[sword-devel] WEB update request

Kahunapule Michael P. Johnson Kahunapule at mpj.cx
Wed Aug 11 02:36:34 MST 2004


At 16:41 11-08-04, Chris Little wrote:
>GBF doesn't offer anything that isn't offered by OSIS.

This is not true. GBF has two advantages over pure, unmodified OSIS:
1. GBF can preserve quotation punctuation correctly. Pure unmodified OSIS cannot.
2. GBF is simpler to convert to and generate than OSIS when a Bible text doesn't need any features that GBF doesn't support. In addition, an open source GBF to modified OSIS converter is available.

Advantage 1 is very significant, but the OSIS standard could easily be corrected to eliminate this defect.
Advantage 2 is actually an argument against supporting GBF and for either supporting modified OSIS or modifying OSIS and the GBF to OSIS converter such that all punctuation is preserved correctly. It does, however, mean that GBF may have a use even after OSIS is made to be acceptable.

OSIS adds all kinds of capabilities that GBF lacks, BUT I'm not willing to give up punctuation mark integrity to gain them.

In other words, OSIS can't replace GBF, yet-- at least not pure, unmodified OSIS. A very slightly modified OSIS could.

>  I don't see a 
>reason to encourage use of multiple formats.

I don't, either, unless one in use is defective, and I'm trying to replace it with a better one, or unless the different formats are better for different uses. OSIS is not and will never be the ONLY Scripture file format standard in use. Even in Sword, you won't exactly use it directly. You will import it into a module format that is more efficient for searching and displaying. The JAARS Translation Editor won't work with OSIS directly. It will work with a database format, and export and import to OSIS.

There is no way SIL will be weaned from SFM to OSIS in any short period of time. There is just too much existing data and existing applications that would be rendered useless by such a conversion. It is challenging enough just to get people on the same page with USFM, using the same markers to mean the same things. Even converting between USFM and OSIS is necessarily lossy at this point, and really can't be done right even for the simple stuff, like quotation marks.

There are many ways that OSIS COULD be made acceptable. They just haven't been implemented. Or have I missed something? Until at least one of them is implemented, I'll consider it a bad thing to follow the OSIS standard exactly.

What is the problem with extending OSIS to allow quotation punctuation to be coded in the text, and <q> to allow the attribute type="x-doNotGeneratePunctuation" to be added to it? I REALLY don't see what the problem is with that, other than the fact that it is overly verbose (but that is easy to fix). Could you explain what the problem is?

I'm still waiting to be convinced that OSIS can handle punctuation quotation correctly in all English versions of the Holy Bible and in a variety of languages with different characters and punctuation rules pertaining to quotations.

Have you considered the idea that I might just have a valid point?



More information about the sword-devel mailing list