[jsword-devel] Hello!

[DM Smith]

On 01/17/2012 01:27 AM, Ryan Meyer wrote:
> Hey guys,
>
> I was checking out the JSword project and I got all the code, got it
> to run, ran some unit tests and took a look around.  I was curious to
> know how active the project is?  I didn't see a lot of current issues
> in Jira.  Is there a need for help in any of the projects currently?
>
>
Welcome Ryan!

JSword is a library used by front-ends.

There is a little activity in its development:
o Previously, I worked on simplifying the effort to internationalize the 
end user text.
o There is significant effort in internationalizing this text. This is 
largely due to the tremendous acceptance of AndBible for Android.
o Currently, I'm in the process of adding support for Alternate 
Versification (Bibles that have a different order of books, different 
chapter and verse numbers, ...). This change will be disruptive until it 
is near done.

The list in Jira is old, but it is still current. There have been very 
few new requests. We just have a lot of old requests to work on.

We have a few frontends to JSword: Bible Desktop, Alkitab, FireBible, 
AndBible and STEP.

I'd suggest that the best place to start is pick a frontend, figure out 
what you'd like to see improved. Or go through the Jira list for JSword, 
BibleDesktop or STEP and find something that interests you. After 
discussion here, give it a shot.

A few thing I see as important in the near term:

One shortcoming of the JSword design that is impacting STEP is that it 
was designed to be used by a desktop application. In the context of a 
multi-user server, its reuse/caching gets in the way.

Another problem we need to fix is the parsing of non-english input for 
the Biblical reference engine. Specifically, JSword reserves '-' 
(hyphen) as a range separator. There are languages where it can be part 
of the book name.

For the next release we need to solve the invalidating of search 
indexes. The fundamental mechanism is present, but it is not nice to the 
end user. Basically, each time a bug is fixed in JSword's search engine 
(Lucene) or JSword uses it differently or Java is upgraded to have a 
different Unicode support, end-user indexes need to be rebuilt.

Our JUnit test cases could use improvement. We should have one for each 
class. This would be a good way to understand the internals.

Hope this helps.

In Him,
     DM