Hi Stan,
Apologies for the long note. The summary is - I'm
suggesting one new item for the list (with a paragraph of
explanation), and suggesting removal of many others that
are out-dated or (I think) inappropriate for this list.
Before getting into specifics - I have one broad, general
comment about how (and whether) DITA-OT should appear on
this list.
DITA-OT itself is just one implementation of the DITA
standard - bundled with or sitting alongside many other
implementations. I don't think it's appropriate to single
out DITA-OT as the only implementation in a list of
overall DITA weaknesses (or strengths). That's fine for a
vendor comparison or "How to choose your DITA tools"
paper, but this list is currently just about strengths and
weaknesses of DITA itself.
There are 8 separate items in the SWOT list that describe
DITA-OT as a DITA weakness. I see many problems with this:
- With so many weaknesses, and no other product
mentioned, it sounds like OASIS really doesn't like
DITA-OT. This is unfair to DITA-OT, and likely to scare
people away from using it (or from using products that
rely on it).
- It takes core toolkit design points - points that made
it (and DITA) successful - and calls those out as
weaknesses. For example, DITA-OT never would have
survived if it also created / maintained both an editor
and a CMS.
- The list really goes out of its way to pick on
DITA-OT. For example, it lists both the general lack of
good samples as a DITA weakness, but then specifically
calls out the samples in DITA-OT as an additional DITA
weakness.
- The list adds to the impression that DITA-OT is
somehow under the control of OASIS.
From all that - I'd suggest a new weakness:
Perception that OASIS owns both the DITA standard and the
DITA Open Toolkit
There is a perception in the DITA community that the OASIS
DITA Technical Committee either owns, maintains, or directs
work on the DITA Open Toolkit. Though widely used, DITA-OT
is just one DITA implementation among several. Some
individuals participate in both, but most DITA-OT
development is done by non-OASIS members, and DITA-OT is not
an official implementation.
Finally - I think all of the following DITA-OT points should
be removed from the list of DITA Weaknesses.
>> Toolkit is just that - a toolkit,
not an application
This is good to know when deciding between (for example)
DITA-OT + file system, or an integrated CMS. But it is not a
weakness of DITA itself.
>> Exclusive reliance on the Open
Toolkit may limit the use of DITA in special environments
I don't understand this (or why it would limit anything).
>> No content editing tool nor CMS
included in toolkit
This is not a weakness of DITA, just as it's not a weakness
of DITA that Oxygen or XMetal do not include a CMS.
>> Toolkit documentation is very poor,
and also fragmented between different sites
This is out of date. The docs have been entirely redone
since this was listed, and all docs are at dita-ot.org.
>> Requires not inconsiderable
customisation, in several XML disciplines: ANT, CSS, XSLT,
XSL-FO
DITA itself does not require any of this. It is only
required of a subset of DITA-OT users.
>> The user experience with many
content and map editors is awful and at best quirky
(Not a DITA-OT item, but still out of date.)
>> Confusion between DTD/Schema version
numbering and Toolkit version numbering
I think this falls under my suggested weakness above. It is
confusing because people think both numbers come from the
same source, and thus should be in sync. Also, again, only
an issue for certain toolkit users - not an issue for DITA
authors who do not use the toolkit, and not an issue for the
many DITA-OT users who have no idea what version they use.
>> Considerable lag time between an
update to the standard and support for that update in the
OT.
Many users derive no benefit from an
update to the standard until the OT supports it, and new
elements and attributes that apparently "do nothing" are
confusing.
This is false. DITA-OT supported 1.2 long before the
standard was done. The development code today has support
for all major 1.3 features; new elements and attributes are
supported in published stable releases.
>> There is not much high-quality
sample DITA content easily available. The samples that
come with the toolkit are not great.
>> Difficult to find well-explained
worked examples
There is no reason to list this twice, other than to
specifically pick on the DITA-OT samples. Suggest deleting
the sentence about the toolkit, and the following bullet.
>> Output using the Toolkit's default
stylesheets Is Not Sexy.
Again, not a weakness of DITA itself. Applies only to the
subset of DITA authors that use DITA-OT, use default styles
instead of those bundled with a product, and need something
flashy.
Thanks for making it to the end...
Robert D Anderson
IBM Authoring Tools Development
Chief Architect, DITA Open Toolkit (
http://www.dita-ot.org/)
"Dr. Stanley
Doherty" ---09/04/2015 21:16:34---Greetings - Thanks to
the people who provided feedback on the "DITA Strengths"
that we distributed f
From: "Dr.
Stanley Doherty" <stan@modularwriting.com>
To: dita@lists.oasis-open.org
Date: 09/04/2015
21:16
Subject: [dita] Request for Feedback: DITA Adoption SWOT
"Weaknesses"
Sent by: <dita@lists.oasis-open.org>
Greetings -
Thanks to the people who provided feedback on the "DITA
Strengths" that we distributed for review three weeks ago.
The Adoption TC (ATC) has rolled up the feedback and will
fold it into the complete SWOT wiki page:
- https://wiki.oasis-open.org/dita-adoption/DITA_Adoption_SWOT_Analysis
The following list represents the current collection of
SWOT "Weaknesses". These have been collected over the past
several years, so some may be out-dated. If you believe
that a "weakness" on this list should no longer be
perceived as such, please provide details to that effect.
We are hoping that we can get from reviewers
paragraph-length descriptions. For example, here's a
sample rewrite:
------------------
Original: Toolkit is just that - a toolkit, not an
application
Rewrite: The DITA Open Toolkit is not a fully implemented,
fully integrated application comparable to what writers
can get with Madcap Flare or FrameMaker. Lack of a
graphical interface and easy-to-configure options makes
DITA-OT inaccessible to writers without a background in
programming.
------------------
Please respond with your feedback by Sunday, September 20.
Again, thanks in advance.
Stan
= = = = = = = = = = = = = = = = =
* Key requirements: DTD/Schema, Toolkit, Instruction
manual, user mailing list, etc. - from as many as 8
separate sites
* Exclusive reliance on the Open Toolkit may limit the use
of DITA in special environments
* No content editing tool nor CMS included in toolkit
* Toolkit documentation is very poor, and also fragmented
between different sites
* DITA adoption can seem more like a paradigm shift, with
a steep learning curve and a complex solution
* Lack of a DITA-compliant, open-source editor makes the
bar to entry higher than it needs to be.
* DITA adoption can appear to be a step backwards when
migrating from the likes of Word or !FrameMaker
* Requires not inconsiderable customisation, in several
XML disciplines: ANT, CSS, XSLT, XSL-FO
* There is a significant geek factor to DITA
* Confusion relating to DTD and Schema - why include both?
* Many TAs don't have required skills and/or may not have
access to required skills
* DITA can be regarded as only being suitable for larger
companies & corporates
* Primary focus on computing needs - h/w & s/w
* Specialization can be a red herring, especially in early
stages of adoption
* The user experience with many content and map editors is
awful and at best quirky
* DITA is perceived as very complex, only for XML
specialists
* DITA is not seen by decision makers as a business
solution, to business problems
* DITA is seen as something that only relates to tech.doc
and/or technical authors
* DITA can produce static HTML files but cannot interface
dynamically with web sites
or web CMSs
* When DITA is introduced, the review cycles can become
much more bumpy because reviewers use other (non-DITA,
non-XML) tools
* DITA is seen by the enterprise as a tech.doc. "black
box" solution, not as an enterprise- wide solution
* Confusion between DTD/Schema version numbering and
Toolkit version numbering
* Updates to the standard happen very slowly
* Considerable lag time between an update to the standard
and support for that update in the OT. Many users derive
no benefit from an update to the standard until the OT
supports it, and new elements and attributes that
apparently "do nothing" are confusing.
* Perceived or real cost and complexity of migrating
legacy content to DITA
* There is not much high-quality sample DITA content
easily available. The samples that come with the toolkit
are not great.
* Difficult to find well-explained worked examples
* Output using the Toolkit's default stylesheets Is Not
Sexy.
* Interoperability requirements are not generally
understood
* Paradigm shift for authors is considerable -- must have
training in a new way of authoring
* Training required for technical support of the model and
the output
* OASIS DITA offers no test suite for DITA compliance. Any
tools vendor could claim<<BR>>DITA compliance
without reference to any objective compliance criteria.
* Technical writers outside the DITA community share a
perception that DITA is developed by elitists for elitists
* Technical writers outside the DITA community share a
perception that adopting DITA is often a precursor to
writers and their groups getting outsourced or offshored
* There are few local users groups to support new adopters
* DITA integration with GIT is poorly documented
* DITA provides little direct control over the output
formatting of individual pages, e.g.page breaks, table
breaks.