Talk:Source code
|
|
Well, hmm. Several issues are raised by the current version of this page:
- Software can be written in a mixture of programming languages,
- Program execution doesn't necessarily require compilation (interpreted languages, etc.),
- An assembler doesn't turn object code into machine code, it turns assembly code into object code. The latter step (object code to machine code) is typically handled by a linker,
- Machine language can be turned back into source code. This is called disassembly.
However, many (most?) interpreted languages compile the code anyways into either a byte code or native code. Smalltalk does this.
The reusability section should talk about maintainability. And it should be noted that people who believe in maintainability (eg, Smalltalk people) believe this to be the primary use of source code. Perhaps a quote to Donald Knuth saying as much is in order that needs to be specified. test
I've been working on rewriting some of this article, especially the first few paragraphs in the current version. Could some of the stuff about mnemonics and reusability be removed, or at least shortened a bit? A lot of this is covered in better detail in the programming language article, and its related pages, and I'd hate to duplicate effort that would be better spent on other articles more suited to these topics. Seems to me that the source code article would be better suited to discussing topics specifically related to source code, such as the legal issues surrounding it, and a discussion of how a bunch of source code becomes a running program, and not go into general programming constructs and concepts. Anyone else have ideas? -- Wapcaplet