Best Practices - MS Integration Experience

From Mpich
Revision as of 16:12, 10 November 2012 by Balaji (talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

This page describes the best practices for MPICH developers to make life easier for developers integrating with our code base. This document is a slightly modified version of the suggestions from Microsoft developers who frequently integrate with the MPICH code base.

  • Avoid renaming and modifying files in a single check-in. Rename the files in one check-in and modify the files in subsequent check-ins.
  • Avoid moving and modifying functions between files in a single check-in. Move functions between files in one check-in and modify the functions in subsequent check-ins.
  • Avoid commenting out or #ifdef out a section of code and rewrite it as new code. Remove the obsolete code and add the new code in the same check-in.
  • Avoid mixing source code beautification, E.g. limiting source code to 80 columns, with actual code change.
  • Avoid using tabs, use spaces instead.
  • Avoid making multiple logical changes in a single check-in. Make separate check-ins for each logical change.
  • Avoid silently breaking a functionality and fixing it after many check-ins. Use a separate branch for the broken code and merge it to the trunk after fixing the functionality.
  • Avoid retaining obsolete fields in a structure, unused variables and unused static functions. Remove the obsolete code in the same check-in where you stop using them.