astro cloud, brainstorming11/22/8
Here is a laundry list of possible areas that could be addressed:
- traditional middleware
- cloud middleware
- security software
- development innovations
specific product categories, existing/known:
- Identity management
- Spring development
- presentation, portals, and CMS
- scalability for cloud deployments
I guess i see a couple of issues with making some analysis of what opportunities are available, and they really come down to these:
- Spring v. JEE6
- Google sponsored
I don't think a decision necessarily needs to be made with respect to Spring and JEE, but eventually it becomes cumbersome to support both, and one or the other makes sense to do as primary. OSS is flat-out the only sustainable model going forward. There is an absolute certainty that Google will enter the enterprise software market within 2009; perhaps not in providing its own app server, but it will continue to sponsor a wide-range of development initiatives, such as Guice (http://code.google.com/p/google-guice/)...the corollary development is that all deployments, whether Google-sponsored or somehow independent, will be moving toward massive cloud implementations...
truly, the only sure bet in 2009 is that all existing enterprise software projects, from Mule to Gigaspaces to Hibernate to MySQL to Intalio to SAP will have to figure out the value proposition of their offering vis-a-vis cloud deployment: the economies of scale are so massively in favor of the cloud, that even pre-existing successful bets on hosting, such as Salesforce.com, will still need to be further re-defined to take in to account the exact nature of the cloud...only those software vendors that have code that fits in the cloud, and business development that understands the scope of partnerships needed to support customer deployments in the cloud will survive the incredible pricing pressure certain to come about through a recession...
any and all existing proprietary hooks, such as Oracle Fusion app server (non-WebLogic), non-Java related development tools, and legacy applications will have absolutely no value and no marketability if they are not able to rapidly re-configure to support the cloud value proposition...Google will have massive advantages by owning infrastructure to build code repositories that take from existing Google Cloud projects...there are literally dozens of potential projects that Google could productize, so that the only determination is not whether Google will have viable enterprise software, but will be which projects they fund, support, and extend to make the most of their massive investments...barring an unforeseen circumstance, all enterprise software vendors have no choice but to subscribe to Google's cloud terms, as the only legitimate stop-gap to a pure Microsoft implementation...
from .Net to Biz Talk to SQL to Live, Microsoft is behind in every category they compete in, by a wide margin; however, if competitors in point categories are not capable at integrating with the Google Cloud, they will have no chance at surviving the downturn, in the face of Microsoft's ongoing dominance and longevity and risk-free reputation...the only thing that is more important than JEE6 in 2009 is Google, and the only thing that can prevent Google from completely dominating every product category from ESBs to Portals to linear scalability boosters is a vendors' ability to out-manoeuvre Google within Google's own infrastructure...it can be done, though very few will succeed...
to me, the only ongoing viable strategy is to build applications for the cloud, as Google appears to be following the playbook of Sun's successful early shepherding of Java as a competitive alternative to Microsoft, which was to not focus on functionality, but instead maniacally adhere to infrastructure-only...the one and only place that has sustainable margins, innovation options, and clarity of market opportunity is in building and supporting cloud applications: everything else will be swallowed by Google...
what are cloud applications?
cloud applications are anything that could be used by a business to facilitate the execution of the acquisition of a product or service that would require large-scale implementation, such as a supply chain order, a customer service record, or a financial transaction of any kind...the most basic premise of a cloud application is that it will involve multiple sources, and is best described and previously known as:
a web services enabled data integration source file.
in other words, the years of enabling SOAP-like web services integration is finally available in deployment on the cloud, for the XML-heavy language barrier to performance is mitigated by the cloud's infrastructure, and functionality...a cloud application will often times be constructed out of available data resources, though will also be built from scratch from available development environments and pre-built code repositories...cloud applications will be horizontal and vertical in scope, and will be easily taken down when the duration of its lifespan has been completed...
there is no functioning business model for how cloud applications will be bought and sold.
there is no existing business that currently addresses the market opportunity that will become apparent with the advent of cloud applications.
there are literally dozens of ways that cloud applications could become a market on its own, and no way to predict which way will be the most efficient; therefore, software vendors and application developers that seek to address the cloud applications market opportunity will either need to be extremely specialized in a particular niche, or very well organized in order to fully support the wide-spectrum of cloud application deployment scenarios that will be tested...
again, there is no way to know how cloud applications will ultimately be transacted, but by the end of 2009, there will be at least enough use cases, for vendors and cloud application developers, alike, to begin to make more reasonable bets on which direction will most likely prove to be the most lucrative...the only known requirement to date of how to address the cloud application market opportunity is to fully understand the broadest scope of the Google Cloud offering, all its features, its future plans, and its offerings for developers...those who can fully articulate how the Google Cloud works will be best positioned to explain how cloud applications will be best implemented...
in short, the most relevant site to the cloud applications initiative is:
though there are other OSS-supported, code repositories, such as java.net, Apache, Sourceforge, etc...there seems to be little rational for choosing over the Google Code site, as the clearing house for all available information on how best to integrate with the Google Cloud is found there...