In Windows DNA, there are two major choices of user interfaces - Win32 clients and browser based clients. During the Internet revolution of the late 90s we saw the emergence of the browser and the Web Server. With the introduction of Internet, information started being available but with limited functionality. With the development of the Windows Distributed Internet Architecture, we started to see Web sites that allowed simple transactions to occur. Clients on browsers could access Web sites that had COM components available to them that allowed them to retrieve information from the database. So now we gained the capability to simulate the environment of the Win32 platform. The client software – the browser – can access information on a server. But as with the Win32 environment, we are limited in the way in which the information is presented to us. Customization is neither widespread nor broadly developed.
Let us look into limitations of these technologies.
Limitations in Win32 Clients
In a client-server environment visual tool such as Visual Basic, are often used to create a rich user interface. The drawbacks is that such client software is difficult to deploy and maintain, requiring and install on every client and a change to every client when an upgrade is needed.
DLL conflicts on the client are frequent because of variations in the version of the operating system and other software installed on the client.
Visual Basic is the most common language used to write middle-tier components. This requires high level of expertise in COM. Since these middle-tire components are implemented using Microsoft Transaction Server on Windows NT or COM+ services on Windows 2000. These components use stateless designs, which can look very different from the stateful designs often used in client-based components.
COM components, in the middle tier must work together, Versioning all the components properly so that they understand each other's interfaces can be a challenge. This requires a highly sophisticated skill level and a well - controlled deployment process.
COM works well on Microsoft platforms. But it suffers from lack of interoperability with other platforms. One of the most important ways functionality can be reused is for a software component to inherit another component, But COM does not support inheritance.
Visual Basic is the most popular language for developing applications with the DNA model, this is used in two major roles - forms based VB Clients and COM components. This VB6 language has its own limitations it doesn’t have the capability of multithreading, lack of OOPS concepts, Poor error handling ability and poor integration with other languages. Hence it makes it unsuitable for development of object-based frameworks.
Today’s applications need to use the Win32 API for a variety of purposes like monitor widows messages, manipulate controls, reading and writing to INI files and socket programming etc. But these widows API are hard to program for variety of reasons, like it is not object oriented and complex calls to the functions with long lists of arguments, since Win32 API is written in C++ language, getting calling conventions right on data
types is messy.
Limitations in DNA-Based Internet Development or Browser based clients
With DNA - based software development, creating software that is accessed by a user locally is done very differently from development for the Internet. The Visual Basic forms for client-server user interfaces versus the use of Active Server Pages for Internet user interfaces. Even though both situations involve designing and implementing GUI based user interfaces the tools and programming techniques used are quite different.
ASP lacks in state management between post backs. Every time a page is rendered, the programmer must make sure that all the visual controls like text boxes, dropdowns have their information loaded. It is the programmer's responsibility to manage the state in the user interface and to transfer state information between pages. This causes developers to have to write a lot of code for the internet user interfaces that is not relevant
to business problem being solved.
If the Internet application is going to run on a group of Web Servers, then considerable additional work is necessary to design a state management system that is independent of particular server.
Browser based clients are somewhat more difficult to create, and offer a more limited user interface with fewer controls and less control over layout of the screen and handling of screen events. It is possible to create rich user interfaces using DHTML, but it requires lot of coding and also browser compatibility issues rises, for which a separate coding or two version of the same page have to be maintained, keeping in mind, the browser we are targeting.
The Internet has caused server-based applications to become much more popular than ever before and has made the connectionless request/response programming model common. But communicating between servers—especially among those running on different platforms—is difficult, and because most substantial Internet applications are Database-Centric, the ability to access awide variety of data sources easily is moreimportant than ever.
As we move on to handheld devices or wireless devices, kiosks or other type of systems, many of which run a different processors and do not use standard operating system. So sharing the data between these devices and communication varies which is not uniform, becomes difficult.
Let us look into limitations of these technologies.
Limitations in Win32 Clients
In a client-server environment visual tool such as Visual Basic, are often used to create a rich user interface. The drawbacks is that such client software is difficult to deploy and maintain, requiring and install on every client and a change to every client when an upgrade is needed.
DLL conflicts on the client are frequent because of variations in the version of the operating system and other software installed on the client.
Visual Basic is the most common language used to write middle-tier components. This requires high level of expertise in COM. Since these middle-tire components are implemented using Microsoft Transaction Server on Windows NT or COM+ services on Windows 2000. These components use stateless designs, which can look very different from the stateful designs often used in client-based components.
COM components, in the middle tier must work together, Versioning all the components properly so that they understand each other's interfaces can be a challenge. This requires a highly sophisticated skill level and a well - controlled deployment process.
COM works well on Microsoft platforms. But it suffers from lack of interoperability with other platforms. One of the most important ways functionality can be reused is for a software component to inherit another component, But COM does not support inheritance.
Visual Basic is the most popular language for developing applications with the DNA model, this is used in two major roles - forms based VB Clients and COM components. This VB6 language has its own limitations it doesn’t have the capability of multithreading, lack of OOPS concepts, Poor error handling ability and poor integration with other languages. Hence it makes it unsuitable for development of object-based frameworks.
Today’s applications need to use the Win32 API for a variety of purposes like monitor widows messages, manipulate controls, reading and writing to INI files and socket programming etc. But these widows API are hard to program for variety of reasons, like it is not object oriented and complex calls to the functions with long lists of arguments, since Win32 API is written in C++ language, getting calling conventions right on data
types is messy.
Limitations in DNA-Based Internet Development or Browser based clients
With DNA - based software development, creating software that is accessed by a user locally is done very differently from development for the Internet. The Visual Basic forms for client-server user interfaces versus the use of Active Server Pages for Internet user interfaces. Even though both situations involve designing and implementing GUI based user interfaces the tools and programming techniques used are quite different.
ASP lacks in state management between post backs. Every time a page is rendered, the programmer must make sure that all the visual controls like text boxes, dropdowns have their information loaded. It is the programmer's responsibility to manage the state in the user interface and to transfer state information between pages. This causes developers to have to write a lot of code for the internet user interfaces that is not relevant
to business problem being solved.
If the Internet application is going to run on a group of Web Servers, then considerable additional work is necessary to design a state management system that is independent of particular server.
Browser based clients are somewhat more difficult to create, and offer a more limited user interface with fewer controls and less control over layout of the screen and handling of screen events. It is possible to create rich user interfaces using DHTML, but it requires lot of coding and also browser compatibility issues rises, for which a separate coding or two version of the same page have to be maintained, keeping in mind, the browser we are targeting.
The Internet has caused server-based applications to become much more popular than ever before and has made the connectionless request/response programming model common. But communicating between servers—especially among those running on different platforms—is difficult, and because most substantial Internet applications are Database-Centric, the ability to access awide variety of data sources easily is moreimportant than ever.
As we move on to handheld devices or wireless devices, kiosks or other type of systems, many of which run a different processors and do not use standard operating system. So sharing the data between these devices and communication varies which is not uniform, becomes difficult.