* Allison Randal <[EMAIL PROTECTED]> [2008-10-26T15:33:16]
> I'd say you already know where the greatest challenges will lie.

That's good to hear.  I hate thinking that I'm 80% done, only to have someone
point out that I totally missed the major problem and that my design is never
going to survive the next 15%.

> - You're not currently paying any attention to the content of the X<> 
> tags. Eventually, someone will want to make a smart formatter for L<> 
> link tags for HTML, PDF, XML, etc, and will need more intelligence in 
> handling these tags. But, you can punt for a long time just by handing 
> the formatter a raw blob of text and letting the formatter parse it. 

Yeah, I'm totally treating all textual paragraphs (text, verbatim, data) like
verbatim paragraphs.  I -think- this is safe, because I don't think this is
ever considered legal:

  =begin møøse

  ...

  =end mE<oslash>E<oslash>se

After I do a bit more work reading and digesting perlpodspec, I hope I can help
make it more explicit in a few cases, especially to make it clear that
formatting codes never affect understanding paragraph syntax.

> - Sooner or later, people will want to use Pod::Elemental or 
> Pod::Eventual as a drop-in replacement for Pod::Parser or Pod::Simple. 
> (A headache, but a worthwhile one.) That's when you'll need perlpodspec.

I'm hoping that I can somehow manage to keep any form of formatting code
parsing on the n+1 tier, so that the product would go:

  pod2sgml is built on
  Pod::Exegete is built on
  Pod::Elemental is built on
  Pod::Eventual is built on
  rock and roll

Actually, I'm hoping to never need to deal with formatting codes, but L</epic
flying ponies>.

-- 
rjbs

Reply via email to