Re: [dita] Review of Lightweight DITA

["Michael Priestley" <mpriestl@ca.ibm.com>]

Hi Dawn, a few quick thoughts - I'l try
to get them in order.

I agree information typing is key for
many DITA scenarios. However, those information types won't always correspond
to Concept, Task, and Reference. One of the activities we went through
in designing and selecting the elements in topic was to look at what we
needed to support specializations in various content disciplines, including
marketing and training. 

One of the earlier iterations of LwDITA
did include specializations for Concept, Task, and Reference - but the
challenge was that those weren't necessarily the specializations that mattered
to the groups we were looking to for adoption. And if we started adding
other specializations, then the specification would quickly stop being
lightweight.

I hope that what we can do is release
a set of example (non-canonical) specializations that do provide information
typing, for various content types specific to different content disciplines
and industries - probably as part of our release of the lightweight specialization
architecture that we want to do next.

So then the question is - why release
LwDITA before we have those sample specializations and lightweight specialization
architecture?

And I think the answer there is - there
is value in a starter set, even at the topic level, being defined now.
One of the reasons is compatibility with other formats, like Markdown and
HTML5.

As Scott said, if it's going to be this
simple, why wouldn't they just use Markdown? And the answer is, for many
- they already have. So how can we provide a standard that brings groups
that are using Markdown back into a DITA community to enable reuse and
common processing? Forcing a set of programmers to use C/T/R and XML isn't
always going to be an acceptable answer.

Finally, just a thought - if you are
happy and productive using full DITA, that's great - but it also means
you're not the target audience for Lightweight DITA. 

Lightweight DITA isn't intended to replace
full DITA, or to be appropriate in every instance that full DITA is. In
most cases, organizations that are using full DITA today should probably
stay with that.

But for organizations that have balked
at using full DITA, either because of an aversion to our out of the box
information types, or an aversion to XML, or an aversion to information
typing of any kind - Lightweight DITA can provide a starting point for
their introduction to a world of structured authoring and an ecosystem
of tools and content reuse. 


Michael Priestley, Senior Technical Staff Member (STSM)
Enterprise Content Technology Strategist
mpriestl@ca.ibm.com



From:      
 Dawn Stevens <dawn.stevens@Comtech-serv.com>
To:      
 "dita@lists.oasis-open.org"
<dita@lists.oasis-open.org>
Date:      
 06/23/2017 06:45 PM
Subject:    
   [dita] Review
of Lightweight DITA
Sent by:    
   <dita@lists.oasis-open.org>

---

Hi all,
In my review of the document and the copious
amounts of comments made already, I don’t know that I have anything more
to add that hasn’t already been said. I completely agree with most of
the comments made by all.

I have my own personal issues that I’m
not sure are worth discussing, since I am sure they’ve been debated before
and I’m sure people much smarter than me and more experienced in this
area have already had these debates. But for the record, here are my bigger
issues. If anyone wants to respond to me privately, I’m happy to get more
understanding; it certainly feels that I’m the only one who is struggling
with these fundamentals, and perhaps it’s just indicative of being late
to the party and not really following any of discussions early on since
JoAnn was our representative.

• IT stands for information typing. There is no such thing in lwDITA — everything’s a topic. To me this is part of the essence of DITA and something I am frequently trying to educate my clients about — why everything just shouldn’t be a concept, why it’s important to determine what type of content you are writing. Now they’re going to have permission to do just what I’ve taught them not to do.
• It seems that what lwDITA boils down to is the elimination of semantic tagging, which again, has been an entire educational process — why is it important that it’s not just a paragraph, but a context paragraph — I’ve worked with my clients to explain why it’s better to tag content not just on what it is or how it looks, but what purpose it serves. When purpose is clear, following the template and structure become easier, and more consistent. Eliminating semantics in my opinion opens the door to just having a topic with a bunch of paragraphs in it, without structure and consistency.
• 100% of my clients use <table>, not <simpletable>, to meet their needs. It seems a shame that this one thing will keep them from an lw solution if they otherwise wanted the simplicity.

Thanks,
Dawn