* Russ Allbery wrote, in an earlier email:
> I'm not sure why you'd express it that way rather than just expressing it
> like this:
>
>     =extend N Y W B,I
>
> In other words, the first argument is the new code, and the subsequent
> arguments are the replacements to use in order of decreasing preference.

This I agree with. It seems more logical.

* Russ Allbery ([EMAIL PROTECTED]) [13 Oct 2002 13:39]:
> Sean M Burke <[EMAIL PROTECTED]> writes:
[...]
> > That's how I handle the fact that the universe of possible useful
> > extension-features is larger than A-Z (minus the letters already
> > used) that are available for Pod formatting codes.

> I understand the problem, but I can't say I really like that solution.
> It means that if you look at a piece of POD and it says Y<...>, you
> have no way of knowing what that means without looking for =extend
> commands.  Feels like action at a distance to me.  And worse, it may
> mean something entirely different than what Y<...> means somewhere
> else.

On the other hand, it allows the authors to do what they like.
The main problem is that of redundancy. e.g. all my Pod will end up
with something like:

=extend M module-name C,I

=extend N method-name C

> My understanding of the point of =extend was to allow us to add new
> interior sequences without breaking all the existing parsers, not to
> let people create their own mini-languages using POD syntax.

Hardly a mini-language. It's more a gracefully degrading syntax. I trust
Sean will have Pod::Simple complain if the =extend sequence doesn't
devolve into some universal Pod interior sequence (or the 0/1).

> The latter is kind of interesting, but I'm not sure it fits with the
> point of POD at all.

The point of POD being "to be simple"? I think it does. It will let me
write what I mean. I refer to modules and methods a reasonable bit, thus
module-name and method-name are appropriate names for me to have. I can
write myself tools to reference these as need be.

All I need to do is write myself a Post-It listing the available single
letter sequences.

It's not as if Pod is becoming pseudo-XML or anything. That's just the
representation used internally by Pod::Simple (and it has made my
writing of Pod::Simple::LaTeX very simple).


cheers,
-- 
Iain.

Reply via email to