Allow Plenty of Time for Large-Scale Software

  • Authors:
  • Ware Myers

  • Affiliations:
  • -

  • Venue:
  • IEEE Software
  • Year:
  • 1989

Quantified Score

Hi-index 0.02

Visualization

Abstract

Estimation of development time for multimillion-line software is addressed. Three methods are discussed. The historical-database approach finds the statistical trend line by regression analysis that relates project development time to size in source lines, and extends it to a ten-million-line size. The software macroestimation methods, Cocomo (cost constructive model) and the software life-cycle model, forecast labor effort and schedule from early estimates of the amount of functionality in the project. The three methods all point to the conclusion that a large amount of time must be allowed.