trm@cbnews.ATT.COM (Tom R. Mueller) (12/15/88)
The X server is a successful example of achieving software reuse by partitioning the software into layers, one of which is completely portable because it is machine independent and others that must be modified when porting to a new architecture. I'm curious about the development process by which this partitioning came about. When was the functionality within the server separated into layers? Before or after the first server was implemented? Was the partitioning based on the implementations of other window systems? What criteria are used to determine the layer for a piece of functionality? When were these criteria developed? Was any formal mechanism used to do the partitioning? If the development was done again, what would be done differently? The layered architecture of the X server can be used as a model for other software systems that need to be ported to a variety of environments. Any insights obtained through the development of the X server may be useful to many software developers. - Tom Mueller CB 0D109 (1-614-860 | 8-353) 5287 6200 E. Broad Street ...att!cblpn!trm Columbus, OH 43213 -- - Tom Mueller (1-614-860 | 8-353) 5287 cblpn!trm CB 0D109