For77, For90, C:大侠们都来举个手

sillybear

大侠们都来举个手！
看看Fortran77, Fortran90, C那个人气好。
我用老掉牙Fortran77格式，虽然Fortran90, C都学了，就是懒得用。

nacaairfoil

情况一样.

microcat

我用For77格式

山泉

当然f90好用

yangguang

还是visual for好阿

streamline

[这个贴子最后由streamline在 2003/05/24 05:04pm 编辑]

没听说C比Fortran更适合数值计算。
Fortran90的动态内存和module，递归，指针确实方便，但据说效率比77低。
Fortran77有丰富的程序、编译遗存！
可查High-Energy Flow Solver Synthesis (HEFSS)的工作！
以下引用“Opportunities for Breakthroughs in Large-Scale Computational Simulation and Design”
Although there are many scientific software languages, the most evolved object-oriented is C++, espe-cially as regards the use of templates. There is, however, a large base of legacy scientific code written in Fortran 77; Fortran 90 (supporting dynamic memory allocation, subroutine/function recursion, and pointers) is now available, and Fortran 2000, not yet available, should contain more object-oriented syntax.There is also an argument that Mathematica is the most appropriate choice because its symbolic manipulation, graphics, text processing, and computing capabilities are bundled interchangeably in one platform  The team could reach no clear consensus, and Fortran 90 was adopted as a reasonable choice for the project. Yet the debate over the most appropriate
scientific computing language continues today in the software industry.
   One interesting finding of the HEFSS team was that Fortran 90 implementations varied widely across machines. Some coding styles that were very useful for modularity had to be discarded because either one or more of the intended compilers did not fully implement the Fortran 90 standard or the execution of the syntax was slower by a factor as high as ten. An example in this regard was the implementation of use modules rather than argument lists to pass in-formation between routines; the use modules were preferred on the basis of modularity but were two times slower in execution. Thus, a practical com-promise was struck between the language and the implementation on current machines.