Source Code Mgmt on Unix

[Richard R Ragan]

We are interested in understanding what organizational
and project strategies UNIX users employ to handle
source code management using RCS and SCCS tools.
We are particularly interested in how they manage a 
project that has many different files (more than 500)
with many different programmers needing access to the code.

1. Do you generally keep all code associated with one
   program in one directory, or do programs reside in
   several different directories?
   If they reside in different directories, how does a
   programmer determine which files are part of the
   program he is is working on?

2. How do you handle the case where more than 1 programmer
   needs to modify code to the same file at the same time?
   Does one wait until the other is finished?
   If not, how are changes reconciled when the second
   programmer checks in his code?

3. When a programmer adds a new function or file to an
   existing program, who is responsible for seeing that
   the makefile is updated to reflect the change?

4. How do you determine which versions of the files go 
   together?

5. Have you found any other available tools to be
   particularly helpful in managing your code? 
   Have you written your own scripts to further enhance
   or simplify the RCS/SCCS tools?

I will summarize feedback. Thanks.
Richard R. Ragan (408) 496-4340
Control Data Corp., Silicon Valley Operations