Some time ago (probably urban story) statistics have been referred to
. .. The average life span of the software is about 3 years
At that time I have come for the following reasons (and I am sure the better are probably better):
-
A new major system (ERP, CRM, etc.) has been implemented and has an "integrated" module to replace the old app in it.
-
There is no single, but no integrated app - but the current app is not adaptable (people have been left out, technology has changed, current IT policies have changed Users do not like the current app.)
-
The company that has acquired the original app from you to customize it for your needs, is missing has gone.
-
Or you do not come well with them any time.
-
The technique for the current app is "obsolete" (according to the structure / Microsoft / consultant / industry expert / new IT manager who is the ear of management).
-
"We are ending (Windows 95 / Windows 98 / Windows 2000 / Windows XP / NT) and we need mailing technology in our application."
-
"We learned a lot (app version N) and we will perform better in the second / third / fourth / n + 1 time."
-
Developers / IT Managers / Division VP / Proprietary Job for Consulting Company.
-
Users do not like it.
-
We've got an acquisition / acquisition / have been acquired by a competitor and they are better.
Some of these are indispensable (eg your company has been bought), but overall it definitely needs to be avoided. Does your organization intentionally fight with this syndrome? What Effective Strategies Will You Recommend?
The average lifespan of the software I write at this time is probably somewhat the day. (I write a lot of scripts, so I can be a aberration. ;-) But the main system that I work is probably 15 to 20 years old. The built-in OS is about 30 years old. Nothing wrong with old or young software. In fact, the best use of software is when it is possible to adapt it to new uses.
Having the layers of intangible between the functional parts is easy to change the functionality in the system. For example, we have gone through many tape libraries on our system and now we will be in future in disk archives Are thinking about going. Since the "archival" part of our system sits behind an abstract layer, we can easily change it in lieu of the rest of the system.
Whenever possible, it is best to use standard parts as well. In this way, if you run in some range, then it is likely that the other will have the same problems and it is more likely that anyone will decide properly.
Comments
Post a Comment