[osis-core] osis_0106! "Throwing a fit!"

[Patrick Durusau]

Todd,

Todd Tillinghast wrote:

>I don't think that the way OSIS_0106 handles ID, IDREF, IDREFS, reference,
>referenceEnd, etc is well thought out enough to publish publicly yet.
>
How rude! ;-) Perhaps, perhaps not. I am willing to look at examples and 
use cases.

>
>1) They are all optional.  This will lead to inconsistent use of various
>purposes.
>
I am not sure how "optional" leads to inconsistent use. Assuming 
everyone (unlikely but possible) agrees that ID on verse milestones 
should be the traditional verse identifier, but some people use it, some 
don't (like Jon Bosak not including verse numbers in his Bible texts on 
the grounds that they can all be calculated), I don't see that as 
inconsistent use.

You CANNOT constrain the USE to which people will put ID or any other 
attribute. If the ID (or any other attribute) appears on an element, 
what use is made of it as recommended, or for some other purpose is 
beyond our control.

>
>2) I would like to put togeather some example cases that seem to be problems
>to me.
>
Would be interested to see those. May try to put together some examples 
here as well.

>
>3) I think there is a very simular and more rubust strategy to accomplish
>the same objective more clearly with better validation possibly with few
>attributes.
>

Possible, possible, although I caution against overloading and 
unnecessary validation. Recall that validation does not mean, "do it my 
way" but means that the document instance conforms to a set pattern of 
usage. Validation is a good bit looser than I think the sort of control 
 you want to exercise over the average user. You may indeed have a 
proposal that if understood and used properly would be better than 
whatever we eventually use. But that is not really the test, it is a 
question of what people can and will use, even if it is less than optimal.

>
>Can we hold off another day on OSIS_02.  I will attempt to produce some
>attractive cases and example fragments.
>

Doubts on the attractive cases, ;-) but yes, and I have meetings all day 
tomorrow so say sunrise on Friday! 20 paces with schemas drawn? ;-)

Note that we really need to get to 1.0 by Monday for a variety of reasons:

1. Troy and Chris need something stable to produce texts against for Rome

2. Steve and I need to write our demos for Rome

3. We need to move to the next stage ( good reason for finishing this one!)

4. We need to have time to get some documentation together for Rome

So, on the bounce folks!

Patrick


>
>Todd
>
>>-----Original Message-----
>>From: owner-osis-core@bibletechnologieswg.org
>>[mailto:owner-osis-core@bibletechnologieswg.org]On Behalf Of Patrick
>>Durusau
>>Sent: Tuesday, April 09, 2002 7:13 AM
>>To: osis-core
>>Subject: [osis-core] osis_0106!
>>
>>
>>Guys,
>>
>>Substantial changes (as promised!) in osis_0106!
>>
>>Getting just a little cleaner day by day. Assuming no one throws an
>>absolute fit, can we bump up the next revision to osis_02? (tomorrow)? I
>>plan on inserting a lot of example material and moving the current
>>revision description to a separate file that I will load with an include
>>statement. Can simply delete the include prior to the 1.0 release.
>>
>>Patrick
>>
>>--
>>Patrick Durusau
>>Director of Research and Development
>>Society of Biblical Literature
>>pdurusau@emory.edu
>>
>>

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
pdurusau@emory.edu