1995Unpublished venueRequires access

Should C Replace FORTRAN as the Language of Scientific Programming?

Linda Wharton

Open publisher page 0 citations

Abstract

Anti-FORTRAN sentiment has recently become more prevalent. Where does the attitude originate? The most probable source is academia, where C and C++ are the languages of choice. Is there a fact based justification for the attitude? FORTRAN and C are evaluated to determine whether C is a better language than FORTRAN for scientific programming. The features of FORTRAN 77, FORTRAN 90, C and C++ are compared, and evaluated as to how well they meet the requirements of the scientific programming domain. FORTRAN was designed specifically for numerical programming, and thus better meets the requirements. Three algorithms in the scientific domain are coded in both FORTRAN and C. They are evaluated on performance, readability of the code and optimization potential. In all cases the FORTRAN implementations proved superior. Is there evidence to mandate that all upgrades and new development should be done in C, rather than FORTRAN? A good computer programmer can solve any given problem in any langua...

About this research paper

What this paper is about

Anti-FORTRAN sentiment has recently become more prevalent. Where does the attitude originate? The most probable source is academia, where C and C++ are the languages of choice. Is there a fact based justification for the attitude? FORTRAN and C are evaluated to determine whether C is a better language than FORTRAN for scientific programming. The features of FORTRAN 77, FORTRAN 90, C and C++ are compared, and evaluated as to how well they meet the requirements of the scientific programming domain. FORTRAN was designed specifically for numerical programming, and thus better meets the requirements. Three algorithms in the scientific domain are coded in both FORTRAN and C. They are evaluated on performance, readability of the code and optimization potential. In all cases the FORTRAN implementations proved superior. Is there evidence to mandate that all upgrades and new development should be done in C, rather than FORTRAN? A good computer programmer can solve any given problem in any langua...

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Anti-FORTRAN sentiment has recently become more prevalent. Where does the attitude originate? The most probable source is academia, where C and C++ are the languages of choice. Is there a fact based justification for the attitude? FORTRAN and C are evaluated to determine whether C is a better language than FORTRAN for scientific programming. The features of FORTRAN 77, FORTRAN 90, C and C++ are compared, and evaluated as to how well they meet the requirements of the scientific programming domain. FORTRAN was designed specifically for numerical programming, and thus better meets the requirements. Three algorithms in the scientific domain are coded in both FORTRAN and C. They are evaluated on performance, readability of the code and optimization potential. In all cases the FORTRAN implementations proved superior. Is there evidence to mandate that all upgrades and new development should be done in C, rather than FORTRAN? A good computer programmer can solve any given problem in any langua...

Key concepts: Fortran, Programming language, Computer science, Programmer, Third-generation programming language, High-level programming language, Domain (mathematical analysis), Programming paradigm

Related papers

Back to paper searchBrowse research topicsOriginal source
Should C Replace FORTRAN as the Language of Scientific Programming? — Research Paper | ScholarLens