[boost] Re: [MPL] Type detection

From: David B. Held (dheld_at_[hidden])
Date: 2003-10-26 01:32:49

"Joel de Guzman" <joel_at_[hidden]> wrote in message
news:003101c39b80$ff8c9bb0$0100a8c0_at_godzilla...
> [...]
> It would be very difficult to debug because of the side-effects.

Could you give a concrete example?

> I'm quite happy with the *functional* nature of the template
> processor, thank you. An imperative template processor will
> open up a can of worms. I don't even want to start imagining
> a step debugger at compile time! Yaiks! No, please don't!

I'm not talking about a template processor.  I'm talking about
a language which supports multiple levels of abstraction
seamlessly.  I'm talking about what you might immodestly
call a "5GL". ;>  And don't just imagine a compile-time debugger.
Imagine a meta-compile-time debugger, a meta-meta-compile-
time debugger, etc.  Imagine a language in which the source
code and compiler form the execution environment of the
meta-compiler.  Imagine what possibilities for reflection and
code generation are possible.  Yes, it's enough to stretch one's
mind quite to goo.  But if we had real meta-exceptions, instead
of mere compiler diagnostics, and a real meta-i/o facility, then
perhaps the idea of a meta-debugger is not so scary or far-
fetched.

> IMVHO, there's nothing wrong with the functional nature of
> the template processor.

No, there's not.  It has served us well, and will continue to do
so.

> Why make it imperative when everyone knows that
> functional is better? C'mon guys! This is a step backwards!

I disagree.  This time, given the choice between A and B, I
choose 'yes'. ;)  I would want this hypothetical meta-language
to fully support FP with native lambda, closures, etc. (maybe
even monads ;).  And not just at run-time, either.  Think meta-
OOP...building up types and meta-types as if they were values.
The person who should be having nightmares is not the
programmer, but the compiler implementor!

Dave

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.521 / Virus Database: 319 - Release Date: 9/23/2003