Wednesday, April 2, 2008

A Time-Saving Programming Tactic That s Doomed to Fail

Suppose you have a software project which is under severe pressure of time. Let us also assume that this time is tight, and you know that & 39; there is a wide range of programming frantic late nights and hot, black coffee. What can you do to go faster this process?
Obviously, the correct answer will depend on the & 39; an individual situations. But I can tell you, what to do when a large number of programmers these crises. They decide to save time, Skipping software for the planning and design phase, coding and begin immediately away.
To a programmer or unruly unge bter that seems to be meaningless. Indeed, the finished product, c & 39; is what really matters, right? The customer does not & 39; occupies & 39; organizational charts, class diagrams, programming or styles of & 39; software architectures. Anything that they want & 39;, c & 39; is something that works.
Sure, it seems to make sense, but & 39; is a foolhardy approach. That way madness. We have all heard that the planning of a & 39; ounces is worth a pound of cure, but in the world of software development, people are all too quickly, c & 39; is & 39; ignore adage .
If real estate developer needs to build a house quickly, said You save time, it ignores the architectural design phase? Has he now decided to forgo simple models and concrete, fixing? Of course not. He knows that the results would be chaotic, and the work is slower, advance real progress without a concrete vision and plan.
Yet c & 39; is the exact procedure that many companies & 39; when he & 39; s & 39; acts software. They tell their programmers to start coding abducted, believe that & 39; is the process of rapid development. Of the projects, relatively straightforward, it might work. On a somewhat more moderate complexity, but such an approach is doomed to failure & 39;. Of course, you can save time at the first. However, without special software for planning and careful planning, problems are bound to be with you before long. Many of these problems will not be visible in the test phase comes around, and then there are late.
And yes, this & 39; is all too common in the world of & 39; economy. Because of the time constraints, a manager from May erroneous order his team to jump on & 39; organization of work and easy to hack. This strategy of & 39; team, and seem productive, but it may draw the chaos on timeline for the project. What is more, the resulting code is often confused a poorly documented, the chaotic confusion. If this software should be in the years ahead, then you have a recipe for disaster.
Mind you, I am not able to say that a heavy & 39;, Up-Front-design is a growing need . When time is scarce, so we & 39; n not a luxury d & 39; complex software design comprehensive literature. However, it should at least put a software architecture out - in detail sufficient to relaunch the process of software development soft and lightweight. Programmers rarely err on the side of over-planning, but more often in the case of & 39; lack of detailed design. Do not make this mistake, if you really want to, you save the time.
About Author:
V. Berba Velasco, Jr., Ph.D. If a leader electrical and software engineering d & 39; CTL, a biotechnology company that ELISPOT analyzers ELISPOT expertise, the immune system and draws PBMCs serum-free culture media. honey caprice



Bookmark it: del.icio.usdigg.comreddit.comnetvouz.comgoogle.comyahoo.comtechnorati.comfurl.netbloglines.comsocialdust.comma.gnolia.comnewsvine.comslashdot.orgsimpy.com

No comments: