*** Forwarding an announcement from Syntext ***

Serna is a very capable wysiwyg XML editor with good support for DITA.
If you need a bunch of cheap (free) DITA editors for developing content,
while you leave the final output to XMetal, Arbortext Author,
FrameMaker, Oxygen, or DITA OT, this is a good choice for content
development, as separate from admin and config tasks.  And you can
always lend a hand with CSS or XSLT files, plug-ins,  etc.

Regards,
Hedley


-------- Original Message --------
Subject:     ANN: Serna Free XML Editor is Open Source Now!
Date:     Wed, 1 Jul 2009 19:20:39 +0400 (MSD)
From: announce-sender@...
To: announce@...



Syntext is happy to announce the release of Serna Free Open Source XML
Editor!  Serna is a powerful and easy-to-use WYSIWYG XML editor based on
open standards, which works on Windows, Linux, Mac OS X, and Sun
Solaris/SPARC.

Download Serna Free Open Source XML Editor at:
<http://syntext.com/downloads/serna-free/>.

Follow us on Twitter! Twitter users can stay on top of all the latest
happenings with Serna at: <http://twitter.com/syntext/>.

Get the latest Syntext news and updates on LiveJournal:
<http://syntext.livejournal.com/>.

If you are passionate about contributing to Serna Free Open Source XML
Editor, here are our requests for you to help make the Open Source and
XML world better:

- Discuss open-source Serna with the community! Please
   share your ideas concerning technology, business or
   partnership at our forum: <http://forum.syntext.com/
   syntext-serna-xml-editor/serna-open-source-discussion/
<http://forum.syntext.com/syntext-serna-xml-editor/serna-open-source-discussion/\
>>.


- Discuss the development of open-source Serna!
   Developers are welcome to discussions in the
   developers' mailing list: <serna-developers@...
<mailto:serna-developers-subscribe@...>>.

- Make a donation! If you love Serna and enjoy using it,
   please support us to provide open-source Serna to the
   world for free and serve people better.  Please
   contribute at: <http://www.syntext.com/donation/>.

- Spread the news! Please mention Serna has gone open-
   source in your blogs, Twitter, and XML mailing lists.
   The free Serna code needs to find its hackers! We
   need to know what they think and how they can
   participate.

We rely on you to help us improve Serna Free XML Editor and keep it free
and running!

The Syntext Team


To subscribe send empty e-mail to: <announce-subscribe@...
<mailto:announce-subscribe@...>>,
<serna-developers-subscribe@...
<mailto:serna-developers-subscribe@...>>.

--

Hedley Finger

28 Regent Street   Camberwell VIC 3124   Australia
Tel: +61 3 9809 1229   Fax: (call phone first)
Mobile (cell): +61 412 461 558   Skype: hedley.finger
Email. "Hedley Finger" <hfinger@...>

Syntext is happy to announce the release of Serna Free Open Source XML Editor!  Serna is a powerful and easy-to-use WYSIWYG XML editor based on open standards, which works on Windows, Linux, Mac OS X, and Sun Solaris/SPARC.

Download Serna Free Open Source XML Editor at: <http://syntext.com/downloads/serna-free/>.

Follow us on Twitter! Twitter users can stay on top of all the latest happenings with Serna at: <http://twitter.com/syntext/>.

Get the latest Syntext news and updates on LiveJournal: <http://syntext.livejournal.com/>.

If you are passionate about contributing to Serna Free Open Source XML Editor, here are our requests for you to help make the Open Source and XML world better:

- Discuss open-source Serna with the community! Please 
  share your ideas concerning technology, business or 
  partnership at our forum: <http://forum.syntext.com/
  syntext-serna-xml-editor/serna-open-source-discussion/>.

- Discuss the development of open-source Serna! 
  Developers are welcome to discussions in the 
  developers' mailing list: <serna-developers@...>.

- Make a donation! If you love Serna and enjoy using it, 
  please support us to provide open-source Serna to the 
  world for free and serve people better.  Please 
  contribute at: <http://www.syntext.com/donation/>.

- Spread the news! Please mention Serna has gone open-
  source in your blogs, Twitter, and XML mailing lists.
  The free Serna code needs to find its hackers! We 
  need to know what they think and how they can 
  participate.

We rely on you to help us improve Serna Free XML Editor and keep it free and running!

The Syntext Team


To subscribe send empty e-mail to: <announce-subscribe@...>, <serna-developers-subscribe@...>.

-- Hedley Finger
28 Regent Street Camberwell VIC 3124 Australia
Tel: +61 3 9809 1229 Fax: (call phone first)
Mobile (cell): +61 412 461 558 Skype: hedley.finger
Email. "Hedley Finger" <hfinger@...>

On 7/2/09 3:54 PM, "Bryce L Nordgren" <bnordgren@...> wrote:

>> Of course, I wasn't seriously proposing that the DITA OT adopt a
> randomzier for <ul> processing.
> Can you provide an example of a processing system (using any technology)
> which does not respect the author's list order?  I can't think of one, but
> I could be missing something.
> <ul> is a de-facto child of <ol> to the processing system.  To the human,
> it's not.  The two target audiences are not the same.

The Learning and Training single- and multi-select assessments do exactly
this: the list of items is specialized from ul specifically because the
order of the possible answers needs to be randomizable in the rendition.

That is, the the traditional use of "A. B. C. D." for multiple-choice
answers is just a way to distinquish the options, it doesn't imply any
ordering. When you render the question you need to be able to present the
options in any order.

I have a client whose online training courses do exactly this: if you miss a
question when it presents it to you again the answer options have been
reordered.

Cheers,

E.

----
Eliot Kimber | Senior Solutions Architect | Really Strategies, Inc.
email:  ekimber@... <mailto:ekimber@...>
office: 610.631.6770 | cell: 512.554.9368
2570 Boulevard of the Generals | Suite 213 | Audubon, PA 19403
www.reallysi.com <http://www.reallysi.com>  | http://blog.reallysi.com
<http://blog.reallysi.com> | www.rsuitecms.com <http://www.rsuitecms.com>

> Of course, I wasn't seriously proposing that the DITA OT adopt a
randomzier for <ul> processing.
Can you provide an example of a processing system (using any technology)
which does not respect the author's list order?  I can't think of one, but
I could be missing something.
<ul> is a de-facto child of <ol> to the processing system.  To the human,
it's not.  The two target audiences are not the same.

> When people list their arguments in a
> debate, they use an ordered construct by saying something like,
> "First, I don't like those shoes. Second, they don't match my
> handbag, and third -- the don't even fit!".

I think the world is not so black and white.  By way of example, here's
something I wrote (lives at
http://www.smoke-fire.us/svn/firelab/FireDB/branches/multistation/doc/concepts/d\
atastorage.xml
)

             <p>The attributes used to describe a station's strategy for
the organization of data are
                 as follows: </p>
             <ul>
                 <li>The base URL (if there is any part in common).</li>
                 <li>The satellite specific parts to the URL (if data for
each satellite is stored in
                     different locations).</li>
                 <li>Are data from the two satellites held in separate
directories or a common
                     directory?</li>
                 <li>Are the files ungrouped, grouped by pass, or grouped
by day?</li>
                 <li>Is the pass divided into segments?</li>
             </ul>

Two list items are related, but not in an ordinal way: "satellite specific
parts of the URL" and "are data from each satellite held in different
locations or a common one".  They should appear together, but since
there's only two items, it's not worth grouping them in a sublist.
Presenting them together is a "weak binder", but it's a binder
nonetheless.  Also, the presence of a "BaseURL" intuitively should occupy
the first slot, even though it's an independent piece of data and it
doesn't matter whether you set it first or last or in the middle.  The
Base URL also "feels" most important and should be presented first. Again,
there is no dependency relationship between it and any other item; but it
should just be first.

Your argument assumes that the "meaning" conveyed to a computer with the
tags is (and should be) identical to the meaning garnered by a thinking
being who parses the content.

In this file, you'll also find an example of where I used <ul> merely to
get bullets (the description of processing levels). This is because the
contents of the list items all start out with processing level (e.g.,
"Raw", "Level 0", "Level 1"...) and numbering the items would have been
counterproductive and confusing.  Note that the argument for using <ol> in
this case implies also that per-list control over list item decoration is
required.  The contents of that particular list makes the "standard"
numbering scheme confusing to the human, not to mention awkward; yet I
doubt that the presense of this one list in the document is an argument
for modifying a "house list style".  Likewise, this one list doesn't merit
defining it's own unique "semantic element" in order to modify the source
code for the processing system so it can handle this one occurence.

--- In dita-users@yahoogroups.com, Bryce L Nordgren <bnordgren@...> wrote:
>
> Where the rubber meets the road, the only practical difference
> between <ol> and <ul> is "Do I want sequence values or bullets?"
> Never once have I thought "Do I want the computer to randomize my
> text?" :)
>
Of course, I wasn't seriously proposing that the DITA OT adopt a randomzier for
<ul> processing. But I disagree that a bulleted list where the author's order is
inviolate, then it's an ordered list.

For example, someone authors a bulleted list where the order is important
because they build the case for some point of view. In such a case, the items
are related and not independent thoughts. In such cases, authors should decorate
them accordingly. It's our natural language to do so. When people list their
arguments in a debate, they use an ordered construct by saying something like,
"First, I don't like those shoes. Second, they don't match my handbag, and third
-- the don't even fit!".

(:>)

--
Phil

Kevin;

Thanks for the clarification ... after the last couple of days of digging and
your excellent explanation, I now know much more about fonts than I ever wanted
to!

;^)

Bob

--- In dita-users@yahoogroups.com, "samk_y2k" <samk_y2k@...> wrote:
>
> Hi,
> I am new to DITA OT/XMetal/RenderX. Here is an issue I am trying to resolve
and may be it is very simple for experts on this forum:
>
> Note: I am using XMetal enterprise 5.5 to publish to PDF output (that uses XEP
renderer and DITA OT 1.4.2.1)
>
> Issue: The odd and even pages have uniform but different page margins. All odd
pages are slightly shifted towards the left while all even pages are shifted
towards the right.
>
> I want to centre align both even and odd pages so that they have equal margins
on left and right.
>
> I looked at DITA_OT\demo\xmfo\customization\fo\layout_masters.xml but it
doesn't seem to be the file that controls these margin attributes.
>
> Any help will be greatly appreciated!
>
> Regards,
> Sammy
>


Hello Sammy,

The file you want to edit is called custom.xsl
It is in the DITA_OT\demo\xmfo\Customization\fo\xsl\ location. For XMetaL 5.5
the full path is usually something like this:
C:\Documents and Settings\<username>\Application Data\SoftQuad\XMetaL
Shared\DITA_OT\demo\xmfo\Customization\fo\xsl\custom.xsl

You will want to add the following lines to the bottom of the file (before the
end of the stylesheet)

<xsl:param name="xm.page.margin.bottom" select="'1in'"/>
<xsl:param name="xm.page.margin.top" select="'1in'"/>
<xsl:param name="xm.page.margin.inner" select="'0.5in'"/>
<xsl:param name="xm.page.margin.outer" select="'0.5in'"/>


Hope that helps,
Amber

Hi,

I know that in DITA-OT1.4.3 the layout-masters.xml file is no longer used: you
seem to have to copy the createDefaultLayoutMasters template from
layout-master.xsl to your [CUSTDIR]\fo\xsl\custom.xsl file and edit the page
layouts there. Worked for me anyway.
>
> Hi Sammy,
>
> samk_y2k wrote:
> > Can anyone help me please..I am hitting a deadline and that's making me
> > anxious!
>
> As far as I know you are editing the right file.  The layout_masters.xml
> file is definitely the one that we edited to adjust margins.
>
> Since you're using XMetaL, you presumably have a support agreement with
> them.  I'd be asking them rather than dita-users, particularly if you
> have a deadline.
>

Hi Sammy,

samk_y2k wrote:
> Can anyone help me please..I am hitting a deadline and that's making me
> anxious!

As far as I know you are editing the right file.  The layout_masters.xml
file is definitely the one that we edited to adjust margins.

Since you're using XMetaL, you presumably have a support agreement with
them.  I'd be asking them rather than dita-users, particularly if you
have a deadline.

Hi,

Can anyone help me please..I am hitting a deadline and that's making me anxious!

Sammy


--- In dita-users@yahoogroups.com, "samk_y2k" <samk_y2k@...> wrote:
>
> Hi,
> I am new to DITA OT/XMetal/RenderX. Here is an issue I am trying to resolve
and may be it is very simple for experts on this forum:
>
> Note: I am using XMetal enterprise 5.5 to publish to PDF output (that uses XEP
renderer and DITA OT 1.4.2.1)
>
> Issue: The odd and even pages have uniform but different page margins. All odd
pages are slightly shifted towards the left while all even pages are shifted
towards the right.
>
> I want to centre align both even and odd pages so that they have equal margins
on left and right.
>
> I looked at DITA_OT\demo\xmfo\customization\fo\layout_masters.xml but it
doesn't seem to be the file that controls these margin attributes.
>
> Any help will be greatly appreciated!
>
> Regards,
> Sammy
>

At 09:05 AM 2/07/2009, Bryce L Nordgren wrote:

>The order might not be relevant to the human reader, particularly _after_
>the information has been assimilated, but order is always relevant to the
>processor.
>
>Where the rubber meets the road, the only practical difference between
><ol> and <ul> is "Do I want sequence values or bullets?"

The essential practical difference is that the <ol> is meant to provide an
enumerated reference which may be nested.  In this respect (as you pointed out
earlier) it is parallel to the structuring of a document...which, amongst a few
other things, is what DITA is about, actually.

Semantically "unordered list" is contradictory.  All lists are ordered.  The
practical difference is more obvious in the Docbook specification.  It doesn't
have an "unordered list" element at all.  It has <itemizedlist> for when the
semantics don't require "do this, followed by this, followed by..." and
<orderedlist> for when they do.  But what's in a name?  Call it <ul> if you
want, but make sure it's clearly understood that we don't *mean* "unordered". 
What we mean is that the child elements don't need to be interpreted as
consequential (sequentially interdependent).

DITA falls short of enabling the semantic distinction between ordered and
unordered lists. If we are to embrace the standard, so we are directed by
Michael Priestley et al., we are to roll over and agree with the standards
committee when they instruct us that <ol> is just an alternative way of
decorating lists and that the only route for getting it to perform its semantic
role is to devise an idiomatically "global" specialization.  That view
thoroughly misses both the point of specialization and the generalised objective
of standards, I believe.  Sure, I want to be able to specialize an <ol> element
if I need to - but only if I need extend its semantics beyond the [currently
absent] standard semantic expectations for ordered lists.

Another commenter here (Gunnar H. Krause) even admonishes that it's disgraceful
to expect such a specification to be in the "free" standard unless one is
willing to pay someone.  Come on.... :-|

Helen

Just a few comments on this to ensure statements are correct:

> Out of the box, FOP and XEP (and Antenna House, I presume) include the
"Base-14" fonts in Type 1 format. In other words, NOT in unicode format.

This statement is unclear or misleading. In reality, the PDF Specification
requires that any PDF-compliant application be able to understand textual
content in a document if specified in any of the base-14 fonts. Therefore,
metrics are used to calculate placement of characters and actual characters
are used within the viewer (not inside the PDF). The "NOT" part of this
statement is not relevant/incorrect. FO Formatters work with Unicode
characters sets, most (if not all) require them.

> , then it will be up to the end-user's browser / SVG viewer to render the
glyphs. This will, of course, use the operating system font system for
rendering. Since most modern OSes have unicode font packages

Of course. You cannot embed fonts in HTML documents. And if the "modern OS"
does not have the font then your document is displayed (most likely)
incorrectly and there is nothing you can do about it. Lest, of course, you
mandate that even one who views your particular document has the exact font
you specified registered in all the OSes they may have.

Contrasted with the fact the in PDF you can embed the exact font (glyph and
shape information) that you need in order to ensure 100% accuracy.

> The fix we discovered is this: specify a font alias in the xep.xml
configuration file that specifies a fall-through font package processing
order. This is exactly what we were already doing in our XSL-FO style sheets
for processing DITA source into FO.

Great. And this is done exactly correctly and done with best consideration
to file size. You should always attempt to use the base 14 if they fit your
needs because no fonts need be embedded in the PDF. But realize that the
font set of Helvetica, Courier, Symbol and ZapfDingbats is limited.

In fact, in every installation of RenderX, you have a set of FO documents in
the examples directory that renders all the characters in these fonts so you
can see what is "in there". For some that is plenty, for others it lacks
what they need. Some people even add the "magic catch-all font" {highly not
recommended} like Arial Unicode (if they have license to do so, it is not
free despite popular thought):

<font-alias name="sans-serif" value="Helvetica, Symbol, ZapfDingbats, Arial
Unicode MS"/>

Arial Unicode is a 22MB font with many, many characters. if you include one
character for this font, it will add significant size to your PDF because
XEP honors the font copyright information and inserts the header of the
font.

You should also note that you can also do the same selection strategy in XSL
FO with font-family="Helvetica, Symbol, ZapfDingbats, Arial Unicode MS" on
an object-by-object basis if needed to handle such things instead of making
it generic in the formatter configuration. you can even do so on a
character-by-character basis by setting the font selection strategy.

Kevin Brown
RenderX, Inc.
(650) 327-1000 Direct
(650) 328-8008 Fax
(925) 395-1772 Mobile
skype:kbrown01
kevin@...
sales@...
http://www.renderx.com

Hi Bob,

While the error is reported, it never caused problems in the past and, once I
found the workaround, does not cause a problem now. I'm still puzzled as to why
the OT looked in a place other than the temp file it created but I'm not going
to be a stickler on this as the reason I was using M15 was to test specific 1.2
features.

I will look to see if I have the SDK installed (probably not) and point the
variable to it in my startcmd.bat file.

Thanks.

Julio J. Vazquez
SDI Global Solutions

--- In dita-users@yahoogroups.com, "bobthomastagsmiths" <bob.thomas@...> wrote:
>
> Hi Julio,
>
> I am glad that you have solved your problem. However, the following line
> in your build messages is troublesome:
>
> Unable to locate tools.jar. Expected to find it in C:\Program
> Files\Java\jre6\lib\tools.jar
> This suggests that your JAVA_HOME environment variable is pointing to a
> Java JRE installation rather than to a Java SDK installation. The SDK
> the includes tools.jar in its lib directory, but the JRE does not.
>
> Regards,
>
> Bob Thomas
>   [Tagsmiths, LLC]  <http://www.tagsmiths.com/>
> +1 720 201 8260
>
>
> --- In dita-users@yahoogroups.com, "Julio Vaquez" <julio_v27612@>
> wrote:
> >
> > I found a solution to my problem. I redefined the temp directory the
> OT uses in my ANT task to look at where the temp files were actually
> being built. This is the statement I added to the ANT file:
> >
> >
> > <property name="dita.temp.dir"
> value="${basedir}${file.separator}temp"/>
> >
> > HTH someone else too.
> >
> > Julio J. Vazquez
> > SDI Global Solutions
> >
> > --- In dita-users@yahoogroups.com, "Julio Vaquez" julio_v27612@ wrote:
> > >
> > > Hi again,
> > >
> > > I'm getting a feeling that I'm somehow doing something that the OT
> doesn't like, but I can't figure it out for the life of me. From the
> appended messages, it looks like DITA is trying to find something in the
> temp subdirectory of the DITA OT directory, but it doesn't exist.
> However, all the files are copied to a temp directory under the
> directory where my source is, so I'm clueless as to what I may have done
> to bollix up the OT.
> > >
> > > I'm appending my task contents and then the entire message stream.
> > >
> > > Julio J. Vazquez
> > > SDI Global Solutions
> > >
> > > Ant Task:
> > >
> > > <project name="sample_xhtml" default="sample2xhtml" basedir=".">
> > >
> > >   <!-- dita.dir should point to the toolkit's root directory -->
> > >   <property name="dita.dir"
> value="${basedir}${file.separator}..${file.separator}.."/>
> > >
> > >   <!-- if file is a relative file name, the file name will be
> resolved
> > >        relative to the importing file -->
> > >   <import file="${dita.dir}${file.separator}integrator.xml"/>
> > >
> > >   <target name="sample2xhtml" depends="integrate">
> > >     <ant antfile="${dita.dir}${file.separator}build.xml"
> target="init">
> > >       <!-- please refer to the toolkit's document for supported
> parameters, and
> > >            specify them base on your needs -->
> > >       <property name="args.input"
> > >
> value="..${file.separator}..${file.separator}..${file.separator}Test-1.2\
> ${file.separator}my_samples.ditamap"/>
> > >       <property name="output.dir"
> > >
> value="..${file.separator}..${file.separator}..${file.separator}Test-1.2\
> ${file.separator}xhtml"/>
> > >       <property name="transtype" value="xhtml"/>
> > >     </ant>
> > >   </target>
> > > </project>
> > >
> > > Build messages:
> > >
> > > C:\DITA-OT15M15>ant -f samples/ant_sample/test15_xhtml.xml
> > > Unable to locate tools.jar. Expected to find it in C:\Program
> Files\Java\jre6\li
> > > b\tools.jar
> > > Buildfile: samples\ant_sample\test15_xhtml.xml
> > >
> > > integrate:
> > > [integrate] Using XERCES.
> > >
> > > sample2xhtml:
> > >
> > > init:
> > >
> > > start-process:
> > >
> > > init-logger:
> > >
> > > use-init.envhhcdir:
> > >
> > > use-init.hhcdir:
> > >
> > > use-init:
> > >
> > > check-arg:
> > >      [echo]
> *****************************************************************
> > >      [echo] * basedir = C:\DITA-OT15M15\samples\ant_sample
> > >      [echo] * dita.dir = C:\DITA-OT15M15\samples\ant_sample\..\..
> > >      [echo] * input = ..\..\..\Test-1.2\my_samples.ditamap
> > >      [echo] * transtype = xhtml
> > >      [echo] * tempdir = temp
> > >      [echo] * outputdir = ..\..\..\Test-1.2\xhtml
> > >      [echo] * extname = .xml
> > >      [echo] * clean.temp = ${clean.temp}
> > >      [echo] * xslt.parser = SAXON
> > >      [echo]
> *****************************************************************
> > >
> > > output-deprecated-msg:
> > >
> > > output-css-warn-message:
> > >
> > > output-msg:
> > >
> > > build-init:
> > >
> > > start-preprocess:
> > >
> > > clean-temp:
> > >
> > > gen-list-flag-check:
> > >
> > > gen-list-without-flagging:
> > >
> > > gen-list-with-flagging:
> > >
> > > gen-list:
> > >
> > > debug-filter-flag-check:
> > >
> > > debug:
> > >
> > > debug-and-filter:
> > >
> > > debug-filter:
> > >
> > > copy-image-check:
> > >
> > > copy-image:
> > >
> > > copy-html-check:
> > >
> > > copy-html:
> > >
> > > copy-flag-check:
> > >
> > > copy-flag:
> > >
> > > copy-subsidiary-check:
> > >
> > > copy-subsidiary:
> > >
> > > copy-generated-files:
> > >      [copy] Copying 2 files to C:\Test-1.2\xhtml
> > >
> > > copy-files:
> > >
> > > conrefpush-check:
> > >
> > > conrefpush:
> > >  [pipeline] java.io.FileNotFoundException:
> C:\DITA-OT15M15\temp\dita.list (The s
> > > ystem cannot find the path specified)
> > >
> > > BUILD FAILED
> > > C:\DITA-OT15M15\samples\ant_sample\test15_xhtml.xml:24: The
> following error occu
> > > rred while executing this line:
> > > C:\DITA-OT15M15\build.xml:59: The following error occurred while
> executing this
> > > line:
> > > C:\DITA-OT15M15\build_preprocess.xml:105:
> java.lang.NullPointerException
> > >
> > > Total time: 5 seconds
> > > C:\DITA-OT15M15>
> > >
> >
>

> IMHO, the semantic difference between ordered lists (<ol>) and
> unordered lists (<ul>) is not that they are rendered with this and
> that or such and so glyphs as lead-ins to line items (<li>). It's
> that the order of the line items themselves conveys meaning, or it
doesn't.

"Conveys meaning" to what audience: the processor or the human reader?
Perhaps the items are actually independent, but it is helpful to expose
the reader to them in a specific order.

> With unordered lists, the order in which bulleted entries are
> rendered is completely arbitrary.

I must say that this opinion contradicts the processing expectations I
have built up over a couple of decades using word processors and markup
languages.  Html browsers, LaTeX, TeX, Microsoft Word, OpenOffice,
KOffice.... every single one of these processing systems respects the
order of the source.  This makes all lists de-facto ordered lists
regardless of how the list items are decorated.

The order might not be relevant to the human reader, particularly _after_
the information has been assimilated, but order is always relevant to the
processor.

Where the rubber meets the road, the only practical difference between
<ol> and <ul> is "Do I want sequence values or bullets?"  Never once have
I thought "Do I want the computer to randomize my text?" :)

Bryce

Did this file ever get committed for 1.4.3?

      Steve

--- In dita-users@yahoogroups.com, "Martin Polley" <martin.polley@...> wrote:
>
> Thanks, Robert.
>
> Martin
>
>
> On Tue, Feb 26, 2008 at 6:23 PM, Robert D Anderson <robander@...>
> wrote:
>
> >   The functionality is not undesirable. I'm not sure why it was not
> > committed
> > - I think that it was on the list to be committed, so something could have
> > been missed; I'll check and see why it was not added.
> >
> > Robert D Anderson
> > IBM Authoring Tools Development
> > Chief Architect, DITA Open Toolkit
> > (507) 253-8787, T/L 553-8787 (Good Monday & Thursday)
> >
> > dita-users@yahoogroups.com <dita-users%40yahoogroups.com> wrote on
> > 02/26/2008 02:48:36 AM:
> >
> >
> > > Hi folks,
> > >
> > > I have been happily using a patched version of conrefImpl.xsl with
> > > OT 1.3.1 for ages (which allowed me to conref specialized elements
> > > into a more general context, e.g., to pull the content of a <step> into
> > a
> > <li>
> > > ).
> > >
> > > Now I have upgraded to 1.4.1, I see that this ability has disappeared.
> > >
> > > Was it decided that this functionality was undesirable? Or was
> > > Deborah's original patch in fact never committed? (I notice that my
> > > original request is still open...)
> > >
> > > Thanks,
> > >
> > > --
> > > Martin Polley
> > > Technical Communicator
> > > +972 52 3864280
> > > <http://capcloud.com/>
> > >
> >
> >
> >
>

Hi Bob,

That line is not really "troublesome". One can still transform successfully with
the DITA OT without having a full Java SDK.

Ant "complains" about not finding tools.jar because Ant thinks it will be called
upon to compile Java at any moment in time, and so it looks for tools.jar
immediately before realizing 'oh yeah, don't have to compile for this task.'

Unless your workflow or Build Manager requires absolutely no warning build
messages, it's OK to get that particular build-time message.

Best regards,
Lee Anne

--- In dita-users@yahoogroups.com, "bobthomastagsmiths" <bob.thomas@...> wrote:
>
> Hi Julio,
>
> I am glad that you have solved your problem. However, the following line
> in your build messages is troublesome:
>
> Unable to locate tools.jar. Expected to find it in C:\Program
> Files\Java\jre6\lib\tools.jar
> This suggests that your JAVA_HOME environment variable is pointing to a
> Java JRE installation rather than to a Java SDK installation. The SDK
> the includes tools.jar in its lib directory, but the JRE does not.
>
> Regards,
>
> Bob Thomas

Success! I reverted back to an older version of the FO.jar and was finally able
to generate a PDF. Again, thanks so much.

Unable to locate tools.jar. Expected to find it in C:\Program Files\Java\jre6\lib\tools.jar

I spent some significant time on discussion forums and poking through various
config guides to find what amounted to a fairly simple fix. I guess this is a
case where drawing the X is very simple ... knowing WHERE to draw the X took
some time (for non-US folks, that's a reference to a anecdotal story about
Thomas Edison).

Quick background:

When an SVG is linked from a DITA topic, the OT XSLTs pass that reference
through and never touch the SVG code. Thus, the OT is "blind" to font
specifications within SVG files. That's fine ... it leaves the rendering of the
glyphs up to the display engine.

If the OT transtype is XHTML, then it will be up to the end-user's browser / SVG
viewer to render the glyphs. This will, of course, use the operating system font
system for rendering. Since most modern OSes have unicode font packages, symbol
glyphs are usually handled correctly.

If, however, the OT transtype is PDF, then the FO processor will be responsible
for rendering the glyphs in the SVG, and here we have a problem. Out of the box,
FOP and XEP (and Antenna House, I presume) include the "Base-14" fonts in Type 1
format. In other words, NOT in unicode format.

Thus, when I processed my SVG files with Unicode symbol glyphs embedded, XEP had
no way to render them, as it couldn't find those glyphs in the Helvetica Type 1
font package.

The fix we discovered is this: specify a font alias in the xep.xml configuration
file that specifies a fall-through font package processing order. This is
exactly what we were already doing in our XSL-FO style sheets for processing
DITA source into FO.

In the OOTB XEP configuration file, these aliases are already present:

<font-alias name="monospace" value="Courier"/>
<font-alias name="sans-serif" value="Helvetica"/>
<font-alias name="serif" value="Times"/>

We simply modified those aliases to add the fall-through order:

<font-alias name="monospace" value="Courier, Symbol, ZapfDingbats"/>
<font-alias name="sans-serif" value="Helvetica, Symbol, ZapfDingbats"/>
<font-alias name="serif" value="Times, Symbol, ZapfDingbats"/>

And since we've got a lot of Windows users creating SVGs with Arial, Times New
Roman and Courier New, we also added these aliases:

<font-alias name="Arial" value="Helvetica, Symbol, ZapfDingbats"/>
<font-alias name="Courier New" value="Courier, Symbol, ZapfDingbats"/>
<font-alias name="Times New Roman" value="Times, Symbol, ZapfDingbats"/>

The result is that now symbol glyphs in SVGs are rendered properly by XEP into
PDF.

If anyone has other ideas for a better way, please let me know!

Regards,
Bob

Hi
I have to create three output from the same topics. I created three DITA maps,
A, B, and C. A has all the seven topics, B has only two out of the seven topics
and C has only three out of the seven topics. My manager want me to send him
three output at a time to integrate with the application. Output file of
A.ditamap should provide a complete toc, B should display only two topics in toc
and D should display only three topics. Also, there is a pagination link in the
topics that comes under A, but he does not want the same links to appear in B
and D, but content should be same. He asked me to create three different folders
in SVN an keep the output in separate folders. I created the files using
ditaval, but I am able to produce only one output at a time and cannot generate
three outputs using the same project files. How do I do it? Please help.

Regards
Julie

The July meeting of the Silicon Valley DITA Interest Group (SVDIG) is
scheduled for Wednesday, July 8 at 7:00pm at NetApp in Sunnyvale, CA.


Topic:  Making a good start: Understanding DITA basics.

Speaker:  Joanne Grey, Sr. Technical Author at Maxiscale

- For technical communicators just beginning in DITA, or those
considering moving to DITA, this presentation discusses the basic
concepts, terms, and information you need to know. You will also learn
of a few real-life "gotchas" from someone who's been there.


Agenda:
7:00-7:30 Mingling
7:30-8:30 Presentation
8:30-9:00 Wrap up

Please RSVP to Scott Prentice by ** 3PM July 7 ** at <sp@...>
if you're planning to attend. We need the list of all attendees ahead of
time for security reasons. If you RSVP late, you'll need to arrive early
(by 6:45) so a badge can be printed for you by NetApp security.

Directions: http://svdig.ditamap.com/bldg4-location.html

NetApp
Building 4 / Aragorn room
1330 Geneva Drive, Sunnyvale, CA 94089
main operator: 408-822-6000

Meet at 7:00PM in the lobby to be escorted in. James Hom will be our
official NetApp escort.

For more information or to download past presentations visit
http://svdig.ditamap.com/

dita-users@yahoogroups.com wrote on 06/30/2009 02:45:23 PM:

> This issue, of developing standardized ways of specifying
> processing parameters for DITA, came up last December in
> this thread:
> http://tech.groups.yahoo.com/group/dita-users/message/12632
>

I was six messages in before I finally figured out what a PI was. :)

> However, as you'll see there, while the test suite itself is
> steadily growing, there's not much yet on the settings and PIs,
> much less DITA-OT plugins to implement them. We have some
> drafts for the PIs, but the settings file has a ways to go. ;-)

I don't see a lot of overlap between the discussion and a settings file,
other than the fact that you'd like to have one (a sentiment that I'd like
to encourage!)  Most of the PIs go straight into the content, right?  I
can't think of how a "page break" instruction would be used in the
settings file...

I guess that means I have a totally blank slate, eh? :)

> One problem is that there has been little outside participation,
> and as a DITA vendor (for Mif2Go, used to convert Frame files
> to DITA (including unstructured Frame)), we're not in a good
> position to produce an *unbiased* informal standard. If you
> have concrete ideas about this, and also the time to work on
> them (in short supply here), you'd be very welcome to join in.

Architectural ideas have been percolating in my head since I hit the
<xref> wall [again] in March and put up all those wiki pages.  This
discussion is the first time I've thought in terms of a separate settings
file, as I was previously thinking mostly of adding elements to the <map>
or <bookmap> elements.  My time is very sporadically available in
unpredictible ways.  In between, I lurk on the list looking for people who
want/need to do the same thing. :) The wiki pages are useful because they
help me figure out where I left off last time (I also had trouble
attracting adherents).

Also, I'm much better at thinking about data models than hamhanding some
XML.  Any DTDs and absolutely all XSLT I write should be reviewed by
someone competent, preferably someone who "gets" the DITA way.  I
understand pieces of it, but other parts of the philosophy either escape
me or strike me as harmful.  So someone who's "in the fold" should
probably temper my blasphemy for the innocent flock. :)

I'll keep in touch with any developments.  Are you using this list or does
that project have its own list?

Bryce

Does anyone know of an environmental cause (access to Java objects, etc.) that
one user can run a build and the args.css parameter gets set in the xhtml
output, and another user runs it and it doesn't? This is 2 differnt users on one
machine.

Any help appreciated, thanks.

Julia

--- In dita-users@yahoogroups.com, Eliot Kimber <ekimber@...> wrote:
<snip>
> I think the only way you could do that in a standard-defined
> way would be to use the DITA 1.2 conref push feature, where you
> conref the section and then conref push the version of the text
> you want into it. But that seems a little cumbersome for what
> should be a fairly simple thing to do.
<snip>

We're basically discussing "conditional text", which for many tech writers is
anything but simple. :).

Cannot one create the library of re-useable sections and use <ph> elements
around the conditional text fragements? Then, assign an approrpiate value to a
filtering attibute in each <ph>, such as:

otherprops="gender:male"
-- or --
otherprops="gender:female"

When you initiate a production process on a map that conrefs the commmon
sections, specify a ditaval file with these two lines set appropriately:

<prop att="otherprops" val="gender:male" action="include" />
<prop att="otherprops" val="gender:female" action="exclude" />

One thing though that I'm not sure about -- are conrefs processed before
applying ditavals or afterwards? If conrefs are processed afterwards, this
scheme will not work.

-- phil

--- In dita-users@yahoogroups.com, Helen Borrie <helebor@...> wrote:
<snip>
> This isn't a styling issue, IMO, but a structural one ...
</snip>

IMHO, the semantic difference between ordered lists (<ol>) and unordered lists
(<ul>) is not that they are rendered with this and that or such and so glyphs as
lead-ins to line items (<li>). It's that the order of the line items themselves
conveys meaning, or it doesn't.

With unordered lists, the order in which bulleted entries are rendered is
completely arbitrary. If a production processor were to apply a randomizer to
line-itme order and place them in different orders each time, the meaning of the
items and the list as a whole should remain the same. By implication, list items
in an unorderd list must be completely independent of each other. Here's an
example.

When buying a pair of shoes, consider the following:
* What color do you like?
* Do they match your hand bag?
* Are they formal or casual?
* Do you prefer aligator or patent leather?
* Do they fit your feet?

The above list is a semantically unordered list, because no matter the order of
the items, the meaning of the set of considerations remains the same. When it
comes to styling, whether the items have asterisks, bullet symbols, or little,
high-heeled shoe graphics as lead-ins should not be specified in DITA source.
Instead of an authoring-time issue, it's really a production-time one.

With ordered lists, the order of items of course conveys meaning. List items in
ordered list are highly correlated; they are not independent. They "build upon"
the ones that came before and "set the stage" for ones that come after. Here's
an example.

To buy a pair of shoes:
1. Go the mail.
2. Visit the shoe store.
3. Pick out a pair that you like.
4. If they are comfortable and your really like them, and they
    are within your budget, buy them.

The use of numeric glyphs in ordinal order is simply a stylistic convention. It
lets the reader know before-hand that the list has this
"build-upon/set-the-stage" semantics. The above ordered list could be rendered
as follows:

To buy a pair of shoes:
- Go the mail.
- Visit the shoe store.
- Pick out a pair that you like.
- If they are comfortable, and your really like them, and they
   are within your budget, then buy them.

The above rendering remains an ordered list, which anyone can determine simply
by reading it, even though a bullet symbol preceds each one. It conveys the same
meaning as the ordered list that uses 1, 2, 3, and 4 as lead-ins. So, to me the
style of lead-in glyphs is a purely stylistic concern with ordered and unordered
lists alike.

<snip>
>The semantics of <ul> and <ol> have certain commonalities, but they
>are not the same.  The bullet image for <ul> items is stylistic,
>without any semantic underlay.  The value of the character subset
>for <ol> enumeration is only semantic and there should be provision >to specify
that as an attribute of <ol>.
</snip>

IMHO, whether lead-ins to items in ordered lists are Arabic numerals, upper-case
Roman numbers, lower-case Roman number, English alpabetic letters, Hebrew
numeral, or Klingon glyphs, the lead-ins per-se convey no meaning. For example,
the below two ordered lists use different lead-ins, but their meanings are
identical.

To buy a pair of shoes:              To buy a pair of shoes:
1. Go the mail.                      a. Go the mail.
2. Visit the shoe store.             b. Visit the shoe store.
3. Pick out a pair of shoes     .    c. Pick out a pair of shoes.
4. Try them on for fit.              d. Try them on for fit.
5. Buy them.                         e. Buy them.

One could argue that the list on the right appears to be part of some larger
list, but that's only because of a stylistic convention in modern, Occidental
writing that the order of precedence for compound ordered lists is 1. a. i., or
er ... , is 1) A) I), or er ... is 1., 1.1, 1.1.1. Hmmmm.

From a DITA perspective, re-use and re-purpose are the driving force behind the
framework's design. Just because you embed an ordered list within an another
ordered list, that doesn't mean that the subornidate list might not be useful as
a stand-alone list in some context other than the one for which your originally
authored it.

If DITA authors mark subordinate lists with number formats, lists could only be
reused at the same subornidation level.

For example, consider the below, compound, ordered list.

My ordered list:
1. Cccccc c cc c.
2. Mmmmm M Mmmm m.
3. Gggg gggg.
    a. Hhhh Hhh.
    b. Pppppp Pp pp p.
       i. Nnnnn n nnnn N.
       ii. Dddddd d dd d.
       iii. Rrrr r rrrr rr.
4. Qqqq qqq qqq Qqqq.

DITA re-use allows one to re-purpose item 3. in the above ordered list. Here's
an example.

To Ggg ggggg:
1) Hhhh Hhh.
2) Pppppp Pp pp p.
    a) Nnnnn n nnnn N.
    b) Dddddd d dd d.
    c) Rrrr r rrrr rr.

Or, one could re-purpose subordinate item b, as in:
To Pppppp Pp pp p:
1/ Nnnnn n nnnn N.
2/ Dddddd d dd d.
3/ Rrrr r rrrr rr.

If DITA authors were to fix the lead-in ordinal styling at authoring time (i.e,
in the DITA source), the opportunity for re-use would be obliterated completely.
(IMHO). :)

-- Phil

I found a solution to my problem. I redefined the temp directory the OT uses in
my ANT task to look at where the temp files were actually being built. This is
the statement I added to the ANT file:


<property name="dita.temp.dir" value="${basedir}${file.separator}temp"/>

HTH someone else too.

Julio J. Vazquez
SDI Global Solutions

--- In dita-users@yahoogroups.com, "Julio Vaquez" <julio_v27612@...> wrote:
>
> Hi again,
>
> I'm getting a feeling that I'm somehow doing something that the OT doesn't
like, but I can't figure it out for the life of me. From the appended messages,
it looks like DITA is trying to find something in the temp subdirectory of the
DITA OT directory, but it doesn't exist. However, all the files are copied to a
temp directory under the directory where my source is, so I'm clueless as to
what I may have done to bollix up the OT.
>
> I'm appending my task contents and then the entire message stream.
>
> Julio J. Vazquez
> SDI Global Solutions
>
> Ant Task:
>
> <project name="sample_xhtml" default="sample2xhtml" basedir=".">
>
>   <!-- dita.dir should point to the toolkit's root directory -->
>   <property name="dita.dir"
value="${basedir}${file.separator}..${file.separator}.."/>
>
>   <!-- if file is a relative file name, the file name will be resolved
>        relative to the importing file -->
>   <import file="${dita.dir}${file.separator}integrator.xml"/>
>
>   <target name="sample2xhtml" depends="integrate">
>     <ant antfile="${dita.dir}${file.separator}build.xml" target="init">
>       <!-- please refer to the toolkit's document for supported parameters,
and
>            specify them base on your needs -->
>       <property name="args.input"
>        
value="..${file.separator}..${file.separator}..${file.separator}Test-1.2${file.s\
eparator}my_samples.ditamap"/>
>       <property name="output.dir"
>        
value="..${file.separator}..${file.separator}..${file.separator}Test-1.2${file.s\
eparator}xhtml"/>
>       <property name="transtype" value="xhtml"/>
>     </ant>
>   </target>
> </project>
>
> Build messages:
>
> C:\DITA-OT15M15>ant -f samples/ant_sample/test15_xhtml.xml
> Unable to locate tools.jar. Expected to find it in C:\Program
Files\Java\jre6\li
> b\tools.jar
> Buildfile: samples\ant_sample\test15_xhtml.xml
>
> integrate:
> [integrate] Using XERCES.
>
> sample2xhtml:
>
> init:
>
> start-process:
>
> init-logger:
>
> use-init.envhhcdir:
>
> use-init.hhcdir:
>
> use-init:
>
> check-arg:
>      [echo] *****************************************************************
>      [echo] * basedir = C:\DITA-OT15M15\samples\ant_sample
>      [echo] * dita.dir = C:\DITA-OT15M15\samples\ant_sample\..\..
>      [echo] * input = ..\..\..\Test-1.2\my_samples.ditamap
>      [echo] * transtype = xhtml
>      [echo] * tempdir = temp
>      [echo] * outputdir = ..\..\..\Test-1.2\xhtml
>      [echo] * extname = .xml
>      [echo] * clean.temp = ${clean.temp}
>      [echo] * xslt.parser = SAXON
>      [echo] *****************************************************************
>
> output-deprecated-msg:
>
> output-css-warn-message:
>
> output-msg:
>
> build-init:
>
> start-preprocess:
>
> clean-temp:
>
> gen-list-flag-check:
>
> gen-list-without-flagging:
>
> gen-list-with-flagging:
>
> gen-list:
>
> debug-filter-flag-check:
>
> debug:
>
> debug-and-filter:
>
> debug-filter:
>
> copy-image-check:
>
> copy-image:
>
> copy-html-check:
>
> copy-html:
>
> copy-flag-check:
>
> copy-flag:
>
> copy-subsidiary-check:
>
> copy-subsidiary:
>
> copy-generated-files:
>      [copy] Copying 2 files to C:\Test-1.2\xhtml
>
> copy-files:
>
> conrefpush-check:
>
> conrefpush:
>  [pipeline] java.io.FileNotFoundException: C:\DITA-OT15M15\temp\dita.list (The
s
> ystem cannot find the path specified)
>
> BUILD FAILED
> C:\DITA-OT15M15\samples\ant_sample\test15_xhtml.xml:24: The following error
occu
> rred while executing this line:
> C:\DITA-OT15M15\build.xml:59: The following error occurred while executing
this
> line:
> C:\DITA-OT15M15\build_preprocess.xml:105: java.lang.NullPointerException
>
> Total time: 5 seconds
> C:\DITA-OT15M15>
>