John Hofman's Design Study
John Hofman (Talk | contribs) (→User Stories) |
John Hofman (Talk | contribs) |
||
| Line 6: | Line 6: | ||
=Design Study= | =Design Study= | ||
| − | == | + | ==Initial Design== |
| + | When I first sat down and started to think about how to divide up an instant messenger client into separate parts, I came up with two different abstractions of the client environment. | ||
| + | |||
| + | ===Chats=== | ||
| + | This design follows how existing instant messengers appear to present the clients structure. The user interface is used to manipulate chats. Chats communicate with the server to update chats in other clients. | ||
UML | UML | ||
| − | === | + | ===Room for Conversation=== |
| + | This design of the instant messenger client models a group of people conversing in a room. The model contains users objects, one for each person who is online. The users can interact through chats. | ||
| + | |||
| + | Users can create, join, leave and post messages to a chat. There is one user, the 'Home' user, that represents the person who is using the client program, the HomeUser is controlled by the | ||
| + | user interface. The other users, OnlineUsers, are controlled by the network. However, the behavior of these 'online users' originated from the user input on their own client program somewhere else on the network. Therefore, whenever the HomeUser interacts with the chats, any actions it makes are also sent to the network connection to update the OnlineUser (kind of puppet) that represents it, in the other peoples client environment. | ||
| + | |||
| + | UML | ||
| + | ===Discussion of Initial Design=== | ||
| + | I prefer the first design. The 'Room for Conversation' is a nice fluffy interpretation of a real world room with people and conversations. But it is complex. | ||
| + | |||
| + | Improvements and Considerations | ||
| + | * The design is | ||
| + | |||
==User Stories== | ==User Stories== | ||
Revision as of 09:46, 2 August 2010
Contents |
My Project
This design study was introduced by my ENEL428 software assignment. The purpose of the assignment is to design a prototype of an Instant Messenger System using concurrent programming. The system is broken into two separate parts a client and a server. This design study is for the client program.
The communication protocol was not specified so I am using a system similar to IRC where logging on only requires a unique username, not a user account. This means that the server only needs to maintain a list of currently online users.
Design Study
Initial Design
When I first sat down and started to think about how to divide up an instant messenger client into separate parts, I came up with two different abstractions of the client environment.
Chats
This design follows how existing instant messengers appear to present the clients structure. The user interface is used to manipulate chats. Chats communicate with the server to update chats in other clients.
UML
Room for Conversation
This design of the instant messenger client models a group of people conversing in a room. The model contains users objects, one for each person who is online. The users can interact through chats.
Users can create, join, leave and post messages to a chat. There is one user, the 'Home' user, that represents the person who is using the client program, the HomeUser is controlled by the user interface. The other users, OnlineUsers, are controlled by the network. However, the behavior of these 'online users' originated from the user input on their own client program somewhere else on the network. Therefore, whenever the HomeUser interacts with the chats, any actions it makes are also sent to the network connection to update the OnlineUser (kind of puppet) that represents it, in the other peoples client environment.
UML
Discussion of Initial Design
I prefer the first design. The 'Room for Conversation' is a nice fluffy interpretation of a real world room with people and conversations. But it is complex.
Improvements and Considerations
- The design is
