[osis-core] Last changes! I Promise!

[Patrick Durusau]

Todd,

Todd Tillinghast wrote:

>Can you explain what is meant TRW and TR?
>
TR is Textus Receptus (? Chris, correct my spelling ?)

and TRW is Textus Receptus Westcott (a particular editon of the Textus 
Receptus)


>
>If there are multiple nameing systems that could be put in the same
>attribute in <w> then we need a way to indicate which naming system we mean.
>If TR is intended as a "naming system" the we should adopt a convention to
>separate the nameing system from the "name".  Something like "TR:12345"
>would seem natural but ":" is not allowed in xs:ID.  What you think about
>using "TR..12345" in the same form as "Bible.KJV..John.1.1".  This would
>give us "strong..1234" for strongs numbers or "strong.extended..9934".
>
You get the naming system for sync since it is referenceType and 
constrained by the rather long regex in OSISScripture.

I understand (or think I do) how separation of naming system and 
references work in general. What I don't see is the payoff, other than 
orderliness, from the proposed regex? If we had a more comprehensive 
proposal, like I expect in the scholarly analysis module, that says a 
Westcott-Hort reference is: Westcott-Hort..followed-by-a-regex, and 
somewhere else there is a mapping of that reference to references in 
other works, so that knowing the name has some defined meaning, then I 
grant that separating in some standard way the name from the reference 
has meaning. Because by knowing the name (note Adam gets to name 
everything and in the boat of Ra (Egyptian religion) the deceased 
addresses everything by proper name) there is a consequence of knowing a 
range of valid references, how those references map to references, etc. 
In other words, it is no longer simply knowing where a nane stops and 
where the reference begins, but you also know something simply because 
you know the name. In the intstant case (IMO) all we know is one is the 
name and one is the reference, which without more, is just a curiousity 
and not useful information.

Steve: If you are online, care to chime in on this? I am ready to 
declare what I sent out earlier the 1.0. Not perfect but I think a very 
worthy effort.

I am running to the store to pickup all my food for the next week or so. 
Back online around 5 PM Eastern time.


Patrick


>
>So I am proposing that we create a SimpleType that looks like
>
><xs:simpleType name="xyz">
>	<xs:restriction base="xs:string">
>		<xs:pattern value="(\w(\.[\w]+){0,}))\.\.(\w(\.[\w]+){0,}))"/>
>	</xs:restriction>
></xs:simpleType>
>
>and use it for the attributes in <w> AND for target.
>
>(I don't think that attributeExtension will work for POS, morph, and lemma.)
>
>Todd
>
>>-----Original Message-----
>>From: owner-osis-core@bibletechnologieswg.org
>>[mailto:owner-osis-core@bibletechnologieswg.org]On Behalf Of Scribe
>>Sent: Sunday, April 21, 2002 3:00 AM
>>To: osis-core
>>Subject: Re: [osis-core] Last changes! I Promise!
>>
>>
>>Thanks tons guys for the accomodations!  I have one more issue to
>>clear up
>>a little-- #2 below.
>>
>>Although a gloss sound kindof like what we are doing, it doesn't quite
>>fit.  To boil down the example:
>>
>>TR:
>><w ID="TRW12345">en</w> <w ID="TRW12346">arXA</w>...
>>
>>KJV:
>><w sync="Bible.TR..John.1.1.TRW12345">In</w> <w
>>sync="Bible.TR..John.1.1.TRW12346">the beginning</w>...
>>
>>
>>This is for what I'm looking.  Something that lets me sync 2 translations
>>together, for things like interlinear display, automatically generating
>>english glosses for the TR, etc.  We discussed this type of syncing
>>mechanism in Dallas, and I was under the impression that various
>>attribute
>>names during the lifetime of the OSIS incarnation process were for this.
>>I'm happy with target, sync, or whatever.  Also happy with freeform
>>sync="TRW12345", rather than a validating reference.  Doesn't matter to
>>me, but would like some mechanism to achieve this.
>>
>>Thanks again for your consideration.
>>
>>	-Troy.
>>
>>
>>
>>On Sat, 20 Apr 2002, Patrick Durusau wrote:
>>
>>>Guys,
>>>
>>>1. On the issue of translation names, Steve and I found while on the
>>>phone a fairly good listing at:
>>>http://geocities.com/bible_translation/english.htm. And, Chris has sent
>>>in some more. I will be adding those, probably early in the morning.
>>>
>>>2. Target references: Steve suggests that listing Strongs number for
>>>lemma, POS (part of speech), etc., is perfectly acceptable. I
>>>
>>guess I am
>>
>>>more accustomed to seeing an actual value there and not a reference to
>>>something else. So, gloss, which is what you are doing, glossing the
>>>translated word with the original, is the place to put the synch. (In
>>>other words, no need to modify the schema.)
>>>
>>>3. Content model question: phrase level markup inside <div>
>>>
>>with no <p>.
>>
>>>Granted. Will be in the final release in the morning. A real gain for
>>>people who do not follow a paragraph markup and no real loss for people
>>>who do. Will be suggested that encoders pick one or the other.
>>>
>>>Would try to send out the mods now but have been up since around 4 AM
>>>and would make too many stupid errors.
>>>
>>>This should settle all the issues that I am aware of!
>>>
>>>Hope this finds one and all in good health and spirits!
>>>
>>>Patrick
>>>
>>>
>>

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