[sword-devel] proposed patch: adding n=X marker content to footnotes and xrefs

[Greg Hellings]

I just wanted to reply to the following by Troy and also to his
comments in #xiphos regarding this topic

>BibleCS, per my preference, only shows hover-over symbols for the note or cross reference, and does not
>include the 'n' label. BibleCS uses different filters though, so not affected.
>
>SWORDWeb will be affected.
>
>My thought on this is similar to strongs. Don't show the numbers. They are left overs from the era of print-
>only.
>
>The exception use case is possibly our NET module which has a commentary accompaniment module
>which might refer to footnotes by label.  If I remember correctly.
>
>So, I would like the choice to not have the label for the frontends I maintain, but if users really want to see
>the label, I could be swayed on this point, wearing my frontend developer hat.

I see both this and previous discussions regarding module-provided
stylesheets as being symptomatic of the same question: Who owns
content and is responsible for its display? Troy, and others here,
seem to be of the opinion that what the frontend developer wants is
the be-all-end-all and that the wishes of the module creator can and
ought to be ignored completely.  If we want to attract major text
publishers, I'm not sure this is the attitude we ought to have.

What I have tried to propose before with CSS, and what Karl has
proposed here with footnote and cross-reference notes is that the
application ought to provide good defaults (generated cross-reference
and footnote markers when they are lacking) but to honor the module
creator's request when it is given. The n=X attribute on OSIS exists
exactly for the purpose of giving the module creator power over his
footnote markers. If our applications ignore that, then it would
properly be categorized as a shortfall in our support of OSIS. Far
from being a "feature" it is, actually, a bug. Publishers are likely
to feel the same in some circumstances.

Lack of support for n=X and for module-supplied styles was a major
reason for the LSDevLinux (portion of Wycliffe I work with) deciding
to use Xiphos instead of any of the other SWORD applications. They saw
that all the other applications completely ignored what they, as the
module creators, wanted to do with THEIR content and chose an
application that did not ignore their wishes. In #xiphos Troy cited
various online publications of Bible texts as being his validation for
ignoring the n=X attribute. While this is perfectly valid for a text
with footnotes restarted on each page (as is typical of print Bibles -
they tend to have footnotes per page and they restart at the beginning
of each new physical page or spread) it is not valid for texts where
endnotes or footnotes-per-chapter are used. In cases like this the
default behavior is supposed to be overridden by the n=X attribute
marker, allowing the content publisher to specify "This is footnote 1,
this is xref A, this is footnote 1b" etc. But Troy wants to completely
disregard that wish.  Why do you think you know better than the module
creator?

The purpose of the application ought to be to provide valid defaults
(generate markers when they are missing), to prevent display of
components which break functionality (e.g. large images on a small
form-factor screen), and possibly to give the user the option of
turning on and off behavior like this. It should not be to render
impotent the content developers. If we start doing that, we will end
up with the likes of WinMo IE from back in the 5/6 version area where
IE displayed nothing but basic HTML and was therefore ignored. Content
publishers couldn't make what they wanted and users couldn't choose to
enable or disable features they wanted. Let's not end up like that and
let's give our module creators the power we advertise (through OSIS
support).

--Greg