If I had to illustrate a framework in common sense terms, it would be like making coffee where we are supposed to mix each ingredient in the right ratio which can prove to be quite challenging. Suppose you had a pitcher of sorts where all the mixing was to occur. The pitcher would then act as your framework.

In computing terms, frameworks are specified for applications which are analogous to the ingredients in our preceding explanation. They provide a visual guide or a wire frame on which applications are built. Writing of software or websites without the inclusion of a framework can prove to be problematic especially if basic tasks like packet encoding and decoding, returning static images and authentication of add ons are to be performed. The redundancy of having to do these tasks over and over again by yourself is reduced by using a framework which does these things for you. The kind of abstraction created by this framework from the small tasks can then allow for changing the generic functionality of the software by adding a bit of code written by the user.

Frameworks can, therefore, be thought of as an ideal platform. One has the option of overriding certain generic functionality. They assume the form of libraries where a program interface suited for the application in question can be used anywhere within the program. There is however a distinct boundary between a framework and libraries which I shall give light on in the succeeding texts and hopefully things will become clearer. From our previous analogy, ingredients are mixed inside the pitcher and so it would naturally follow from common sense and logic that code pertaining to the application we are writing falls within our application. On the flip side, the code is external to libraries. For the case of the framework, one can exercise control over when a particular set of code is called onto the framework while for the case of libraries, the program itself delegates.

In broader applications, a framework can be used for verification, validation and uncertainty the uncertainty of an application. Incorporating uncertainty in the model is accrued to its mathematical nature. For this reason, the model can be adjusted to include predictions of mathematical error in the foreboding uncertainty. Certain types of mathematical errors due to iteration and round off can be approximated by using verification techniques. The uncertainties can then the quantified and the final decision made and presented.

Other than the aforementioned applications, further research has led to the development of programming environments where the execution of a large number of applications running concurrently has been made possible. Earlier in history, this would not have been plausible. The possibility of certain algorithms being executed at the same time with some interacting with each other allows the implementation of more than one applications. Such environments, however, can be made more user-friendly by creating a more straightforward method on the user input side. Were well aware that not everyone, especially in developing nations, is tech-savvy and therefore this would go a long way in creating an easier time for those who are not well versed with computer technology.