 |
privacy |
||
|
Hangout for experimental confirmation and demonstration of software, computing, and networking. The exercises don't always work out. The professor is a bumbler and the laboratory assistant is a skanky dufus.
Blog Feed Recent Items 
The nfoCentrale Blog Conclave nfoCentrale Associated Sites |
2005-04-23Conquering the Business-Application Life CycleACM News Service: Together--At Last. This blurb identifies the creation of systems for application-lifecycle management to combat the current wasteful, late, and failed corporate projects. The commoditization of application development tools is cited as the basis for failure in the process (?) and this appears to be the driver for moving away from the saturated low end to integrated platform approaches: IBM Rational's Unified Process (RUP), Borland's Software Delivery Optimization Strategy, and the rapidly approaching Visual Studio 2005 Team System platform. "ALM Platform" is the latest buzz. I think I need to understand better how these dots connect. I wonder how one can avoid going through the equivalent of a SAP integration for Application Lifecycle Management. That's way too high-end for my taste. Gavin Clarke's 2005-04 Computer Business Review article, "The Day of the Suites" provides deeper background. There is more effort to put modeling, such as MDA, into context, and also distinguish platforms for lifecycle management as opposed to the IDE focused on the programmng code-compile-test cycle. The writer's bottom line:We are moving into a world of fully integrated suites where the IDE, once a by-word of integration, looks increasingly anachronistic. Customers should be wary, however: this headlong rush to appear integrated is subject to some marketing hype, and integration is a subjective term that varies from vendor to vendor.Clarke's 2005-03 "Together At Last" article begins with a recounting of the US FBI's case-management system project failure where even the cost of exiting is estimated in the $100 millions. After building a case for the gap between development-methodology coverage and the pain centers of application development, the case for ALM is stimulated by three colliding factors. Commoditization at the programming level has forced development-system vendors to look upscale. Recognition that the vendors must align their offerings with business needs leads to extension of platforms and tools to embrace the life-cycle. And new regulations on business operations are compelling even more attention to system lifecycle management and operating integrity in the business problem space, not just the IT solution-instrument space. Computer Business Review has more to offer on the application-development scene. There is also more over-all perspective to be gained from their other coverage of business computing analysis topics. My radar was triggered on this because I have my attention on life-cycle considerations in raising open-system trustworthiness. This is the squishy end of the trustworthiness problem and I despair over finding a clean, simple characterization that I can use as the context for framing the system deployment and software development levels. I believe the attention to lifecycle factors, will be critical, along with heightened attention to cradle-to-grave coordination of risk management. It is time to draw pictures and not trying so hardto find the perfect words. The only way I can see avoiding overwhelm and despair over the intricacy of system-lifecycle considerations is by taking on increased dependability and trustworthiness as a journey, not an end, with exploration, correction, and continuous improvement the name of the game. I'd like to see an easy on-ramp for building up trustworthy software-development practice. I want to hold onto the prospect of traveling light and improving the process while moving from one success to the next. I am a little wary of integrated platforms at the present (dare I say eroding) stage of application-development maturity,because of what strikes me as a high cost of engagement and no little desparation of the platform vendors to find a market. 2005-04-21FLINT for bug-free, secure, and reliable software. That should cover it!ACM News Service: Studies Recharge Computer Science. Susie Poppick's 2005-04-20 Yale Daily News article recaps a number of activities in the Yale Computer Science Department. What caught my eye is the description of Zhong Shao's FLINT project. I'm also a little wary of the blurb's use of "with an eye toward commercialization." The article is clear that the project is oriented toward certification of commercially-marketed software. That's reassuring although I don't think that's meant as a limitation. The FLINT approach involves writing certified software, accompanied by a logical proof (of what, pray tell?) that is amenable to mechanical verification. I've always been curious how that could work, and I am more curious since it figures into my immediate interest in raising open-system trustworthiness. There are no links in the article, but a quick search provides some hot leads. Of greatest importance is noticing that the methodology depends on use of a common intermediate language for expressing programs, and it is based on strong typing and safety qualities of the chosen intermediate form. I know where I might be able to use that. Some day. To dig into as part of my immediate trustworthiness preoccupations, there is FLINT's apparent use of SML/NJ (an open-source implementation of Standard ML, where ML is Robin Milner's Meta-Language), as well as efforts on secure programming and proof-carrying code to dig into. The direction is interesting, though I am willing to start with code-carrying attestation, with code-carrying proofs a "nice-to-have when we can finally get it." |
||
|
You are navigating Orcmid's Lair. |
template
created 2004-06-17-20:01 -0700 (pdt)
by orcmid |