[Docstring-checkins] CVS: dps/spec dps-notes.txt,1.30,1.31

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Update of /cvsroot/docstring/dps/spec
In directory usw-pr-cvs1:/tmp/cvs-serv4387/dps/spec

Modified Files:
	dps-notes.txt 
Log Message:
updated

Index: dps-notes.txt
===================================================================
RCS file: /cvsroot/docstring/dps/spec/dps-notes.txt,v
retrieving revision 1.30
retrieving revision 1.31
diff -C2 -d -r1.30 -r1.31
*** dps-notes.txt	11 Mar 2002 03:45:15 -0000	1.30
--- dps-notes.txt	13 Mar 2002 02:48:34 -0000	1.31
***************
*** 87,103 ****
    SGML-ID-friendly id's and a one-to-one mapping for later lookup.

!   - Add conversion to nodes.document record-keeping code? Or as a new
!     transform?

!   - Use a "name mangling" scheme (like name "section title" becomes id
!     "section_title", name "1" becomes id "footnote1")?  Or use an
!     arbitrary sequential id?  How to do inter-document references in
!     that case?

!     - Perhaps combine name mangling with sequential ids?

!   - ID everything?  Or only named elements?

!   - How to do inter-document references in *any* case?

    - Perhaps keep a name->id mapping file?  This could be stored
--- 87,104 ----
    SGML-ID-friendly id's and a one-to-one mapping for later lookup.

!   - Implement a "name mangling" scheme: name "section title" becomes
!     id "section_title", name "1" becomes id "footnote1" (or just
!     "1"?).

!   - Whenever possible, use the "name mangling" scheme.  When that's
!     not possible (duplicate names, no name), and use an arbitrary
!     sequential id.

!   - ID as required.

!   - Remove dependency on names as sole distinguishing characteristic.
!     Use IDs instead.

!     - Perhaps get rid of "name" attributes altoghether?

    - Perhaps keep a name->id mapping file?  This could be stored
***************
*** 105,108 ****
--- 106,112 ----
      new entries.  ("Persistent ID mapping"?)

+   - When resolving reference IDs from names, we must check for
+     non-existent mappings.  In the Writer or in a transform?
+ 
  - Considerations for an HTML Writer [#]_:

***************
*** 121,127 ****
    __ http://www.webreference.com/dev/html4nsie/index.html

- - The Reporter needs another threshold writers. Use warninglevel as a
-   default.
- 

  Coding Conventions
--- 125,128 ----
***************
*** 405,408 ****
--- 406,410 ----
    transforms. (CamelCase only or unrestricted?) Lazy indentation?
  - Web Page: As standalone, but recognize meta fields as meta tags.
+   Support for templates of some sort? (After <body>, before </body>?)
  - FAQ: Structured "question & answer(s)" constructs.
  - Compound document: Merge chapters into a book. Master TOC file?
***************
*** 477,516 ****
  - Multiple files & directories
  - Objects in memory
- 
- 
- Mixing Automatic and Manual Footnote Numbering
- ==============================================
- 
- [David]
- I'm re-examining (for validity) what I wrote in the spec:
- 
-     Automatic footnote numbering may not be mixed with manual footnote
-     numbering; it would cause numbering and referencing conflicts.
- 
- Would such mixing inevitably cause conflicts? We could probably work
- around potential conflicts with a decent algorithm. Should we?
- Requires thought.  Opinions?
- 
- [Tony]
- Well, I read that paragraph in the documentation, and decided that it
- was in the category of "don't, in practice, care" so far as I was
- concerned. This is the same category I put the forbidding of nested
- inline markup - quite clearly one *can* do it, but equally clearly
- it's a pain to implement, and not a terribly great gain, all things
- considered.
- 
- It's a category with the subtext "examine for correctness after we've
- had some experience of people *using* reST in the wild".
- 
- Thus, given there are lots of other things to do, I would tend to
- leave it as-is (especially if you are able to *warn* people about it
- if they do it by mistake).
- 
- To my mind, being able to do ``[#thing]_`` probably give people enough
- precision over footnotes whils still allowing autonumbering - the
- *only* potential problem is when referring to a footnote in a
- different document (and that, again, is something I would leave fallow
- for the moment, although we know I tend to want to use roles as
- annotation for that sort of thing).

--- 479,482 ----