Difference between revisions of "Minutes for May 26, 2009"

From TEIWiki
Jump to navigation Jump to search
(@type/@key persName/orgName: clarifications)
 
(5 intermediate revisions by one other user not shown)
Line 1: Line 1:
DRAFT DRAFT: NEEDS CLEAN UP SO DON'T READ NOW!
+
[[Category:SIG:Libraries]]
  
 
'''In Attendance:''' <br>
 
'''In Attendance:''' <br>
Line 21: Line 21:
 
* Kevin will re-word these so that they not rhetorical questions but examples of kinds of info. one would include in the editorialDecl.
 
* Kevin will re-word these so that they not rhetorical questions but examples of kinds of info. one would include in the editorialDecl.
  
== Use <tt><ref></tt> for ToC example in Level 3 ==
+
== Use <tt>&lt;ref></tt> for ToC example in Level 3 ==
* level 3 in ToC should use to <tt><ref></tt> tag for specific pages.  Use <ptr> when there's no
+
* level 3 in ToC should use to <tt>&lt;ref></tt> tag for specific pages.  Use <tt>&lt;ptr></tt> when there's no
page numbers, but <tt><ref></tt> examples in the P5 Guidelines has page numbers.   
+
page numbers, but <tt>&lt;ref></tt> examples in the P5 Guidelines has page numbers.   
 
* Why do we bother encoding ToC anyway?  Michigan doesn't encode the ToC.   
 
* Why do we bother encoding ToC anyway?  Michigan doesn't encode the ToC.   
 
* Add a note you may not want to bother encoding the ToC (optiional) if system will generate this automatically, but you plan to manually encode the ToC, here's an example.   
 
* Add a note you may not want to bother encoding the ToC (optiional) if system will generate this automatically, but you plan to manually encode the ToC, here's an example.   
 
* Same for list of illustrations occurring in prefatory pages (add to level 3 description) about possibly removing these as well.
 
* Same for list of illustrations occurring in prefatory pages (add to level 3 description) about possibly removing these as well.
  
 +
== @type/@key persName/orgName==
 +
* Added type values for marc100 and marc110 (to better represent author names + dates).
 +
* Did not add marc 111 or marc 130. 
 +
* Have a key attribute that would reference an authority file.  Key is not really an ID/REF type. 
 +
* Need a taxonomy element in the Header. 
 +
* What's the difference between lccn-n78-#### (OCLC World Cat Identities) and lcnaf #? Should we align these?
 +
* Explain the need to develop a local ref/key scheme for referencing authority files under General Recommendations.  Also add ref/key elements under "General Guidelines for Attribute Usage".
  
--@type/@key persName/orgName
+
== Linking to outside Metadata schemes==
Added type values for marc100 and marc110 (to deal with author names + dates). Did not add marc 111 or marc 130. Have a key attribute that would reference an authority file. Key is not really an ID/REF type.  Need a taxonomy element in the Header.
+
* Syd doesn't agree:  There's nothing at all that stops us for incorporating non-TEI metadata in the Header.  Since we are making ODDs we can support the embedding of metadata schemes and values. 
 +
* But METS handle this so should we even embed metadata? 
 +
* Need time to provide thoughful recommendations for embedding 
 +
* Linking to other schemes can be hacked (but then we need to provide recommendations for what a minimum header looks like)
 +
* TEI document has an xml:id that references a METS or some other file, but it doesn't provide sufficient context.
 +
* Melanie recommends that the Header subgroup reforms to work on these issues.
 +
* Neither embedding or linking recommendations will be part of version 3 of the best practices
 +
* Kevin will revise this text slightly to clarify and to indicate future revisions will handle this.
  
What's the difference between lccn-n78-#### (OCLC World Cat Identities) and lcnaf #? Should we align these? Need to develop a local scheme for referencing authority files under General Recommendations.  Also add ref/key elements under GR sections.
+
== Rendition in header ==
 +
* Use CSS instead.  Use the rendition element in the Header; css rules in the Header.  Method of indirection. 
 +
* Inline CSS versus defining CSS in the Header?  What are the pros and cons?
 +
* Rendition placed in external file and xinclude it in the Header or define it in the Header; minimizes errors and keystrokes.  Maximizes single renidtion reference for a set of documents.
 +
* Level 3 use rend (with inline CSS) and not rendition and level 4 say that rendition have X advantages?
 +
* Worth saying that rend SHOULD NOT be used in the Header
 +
* Place under general attributes message and support both
 +
* CSS inline (easier to process) or level of indirection (minimize errors) in the Header
  
Linking to outside Metadata schemes
+
== Level 5 ==
--Syd doesn't like it:  There's nothing at all that stops us for incorporating non-TEI metadata in the HeaderSince we are making ODDs we can support the embedding of metadata schemes and values.  METS handle this so should we even embed metadata. 
+
* Axed the examples and mimicked the sections for the other levels.   
--Linking doesn't exist as fully (but then we need to provide recommendations for what a minimum header looks like)
+
* Header and general recommendations still apply for Level 5
--TEI document has an xml:id that references a METS. 
+
* Syd: feels that it's more than just the Header, level 4 + extra stuff + general recommendations.
--Melanie recommends that the Header subgroup reforms to work on these issues.
+
* Glen will clairfy and add the other bits.
--Not part of version 3.
 
--Kevin will revise this text slightly.
 
  
Rendition in header?
+
==Hyphenation==
--Use CSS instead.  Use the rendition element in the Header; css rules in the Header.  Method of indirection. 
+
* Kevin recommends the Tite way (to facilitate conversion)
--Inline CSS versus defining CSS in the Header?
+
* Should be an editorial decision?
--Every time a new style is discovered in the document; add to the Header and represent in CSSOnly create the short cut when used a lot.
+
* Improve recall when searching for a corpus (Tite rationale).
--Rendition placed in external file and xinclude it in the Header or define it in the Header; minimizes errors and keystrokes.
+
* hyphenation in the header needs to be re-worded
--Level 3 use rend (with inline CSS) and not rendition and level 4 say that rendition have X advantages.  
+
* Perry disagrees; Tite won't represent the majority of projects already created and many projects won't conform to thisPerry and Syd don't feel this is the best practice.
--Worth saying that rend SHOULD NOT be used in the Header
+
* Level 1/2 generated in OCR, OCR software can remove soft hyphens. Abbey FineReader (can remove soft hyphens).
--Place under general attributes message.
+
* Defer issue to next call or email discussion.
--CSS inline (easier to process) or level of indirection (minimize errors) in the Header
+
* Subgroup needs to form
  
Level 5
+
==Conference Call Schedule==
--Axed the examples and mimicked the sections for the other levels. 
+
* Schedule next call a week from today to wrap up the pending issues.
--Header and general recommendations still apply for Level 5
+
* By July, calls can ease up or stop entirely
--Syd: feels that it's more than just the Header, level 4 + extra stuff + general recommendations. 
 
--Glen could add the other bits. 
 
 
 
Hyphenation
 
--Recommend the Tite way (to facilitate conversion)
 
--EOL hyphens in Tite (distinguish between hard and soft hyphens; use soft hyphen document if there weren't a line break; and use ASCII for hard hyphen).  Preference to hard hyphens.
 
--Should be an editorial decision. 
 
--Improve recall when searching for a corpus (Tite rationale).
 
--hyphenation in the header needs to be re-worded
 
--Perry disagrees; Tite won't represent the majority of projects already created and many projects won't conform to this.  Perry and Syd don't feel this is the best practice.
 
--Level 1/2 generated in OCR, OCR software can remove soft hyphens.  Abbey FineReader (can remove soft hyphens).  Defer to next call.
 
--Subgroup needs to form
 
 
 
Schedule next call a week from today to wrap up the pending issues.
 

Latest revision as of 15:03, 27 May 2009


In Attendance:
Syd, Kevin, Michelle, Matthew, Rich, Melanie, Natasha, Perry and Glen

Note taker: Michelle

Level attribute on title element

  • use of level on title; Guidelines ambiguous; got clarification via SourceForge
  • Guidelines said the attribute was required when applicable (supposed to

be recommended when applicable)

  • Usage note now clarified; don't have to use it if value can be inferred (e.g. title in a series, don't need title level="s")
  • Do we have in all cases, or only when it can not be inferred?
  • Perry thinks we should avoid inference route; easier to process
  • So have it in all cases

Pending revisions

  • AACR questions pulled from the header element description (for editorialDecl).
  • Rhetorical questions ... could be confusing.
  • Kevin will re-word these so that they not rhetorical questions but examples of kinds of info. one would include in the editorialDecl.

Use <ref> for ToC example in Level 3

  • level 3 in ToC should use to <ref> tag for specific pages. Use <ptr> when there's no

page numbers, but <ref> examples in the P5 Guidelines has page numbers.

  • Why do we bother encoding ToC anyway? Michigan doesn't encode the ToC.
  • Add a note you may not want to bother encoding the ToC (optiional) if system will generate this automatically, but you plan to manually encode the ToC, here's an example.
  • Same for list of illustrations occurring in prefatory pages (add to level 3 description) about possibly removing these as well.

@type/@key persName/orgName

  • Added type values for marc100 and marc110 (to better represent author names + dates).
  • Did not add marc 111 or marc 130.
  • Have a key attribute that would reference an authority file. Key is not really an ID/REF type.
  • Need a taxonomy element in the Header.
  • What's the difference between lccn-n78-#### (OCLC World Cat Identities) and lcnaf #? Should we align these?
  • Explain the need to develop a local ref/key scheme for referencing authority files under General Recommendations. Also add ref/key elements under "General Guidelines for Attribute Usage".

Linking to outside Metadata schemes

  • Syd doesn't agree: There's nothing at all that stops us for incorporating non-TEI metadata in the Header. Since we are making ODDs we can support the embedding of metadata schemes and values.
  • But METS handle this so should we even embed metadata?
  • Need time to provide thoughful recommendations for embedding
  • Linking to other schemes can be hacked (but then we need to provide recommendations for what a minimum header looks like)
  • TEI document has an xml:id that references a METS or some other file, but it doesn't provide sufficient context.
  • Melanie recommends that the Header subgroup reforms to work on these issues.
  • Neither embedding or linking recommendations will be part of version 3 of the best practices
  • Kevin will revise this text slightly to clarify and to indicate future revisions will handle this.

Rendition in header

  • Use CSS instead. Use the rendition element in the Header; css rules in the Header. Method of indirection.
  • Inline CSS versus defining CSS in the Header? What are the pros and cons?
  • Rendition placed in external file and xinclude it in the Header or define it in the Header; minimizes errors and keystrokes. Maximizes single renidtion reference for a set of documents.
  • Level 3 use rend (with inline CSS) and not rendition and level 4 say that rendition have X advantages?
  • Worth saying that rend SHOULD NOT be used in the Header
  • Place under general attributes message and support both
  • CSS inline (easier to process) or level of indirection (minimize errors) in the Header

Level 5

  • Axed the examples and mimicked the sections for the other levels.
  • Header and general recommendations still apply for Level 5
  • Syd: feels that it's more than just the Header, level 4 + extra stuff + general recommendations.
  • Glen will clairfy and add the other bits.

Hyphenation

  • Kevin recommends the Tite way (to facilitate conversion)
  • Should be an editorial decision?
  • Improve recall when searching for a corpus (Tite rationale).
  • hyphenation in the header needs to be re-worded
  • Perry disagrees; Tite won't represent the majority of projects already created and many projects won't conform to this. Perry and Syd don't feel this is the best practice.
  • Level 1/2 generated in OCR, OCR software can remove soft hyphens. Abbey FineReader (can remove soft hyphens).
  • Defer issue to next call or email discussion.
  • Subgroup needs to form

Conference Call Schedule

  • Schedule next call a week from today to wrap up the pending issues.
  • By July, calls can ease up or stop entirely