I used to use Prolog a lot on R&D projects at SAIC but since becoming a consultant, few customers have wanted work done in Prolog (it has been about 5 years since anyone has paid me to do Prolog).
Prolog is awesome for some types of problems like graph operations, ad-hoc rule systems, planning, etc.
If you want to play/learn, I would recommend Swi-Prolog. Jan Wielemaker does a great job writing and maintaining Swi-Prolog. The semantic web libraries and other Swi-Prolog libraries are very good stuff!
Thanks for weighing in - I am a somewhat frequent visitor to your blog and websites. This type of discussion (eg, using uncommon or less well-known languages to be more productive or reason about an approach to a problem domain) comes up here on Hacker News frequently.
I have been searching for new work opportunities that would involve more freedom in programming language choice (specifically allowing me, as the developer, to explore tools like Prolog, Lisp, and Ruby as appropriate). Sadly, this has been somewhat fruitless in my geographic region. There is a (probably well-founded) desire on the part of companies to use what we can refer to as "lowest common denominator" tools (Java, C#, Oracle, MS SQL Server) with the occasional upstart using Ruby/Rails. When I say LCD tools, I don't mean to denigrate those languages/platforms, merely point out that companies seem more comfortable knowing that there is a large pool of potential programmers available with those skills.
PG has weighed in on the side of starting your own company and using more powerful languages as a competitive advantage (the whole blub/magic of Lisp set of essays). It seems that few companies see programming language as competitive advantage or (based on your comment) be interested in hiring consultants such as yourself to implement higher level software that leverages these uncommon languages.
I'm curious - have you had projects where you felt that Prolog or Lisp represented a "better" choice than Java or Ruby but had customer requirements that specified language choice upfront? Have you "pitched" Prolog or Lisp to customers, maybe saying something like "ah, using Prolog here would require only 20 hours of my dev/consultant time, but using Java will require an 80 hour solution"?
I'm quite interested in using Prolog, Scheme, or Lisp at the "day job" so hearing any of your thoughts might help better prepare my own internal pitches . . .
Hello Tom, first I should say that Ruby is my favorite language and fortunately there is a lot of demand for Ruby development. I don't really pitch customers on language since almost everyone I work for is very tech saavy and know what they want they want to use.
If you want to work in a particular language such as Common Lisp I suggest writing a useful open source project and that could lead to work in Lisp, and at the least you will learn something and have fun.
Prolog is awesome for some types of problems like graph operations, ad-hoc rule systems, planning, etc.
If you want to play/learn, I would recommend Swi-Prolog. Jan Wielemaker does a great job writing and maintaining Swi-Prolog. The semantic web libraries and other Swi-Prolog libraries are very good stuff!