Comment by lispm | Hacker News Reader

Comment by lispm | Hacker News Reader

Parent Full thread

COBOL, Fortran, C++, Ada... were the scripting languages of what exactly?

If I were staying in your argument, then Lisp was the language of symbolic AI. The language where things like symbolic mathematics (Macsyma, Reduce, ...), planners, theorem provers, natural language understanding, reasoning systems, expert systems, knowledge-based systems (Cyc, ...) were explored and first tried to commercialize (mid 80s). Symbolic AI funding went away with the cold war ending...

Lisp was actually invented and designed for this. John McCarthy was designing Lisp as a vehicle for (his) AI research.

There might still be a lot of Lisp users. Even among AI researchers (and related domains) there are many using Lisp tools which you probably have never heard of - like ACL2 (a theorem prover), PVS (a theorem prover used at NASA), ACT-R (a cognitive architecture with more publications written about then you can ever read). The usage of Lisp may even be constant in these domains. There was an explosion of programming in general. Outside these domains Lisp never really participated in this explosion (with a few exceptions like Lisp in Autocad, Lisp in Emacs, ...).

But should it? Lisp's almost unbounded flexibility is obviously not needed by the average programmer. It is actually a something which makes a bad programmer worse.

'Lisp Machines' were just one development/deployment platform - 10000 sold in a decade. Lisp was used before Lisp Machines and after Lisp Machines.

> It never became the scripting language of something else that was exploding in popularity.

Autocad sold millions of copies and is still widely used today. Several Autocad 'clones' support the use of Lisp for CAD scripting.

"COBOL, Fortran, C++, Ada... were the scripting languages of what exactly?"

Isn't Ada the scripting language of the U.S. Department of Defense? As far as I know, most real world Ada usage is in DoD funded projects.

C++ won the battle over Objective-C to become C's successor for many applications, so fills a similar ecological niche.

First sentence from History section of Fortran Wikipedia page:

"In late 1953, John W. Backus submitted a proposal to his superiors at IBM to develop a more practical alternative to assembly language for programming their IBM 704 mainframe computer."

So Fortran was the scripting language of the IBM 704.

COBOL's Wikipedia page indicates it was championed by a committee of all the major computer industry players of its day, making it the Java of its time. So COBOL and Java may be exceptions to this "scripting language of..." rule, and instances of the "successful language created by a committee" rule. (Common Lisp, of course, being an example of "unsuccessful languages created by a committee".)

Ada is not a scripting language at all.

Ada is used as the implementation language of many real-time applications in several domains.

"Isn't Ada the scripting language of the U.S. Department of Defense?"

It's a language for writing airplane control systems that fail when you cross the international date line. ;)

I think you misunderstand scripting here. Scripting languages are interpreted, and can be quite a bit simpler to implement and code in than compiled languages.

Ada, C++, et al., are compiled languages.

The "scripting language of" argument is normally presented with the first example of C and UNIX - scripting is used here to describe the purpose it's being used for, not the nature of the language itself.

"Scripting languages are interpreted"

Uhm, no. There is no such implication.

COBOL and Fortran exploded with the IBM Mainframe and the AS400 midrange computing platform. Both platforms still exist in modern versions under different names now (IBM has changed their names so many times it ridiculous). LOTS of legacy COBOL still has our financial and insurance markets working smoothly.

> LOTS of legacy COBOL still has our financial and insurance markets working smoothly.

Is there any proof for such a statement? Has anyone seen business applications written in COBOL running on any thing then maybe a few mainframes in a few US agencies/corporations that haven't been replaced simply because they aren't broken (yet!)?

I know there's tons of Fortran and some PL/I code that will never be translated in any other language anytime soon because it's just good the way it is, but COBOL? I'm starting to think that this whole "COBOL will still be around in 100 years" is an urban myth...

If you would spend 2 minutes to do a google search for businesses still using COBOL you'll find your answer. My statement is based on professional experience in the IBM Midrange field talking with people who still work on these systems. Use this link as a jumping off point http://www.guardian.co.uk/technology/2009/apr/09/cobol-inter... (not much has changed in this space in the last 4 years, this space moves extremely slow when it comes to change at the systems level)

If you use a charge/credit card, are an electric customer, had prescription filled, an insurance claim adjudicated, etc.…, your transaction was processed through some COBOL code.

Yes, corporate system overlords would like to put these systems to rest, but that a task easier said than done, with a great deal of business intelligence buried in the code.

COBOL when business people needed computers but didn't want engineers.

Fortran when engineers wanted something nicer than ASM.

C++ when object oriented was taking off but people had already invested in C.

Ada for the military.

His argument still holds up for these languages.

If you call them 'scripting language' and accept that, then the idea of a 'scripting language' is useless.

Ada is the implementation language of all kinds of real-time software. It's not a scripting language at all.

C++ is also not a scripting language. Games for example have their core written in C++ and might have an added scripting language.

That a language is used in a domain, does not make it its 'scripting language'.

The term 'scripting language' is useless. If you are getting hung up on that term then you are missing the point.

Lisp is still the scripting language of Emacs. And I'm a Lisp user too! (well, Scheme, actually, but Clojure is tempting me back to the JVM).

I've always been curious about McCarthy's motivations and whether he was informed by Church's work on the lambda calculus. I heard he was schooled in functional analysis and symbolic differentiation and operator theory were what drove the invention of lisp, AI came later. I'm sure I could research it but I'm too lazy :)

That sounds like a new version.

McCarthy says in his 1960 paper about Lisp just in the first paragraph of the introduction:

> A programming system called LISP (for LISt Processor) has been developed for the IBM 704 computer by the Artificial Intelligence group at M.I.T. The system was designed to facilitate experiments with a proposed system called the Advice Taker, whereby a machine could be instructed to handle declarative as well as imperative sentences and could exhibit ``common sense'' in carrying out its instructions. The original proposal [1] for the Advice Taker was made in November 1958. The main requirement was a programming system for manipulating expressions representing formalized declarative and imperative sentences so that the Advice Taker system could make deductions.

So it was developed by an AI group for an AI software (Advice Taker).

well sure, McCarthy worked on that when at MIT, right? That AI group was also doing a computer algebra system, so symbolic differentiation was a current problem.

He is credited with the invention of Lisp though, and had a lot to do with Algol. I'm just curious when the connections between lisp and lambda calculus came about. Did it come later with the Scheme work at T at Yale or was it there also in the earlier work on Lisp.

I asked a former colleague about this, Jim Griesmer (who worked on getting the 704 lisp running on the 360 machines), and it was he who mentioned McCarthy's background in analysis.

EDIT: ha!, a little googling reveals[1] this story, not very definitive but interesting nonetheless. I should add that Jim Griesmer also told me that Minsky wrote his thesis on turing tapes or some such thing, because his advisors thought AI wasn't yet a developed field.

[1] http://brainbrain.blogspot.com/2006/06/dinner-with-john-mcca...

McCarthy himself said of his pre-Lisp research:

"To use functions as arguments, one needs a notation for functions, and it seemed natural to use the -notation of Church (1941). I didn't understand the rest of his book, so I wasn't tempted to try to implement his more general mechanism for defining functions. Church used higher order functionals instead of using conditional expressions. Conditional expressions are much more readily implemented on computers." - History of Lisp: http://goo.gl/EDuhl

I've always liked the fact the he did so much with a book he didn't understand the whole of. :-)

Reply on news.ycombinator.com