[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: DITA Translation SC meeting minutes - Monday 27 February 2006
Hi all, Here are the minutes from today's Translation SC meeting. Please could those of us who have action items start working on them ASAP so we can have feedback in time for next Monday's SC meeting. Best Regards, Gershon --- Gershon L Joseph Member, OASIS DITA and DocBook Technical Committees Director of Technology and Single Sourcing Tech-Tav Documentation Ltd. office: +972-8-974-1569 mobile: +972-57-314-1170 http://www.tech-tav.com
MEETING MINUTES -- 27 February 2006 -- DITA TRANSLATION SUBCOMMITTEE
(Minutes taken by Gershon Joseph <gershon@tech-tav.com>)
Date: Monday, 27 February 2006
Time: 08:00 - 09:00 PST
DITA Translation Subcommittee resources:
- SC Web site: http://www.oasis-open.org/apps/org/workgroup/dita-translation/index.php
- Mailing list: dita-translation@lists.oasis-open.org (non-OASIS members
please email Gershon or Don and we'll post on your behalf)
- Roll call -> Don Day, Nancy Harrison, Rodolfo Raya, Robert Anderson,
Kevin Farwell, Chris Wong, Gershon Joseph, Andrzej Zydron, Charles Pau,
JoAnn Hackos (chair), Sukumar Munshi, Christian Lieske, Patrick Klassen,
Bryan Schnabel
- Review/approve minutes from previous meeting (20 February 2006):
http://www.oasis-open.org/apps/org/workgroup/dita-translation/email/archives/200602/msg00001.html
Don moves to approve, Nancy seconds.
- JoAnn -- Any other urgent issues to discuss today before we move to the next
on the agenda?
- Andrzej -- How should SC members communicate?
- Don -- It was officially easier to set up the liaison as a DITA
subcommittee with the XLIFF TC. Non XLIFF TC members can also join the
DITA Translation SC. Non-OASIS members will be sent emails directly (not
via the SC's official email alias). Non-OASIS members will be observers
and can provide input and feedback to the SC via the SC's comment form
or via another OASIS member. Non-OASIS members should send emails to
Gershon or Don, who will forward them to the SC members and observers.
- Don updated SC on revised DITA roadmap for DITA 1.1, which includes a
placeholder on input from the SC for changes to the DITA spec required
for translation. We therefore need to provide this feedback urgently.
- SC Business Task 1: Review and rollup of recommended specification updates
for DITA 1.1.
- DISCUSSION
- Don -- DITA was originally designed using the following papers for
guidance:
- Richard Ishida's paper on DTDs for translation
- Yves Savorel document on xml:lang
1. Is our present markup correct?
2. Is our present documentation correct?
[BACKGROUND: DITA uses xml:lang and the translate attribute for translation.]
- Andrzej -- Has anybody actually used xml:lang or translate attribute?
- Robert -- Rare usage on xml:lang and translate attribute, e.g. for
API names and product names.
- Kevin -- Uses it for localization projects. It defines the language
of content in order to process generated text.
Does not think the lang attribute should be used on anything that's
not going to be translated.
- Andrzej -- Is xml:lang used on the source document prior to translation?
- Kevin -- xml:lang is not very well designed, so we adjusted the DTD
to support a specific subset of languages and symbols were hard-coded
so users had to choose from the available list.
Defaults as English. Used on source documents, but not actively
set by the users.
- Robert -- IBM only uses xml:lang at the topic level.
- Andrzej -- Should restrict xml:lang to the topic level.
- Christian -- Suggests comparing the description in the DITA spec to
the description in the official ITS W3C spec of xml:lang.
Also, we should check whether the xml:lang attribute appears on
all types of entities that appear in DITA content, e.g. graphics
and references to graphics.
Should DITAMAP carry the xml:lang attribute?
- Robert -- Yes, DITAMAP does. Also, DITA-OT lets the user set the
default xml:lang value, so it's easy to change.
- Don -- Not all editors insert predefined attributes. Therefore, we
would have to set a default in the DTD, which is not the way to go.
Therefore, the spec should state that the default is open and tools
must set the default as needed. Also, translation tools should set
xml:lang as relevant for the translated document.
- Robert -- xml:lang will be universally available in DITA 1.1.
- Kevin -- if it's not translatable text, it should be language neutral,
e.g. <table> and <row>. xml:lang should not be allowed everywhere,
since authors may set a <table> to one language when the <topic>
is set to another language.
- Robert -- xml:lang should be set on <topic> and allowed to inherit,
to prevent setting it on all elements (e.g. <title> and <shortdesc>,
which is cumbersome).
- Andrzej -- e.g. of xml:lang in <table> is a table of equivalents e.g.
Farsi and English, since cells would be in different languages.
Most users will need only xml:lang on topic level only, but a
small percentage of documents may have multilingual content that
needs explicit markup within the elements inside the <topic>.
Recommend as best practice having xml:lang at topic level.
-- ACTION ITEM -- Richard Ishida to ensure DITA xml:lang spec is correct according
to W3C ITS spec regarding xml:lang.
[End of discussion on xml:lang; start of discussion on translate attribute]
BACKGROUND - translate = yes|no; attribute is universally available
on all elements.
- Andrzej -- ITS worked on scoping rules allowing rules to be
specified outside of the document itself. Practically, what's in
DITA 1.1 is sufficient for any processing tool to skip text that
must not be translated.
- Christian --ITS has mechanisms to override the translate attribute.
Also, not good practice to not put translatable content into
attributes.
- Don -- work is in progress to replace translatable attributes with
elements in DITA in the future.
- Christian -- ITS addresses the schema developer who should use ITS
inside the DTD/schema for translation parts of the DTD.
- Andrzej -- What we have already [in DITA] is sufficient.
- Don -- question to Robert - does the table of inline elements
address translatable attributes?
- Robert -- Very few.
Andrzej -- Let's publish the table as an additional note or addendum
with the [DITA] spec as guidance to localizers.
- Don -- Agreed, and the table would be a good basis for the ITS
recommendation.
-- ACTION ITEM -- Robert and Chris to check for translatable attributes
and update their table accordingly.
- Christian -- Would like to discuss the overlap between DITA translate
attribute and what ITS is doing with the ITS group.
-- ACTION ITEM -- Christian to discuss overlap between DITA translate
attribute and ITS work with ITS TC and get us a response for next
week's SC meeting.
- Christian -- Felix may be able to contribute to SC via conf call.
-- ACTION ITEM -- Gershon to invite Felix to participate in next week's
SC call.
- QUESTION: What about markup support for ruby/rubi and text direction
in the DITA spec?
- Don -- Perhaps we should re-address support for rubi and dir=rtl|ltr.
-- ACTION ITEM -- Gershon to write proposal for role of internationalization
attributes, in particular discussion of rubi and directionality.
Proposal must include use cases for authoring. Kevin to work with
Gershon (limited time commitment) and will review Gershon's proposal.
Christian asks to be involved too.
- Don -- Could any attributes we want to add be universal attributes?
- Meeting adjourned. The following agenda items will be rolled into next meeting:
Discussion items
1. Continue refining the typical workflow for DITA content
2. Do we need to discuss packaging of content for how material enters the
translation workflow?
Note:
questions related to packaging are of interest to people
working on the process. e.g. do we only ship source material and source
language or also rendered versions of docs so the translators can see the
source language in rendered format?
How much context do authors and translators need?
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]