maintenance and support, or via on-demand subscriptions. By targeting just RHEL, developers can then deploy directly to dedicated servers, to virtualized instances of Linux, via optimized software appliances (apps plus OS plus required stacks), as well as via several choices for on-demand hosting (including Amazon).

Red Hat calls this plethora of deployment models and approaches as its new Linux Automation strategy. Red Hat is also interestingly getting chummy with Sun Microsystems on enterprise Java support, now that Java is open source.

What's more, through its burgeoning embrace of virtualization options, Red Hat plans to aggressively support Windows instances on Red Hat, so that any Windows 32-bit applications (from across many versions of Windows) can be supported virtually on RHEL. The leading Linux supplier also plans to work inside of Microsoft's pending Viridian hypervisor, which is based on the Linux-based Xen hypervisor (at least for now).

On a large business level, Red Hat is seeking to bring more open source deployment options to global developers of nearly every ilk. For operators, Red Hat hopes to simplify their infrastructures while offering strong performance and lower total costs through higher utilization and capacity management on top of lower licensing and hardware costs. Red Hat says that RHEL is gaining ground quickly on mainframes, including IBM's System Z.

My take on these announcements is that Red Hat wants to take its Linux distribution clout far beyond the market for dedicated servers and blades at individual enterprises, and become the de facto industry standard for how hosting organizations, telecommunications providers, entertainment providers, and on-demand ISVs deploy all their applications over the next 10 years.

Red Hat is banking on the trends around virtualization, clustering and utility computing, multi-core hardware/parallelism, and the increasingly advantageous on-demand subscription economic models to become the low-cost, high-performance enough foundational supplier.

With virtualization performance gaining ground, and allowing many types and kind of server applications to run on instances of many OSes, then more (and nearly) all things become equal such that a pure price-performance comparison can be made for providing applications and services. Whether you are building, hosting, providing, or monetizing around those services -- you will eventually move to the best price-performance deal.

The race is on then: Windows, Solaris/Open Solaris, Red Hat Linux -- on various types of platform iterations -- for the best story when it comes to hosting price-performance and deployment ease experience. Red Hat is aiming high on this one. And it has a pretty good shot at it doing very well.

Red Hat is boldly predicting it can, by 2015, double its market share and support more than half the worlds server instances. And they didn't even use the "grid" or "utility" words, not once.

Enterprise mobile remains up for grabs despite Google, Android and OHA

What do the major players missing from the Google Android and Open Handset Alliance (OHA) have in common? An abiding interest in enterprise mobile, and effectively integrating IT back-end resources for delivery to mobile B2E, B2B and B2C endpoints.

So-far, the non-OHA-committed players include Apple, Microsoft, Sybase, Adobe, Oracle, BEA Systems, Red Hat, Sun Microsystems, IBM, Verizon Wireless, AT&T, SAP, and HP. When viewed through this prominent global crowd, the OHA roster seems a bit flimsy for satisfying the major business activities/corporate opportunity.

Many questions arise on what will happen in an OHA world on the business side of the aisle. For example, should Google and the OHA members be satisfied with the B2C play? Will the current B2C OHA, if successful, play a coalescing role for B2B and B2E architectures? Could a separate business-oriented OHA or equivalent be in the offing?

For mobile especially, why have separate architectures for B2C and B2B? Why should enterprise mobile SaaS be constructed any differently from mobile Web SaaS? Fragmentation is the problem, not the solution.

If Android is to progress as an open environment into PC-like devices, then the architectural approach should be common. Respondents to my ZDNet blog poll say 58% to 42% that open mobile devices threaten closed PC models.

Well, of course, many things need to happen for an "open" business-ready mobile architecture: security, control, management, governance and mission critical reliability come to mind. The level for such risk-avoidance measures would be lower for a B2C and mobile commerce approach, at least at the outset.

So perhaps what is needed is a two-tiered approach to Android/OHA. One level, for Web-facing and consumer-type activities (supported increasingly by ad and mobile commerce revenues), will arrive Nov. 12. But how about a second level for a more enterprise-calibre stack, one with the concerns of CIOs addressed (supported more by a licensed, royalty, maintenance/support or subscription revenue model). Come on, guys, let's see the business version!

In fact, this could well mirror what has happened with free open source and so-called commercial open source. You take the same code base, the same adherence to openness, standards and interoperability -- yet take it to market on two levels. Android may very well need an more mature brother: Robot.

Android can be the mobile consumer-facing approach. Robot can be the workhorse for the business-class needs. This makes a great deal of sense and would allow for the common community of development while satisfying two quite different architectural integrity needs. It also allows for more traditional business monetization for Robot, one that would jibe with enterprise licensed and commercial open source selling. The Apache license could work for both, another essential commonality.

I think that Sybase, IBM, Oracle, Apple, Sun, Red Hat and the business revenue-hungry networks might go for it. Only Microsoft might be left trying to figure what in heaven's name to do about this whole thing. (What? Windows Mobile Open Live Software and Services .NET?)

Make no mistake. There is growing demand for enterprises to get mobile to their employees, partners, and customers. Sybase is betting the company on it. Based on a poll I presented in a recent ZDNet blog here, a whopping 78% of respondents think that IBM and Apple should work together to bring a Lotus Notes client to the iPhone.

I say three cheers for Notes on iPhone! It would be a great solution and drive Apple deeper into enterprises, while extending the shelf life of Notes/Domino. But why not build it via a Robot architecture, with Android at the core, and allow these mobile devices to be common endpoints for corporate and consumer activities, such as Notes/Domino, based on a common -- yet two-tired -- open/commercial middleware stack?

Windows Mobile didn't get the job done. Java ME did not get the job done. Embedded Linux did not get the job done. How about a Robot brother to Android?

Monday, November 5, 2007

Google's Android approach threatens no less than the personal computer itself

Google's announcement of mobile software platform Android pretty much disrupts and disintermediates a large swath of the edge of the Internet that connects via closed, non-PC devices to ... well, a fairly limited amount of content, apps, and data.

Google with Android and the Open Handset Alliance, however, may blow open a marketplace through a common open platform that can then provide a lot more content, apps, data, media, and services. And that will feed the demand by developers, users, and ultimately advertisers that open platforms be provided on mobile devices.

At the same time, the boundaries between laptop, PC, converged device, entertainment device are eroding and blurring. What will determine what the use will be for the content and apps, the services and the media? Not the location. Not the network. As the device user goes, so goes the options for its use. As long as there is broadband, a critical mass of apps and open services -- the device can be the size of an iPhone and do it all.

And that gives it serious advantages over a PC. The PC is locked down, and not nearly as versatile as a fully open, full-function mobile converged device. What's more, the services and content will begin to matter more, and drive the user behavior -- not the device itself, once it's made open (and maybe even free). The business model that favors the media over the closed platform will usually win. The business model that favors the platform over the limited and choked content will not.

Google through the Open Handset Alliance plans on Nov. 12 to unveil an operating system, middleware and mobile applications (and early look at the Android SDK). The goal is to foster ease and volume in binding together content providers and devices aka users. It's write once-run anywhere all over again. Not all the carriers are in, as TechCrunch points out, and most that are come from outside the U.S. where handset choice has been greater.

The crowd of members to the alliance is impressive. It will also be curious to see how Apple groks Android, and if its open API plans will marry or mesh with Open Handset Alliance plans. My guess is that Apple will need to adopt this, but may take its sweet time.

The energy and potential here with Android and community reminds me a lot of Java in the early days, and that's not a surprise given where Google honcho Eric Schmidt spent considerable time in the 1990s: inventing and promoting Java. Eric must love the very notion of "disrupts and disintermediates." Only this time its not to ward off just Microsoft, but to ward off the possibilities of future Microsofts.

No one provider, handset maker, or carrier is a kingmaker in the mobile market, not even Microsoft. The Win-Tel monopoly never made it to the handheld. The mobile market in many regions is unformed enough that Google and its partners can have a chance at keeping it open enough so that the loosely coupled content model may ultimately outshine the current dominant PC model as defined by Microsoft for some 20-plus years.

And the Java connection is more fitting than Eric's dual roles: Android and the Open Handset Alliance may very well presage -- if successful -- the disruption and disintermediation of the PC itself. It also explains why readers think that Java SE (and not ME) on converged devices makes sense.

So over the next couple of years, Android-supported mobile devices will spawn the applications ecology that creates all the hens to lay all the eggs that will best hatch into the chickens that come home to roost. The Google Trojan Horse Android could make it a lot easier for developers to thrive in the mobile space. An open iPhone or similar type of device can grow in its category to encroach on the PC. PCs will become notebook PCs that begin to act a lot more like an Android, or ... lose developer and media outlet (and ad dollars) allegiance.

If my vision is nearly correct (timing is always a tough one to call), more of the content designed for an Android and Open Handset Alliance-type device will also be used on a PC, the UI can be really all about Web services. And Microsoft will, as with the Web, Java, and SaaS, have to capitulate and adopt or support Android.

We're already seeing encroachment of what's known as the converged mobile device -- personified best so far by the Apple iPhone -- into the domain of the PC. If you were to hook up an iPhone to a monitor, mouse, and keyboard ... well, you have a PC. As long as it connects via wireless broadband, uses a browser to reach all the rest of the Google-navigated web content, and Google apps, and Apple's content (per per click) too.

Desktop PCs will be for large enterprises and the un-imaginative. A successful Android approach means that Windows Mobile will face daunting and probably insurmountable odds. It means the Windows PC will face new competition, and not just like Mac OS X -- the Windows franchise will face competition of a categorical nature, a game changer: The open mobile device ecology. And it's because Microsoft was not able to capture enough of the mobile market and lock it into Windows and its Visual Studio developers in time.

As Linux is at the core of Android, there's already an open source approach. That should be extended up and down the Android stack, and also account for a share of the applications. Google should make sure that money can be made by content producers, and that Google's ad revenues are shared, just as with AdSense on the web. Carriers will need to move to these Android devices and find a model based on content subscription and use. In effect, the mobile platform goes to the Internet model, and not just for limited browsers use.

The Google Android platform and the Apple iPhone have a lot in common. In effect, the two global innovators of Apple and Google are placing different bets on diverging paths to a similar end point. As such, they probably are complementary in the long run. And that spells trouble for Microsoft, the mobile carriers, and the closed handset makers.

It's also possible that Apple's best interests and Google's will diverge at some later point. How open will they go? If it threatens the PC, it could also mean Apple's platform model comes under pressure.

Meanwhile, however, an Android-supporting iPhone may be about the best mobile experience on the planet for a long time in the not too distant future. Hook it up to a dock and its the best PC experience too. Write once, run anywhere, do anything, anywhere -- that's the potential we're looking at. It's hard to see how a closed Microsoft Windows Everywhere approach -- while still hugely successful on the PC for now -- can lock in at the required level on the mobile device. It's easier to see open mobile devices usurping the PC.