[pmwiki-users] Some experiments with whitespace indentation

[Patrick R. Michaud]

On Mon, Oct 17, 2005 at 03:38:52PM +1300, John Rankin wrote:
> On Monday, 17 October 2005 2:40 PM, Patrick R. Michaud <pmichaud at pobox.com> wrote:
> >> 4. you may wish to consider adjusting p.vspace within lists:
> >
> >My basic thought is that most people expect vertical spaces
> >between paragraphs to remain the same, regardless of whether
> >they're embedded in lists.  But perhaps it can be handled with
> >CSS:
> >
> >   li p.vspace { margin-top:1em; }
> >   p.vspace + li { margin-top:1.33em; }
>
> OTOH 'most people' aren't (and don't want or need to be) page
> layout experts. But an administrator may wish to put in place 
> rules which automatically represent the author's intentions in
> a readable manner. And if the rules change, then one css change
> can raise the quality the entire site.

Oh, I agree.  I'm just thinking that this sort of CSS styling
decision belongs to skin designers and not built-in to the markup
itself (except to the extent that we may need to render something
besides <p class='vspace'></p> to render vertical space in lists.

> >Or if that doesn't work we can see about adding some more tags
> >(<div>, <span> or something) to the continuation text to give it 
> >some more styling options.
> 
> I think it can be achieved with <p> tags around continuation 
> paragraphs. 

Yes, as long as we don't mind the linebreak that occurs when
someone writes:

     # This is a list item
       that is continued on the next line

> My feeling is that one ought to use <div> and <span> 
> only as a last resport after other possibilities have been 
> exhausted. If the semantics are correctly represented, the
> presentation will take care of itself.

Yes, but with <p>...</p> we're assuming that continuation
text is semantically a "paragraph", and often it is not.
I've often mentioned here that HTML's definition of <p>...</p>
makes it nearly impossible to correctly represent true 
paragraph semantics when mixed with lists, without that we're
guaranteed to get *something* wrong.

> <snip>
> >we need to be more concerned with things like:
> >
> ># an item
> > its continuation
> ># another item
> >  its continuation
> >
> >where an author might be surprised that "its continuation" isn't
> >treated as one here.  Perhaps we allow both the length of the
> >list markup and the length of markup with spaces to be used as
> >the indent levels for a continuation.  Or maybe anything in
> >between the two?  
> 
> Ideally, anything in between the two. But...
> 
> # An item
> 
> ## A sub item
> 
>   A continuation item
> 
> Is this A continuation item of An item or A sub item?
> 
> This was why I wndered if it should be based on the list markup
> length.

Excellent example -- clearly it's too ambiguous for us to
use multiple column positions for indent level.  So, I think
I'd fall back to wanting to indent to the same column as the
list item.  Otherwise it just doesn't visually align for me
at all (especially with several continuation paragraphs in
a non-nested list).

Pm