Meeting Notes



Tuesday, February 1

The group met shortly after class to delegate the duties and to discuss other appropriate matters concerning our project.  We agreed that David would be the producer, Lee would be the technical director, and Joe B. and Joe R. would be in charge of designing and keeping the project web page up to date.  We also discussed weekly meeting times, weekly reports, and contacting our client Carl Hatchell.
 

Thursday, February 3

The group met for one of our weekly meetings to discuss the code of the Hoops Scheduler written during the Comp145 class during the fall 99 semester and to finalize our meeting time with Carl Hatchell.  After finalizing the meeting time with our client for 3pm on Sunday, February 6, we determined that we would have to wait until after our client meeting to determine our approach concerning the previous code of the Hoops Scheduler.  We made this decision for two reasons.  The first reason was that we did not have all the code needed to compile the previous version of the Hoops Scheduler.  The second reason was that the executable file that we found for the old Hoops Scheduler did not work properly.  We also continued to discuss the design of our project web page.
 

Sunday, February 6

 The group met with our client Carl Hatchell to discuss his requirements for the Hoops Scheduler.  We also reviewed the previous version of the Hoops Scheduler written during the Comp145 class during the Fall 99 semester to point out its many problems which included the inability to work outside Sitterson Hall and a non-user-friendly graphical interface.
 

Monday, February 7

David and Lee met to write our preliminary contract for our project, which we named the Last Choice Hoops Scheduler.  When we were done writing the preliminary contract, it was given to Joe B. to add to our project web page.  Afterwards we continued to test the executable file from the previous version of the Hoops Scheduler written during the Comp145 class during the fall 99 semester.  From this testing, we came to the conclusion, to be brought up at the next weekly meeting, that the group's best approach may be to rewrite the program from the ground up, reusing some code from the previous version of the Hoops Scheduler.
 

Tuesday, February 8

The group met to discuss the original HOOPS code and our preliminary approach to producing a working version.  Joe R. had previously conducted a test of the software with two leagues, each with nine teams.  He distributed printouts of the results the previous scheduler had produced, followed by a discussion to pinpoint problems with the software.  We identified two (mostly disparate) problem areas: Scheduling algorithm problemsand GUI problems.

Wednesday, February 16

The entire group met with three goals in mind:  finalizing our contract, creating a WBS (work breakdown structure) for our project, introduction to creating projects in Visual C++, and continued discussion of preliminary testing of the Version 1 HOOPS code.  We discussed further refinements of the contract, and Dave will get feedback from Hatchell this weekend on our final draft.  Joe B. sketched a rough draft of the WBS, omitting some details,  and the group decided to individually consider what details should be added and how long each task would take.  The (final) preliminary WBS will be discussed and agreed upon by all in the next week, as will the corresponding temporal chart.  Joe R. led a short discussion on a more thorough test of the Version 1 HOOPS code, in which a few more problems with the scheduler were noted.  Afterwards, we all moved to the lab for a short overview, by Lee and Joe R., on how to set up and run a project on Visual C++ 6.0.  We also noted some possible (later) refinements for the GUI.

Thursday, February 24

We completed the preliminary PERT chart using MS-Project, and once design/division of labor if finalized, we will add color-coding to the corresponding GANT chart reflecting each members' responsibilities. However, we have encountered difficulties in posting a copy of the PERT to our web page; MS-project exports project information as an HTML table, and the GIF capture option it has only captures what is in the view window of the application. This is a minor task to complete in the next week.

We discussed the user manual, and decided a total re-write of the Version 1 user manual was necessary, as neither its format or content is desirable, nor will the GUI of our version be the same as that described in the old manual. The manual should contain (preliminary) sections including, but not limited to, the following: creating a camp, creating leagues, creating teams, adding/removing courts, specifying court availability, and viewing reports. Each of these sections should be accompanied by screenshots of the corresponding dialog boxes from the GUI, with detailed explanations of every button, drop-down menu, etc.

Tuesday, February 29

The focus for today's meeting was on our interface design and the corresponding user manual. We discussed options for our GUI refinement and decided that the tab-pages style of Version 1.0 was not user-friendly at all. Instead, we have opted for a conventional drop-down menu bar primary interface, perhaps with a pop-up navigation tool bar as well for frequently-used options. We discussed modifications for the various input and editing features as well as extensions to make schedule modifications easier. The conclusion of our discussion was to make a preliminary layout of the new user manual.

Division of labor has finally become a reality: Lee will be the project lead on GUI development and Joe R. the project lead on scheduling algorithm refinement. Joe B. will continue to maintain the web page/update online documentation. Dave will coordinate efforts between the other three team members, contact the client with progress reports, and continue with other duties as producer. Both Dave and Joe B. will develop code with Joe R. or Lee, depending on the rate of progress in their respective areas.

Monday, March 6

We completed the preliminary user manual over the weekend, and have assigned tasks to members for our project schedule. Presentations are this week, so we met briefly today to discuss the format for our presentation. Joe R. and Dave volunteered to lead it, and they met later in the evening to prepare further. We discussed a meeting with Hatchell, but decided to postpone it until after spring break.

Monday, March 20

Joe R. and David met with Coach Hatchell for a couple of hours to have the final contract signed and to review our course of action concerning modifications to the camp scheduler. A majority of the time was spent running a demo and determining what features Coach Hatchell approved of and what and how other featrues needed to be changed in order to better accomidate the client.

Thursday, March 23

Joe R. and David met for several hours to continue to interpret previous version of camp scheduler and began to make changes to the part of the algorithm that schedules games for Carmichael Auditorium. They also successfully modified the amount of time alloted per game (per Hatchell's recommendation), reducing the number of errors by 5-15, depending upon the camp setup.

Tuesday, March 28

Status Report:
All testing of the Version 1.0 software has been completed, and the preliminary design of the GUI is ready for implementation. Joe R. and David are making progress in reducing the number of errors generated by the scheduling algorithm. Progress on the GUI modifications is slightly behind schedule, due in part to revisions required by the client.
The whole group met to discuss the modifications that Hatchell wanted made to the GUI and discuss our implementation plan. Hatchell wants to retain the tab-page style appearance of the GUI, but with serious modifications. The new planned modifications are listed on the GUI Problems page , at the bottom. We had originally planned on a drop-down menu style GUI, which Lee had begun coding--unfortunately this approach had to be scrapped at the behest of the client.

In addition, we discussed a system for code modification so that all changes are documented. All changes must be noted in a code modfications text file, as well as documented with comments where they appear in the code. All new code must be tested thoroughly with the existing system before being uploaded to our common code directory. Of course, we have multiple backups of the existing code.

Friday, April 8

Joe R. and Dave continue to meet (three times a week) and work together on error reduction in the scheduling algorithm. Lee and Joe B. are working on functionally disparate areas of the GUI to complete the modifications in as rapid a manner as possible. They met Friday to integrate their respective modifications into an updated interface.

Tuesday, April 11

Status Report:
The scheduling algorithm continues to pose problems for our team. Error reduction may have to be accomplished by a "fixer" subroutine that corrects problems created by the main algorithm. The GUI is nearly complete--modifying the format of the reports is the only critical thing remaining. Secondary requirements for the GUI are to be postponed and GUI development suspended at the end of this week so Lee and Joe B. can help Joe R. and Dave with the algorithm.

Dave, Joe R., and Lee met with hatchell Monday night to show him the GUI modifications and get clarification on some scheduling issues. He was, in general, pleased with the interface--the formatting of the reports is the only problem he has with it. We discussed immediate goals for the next week: Joe B. is to try to re-format the reports pages and make them manually editable, as well as correct a few minor problems with buttons, boxes, etc. Lee is shifting gears to start learning the scheduling algorithm, and Dave and Joe R. are to continue with error reduction in the algorithm.

Tuesday, April 18

Status Report:
The interface now contains the various schedule views that Hatchell wanted to see, plus the manual editing abilities he needed in the Master schedule. Unfortunately, the algorithm continues to present problems for our team, mostly due to the limitations inherent in the original design. The use of linked-lists for data structures throughout has hindered our efforts to modify the schedule internally after it is first created. Our goal, then, is to reduce the most critical errors rather than trying to perfect the algorithm.

We had a brief meeting so we could assess the new requirement Hatchell informed Joe Russ of at the last client meeting: There must be a view that displays all the games for a single league AND fits completely on one page with a key at the top. This is necessary because many copies must be made of the league schedule for distribution to players, coaches, referees, etc., and the copying costs will be double if it doesn't fit on one page. This requirement will necessitate a new Document View page and some format tweaking--the GridCtrl cannot be used, because it is impossible to adjust the font size within the GridCells. We also discussed meeting times for the upcoming week to prepare for our presentation/demo, and to update our manuals, etc.

Wednesday, April 26

Status Report:
The league schedule preview page, accessible from the tool bar in the main window of the application, is now functional barring some minor formatting issues that can be resolved after presentations. All other minor windowing problems have been corrected, and the algorithm has been modified to reduce errors as much as is allowed by the current implementation structure.

We met for a couple of hours late this evening to discuss our respective areas in the presentation and demo, which we agreed to volunteer for on the first day (Thursday, April 27). We also briefly discussed our upcoming client meeting (Thursday, April 27 @ 7:00 p.m.) and exactly how we were going to deliver the software to the client (InstallShield, set it up for him, etc.).

Friday, April 28

Our presentation and demo was a success--we all felt that we communicated the purpose of our project effectively and demonstrated the software's capabilities in an easily-understood manner. Our client meeting last night (Wednesday) went fairly well. Hatchell was pleased with both the appearance and functionality of the application. However, he did request the ability to add comments on the schedule reports so that referees can be assigned to courts/times. He also wanted to see a better time distribution of games in the schedule so that less referees could be hired for the camp (the less games that are played at the same time the better), and a more error-free schedule in general, so we'll make those changes in our spare time early this summer.

Tuesday, May 2

We had our final meeting today to discuss the final changes and documentation we need to finish before our project is reviewed. We also discussed the structure and content of the group paper we are to turn in. We split our final duties up as follows: Lee is to revise the implementation manual and add new class documentation. Joe B. is to update our meeting notes and code modifications pages, update the user manual, and make sure all the web page links work correctly. Joe R. is responsible for creating a setup program to install our software and the help files the help browser requires, as well as contact our client personally to install it on his machine. Dave volunteered to write the group paper based on the group's input, and make all our code accessible from the project web page.
That's all, folks.