Web lists-archives.com

Re: text editors




Hi,

Nicholas Geovanis wrote:
> > > In other words, lisp and prolog (and clojure and guile and scheme....) give
> > > the "feeling" that "elegant code" can be the best software representation
> > > "in coding". Why? Because "the correctness is almost blatant", as stated. In
> > > other words, semantics collapses to syntax. As in mathematics.

I wrote (mainly towards the last sentence about math):
> > The attempt to create a language where semantical correctness results
> > from syntactical correctness was killed by Goedel's incompleteness
> > theorem.

> That is correct in the correct context [...]
> I used the phrase "collapses to". I shunned the word "attains".

So you hope for approximation of programming languages to a situation
where they enforce correct semantics ?
Well, my conclusion from model theory and Goedel was that you can either
have it certain or interesting, not both.

I became the C type of a programmer. Implementing straight ahead. Always
trusting the ability to solve the detail problems. Bad hair days included.
gdb and valgrind are my friends.

I cannot see much clarity of expression in Lisp or C++. Java seemed better
until i got in contact with the spaghetti inheritence of its class libraries.
In general i object the idea of separating model from implementation and
then hiding the latter.


John Hasler wrote:
> I think that you confound soundness and completeness.

The liar paradox is not sound. It demonstrates the fundamental risk
that a fine looking syntactical contraption turns out to be utter nonsense
with undefined consequences. (What Goedel did was to link that kind of
paradox to the properties of natural numbers and to math being performable
by countable many finite sentences composed from a countable alphabet.
So if you want numbers and math books, you also get Epimenides, the most
credible of all cretians.)


Have a nice day :)

Thomas