[Openmcl-devel] ccl manual (was Re: trace on recursive functions)

[Ron Garret]

On Dec 14, 2009, at 1:13 PM, Philippe Sismondi wrote:

> Ron - does the usefulness of Lisp really imply that one should re-do markup languages using it? Perhaps SGML and XML are ugly and sub-optimal. So, I guess, is HTML. But there is a wealth of new problems to solve (with Lisp!) without rolling our own that area.
> 
> I'm a recent convert to Lisp. I think it will solve lots of my problems, once I am reasonably adept with it. But not ALL of my problems. I am willing to bite the bullet and go with the crowd on some things, like markup.
> 
> It's near xmas - maybe Erik Naggum (rest in peace, Erik) will appear to chastise me for my idiocy. But I'm ok with a certain amount of cruft, if it sorta works and I can move on to other things.

I am certainly no fan of reinventing the wheel for the sake of reinventing the wheel.  But markup is a special case, for two reasons.  First, SGML/XML/HTML *are* S-expressions, only with a more cumbersome syntax.  There is a formal equivalence between:

<foo>...</foo>

and

(foo ...)

So it's trivial to go back and forth between the two.

Second, if there were an existing markup language that everyone agreed on was the Right Thing in this case then I'd say go ahead and just use it.  (When I wrote my wiki -- and yes, I wrote a wiki, which is why I know how hard it is to write a wiki -- I used Markdown for exactly that reason.)  But in this case there isn't.  Some people like this, other people like that.  But there is one thing that everyone here should be able to agree on and that is that S-expressions are cool.  That, plus the fact that there is a good impedance match here, to me means that this is as good a problem as any to collectively tackle.  If not this, then what?

rg