Subject: Re: Is there a useful distinction between "programming" and "scripting" languages? From: Erik Naggum <erik@naggum.no> Date: 19 Sep 2002 04:16:53 +0000 Newsgroups: comp.lang.lisp Message-ID: <3241397813684563@naggum.no> * Erann Gat | That makes a language's status dependent on the context in which it is used. | Lisp running on Unix would be a scripting language on that definition (but | not Lisp running on a Lisp machine). This just goes to show how differently we conceptualize. I do not see a runtime system as an application in itself. If, however, the application merely runs inside the Lisp system which is also an application from the operating system's point of view, the question becomes whether the Lisp application does the same as the LIsp system. E.g., Emacs Lisp usually only enhance the operation of Emacs and relies heavily on the Emacs "machine" to do its work. Therefore, Emacs Lisp is a scripting language with programming language capabilities, but its programs could not run without Emacs running first, and Emacs is a stand-alone application. A Common Lisp system may only provide the environment for a Common Lisp program but would not do anything useful by itself. In a sense, it has the same capacity as a shell to launch applications to run "in" the system, but the same way that the interactive loop acts as a job control language the way the shell does for Unix programs, the applications launched from the Common Lisp REPL are as distinct from the REPL as Unix programs are from the shell. To refine the distinction I see, a programming language has to assume control over the execution of the application and terminate execution when it returns, whereas a scripting language is subject to the control exercised by another application, which does not necessarily terminate when the script terminates. Now, I believe there is a distinction from some deep-rooted sense of there being a difference in the amount of machinery needed to support a language in addition to the notion of an intention of the language to be compiled into native machine code. Another point of view is the intended efficiency of the resulting code. Scripting languages are not expected to be compiled into native code that assumes full control over the machine for an extended period of time and only calls upon the support system, but rather the reverse: a script is run only occasionally as part of another application that has such full control. However, I am sure there are people who consider Common Lisp to be a "glue language" when they observe that low-level functionality is implemented in C and the higher-level application is written in Common Lisp. -- Erik Naggum, Oslo, Norway Act from reason, and failure makes you rethink and study harder. Act from faith, and failure makes you blame someone and push harder.