https://bugs.documentfoundation.org/show_bug.cgi?id=160686

--- Comment #46 from Eyal Rozenberg <[email protected]> ---
(In reply to ajlittoz from comment #45)
> Stumping on this bug report totally unexpectedly, but since discussion is
> very long I didn't read everything.

I'm afraid perhaps the long argument with David makes readers lose track of the
main point made already in the opening comment... the consideration of page
position relative to a book spine is a mechanism intrinsic to a _single_ page
style, and does not need a pair of completely separate page styles to handle.

> I wonder if Eyal's complaints does not come from its RTL context (mentioned
> in one of the early comments). I think this is the core of the problem.

No, that's not the core, it's peripheral; I've repeated the core of the problem
above. It's just when also considering RTL, the conceptual problems with the
"Left Page" and "Right Page" styles become more glaring, and the confusion of 
users with their availability is more serious (especially when it is confounded
by the atrocious semantic difference of Page Styles from other styles, in that
when you format a sequence of pages, you change the named style rather than
applying a DF; see bug 161078.)

> In the common LTR context of most of us, pages follow a sequence right(1),
> left(2), right(3), left(4),… We usually start our chapters on a right page.
> This is dictated by our reading order and binding side (left when book is
> closed)

You are not defining what a "Left" or "Right" page means... so, let me assume
you mean pages which, when the book is open to them, are located to the
left-of-spine and right-of-spine respectively.

Even with this formulation - your claim is often not true. The first page in a
document may be either a left-of-spine or a right-of-spine: 

* The author may intend for the left page to correspond to the book's front
cover (so, right-of-spine); 
* but it could also be the printing on the back-side of the cover
(left-of-spine); 
* or a page that's glued to the back-side of the cover (left-of-spine); 
* or the page after that one (right-of-spine); 
* or if that's considered necessarily empty, then the next page after that
(left-of-spine).

But again, that is a minor point and not the core of the bug.

> Alternation left-right makes sense in both cases; you simply start
> differently.

If the pages are contiguous when bound, this alternation makes sense - but not
an alternation of page _styles_; just an alternation of the side on which the
gutter space is added.

> BUT, left pages are odd-numbered in LTR

No, left-of-spine pages are not necessarily odd-numbered, because:

1. Pages may not be bound contiguously when printed;
2. Pages are not necessarily numbered at all;
3. The numbering of pages may not begin with the first page;
4. The number of pages may not begin with page number 1

Note also, that if you were to make this statement about "Left Page"-style
pages, then there would also be:

5. In LibreOffice Page numbering is not an aspect of the Page Style.

> Writer must be informed about the cultural context of the document.

Now, that might be an interesting avenue to pursue. But whether it is or it
isn't - we should still remove the "Left Page" and "Right Page" styles, so that
users either edit the default style to introduce an alternating-gutter-position
layout, or we could have a _single_ style named "Book Page" or
"Alternating-Gutter Page" etc., with this feature.

> The simplest tempoary workaround is an option to remove side-parity 
> constraint.

If I interpret your suggestion correctly - I would tweak it so as to apply it
to single page styles, it would be a suggestion that the Page Style dialog,
Page pane, would offer an explicit choice of whether the first page is to the
right-of-spine or to the left-of-spine (if there is a spine and a gutter
defined). And, again, it wouldn't resolve the bug. But I might have
misunderstood the suggestion.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to