View this PageEdit this PageAttachments to this PageHistory of this PageHomeRecent ChangesSearch the SwikiHelp Guide
Hotspots: Admin Pages | Turn-in Site |
Current Links: Cases Final Project Summer 2007

Case Study Milestone 3

Untitled Document Project Goal:

To make a detailed design for the rest of the milestones (P4 - P7) with CRC card analysis, UML diagrams & Scenarios. To assign responsibilities to each group member.

Purpose of the Project:

To learn how to use ECODE.

Problem:

To have a believable and implementable design about the milestones involving Wonderland without knowing anything about Wonderland. To use a completely new software ECODE to turn our project in.

Solution:

We started finding out stuff about Wonderland early.

Design Process:

We assigned tasks to group members according to their strength, for example Umang is good with File I/O in squeak & Deepak is comfortable with wonderland. Initially when we were designing M6 and M7 together our aim was to make a design that could easily be implemented to any map. To retrieve the coordinates of the buildings and to trace the route, we had the option of using either a web crawler or pull the coordinates out of a text file. The problem with the web-crawler was that the coordinates of the georgia tech campus were available but to it would have been hard to find coordinates of an imaginary campus which were to create in M7, therefore we decided to to go ahead with text files. It was a slightly tedious job to create the text files , but once they were created it was very easy to implement them. We used the map from the alumni website to place our buildings and trace the route. An issue with the campus map on the alumni website was that the roads on the map werent straight lines, there were plenty of curves on the map and this had to be negotiated wisely as we could only draw or move the camera only in a straight direction. We countered this problem by creating another text file. We divided each curve on the map to several nodes , therefore when we travel between each of these nodes it would give us an illusion of moving in a curved path , this way we could use the map on the alumni site thereby making the tour more realistic. We entered the coordinates of the node into a text file and retrieved the information from this file to trace the route between any 2 buildings.
This aspect of using text files made our design extendable to any map because all we needed was a set of text files for the respective campus. We were positive that this implementation would work , however were still a bit skeptical about the chances of the design working as we had no knowledge about wonderland to begin with, but a little but of experimentation assured us the chances of our design working. We were a little skeptical about ECODE too , as there were way too many complaints by students, but once we started using it suprisingly we didnt have a single problem using it and our design went of very smoothly.

Coding Issues


There were no coding issues.


Conclusion

We believe that this was the most important milestone. We saved ourselves a lot of work and pain because of our good design that allowed for everything that future milestones demanded of us. because of our design we did not have to reinvent the wheel every new milestone, and that made designing the whole thing beforehand really worth its while.We might have changed the design slightly over the next few milestones as got know Wonderland better, but the essence of the design was the same all throughout and this helped us extend our design with from M6 to M7. ECODE turned out to be a very useful over the course of the next 4 projects, we could not have imagined making those CRC cards and scenarios on paper, we actually landed up doing that for M6 and then realised how useful ECODE was.



Link to this Page