Re: [dita-learningspec] Comments on work document - not so sure aboutembedding topics

[john_hunt@us.ibm.com]

France,

Thanks for the good comments. Here's some
replies.

>>I'm not getting the specialization
for learning content. Shouldn't that be just regular topics that can be
ordered in map instead of embedded into a learning object? My concern is
for reuse in reference guide vs training guide. If learning topics exist
because we need to add some elements to regular topics, can't we just propose
to add these elements to all topic types? Adding elements, unless we make
them mandatory should not bring up any backward compatibilty issues. <<

JH: First, I don't think we've ruled out
including additional topic types in a map. Phase 2 of the design plan addresses
the learning topic types we need to design, and Phase 3 addresses the design
of a specialized map. I'd be supportive of allowing for direct use of task-concept-reference
(etc) topics in the learning map.

Second, I also think that the learning design needs a learning content
type, to enable a place to author and include additional types of content,
as well the option to provide a learning-specific context for the content,
such as objectives, summaries, keypoints, perhaps. 

It also seems to me that changing the design
of existing types by adding new elements makes these topics less reusable
than enabling them to be embedded/nested as content inside a new type.
(Up for input and discussion), but we also may want to enable nesting several
related task-concept-reference topics into a single learning content topic,
which provides a single learning context for those nested topics. 
 
>>I do agree with the framework topics.
These topics are interesting to the learning process and would probably
not contain anything 'informational' in nature, they just provide context
for the learing environment. I can easily imagine that some of them can
be created automatically (overview, summary, assessment), at least in part,
from content in the core topics. That could be possible if topics are built
correctly, with a learning short description, for example. <<

JH: Yes - I agree: these framework topics
should be relatively straightforward specializations of topic. For a starting
example, see http://wiki.oasis-open.org/dita/LearningContent/LearningOverviewTopicType.
 
>>What I don't like is the idea of
the bucket topic where other topics are 'embedded'. What is the added value
of this bucket? Doesn't it limit reuse opportunities with non-learning
based documents? Isn't it better to add required learning elements to all
basic topic types (as non-mandatory for backward compatibility). This way
topics created for learning can be reused for other types of documents
by hiding the learning specific stuff in processing. And vice versa. If
the 'bucket' is for organizational purposes only, why don't we just have
a section our map model that could become a 'learning content section/chapter/module/whatever
name we want to give it' where topics are are referenced with the standard
topicref mechanism?.<<

JH: Yes, I like the idea of a specializaed
map section for learning content where you can directly topicref task-concept-reference-etc
content. I also think we need a learning content topic, as well, to enable
additional learning content. Then the issue is: do we enable for nesting
of task-concept-reference infotypes in this learning content topic. I anticipate
a good design discussion on that. Live samples will help a lot.

>>Proposal 2
I already have a specialization for exercises,
if you guys would like to use it to get started, I'd be happy to provide
it. <<

YES! Please do.
 
>>Proposal 5
I'm still not totally getting SCORM and how
it integrates with DITA. I see it as an extra layer on top of topics and
maps... I'm going give all my trust to those who understand it better than
me :-) <<

JH: I see SCORM primarily in terms of output
processing. Just as we can now process to HTML, PDF, Eclipse Help, etc,
with learning, processing out to standards-based SCORM is a priority. At
the minimum, SCORM is a packaging layer on top of default HTML processing.
Then it can grow from there to include navigation/sequencing, metadata
(IEEE LOM), and more. Our challenge will be in identifying those sub-sets
to support!

>>Overall comment
It is my impression that training requires
more linearity than reference guides. People need to understand how different
tasks are linked to each other, and how they can be put together in order
to perform metatasks. (Different between system task and user task, where
the user task may require performing 2 or more system tasks in a specific
order.). I'm not sure we have anything to answer that issue. Maybe that
is what the learning content specialization is tying to answer? <<

JH: Good observation, and a key challenge
in applying DITA best practices for intent-based, topic-oriented authoring
to learning content. This is part of the debate around learning objects.
And, yes, I hope this is what the learning topic and map specializations
can answer - how to enable good structured topic-based content for learning
and training that also supports good instructional design. 

Again, samples, samples, samples - let's
gather really solid samples of good learning content, and see how this
DITA learning design holds up against it.

John Hunt, IBM

From:	"France Baril" <France.Baril@ixiasoft.com>
To:	<dita-learningspec@lists.oasis-open.org>
Date:	02/09/2007 04:32 PM
Subject:	[dita-learningspec] Comments on work document - not so sure about embedding topics

---

Hi, I sent this to John a few weeks ago,
I'm finally sending it to you guys. I'm doing a lot of travelling and am
unable to attend most meetings, but feel free to contact me by email if
you have comments about this.
 
Question related to proposal 1 and 2. 
I'm not getting the specialization for learning
content. Shouldn't that be just regular topics that can be ordered in map
instead of embedded into a learning object? My concern is for reuse in
reference guide vs training guide. If learning topics exist because we
need to add some elements to regular topics, can't we just propose to add
these elements to all topic types? Adding elements, unless we make them
mandatory should not bring up any backward compatibilty issues. 
 
I do agree with the framework topics. These
topics are interesting to the learning process and would probably not contain
anything 'informational' in nature, they just provide context for the learing
environment. I can easily imagine that some of them can be created automatically
(overview, summary, assessment), at least in part, from content in the
core topics. That could be possible if topics are built correctly, with
a learning short description, for example. 
 
What I don't like is the idea of the bucket
topic where other topics are 'embedded'. What is the added value of this
bucket? Doesn't it limit reuse opportunities with non-learning based documents?
Isn't it better to add required learning elements to all basic topic types
(as non-mandatory for backward compatibility). This way topics created
for learning can be reused for other types of documents by hiding the learning
specific stuff in processing. And vice versa. If the 'bucket' is for organizational
purposes only, why don't we just have a section our map model that could
become a 'learning content section/chapter/module/whatever name we want
to give it' where topics are are referenced with the standard topicref
mechanism?.

Proposal 2
I already have a specialization for exercises,
if you guys would like to use it to get started, I'd be happy to provide
it. 
 
Proposal 5
I'm still not totally getting SCORM and how
it integrates with DITA. I see it as an extra layer on top of topics and
maps... I'm going give all my trust to those who understand it better than
me :-) 

Overall comment
It is my impression that training requires
more linearity than reference guides. People need to understand how different
tasks are linked to each other, and how they can be put together in order
to perform metatasks. (Different between system task and user task, where
the user task may require performing 2 or more system tasks in a specific
order.). I'm not sure we have anything to answer that issue. Maybe that
is what the learning content specialization is tying to answer? 
 
France Baril

--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.441 / Virus Database: 268.17.32/677 - Release Date: 2/8/2007 9:04 PM