OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

docbook-tc message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Results of survey of DocBook 5.0 tools

Here are the results of the survey I did on the mailing list.  I list each 
question, the counts for each choice, and then any comments regarding that 

> I'm doing an informal survey for the DocBook Technical Committee regarding 
> schema usage for DocBook 5.0. We would like to know among DocBook 5.0 
> users:
> 1.  Which schema type are you using (RelaxNG, DTD, XML Schema) and why.

14 RelaxNG

Seems the most powerful way of defining a valid DocBook file.

RelaxNG, because it's simple, easily extensible, the source format for
DocBook, and supported by our tools.

Currently using DTD, because that's what we've done and what works

RelaxNG. It is easy to customize and we can keep up with future 

RelaxNG. The main reason is that nxml-mode supports it

RelaxNG, because it

- is quite powerful and flexible
- is very easy and intuitive to read and write
- is straightfoward to customize
- works with James Clark's nxml-mode for Emacs

RelaxNG (compact syntax) because it works with Emacs and nxml-mode
and because of the power of Schematron rules.  I don't understand
the details, but RelaxNG seems most useful for future work.

 relax NG - 'cos it's easier to work with

RelaxNG, because it's easy to work with and powerful.

DTD, because it provides named character references.

RELAX NG because of its explanation of elements which are shown as a tool
tip in some XML editors (for example, in oXygen) and for its ease of use.

> 2.  What editing tools are you using, and why.

7 Oxygen XML Editor
7 emacs+nxml
1 plain text editor(EmEditor, vim)
1 UltraEdit
1 BBedit

Oxygen. Our team likes it because the author view is a nice wysiwyg editor.

emaxs+nxml, XXE. because they have great support for RelaxNG, and match
our needs.

Oxygen because that seems to be the best for our purposes: great editing, 
ability to transform from the tool (for testing though; for prod we use an 
ANT script outside of Oxygen).

XMLmind XML Editor. It supports DocBook 5 and XInclude out of the box, it 
can be easily customized and extended, editing DocBook documents is 
intuitive and fast as soon as you learn the basics.

nxml-mode. I mainly use it because it offers on-the-fly validation and
element/attribute completion, and because it runs in Emacs. This e.g.
allows me to manage citations and bibliographies effectively with an
Emacs frontend to my bibliography database.

Emacs, because I use Emacs for almost everything.

nxml-mode, because it does many things for me, most importantly

- on-the-fly validation
- tag autocompletion
- all the basic stuff (syntax highlighting, auto-indentation, ...)

For me, its most serious shortcoming is the absence of built-in
support for XIncludes (together with a choice of "shallow" and "deep"
validation).  I help myself with customized RelaxNG schemas where I
allow xi:includes wherever I need them in practice.

UltraEdit, a familiar tool; I avoid smarter tools such as XmlMind while I'm 
still trying to master the underlying technology. (I drive a 
manual/stick-shift too.)

GNU Emacs + nxml-mode because I use Emacs (or vi) for all editing.
It works well for DocBook, XHTML, and other XML editing, and it runs
very well on all machines that I use (FreeBSD, GNU/Linux, Mac OS X).
I've used Emacs for 2 decades and see no reason to change editors.

nxml-mode in GNU Emacs, because I'm an Emacs user anyway

- Emacs + nXML for the one and only(?) open source editor for
  editing with RELAX NG schema
- oXygen for its different editing modes (like the author mode as a
  kind of WYSYWYG), XSLT debugging and lots of other good features

> 3.  What validation tools you are using, and why.

4 Oxygen
3 Sun MSV
7 jing
1 SaxonSA
7 xmllint
2 emacs+nxml

(a) Oxygen built-in validation
Convenience while writing text

(b) Sun MSV
Part of my ANT build environment
I use ANT instead of transformation scenarios in Oxygen

- nxml-mode for on-the-fly, "shallow" validation (using schemas that
  allow xi:includes)

- xmllint --xinclude followed by jing for "deep" validation (within an
  Emacs compilation buffer, of course)

  Why? The main reason is that I heavily use the xpointer() scheme
  (http://www.w3.org/TR/xptr-xpointer/) in my xi:includes, which is
  not widely implemented. This essentially limits me to libxml.

  Jing can use the same compact-syntax RelaxNG schema as uses

xmllint, for xincludes

xmllint, because it's fast and just works for me

Normally I use xmllint to validate my DocBook4 documents. However, it
seems it has some problems with the RELAX NG schema of DocBook5. So I use
it mainly to resolve XIncludes (rarely XPointers) and validate it through
jing or msv.

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]