[sword-devel] Concerns about Alternate Versification

[Jonathan Morgan]

On Tue, Jan 6, 2009 at 3:28 PM, Greg Hellings <greg.hellings at gmail.com> wrote:
> On Tue, Jan 6, 2009 at 3:39 AM, Jonathan Morgan <jonmmorgan at gmail.com> wrote:
>> On Tue, Jan 6, 2009 at 3:05 PM, Chris Little <chrislit at crosswire.org> wrote:
>>> Basically, alternate versification support is phase 1. Mapping between
>>> versifications is phase 2.
>>
>> And I'm saying I don't believe it is ready for use without phase 2.
>
> And what I heard when you say that is that you don't believe Genbooks
> are good for Phase 1, however all your arguments are that Phase 2
> isn't implemented yet.  Genbooks are excellent for this type of thing
> - they already exist in SWORD, they allow for arbitrary key content,
> they allow for each Bible to have its own system of versification...
> which is the right way to implement a Bible.  Having any sort of
> programmer-dictated requirement for versification is inherently wrong.

They were two entirely unrelated arguments (and possibly unwise to put
into the same email), and I have withdrawn the Genbook one.

> Providing the mapping is, except perhaps for those versions that do
> adhere to the major "standards," a job that only the module creator
> can do.  Once that is implemented, all of your concern over whose
> reference this is and what it points to can run through a reference
> translator if the front-end developer desires or can be left in its
> provided form and placed into the user's preferred module - regardless
> of whether the results make sense.  That problem is, I believe, best
> left to the front-end developer to decide.  Getting module-creator
> preferred target module should come from the library, but whether or
> not to use that is up to the front-end.

I still say that there is a difference between a version specific
reference and a versification specific reference (even if it is
referring to the versification of a particular Bible), and from my
perspective it would be preferable to differentiate the two.  Just to
illustrate:

1. If I were highlighting or relying on a peculiarity of a particular
version, then I want a version specific reference (e.g. the ESV uses
this Greek text to show us meaning X in Genesis 29:22 (ESV)).  Mapping
to the user's preferred text doesn't make sense, since I am possibly
highlighting an unusual translation in the ESV that is unlikely to
occur in the user's preferred text.

2. If I had a commentary that was based on the KJV versification
because that was what was around when the commentary was written, then
I generally don't want the KJV, since it is commenting on the text in
that verse in the KJV versification, not the text in the KJV of that
version (sometimes it may be making KJV specific references, but
generally it won't be).

> From what I hear from all of your comments, once the inter-version
> mapping exists, your problems will be handled.  Whether that's
> implemented as an offset from our concept of "standard" or provided on
> a case-by-case basis by the module creators doesn't seem to be decided
> yet.

Yes.  I would like to know what the interface is, but was more
concerned by the fact that it wasn't being discussed after initial
sporadic discussion and I was concerned that it was not being
addressed.

Jon