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

[Greg Hellings]

OK, we seem to be talking past each other. You are using Lockman and
the NASB as your example because of its prevalence all over the web.
I'm specifically not talking about Bibles. As I stated in my previous
email, most of those have footnotes which start over on every page,
even in print. If you get a basic NASB or NIV or other modern
translation with just translator's footnotes, all the versions from
every publisher I've ever used have footnotes, labeled with letters,
that begin anew on every page and ignore verse, chapter and even book
boundaries.

I'm talking about general books and commentaries and to that point...

On Sat, Feb 11, 2012 at 11:43 AM, Troy A. Griffitts
<scribe at crosswire.org> wrote:
> And I still disagree with you that module author wish to retain their own
> style (or should want to retain their own style).  I hear from module
> publisher that they simply want their content presented professionally and
> represents their intention for the content.  None have ever tried to dictate
> to me specific visual attributes they need attached to content.  Think of
> all the other attributes a print publisher attaches to content: 2 column,
> center-row footnotes, page content start and end summary lines at the top.
>  We write software where these things don't make sense.  I would tend to
> place footnote labels in this bucket.  I understand why others disagree with
> me.  I want to give us both the ability to display things how we think best.

I have about 950 modules (none Bibles) that disagree with your claim.
They're from all manner of publishers, but they all come through
Wycliffe. True, not all of them will be available for public
consumption due to copyright, but those to which Wycliffe is the
publishing holder are still potentially going to be released for
consumption.

When we were evaluating software, and I was getting my first real look
at SWORD development we lined up BibleTime and GnomeSword against one
another to compare their features - since we were only interested in
Linux and this was back in 2005. Key problems they had me tackle when
I first got there were:
1) Styling was not being preserved, even though a CSS file was
included in the ThML header and that same file styled an HTML file
properly - we handled this by using an XSL stylesheet to insert
style="blah" tags on every ThML element. Styling was extremely
important to them - they nearly abandoned SWORD to develop their own
library because they couldn't get SWORD to preserve their styling
until they moved to inline inclusion of a style tag. This styling is
not preserving print styles, but rather styles that are able to be
applied to the content in another display application (Logos/Libronix)
- and it is very necessary to proper display of the contents.

Some of it is as innocuous as font sizes and weights, others were
standardizations of titles (turquoise background, centered and italics
on by-lines and the like, borders). Still more of it was very
important for display - selection of different fonts for Greek vs
Hebrew vs Aramaic displays. OSIS provided no way to specify these
things and SWORD refused to support an external stylesheet. So we were
forced to use an HTML-based application that had a solid rendering
engine with good language support and poorly designed ThML modules
with inline style content. This was not a small issue and, at the
time, Gnomesword/Xiphos was the only option that had all of this
support and possibly BibleTime.

2) Footnote markers were not being preserved in BibleTime, no matter
the content type. This was a deal-breaker for them and they insisted
on using Gnomesword/Xiphos unless they create a custom patch that
allows BibleTime to honor the n=X setting for footnotes.

I've documented this before, but you continue to insist that no
publishers anywhere care about the display of their modules so long as
the content looks right. But I'm telling you - this is not an
academic, "What if..." or an "I prefer..." This is a real module
creator and SWORD software user who has demanded these features, and
is limiting their use to only Xiphos because it remains the only
application on Linux which supports both of these "killer features"
that they require. So you waving the Lockman example - which has
already been conceded on the basis of even print Bibles going by the
same motif - is a moot point. At the very least honoring the n=X (and
it might be in the on-hover box that it pops up, I dunno, not
necessarily on the inline marker) by default and allowing a user the
ability to disable them would be behaving "as advertised" on the OSIS
label. Claiming to support OSIS and then willfully behaving in a way
that the specification claims is not behaved amounts to a bug. We are
not following the spec we purport to follow.

--Greg