[sword-devel] Chapter-centric browsing (was: iPhone frontend alpha screenshots)

[Jonathan Morgan]

On Wed, Nov 26, 2008 at 10:52 AM, Troy A. Griffitts
<scribe at crosswire.org> wrote:
> Finally back from ETS/SBL and staying in the States for a while.  Hope
> to devote quality/quantity time to SWORD stuff again starting in the
> very near future.
>
> a few comments on this thread:
>
> Frontends are welcome to show context however they want.  The engine
> doesn't favour chapter context-- as Ian has already mentioned.
>
> All things considered, I like chapter context.  I know the arguments...
> they aren't very good linguistic context containers.  Of course, I agree.
>
> But practically speaking, as others have mentioned, consistent context
> for a verse that is easily understood by users is valuable.  They should
> know the breaks are false, and if not, they should learn cuz we're not
> going to change the world for them.

Isn't that what computers are for? :)

> I like the idea of scrolling past (or before) a chapter break in a
> display window.
>
> Anyway, my opinions aside.  If there is a simple mechanism we can
> implement in the engine to facilitate frontend developers who wish to
> support other boundaries, then we should accommodate.
>
> I think something similar to:
>
> module.setKey("jn.3.16");
> module.setKey(module.getContextBegin());
> SWKey contextEnd = module.getContextEnd();
> do {
>  cout << module.RenderText();
>  module++;
> } while (module.getKey() != contextEnd && !module.Error());
>
>
> So, in conclusion, we would be proposing the addition of
> getContextBegin() and getContextEnd() on SWModule, which would return
> SWKey references in the module to where the author chose to break
> context (paragraphs, etc.).

This is really going to depend on the module how useful it is.  Any
VPL style Bibles are going to do very badly with this (not sure if we
have any such), since it won't be possible to get any useful content
out of them.  Similarly those with no paragraphs.

I suspect you would get better results if you had Bible-independent
sectioning as Eeli suggested, which would be used by this context
start and context end support.  This independent sectioning would only
be used if the Bible had no internal sectioning.  It would probably be
more work too, though.

Another thing to consider is commentaries.  This isn't hugely
important since most commentaries are displayed a verse at a time, but
ones like Matthew Henry's comment on sections, and do it with linked
verses, and if you were wanting to support such context with
commentaries (as the interface given above might suggest) then you
would want it to be in the sections the commentary was in.  It might
also be nice to indicate with the Bible which sections linked
commentaries apply to, maybe with lines down the side to indicate
extents or something similar (though this kind of structure only
really works well with VPL style layouts, since otherwise a line in
the margin could apply to only part of the text line, ...).

Jon