SCENARIO-BASED CLAIMS ANALYSIS OF MOOSBURG
SCENARIO 5:
John, one VPI Student, wants to talk on line with his girlfriend Samantha who lives in Chicago. Samantha has AOL as an ISP (Internet Service Provider) but no account on any Unix machine. John has no access to a PC so they can't use some talk software. IRC is not very convenient so John gives Samantha rendezvous in MOOsburg in front of Arnold's. He wants to show her around Blacksburg. John logs into MOOsburg and goes north until he "reaches" the center of Blacksburg with Arnold's. Two minutes later, Samantha logs into MOOsburg. John has told her to go North. Thus, she clicks on North (the link on the web page) many times. Then she doesn't know what to do. By using the help John knows how to talk. He writes "hello Sam. Samantha reads the message and type Hello John but gets the error message: "I don't understand that". John waits a little for an answer and then realize that maybe Samantha doesn't know how to talk in MOOsburg. He spends a few minutes explaining to Samantha some basic commands. After that, John wants to show Main Street to Samantha and goes South. Samantha says "wait for me" but nobody can hear that because now she is alone. Then, she goes South. After the description of the place, there is a sentence from John: "and here..." she realizes that she missed the previous description. They talks together.
CLAIMS ANALYSIS OF SCENARIO 5
- Use of the mouse and the keyboard:
... is a good idea: the users can use the keyboard only if they are very familiar with that, or else they can use the web-based links/icons to move.
... but to see a picture, the user needs to use the mouse.
... but a novice user will have to switch between the mouse for the movements, look/help and the keyboard to talk. The user needs to click on the text frame and then move his/her hand to the keyboard. The GOMS theory could show that the time spent here is very costly.
A mouse-only interface isn't a good solution because it's impossible to find enough icons to represents the words used by people. But some features could be added to avoid a very intensive use of the keyboard. Some smiley icons could be used to emote something. The user should use the keyboard only for talking if he wants to. The learnability of a system using graphical representations (icons) instead of text commands is really better.
- No feedback about who is listening or what one user says:
... avoids putting too much things in the telnet frame or in the web-frame. When the user talks, every people in the same room can hear him/her.
... but the user needs to be careful to notice those messages as "user xx goes North" before he said something to this user. The information is available but it tends to be confusing.
Something like "You say xxx to user1, user2, ..." could be used as a feedback mechanism.
- No history in the rooms:
... avoids putting too much text information in the telnet frame.
... but it's not easy for someone who enters a room to understand what is the current talk about.
Maybe a history button could be available in the toolbar. Clicking on this button could open a new window with the last 5 minutes' events/talks in the room.
- No feedback about new people arriving in the web-window:
... keeps the bandwidth use low.
... avoids being redundant with the telnet frame.
... but the users won't see that some new people could hear them.
... but this will decrease the social aspect of MOOsburg.
Maybe a frame with the names of the peoples in the room could solve the bandwidth and the social problems.
