Beware of Counting LOC.
Saved in:
| Title: | Beware of Counting LOC. |
|---|---|
| Authors: | Armour, Philip G.1 armour@corvusintl.com |
| Source: | Communications of the ACM. Mar2004, Vol. 47 Issue 3, p21-24. 4p. |
| Subjects: | Software productivity measurement, Source code, Function point analysis, International Function Point Users Group (Organization), Information retrieval, Computer science |
| Abstract: | The most common software sizing metric is source Lines of Code (LOC). Counting LOC implies trying to size a system, by counting the number of compiled source codes and lines that have been created. The need to count and measure LOC is to evaluate the power of the computer system, many times to determine the size of the system that is being built. The situation offers a dilemma, as LOC cannot be reckoned before the system is built, and once the system has been built, estimation of LOC does not remain necessary. This institutes the necessity to seek other ways to measure the power of computer system, and Function Points is one such way. This approach, adopted by the International Function Point Users Group involves counting and weighting input, output and data storage elements. Such counts are more helpful; as they can be reckoned at the time estimates are required in the project. However, this system is not free from drawbacks either. Function Point analysis may not always count transform behavior, state behavior, platform interaction, design dependence and a number of other attributes that legitimately affect the size of the system. |
| Database: | Engineering Source |
| Abstract: | The most common software sizing metric is source Lines of Code (LOC). Counting LOC implies trying to size a system, by counting the number of compiled source codes and lines that have been created. The need to count and measure LOC is to evaluate the power of the computer system, many times to determine the size of the system that is being built. The situation offers a dilemma, as LOC cannot be reckoned before the system is built, and once the system has been built, estimation of LOC does not remain necessary. This institutes the necessity to seek other ways to measure the power of computer system, and Function Points is one such way. This approach, adopted by the International Function Point Users Group involves counting and weighting input, output and data storage elements. Such counts are more helpful; as they can be reckoned at the time estimates are required in the project. However, this system is not free from drawbacks either. Function Point analysis may not always count transform behavior, state behavior, platform interaction, design dependence and a number of other attributes that legitimately affect the size of the system. |
|---|---|
| ISSN: | 00010782 |
| DOI: | 10.1145/971617.971635 |
