En/Je/On 2012-12-31 13:13, Thomas Harte escribió / skribis / wrote : > Would it be fair to describe Forth as the procedural analogue of Smalltalk?
I don't know Smalltalk. I think I read a comparation some time ago. AFAIK they internally work quite differently. > I'm thinking specifically about strict left-to-right evaluation, Right. Forth has no syntax. That's one of its strongest features. Words are executed in the order they are received (from a keyboard, a file, a serial line or whatever). > words having meaning only by definition Right. > and runtime components doing the things that are usually specialised > syntax like branches and loops. Words that create control structures are ordinary words in Forth. They compile lower-level words and branch information on the word currently compiled. Thus you can write your own special control structures on the fly if you need, anywhere in a Forth program, and use them right on the next word you create. In fact, actually you don't write a "Forth program" at all -- you write your own dialect of the language suitable for resolving your particular problem. I think the following online/downloadable books by Leo Brodie could be helpful in order to solve your doubts: The "Forth way" to solve problems (no Forth knowledge required, useful also for programmers of other languages): http://thinking-forth.sourceforge.net (The simple Forth "compiler" is caricatured on Figure 4-7, page 103, and explained on page 102). A good classic tutorial for beginners: http://forth.com/starting-forth/ Marcos -- http://programandala.net