Monday, June 16, 2008

'Instant replay' helps software developers fast-forward to application problem areas

Fixing software bugs is often easier than finding them. Stepping up to the plate to address this problem is Replay Solutions, which today announced general availability of ReplayDIRECTOR for Java EE, a TiVo-like product that allows instant replays of applications and servers at any stage of the application lifecycle.

ReplayDIRECTOR, which was released in beta by the Redwood City, Calif. company in March, makes deep recordings of applications and servers -- notably non-deterministic inputs and events that affect the application. Engineers can then fast forward directly to the root cause of the problem.

The idea behind the technology is that it allows companies to drill down into source code quickly, eliminating unnecessary IT costs and time spent searching for issues that can't be replicated or easily detected. The software is designed to cut through the complexity that IT departments face with shorter release cycles, multi-tier applications, and dispersed development teams.

According to Replay Solutions, every line of code that an application executes while ReplayDIRECTOR is recording will be re-executed in precisely the same sequence during playback. No source code changes are required and recordings can be played anywhere, without requiring the original environment, inputs, databases, or other servers, all of which are virtualized during replay.

As virtualization becomes more common, these replay approaches may be necessary as instances of apps and runtimes may come and go based on automated demand response provisioning. These left-over breadcrumbs of what once happened in a virtualization container will be quite valuable to then prevent recurrences.

I'm sure innovative developers and testers will come up with other interesting uses, especially as apps and services become supported in more places, inside and outside of enterprises. Got compliance?

Designed to deploy in any environment and have a minimal effect on the environment, ReplayDIRECTOR allows applications to run at near full speed while recording and faster than full speed during re-execution. It also has minimal performance impact, and can run in a production environments as an "always on" solution.

ReplayDIRECTOR for Java EE is available now. You can find more information at the company's Web site.

Saturday, June 14, 2008

Kapow takes a jab at challenge of creating mashups from JavaScript and AJAX sites

Kapow Technologies, whose solutions helps companies assemble mashups by harvesting and managing data from across the Web, has enhanced its approach to overcome the obstacle many businesses encounter when targeting sources with dynamic JavaScript and AJAX.

The Palo Alto, Calif. company's Kapow Mashup Server 6.4, which it unveiled this week, features extended JavaScript handling, a response to the burgeoning number of AJAX-based Web sites. [Disclosure: Kapow Technologies is a sponsor of BriefingsDirect podcasts.]

The Web 2.0 Edition, one of four editions of the new Mashup Server, now includes support for Web Application Description Language (WADL), making it easier for applications and mashup-building tools to discover and consume REST services. The WADL support also helps developers leverage the Kapow Excel Connector, an Excel plug-in provided by StrikeIron.

The Portal Content Edition, which enables companies to refurbish existing portal assets, has several enhancements to the web clipping technology for development and deployment of JSR-168 standards based portlets. It now provides the ability to make on-the-fly changes to clipping portlets that enhance portal functionality, while adding a portlet deployment mechanism on major portal platforms such as IBM WebSphere, Oracle Portal and BEA WebLogic.

Last January, I did a podcast with Stefan Andreasen, founder and CTO of Kapow. Andreasen described the mashup landscape. You can listen to the podcast here or read the full transcript here. I also blogged last April about Kapow's Web-to-spreadsheet service. At that time, I said:

Despite a huge and growing amount of “webby” online data and content, capturing and defining that data and then making it available to users and processes has proven difficult, due to differing formats and data structures. The usual recourse is manual intervention, and oftentimes cut-and-paste chores. IT departments are not too keen on such chores.

But Kapow’s OnDemand approach provides access to the underlying data sources and services to be mashed up and uses a Robot Designer to construct custom Web harvesting feeds and services in a flexible role-based execution runtime. Additionally, associated tools allow for monitoring and managing a portfolio of services and feeds, all as a service.

In addition to the Web 2.0 Edition and the Portal Content Edition, the Kapow Mashup Server is also available in the Data Collection Edition and the OnDemand Edition.

All editions are available now. More information can be found on the Kapow Web site. Product pricing is based on a flexible subscription offering.

SOA Software, iTKO team up to offer SOA lifecycle management and QA

SOA Software and iTKO have teamed up to offer enterprises continuous management and quality assurance across the entire lifecycle of service-oriented architecture (SOA) applications.

The new offering incorporates the LISA Testing, Validation, and Virtualization Suite from Dallas, Tex.-based iTKO and Policy Manager and Service Manager from Los Angeles-based SOA Software. The two companies say the combined solution will provide protection across the entire design, development, and change lifecycle.

Among the benefits of the combined solution are:
  • Continuous compliance and quality automation from concept to production support for SOA, with LISA validation natively executed as part of the workflows within SOA Software Policy Manager.

  • Visibility into SOA policy compliance levels, with all tests, test results, endpoint data, and models viewed in a single repository.

  • An increase in the types of SOA policy that can be modeled and validated, ensuring reliable service level outcomes.

  • Service virtualization of endpoints, locations and binding properties from SOA Software combined with simulation of service behaviors and data from iTKO.

  • Enhanced runtime validation of live SOA applications for both functional and performance purposes.
The joint solution is designed to meet the needs of enterprises seeking to manage complex, heterogeneous service assets to ensure that business requirements are met, while mitigating the risk of inevitable change in underlying systems such as enterprise service bus (ESB)/messaging, databases, mainframes and other custom and legacy applications.

I took a briefing recently on LISA and was really impressed with the approach and value. It's worth a look if you're not familiar with iTKO.

Etelos puts more 'sass' into SaaS with four additional hosted Web 2.0 offerings

Etelos, Inc. has beefed up its software-as-a-service (SaaS) offerings with the addition of four Web 2.0 stalwarts to its Etelos Marketplace. Users can now take advantage of WordPress, SugarCRM, MediaWiki, and phpBB as hosted solutions from the San Mateo, Calif. company.

The new additions are designed to help enterprises, small businesses, bloggers, and individual users connect with customers and other online communities on an on-demand basis. Users can set up a blog or a wiki with nothing more than a browser and Internet access. Technical details are handled by Etelos.

Founded in 1999, Etelos has been a go-to place for open-source developers eager to get their apps into the marketplace without having to go into the software distribution business. It also provides one-stop shopping for businesses looking for those apps, offering common user management, billing, support, and security.

6th Sense Analytics adds new features for collecting development productivity metrics

6th Sense Analytics, which collects and provides metrics on software development projects, this week announced several enhancements to its flagship product. These enhancements provide a more user-friendly interface and organize reports into workspaces that more closely align with the way each user works.

The Morrisville, N.C. company targets its products at companies that want to manage outsourced software development. It automatically collects and analyzes unbiased activity-based data though the entire software development lifecycle. [Disclosure: 6th Sense has been a sponsor or BriefingsDirect podcasts.]

Among the enhancements to the product are:
  • Reports can now be scheduled for daily, weekly or monthly delivery by email, reducing the number of steps required to access reports, providing easier integration into customer work routines.

  • Users can now select specific reports providing the ability to see only the information pertinent to their needs.

  • The registration process has been streamlined. After inviting a new user to a team, the user’s account is immediately activated and the user is sent a welcome email that provides details for getting started including instructions for desktop installation. The action of removing users has also been simplified.

  • Reports are now relevant to any time zone for customers working with resources across a country and on multiple continents.
I've been following 6th Sense Analytics since they first emerged on the scene. Last year, I had a podcast with Greg Burnell, chairman, co-founder and CEO, as he explained the need for metrics in outsourced projects. You can listen to the podcast here and read a complete transcript here.

Last August, I reported on the first metrics that 6th Sense Analytics had released to the public. Those findings confirmed things that people already knew, and provided some unexpected insights. I saw a real value in the data:

And these are not survey results. They are the use data aggregated from some 500 active developers over past several weeks, and therefore make a better reference point than “voluntary” surveys. These are actual observations are on what the developers actually did — not what they said they did, or tried to remember doing (if they decided to participate at all). So, the results are empirical for the sample, even if the sample itself may not yet offer general representation.