Re: TEXT: (equations)

["John E. Clifford" <pcliffje@CRL.COM>]

>From Graeme:
> Given certain restriction I do not see any real difficulty in being
> able to speak equations of any degree of complexity - assuming the
> reader is sufficiently numerate (algebra-ate?!).  I presume
> that before he got his computer and while he could still speak,
> Stephen Hawking had to do just that.  For most people though, normal
> algebraic notation is perhaps a mess because it is not sequential -
> for instance because of the use of brackets.  The few complex
> equations I *really* understand I don't read sequentially, but view
> as pictograms.
>
> Equations in lojban could be expressed in REVERSE POLISH notation, as
> used in some Hewlett-Packard calculators and the Forth programming
> language.  This does away for the need for parentheses.  For complex
> equations the information would need to be "chunked", in much the
> same way as over long sentences can be re-written as a number of
> shorter ones.  This by the way is good Forth programming style, which
> is in contrast to the multiply nested code often found in other
> programming languages.
>
> Is there anybody else out there interested in the parallels between
> lojban and Forth?  Or lojban and object oriented or declarative
> programming?

Lojban already has the facility for regular Polish notation, which has
all the parenthesis-less advantages of RPN plus the psychological
advantage of telling you what the hell is going on as the data comes in.
RPN is nice for calculators ONLY because it eliminates the need for an
"execute" command at the end of its function strings.  The branching
depth rule means that RPN formulae are usually unintelligible as read
(regular Polish are also less readable than infix but more than RPN,
when the parenthesis problem is factored in regular Polish comes close
to infix in most normal cases). pc>|83