Re: names of vertical spacing dimensions

[David Kastrup]

David Kastrup <address@hidden> writes:

> Mark Polesky <address@hidden> writes:
>
>> David Kastrup wrote:
>>> The main problem I see with that naming scheme is that it
>>> does not reflect score sheet design, but the current
>>> implementation.
>>>
>>> [...]
>>> 
>>> So the proposed scheme ties something presented as document
>>> spacing parameters into internal details of their
>>> implementation.
>>
>> What would you propose to resolve that?
>
> I don't think I can propose something that would not move seriously into
> GLISS domain.  I don't see how one could sensibly manage this in a
> natural, designer-intuitive way without a spacing system that offers
> some sort of inheritance/fallback/hierarchy where you can get consistent
> design by specifying few parameters, but have an option to specify more
> specialized spacing independently/additionally and/or combine several
> simultaneously triggered spacing parameters (i.e., taking their
> maximum).
>
> The usual kind of document spacings fall into several kinds depending on
> a hierachy level.  If we say that a high hierarchy level corresponds to
> low letters, low hierarchy to later letters, you may have
>
> inter-b-spacing for b-b
>
> before-b-spacing for c-b, d-b, e-b
>
> after-b-spacing for b-c, b-d, b-e
>
> But after-a-spacing for a-b.
>
> I am not sure that this sort of pure hierarchy is good enough, or
> whether one needs some max/min scheme.

It is also obvious that we have pagebreak possibilities associated: c-b,
d-b, e-b are good breakpoints, b-c, b-d, b-e are awful breakpoints.

It is also obvious that titles have higher priorities than the material
they are titling, but some markup postscriptum to a score has _lower_
priority and should not be moved to a separate page.

With regard to sane document design, conflating all forms of markup
including titles is not going to lead to happy campers.

We need different categories for markups with different functionality
within the document.  On the plus side:

> The basic point is that for x different document element levels, we
> get along more or less with a hierarchy and 3*x settings rather than
> x^2 flat settings.

So my fear is that the new scheme is both strictly logical, and not
useful for specifying a coherent document layout.

-- 
David Kastrup