Difference between revisions of "Talk:Best Practices for TEI in Libraries"

From TEIWiki
Jump to navigation Jump to search
(→‎hyphenation: can't have both editorialDecl/hyphenation and editorialDecl/p!)
(→‎Rendering better documentation from the ODDs: noted things that aren't issues any more)
 
(192 intermediate revisions by 8 users not shown)
Line 1: Line 1:
−
== Possible expanded filename recommendations ==
+
The following are things to do to the BP before making an official "release".  There is a separate list of [[Future changes to Best Practices for TEI in Libraries]].
  
−
Standardized file naming for a particular encoding project is key for reliable online storage and delivery of these files.  Consider the following best practices when determining the file name scheme for your project:
+
== Handling of hyphenation ==
  
−
* Each filename should contain an identifier that uniquely specifies a single digital object within the parent collection (e.g., a parent collection of text, images and other related materials)
+
Syd will post to TEI-L asking how to distinguish "his-tory" and "run-on" when they break across lines. Both would take break="no", but how to allow for searching of "history" and "run-on"?
−
* Each filename should be fully specified. It should not just be a sequence number that is dependent on location within a directory structure for context
 
−
* Filenames should not include spaces
 
−
* Filenames should following a predicatble case constructions (e.g., all lowercase, camelCase, etc.)
 
−
* The first character of the filename should be an ASCII letter ('a' through 'z' or 'A' through 'Z') to comply with current restrictions on identifiers by many programming and metadata languages such as METS
 
−
* The "base" filename may include only ASCII letters ('a' through 'z' and 'A' through 'Z'), ASCII digits ('0' through '9'), hyphens, underscores, and periods. Refrain from using other characters and limit period usage to only once (to separate base name from file extensions).
 
  
−
For those saving files to CD-ROM for storage or file transfer, file naming should follow ISO 9660 conventions: 8-character filenames, 3-character extensions, using A-Z, a-z, 0-9, underscores and hyphens.
+
: Syd re-read P5 and emailed Kevin that he thinks we can handle this without a wider call.  Kevin agreed with Syd on a solution and made the revisions to main-driver.odd and level4.odd.  Asked Syd to check over the changes to [https://github.com/sydb/TEI-in-Libraries/commit/31a8730b9d0af093099a18ab57078d4b8577c23b#BestPractices/main-driver.odd main-driver.odd] and [https://github.com/sydb/TEI-in-Libraries/commit/9e4f98caeda0446ebb3397f619c83b534ccd1fca#BestPractices/level4.odd level4.odd]. Then will update the SIG on our handling of this issue and on the final work on the BP as a whole. ([[User:Kshawkin|Kshawkin]] 11:28, 13 September 2011 (EDT))
  
−
== @type and @key on persName and orgName ==
+
:: Syd said everything looked fine but made an additional suggestion, which Kevin [https://github.com/sydb/TEI-in-Libraries/commit/8c60fef9dd3b9641b878f16d0370b05d4b838856 just committed].
  
−
In the best practices document, the author element is described as follows:
+
== Bug fixes ==
  
−
One or more author elements (one name per element) are used to encode the name for the personal author or corporate body responsible for the creation of the source document, even if this creator is not the main entry in the catalog record. Use <persName> or <orgName> when applicable. Whenever possible, establish or use the form of the name from a national name authority file.
+
* Fix schema bug where list element is not allowed except as a child of p.
 +
: Syd put model.inter back into model.common at Levels 3 and 4.  This fixed the problem. <b>Syd wonders whether this change will have any side-effect; he will investigate further.</b>
  
−
Since the forms of names used in name authority files have a rigid form that doesn't look like a name in the TEI sense (strings like "Welles, Gideon, 1802-1878" offend the sensibilities of XML folks), during the Raleigh meeting we decided that name authority records given in the header should have a type attribute, similar to that used on <title>.  So on fileDesc/titleStmt/author and fileDesc/sourceDesc/biblStruct/monogr/author the following values of @type would be allowed:
+
== Copyediting ==
  
−
<pre>
+
Change all instances of "must" to "should" and "required" to "recommended", and all existing "recommended" to "optional".  All this is in accordance with RFC 2119 and with the BP's general policy of not requiring anything but simply being "best practices".  Clarify that the ODDs require things just to encourage conformance.
−
marc100
+
: done ([[User:Kshawkin|Kshawkin]] 20:47, 5 September 2011 (EDT))
−
marc110
 
−
</pre>
 
  
−
We would not allow marc111 or marc130 because these MARC fields, while used for main entries in cataloging, are not authors in the TEI sense. As explained in the description of the author element, this element should be used for personal authors or corporate bodies, not necessarily main entries.
+
Move caveats before Level 1 Example to an appropriate place in main-driver.odd.
 +
: done ([[User:Kshawkin|Kshawkin]] 20:47, 5 September 2011 (EDT))
  
−
We also decided to recommend use of @key, as in the "Level 4 Name Tagging" section, to reference authority file records. In "Level 4 Name Tagging", it says, "the key attribute points to the unique key in the database table or, as with the ref attribute, the key attribute can point to the xml:id value in the external file".
+
Copyedit element recommendations at each level to avoid awkward and misleading syntax.
 +
: Not finding anything in particular to fix. I think I was just tired when reading these in person with Syd.
  
−
However, @key does not take an IDREF data type (as it was called in the old days), and once I tried to create examples, I realized we don't want to be in the business of adding a <taxonomy> for each authority record referenced elsewhere in the header.  So I think what we want is this:
+
Remove ODD markup in level specifications that Syd added from P5 so as not to include examples that contradict rest of Guidelines.
 +
: Commented out gloss, desc, exemplum, remarks, and listRef in level1.odd for Syd to check over before I do in other levels. ([[User:Kshawkin|Kshawkin]] 20:47, 5 September 2011 (EDT))
 +
:: Syd said he un-commented out these elements.  He'll review the 79 elements in the various ODDs to check for conflict with the BP prose.  That way we can include them with our release. ([[User:Kshawkin|Kshawkin]] 10:46, 3 October 2011 (EDT))
  
−
<pre>
+
Edit main-driver.odd to review metadata for the BP as a whole (in both the teiHeader and front elements): editors (stated more prominently than in the appendix), copyright, version number, etc.  Check with Michelle on whether we should revise acknowledgment of DLF support.
−
<author><persName type="marc100" key="lccn-n78-95332">Shakespeare, William, 1564-1616</persName></author>
 
−
<author><orgName type="marc110" key="lccn-n50-63455">National Organization for Women</orgName></author>
 
−
<author>(unknown)</author>
 
−
</pre>
 
  
−
plus this elsewhere in the header:
+
: In teiHeader, added Michelle as <editor> and copyedited name of SIG.  Having been in touch with DLF recently, I'm pretty sure our statement of support is sufficient. 
  
−
<pre>
+
: As last step before releasing, add <code>&lt;editionStmt>&lt;p>Version 3.0 (October 2011)&lt;/p>&lt;/editionStmt></code> [[User:Kshawkin|Kshawkin]] 20:27, 29 September 2011 (EDT)
−
<taxonomy xml:id="lccn"><bibl>Library of Congress Control Number</bibl></taxonomy>
+
:: Done
−
</pre>
 
  
−
Sound right?
+
== Rendering better documentation from the ODDs ==
  
−
== The_TEI_Header_and_Other_Metadata_Schemas ==
+
Tables should have borders on cells (or some clear path for indicating in ODD document that you want borders).
 +
: Syd said this can be overriden by giving values for cssFile and/or cssSecondFile parameters to odd2html.xsl.  Not sure how to "send" these through roma2.
 +
:: Sebastian wrote in an email to use, e.g., <tt>--docflags="cssFile=foo.css"</tt> but warned that this was from memory ([[User:Kshawkin|Kshawkin]] 16:08, 22 August 2011 (EDT))
  
−
I added a paragraph to the section called [[Best_Practices_for_TEI_in_Libraries#The_TEI_Header_and_Other_Metadata_Schemas|The TEI Header and Other Metadata Schemas]] explaining why you can't link to outside metadata the way you'd like to. Does it sound right to everyone?
+
Omit exemplars from output — preferably just those from the P5 source, not the customization ODD file.
 +
: Done. See the [https://github.com/sydb/TEI-in-Libraries/blob/master/BestPractices/odd2odd.xsl.patch stylesheet patch file].
  
−
== rendition in header? ==
+
<code>&lt;editor></code>s are not showing up in HTML version of ODD: need to figure out how to make this happen.
  
−
I changed removed mention of rendition ladders and replaced recommendation and examples with CSS.  Should we add rendition element to the header?  P5 talks about using this element to define local styles in terms of CSS, but we want to put CSS right inline.  So is there any point in using the rendition element in the header?
+
: Done
  
−
== hyphenation ==
+
Check that editionStmt gets rendered in HTML version of ODD.
  
−
One of the comments we received on the Best Practices draft was a request to recommend on a specific way for handling end-of-line (and end-of-page) hyphenation within the <tt>body</tt>.  We have not discussed this at all in the Best Practices!
+
: Done
  
−
We should have a new section under "General Recommendations" explaining to do end-of-line hyphens like in Tite (to ensure that texts can be converted to Level 3 automatically).  See http://www.tei-c.org/release/doc/tei-p5-exemplars/html/tei_tite.doc.html#e-o-l .  End-of-page hyphens should be done the same way, with the <tt>pb</tt> tag intervening, e.g., <tt>obfus{U+00AD}&lt;pb n="33" facs="00000037.tif"/&gt;cation</tt>
+
== Documentation of ODD processing ==
  
−
P5 offers the <tt>hyphenation</tt> element in the header to document your method. We could give the following in the header element recommendations:
+
Syd will write down the command-line code needed in order to generate HTML files from the ODDs and add this to https://github.com/sydb/TEI-in-Libraries/blob/master/README .
 +
: Done
  
−
<tt>&lt;hyphenation eol="hard"&gt;&lt;p&gt;End-of-line hyphenation silently removed where appropriate.&lt;/p&gt;&lt;/hyphenation&gt;</tt>
+
== Before release ==
  
−
However, if we use this element, we would need to find another way to encode all of the prose description that is currently in p elements inside the editorialDecl. What should we do?
+
Update [http://www.tei-c.org/SIG/Libraries/teiinlibraries/ the official HTML version] to remove "Expected release October 2011."
  
−
For the record, if I weren't trying to ensure this compatibility, I would have distinguished three cases:
+
Update copy of main-driver.html linked from there.
−
 
 
−
* hyphenation that occurs at a line or page break but would never occur normally: use U+00AD
 
−
* hyphenation that occurs at a line or page break but might have occurred normally: use U+002D
 
−
* hyphenation that should always be present, such as "re-creation" (to create again, as opposed to leisure activity): use U+2011
 
−
 
 
−
== profile/langUsage/language as empty element? ==
 
−
 
 
−
I added the language element to the header.  The ident= attribute is required, but the element may be empty (despite no such examples given in P5).  It seems entirely redundant to me to have content for this element. Should we prescribe having no content for it (and just a value for ident=)?
 
−
 
 
−
== idno in sourceDesc ==
 
−
 
 
−
I substantially rewrote the description of the idno element in sourceDesc. Does this sound right to everyone?
 

Latest revision as of 02:31, 24 October 2011

The following are things to do to the BP before making an official "release". There is a separate list of Future changes to Best Practices for TEI in Libraries.

Handling of hyphenation

Syd will post to TEI-L asking how to distinguish "his-tory" and "run-on" when they break across lines. Both would take break="no", but how to allow for searching of "history" and "run-on"?

Syd re-read P5 and emailed Kevin that he thinks we can handle this without a wider call. Kevin agreed with Syd on a solution and made the revisions to main-driver.odd and level4.odd. Asked Syd to check over the changes to main-driver.odd and level4.odd. Then will update the SIG on our handling of this issue and on the final work on the BP as a whole. (Kshawkin 11:28, 13 September 2011 (EDT))
Syd said everything looked fine but made an additional suggestion, which Kevin just committed.

Bug fixes

  • Fix schema bug where list element is not allowed except as a child of p.
Syd put model.inter back into model.common at Levels 3 and 4. This fixed the problem. Syd wonders whether this change will have any side-effect; he will investigate further.

Copyediting

Change all instances of "must" to "should" and "required" to "recommended", and all existing "recommended" to "optional". All this is in accordance with RFC 2119 and with the BP's general policy of not requiring anything but simply being "best practices". Clarify that the ODDs require things just to encourage conformance.

done (Kshawkin 20:47, 5 September 2011 (EDT))

Move caveats before Level 1 Example to an appropriate place in main-driver.odd.

done (Kshawkin 20:47, 5 September 2011 (EDT))

Copyedit element recommendations at each level to avoid awkward and misleading syntax.

Not finding anything in particular to fix. I think I was just tired when reading these in person with Syd.

Remove ODD markup in level specifications that Syd added from P5 so as not to include examples that contradict rest of Guidelines.

Commented out gloss, desc, exemplum, remarks, and listRef in level1.odd for Syd to check over before I do in other levels. (Kshawkin 20:47, 5 September 2011 (EDT))
Syd said he un-commented out these elements. He'll review the 79 elements in the various ODDs to check for conflict with the BP prose. That way we can include them with our release. (Kshawkin 10:46, 3 October 2011 (EDT))

Edit main-driver.odd to review metadata for the BP as a whole (in both the teiHeader and front elements): editors (stated more prominently than in the appendix), copyright, version number, etc. Check with Michelle on whether we should revise acknowledgment of DLF support.

In teiHeader, added Michelle as <editor> and copyedited name of SIG. Having been in touch with DLF recently, I'm pretty sure our statement of support is sufficient.
As last step before releasing, add <editionStmt><p>Version 3.0 (October 2011)</p></editionStmt> Kshawkin 20:27, 29 September 2011 (EDT)
Done

Rendering better documentation from the ODDs

Tables should have borders on cells (or some clear path for indicating in ODD document that you want borders).

Syd said this can be overriden by giving values for cssFile and/or cssSecondFile parameters to odd2html.xsl. Not sure how to "send" these through roma2.
Sebastian wrote in an email to use, e.g., --docflags="cssFile=foo.css" but warned that this was from memory (Kshawkin 16:08, 22 August 2011 (EDT))

Omit exemplars from output — preferably just those from the P5 source, not the customization ODD file.

Done. See the stylesheet patch file.

<editor>s are not showing up in HTML version of ODD: need to figure out how to make this happen.

Done

Check that editionStmt gets rendered in HTML version of ODD.

Done

Documentation of ODD processing

Syd will write down the command-line code needed in order to generate HTML files from the ODDs and add this to https://github.com/sydb/TEI-in-Libraries/blob/master/README .

Done

Before release

Update the official HTML version to remove "Expected release October 2011."

Update copy of main-driver.html linked from there.

Retrieved from "https://wiki.tei-c.org/index.php?title=Talk:Best_Practices_for_TEI_in_Libraries&oldid=10260"