Comp 145, Team 8
Week 4 Meeting Minutes w/ Dr. Brooks
Outside Room 214
Attendees: Dr. Brooks, John Ehrhardt, Cary Hall, and Andy Hans
Discussed the architecture.
How the APL interpreter would load in the specified computer definition in APL. Then the client would load a program into the running simulation through the interface. After loading, the client would execute and monitor the program through the interface.
Dr. Brooks with consult with one of his associates about which interpreter would provide the best solution.
CapLib2 - open source C++ library.
SharpAPL - separate Linux process that connects through sockets. Issue of the latency through the sockets for passing entire memory segments and other large pieces of information to the interface needs to be investigated.
APL2000 - expensive IBM solution ($2000)
Try to get more graduate students to reply to the email survey.
Which OS and language
Linux programmed in C++ using FLTK for windowing
Dr. Brooks agreed upon the feature set proposed in "Plan A" as long as breakpoints are the highest rated non-required feature.
It is necessary to read Dr. Brooks' book - Chapter 9 to know the naming conventions for registers, memory, etc. in his APL definitions of the different computer architectures.
It was noted that programming an APL editor would be somewhat wasteful of our time. We should simply find an APL editor with ASCII functionality for the graduate students to use if other architectures are going to be defined.
Do not worry about writing an assembler that can be used for all the different instruction sets. It is too much to ask of a 145 group. Save for another project. Let the graduate students deal with writing in binary/oct/hex or let them write their own assembler(s).