[sword-devel] summary

Chris Little chrislit at crosswire.org
Fri Nov 28 23:42:21 MST 2008



Matthew Talbert wrote:
> But these issues are affecting me. I started working on Easton's as
> practice for me. I first was going to try TEI as that was what was on
> the wiki and appeared to be recommended. However, there is scarce
> documentation, and the xsd is of little value to me. I understand what
> it is, but it is quite overwhelming, and in my opinion meant to be
> read by a computer, not a human.

The schema is definitely intended to be read by a computer, not a human. 
  To learn how to write TEI, you basically have to learn the same way 
everyone else did: online documentation. We have a fair start on that in 
our Wiki, and I'm working to fill out some additional info (especially 
CrossWire-specific details like standard rend values). But the best 
resource is and shall likely remain the TEI P5 documentation. In 
particular, I would recommend that those wanting to delve into TEI 
dictionaries begin at the <entryFree> page:

http://www.tei-c.org/release/doc/tei-p5-doc/html/ref-entryFree.html

CrossWire specific matters, like how to incorporate osisRefs into TEI 
docs are covered by the Wiki.

Aside from that, there are sample docs that I've posted links to on this 
list. And the TEI P5 docs have examples for virtually everything you 
could think of.

 > Also, I wasn't sure which part of the
> TEI specification applied and what didn't.

Almost all of it applies. If anything additional is needed, I will add 
it to the schema, just let me know.

> Furthermore, I couldn't
> find any modules, or examples that used it and I was told on IRC that
> there weren't any.

Webster1913 has been available (as a module in beta as well as a TEI doc 
that I linked to) since June. The new StrongsGreek, StrongsHebrew, and 
Strong (which combines the two and uses a different keying system) have 
been posted in the last few days (both as modules and a TEI doc of the 
latter).

> Beyond that, one of the main concepts that I was
> trying to implement was self-referencing links, and again I was told
> on IRC (both times on the #sword channel) that TEI and OSIS support
> wasn't finished yet.

Indeed, we still have plenty of work to do on both OSIS and, especially, 
TEI support. But, as has been pointed out here before, linking is 
essentially a front-end issue. All the same, I guess there's nothing 
wrong with pushing front-end processing back into the library (e.g. 
translating references to sword:// URIs in the HTMLHREF filters), 
assuming no one else using those filters has an objection.

--Chris



More information about the sword-devel mailing list