[sword-devel] Av11n and coverage

[Greg Hellings]

2012/1/6 Костя и Алёна Маслюк <kalemas at mail.ru>:
> While there is no support for module-supplied v11n in engine, no sense to
> making and testing modules. Test module for this task could be any
> incomplete module (without NT or OT).

And there you have the problem.
- The engine has no reason to support it if there is no material
(remember, we're a community of volunteers who all implement whatever
happens to be our pet peeve at the time - not a commercial entity
doing what is best for sales or marketing).
- Material creators have no reason to create materials if the engine
doesn't support it.

Eventually one or the other side will have to give in and do their
part of the work first. This is most likely to be a module creator who
gets their hands on an OSIS file with a completely custom module that
the engine developers don't feel like supporting with the current
av11n system. The current system is flexible enough, however, that
anyone creating a module could create the necessary header file and
submit that for inclusion in the engine rather than forcing someone to
implement a whole new feature in the engine. Ergo, there is not much
reason to push the support for now, until someone demonstrates a need
for it.

>
> As for me this task is not very difficult and well discussed to start
> working over implementation. But two things, implementation of av11n
> mappings affects on storage format for module-supplied v11n. Policy for
> custom v11ns should be strict because we can't allow countless amount of
> very similar v11ns.

The policy already is strict. But since the source is available,
anyone can create an av11n header for their needs, submit it along
with the module, and if the person in charge of module maintenance
thinks the new v11n is necessary, they will submit that to the engine
along with the new module. If they don't think so, they'll work with
the module developer to figure out an acceptable alternative. There is
nothing stopping you from creating such a module, testing it with your
custom v11n and submitting it as necessary.

Inter-v11n mapping is, at this point, moot since it doesn't exist.

>
> Returning to the subject, BibleTime do coverage calculation on testament
> level, but book coverage can be useful too. BibleTime Mini have continuous
> display and display stops if there is absent significant part of text like
> chapter or book, coverage ranges could solve this.