Talk:Fifth-generation programming language

Are we sure of this entry? I've never heard of fifth generation languages being defined as constraint solving languages.

You right: 5GL will change the world, so we will cannot miss that. ;-) Vlisivka (talk) 11:26, 14 April 2010 (UTC)

I have. There's the Japanese fifth generation computer systems project, this link here, FOLDOC says this, and the ALP mentiones it. Also, it's defined here as "sometimes considered to be a 4GL with a knowledge-based system built in". I've also heard it mentioned in the Prolog community and I believe in the book Principles of Constraint Programming by Apt. However, I have also seen it defined as a language that uses a graphical design interface (like VB or Delphi) but I've only seen it mentioned as this once here. - DNewhall
 * The last definition you provided doesn't make sense and the contributor is a Brian Henry... who is this guy....--Slamcool (talk) 18:40, 4 March 2008 (UTC)
 * Why aren't any of these references in the article? From searching google, it seems like this is an infrequently used term that largely died in the 80s. Mbelisle 05:43, 2 July 2007 (UTC)

Yes, there are constraints and then other strategic methods that can be applied. This page needs to be further developed. jmswtlk 22:31, 9 November 2006 (UTC)

A few of the programs that are referred to as 5th generation were released before those in the 3rd generation, is their any references that show how to properly place a language based on features over when it was release or is it entirely the concepts/paradigm it supports. WikipedianYknOK 21:52, 18 July 2007 (UTC)

This page is completely bogus. It needs to be killed. John Nowak (talk) 01:35, 21 February 2008 (UTC)

This page should be removed as it is a complete overlap with the Declarative programming page. I am not even sure that there is such a concept as 5GL. Tarik April 27, 2010

This entry definitely belongs here; this terminology was influential if not exactly common in the 1980s and early 1990s. Rdviii (talk) 03:18, 10 June 2019 (UTC)
 * Firstly, cite your sources for it being influential. How does a term that gained almost no usage count as influential?  What did it influence? Secondly, even if it had some limited value in the 80s, it doesn't now. The page should reflect that, not pretend the term currently has validity.

Ref. 4 no longer links to the original source. Rdviii (talk) 03:18, 10 June 2019 (UTC)

This is a vacuous page, making claims for a poorly defined category that never gained wide buy-in or usage. If it is to remain at all, it needs serious updating to frame it in an honest description of the history and low credibility of the concept. Itsbruce (talk) 08:48, 4 November 2019 (UTC)

Programming language generations
This group of articles has problems. Please see current discussion at Talk:Second-generation programming language. Trevor Hanson 20:12, 7 October 2007 (UTC)

The problem with "generations" of programming languages is that they are recognized in hindsight; We look *back* and we see which ideas and which shifts in thinking were the important things that marked transitions from one generation to another. 5GL (and to some extent 4GL) are attempts to hopscotch this process by looking *forward* and designing languages around ideas we think or hope will be important and mark a transition from one generation to another -- but there are a hundred wrong or unimportant or unhelpful ideas, or ideas that actually increase complexity without helping people manage it, for every good one.

It's a bunch of scientists trying to predict what the next big scientific breakthrough in their field is going to be. The problem is that everyone *wants* it to be the idea they're working on, and many are working on the idea they're working on *because* they believe it will yield the next major breakthrough. Anyway, the terms are loaded and suffer from the "insert a higher number to get better marketing hype" problem, which means everybody with an axe to grind or a product to sell is going to try to claim it.

Most of what Wikipedia's current set of articles refers to as "4GLs" are in the industry called DSLs, for "Domain-Specific Languages." The current article's insistence on constraint solving languages as THE fifth generation reminds me of hype from the middle '80s touting database-connected languages as THE fourth generation. Which, if you recall, turned out to be a good idea but not the only good idea, and is not now considered the defining characteristic of 4GLs.

I say this as a programming language geek. I am a person with a CS degree. I have studied programming language design, implemented a programming language, and have been the facilitator for the IEEE committee for standardizing a programming language. I actually have credentials in this area.

Zrebbesh (talk) 18:26, 23 September 2009 (UTC)

Standard Definition
Prolog is not 5GL. Prolog can be written in Lisp and Lisp can be written in Prolog, they are both 4GL. Prolog interpreters are commonly written in a 3GL and are therefore 4GL. Prolog is similar to SQL which is a 4GL. Anything that parses a language into 4GL code is 5GL. This new 'graphical' definition is fine if the icons generate 4GL code, but the original goal of 5GL was natural language parsing, i.e. artificial intelligence that talks in English. It is unknown whether natural language is a 5GL.

An approach to 5GL is parsing sentences into SQL.

INPUT : "What is the capital of Italy?" SQL: "select city from capitals where country = "Italy""

RESULT: [City] "Rome"

OUTPUT: "City is Rome"

~ Herc (BInfTech) —Preceding unsigned comment added by 124.180.120.188 (talk) 13:35, 23 October 2009 (UTC)

There is such a system: MASQUE/SQL - written in Prolog. As for whether or not one language can be written in another, that is more of a question of Turing completeness: any language can be written in any other language that is Turing complete. C can be written in assembler, and assembler can be written in C. It does not mean C is a 2GL. Markdwilson (talk) 13:15, 23 April 2011 (UTC)

The page does not list a date range that defines the fifth-generation. The other generations are all list them (for example, 4G as 1970s-1990s). — Preceding unsigned comment added by 98.210.171.14 (talk) 21:59, 13 January 2012 (UTC)

Claims to meet 5GL-requirements
This is my first contribution so I decided not to edit the page myself... Anyhow, it's about the weasel words "A number[who?] of software vendors currently claim that their software meets the visual "programming" requirements of the 5GL concept." Looks like one of them is WorkXpress claiming that their product (of the same name) is "the first complete 5th generation programming language tool" TalkmasterGIP (talk) 08:24, 9 August 2012 (UTC)

Verilog is a 5GL
Verilog is a 5GL 70.173.148.152 (talk) 03:09, 23 December 2022 (UTC)

chatgpt
could chatgpt's code generative and review iterative process be considered 5GL?

24.186.215.100 (talk) 17:03, 9 January 2023 (UTC) QuakePhil


 * Agreed, LLMs could very well be. I don't think we are there yet, issues remain around reliability and reproducability. In any case, until scholars and so on report such things in peer reviewed papers and the like that we can cite, this would just be an opinion/ original research, so we can't add it to the article. If there are news articles where chatgpt is discussed as a 5GL candidate, that could be added, though. 139.61.117.252 (talk) 14:09, 29 June 2023 (UTC)