Session Abstract: Customers, prospects, and partners use the Web to connect to enterprise applications at an increasing rate, underscoring the need for high developer productivity that achieves superior time-to-market relative to rivals. Cloud computing provides a powerful combination of value and cost drivers, and a growing number of Web-facing enterprise applications will find a home there, heralding a new era in enterprise Web development and execution.
Key Issues:
- What’s driving the convergence of Web and cloud AD, and what does the market offer today?
- How will future enterprise-developed solutions be architected, and how will the software development life cycle evolve?
- How should this be factored into near-term AD decisions on development environments, tools, methodologies and staff/skill planning?
Eric opens: By 2014, about a 1/3 of new web applications will be developed on platforms in the cloud. Platform that is horizontally scalable, designed for multi-tenancy.
Drivers of this enterprise shift: developer productivity, time to market, cost of ownership. Eric shares data point that 80% of cost for custom application development is salary. So, imagine if you can make developer’s more productive. What else can you get done?
What is converging in the clouds: runtimes. tools, techniques and patterns.
SEAP – SaaS-Enabled Application Platform, such as Google AppEngine. SEAP container has variety of capabilities, such as AppDev, XTP, Multi-Tenancy (MT), AppServer and Metadata. SEAP is part of an APaaS – application platform as a service.
Trends in enterprise adoption. SaaS is a gateway to APaaS. Steps:
- Start with SaaS
- Expand to APaaS with customizations (Force.com)
- Dabble in On-Premises SEAP
- Tiptoe into low-risk APaas Apps
- Trust APaaS with the Mission Critical … Not done in enterprise yet, but many startups are building their products on top of Google AppEngine. (Cites OrangeScape). Some VC’s are demanding a cloud start.
Eric has a good slide up now, comparing cloud application platforms. The platforms are: Google AppEngine (GAE), Force, Azure, Zoho, Heroku and Intuit. The features are browser-based IDE, Eclipse-based IDE, Generic Business Components, Proprietary Language (all but GAE & Heroku), Relational Database (all but GAE).
Next side: Enterprise Cloud AD by Project Type in 2013: 25% new SaaS, 12.5% new Services, 12.5% Cloud Bursting, 50% SaaS Extensions. Unfortunately, I’m not sure if the 25% of new SaaS is enterprise specific apps, built using APaaS, or SaaS from the cloud. Based on the talk, I’d guess the former, but he didn’t say this specifically. (Or, I didn’t hear him).
How do you factor this information into near-term decision process?
Capabilities & Skills Tradeoffs
- Proprietary vs. Open
- Libraries vs Components
- A Range of Programming Models: 3GL, 4GL, beyond?
An example, Force.com is high productivity, low flexibility, medium programming skills, low openness.
Barriers to Adoption, Pain Points identified by CIOs:
- On-premises integration <—CIOs say this is their biggest problem
- unique usage scenarios <—can’t benefit from jumpstart (available) capabilities
- security and SLAs <—Eric wants bonded insurance policies
- buyer confusion <—sheer number of vendors, pace of new offerings, inconsistent capabilities (apples & oranges)
- governance concerns <—plug into my existing SDLC, change management practices etc
Eric is advocate, but says can’t ignore the risks, the barriers that exist today. Even if all technology and risk barriers are resolved, there is still the cultural adjustment, giving up control, not having physical infrastructure, managed at your location.
Comments