L2/00-129

 

Asking for advise regarding XML in standard development

 

This is a request for support in looking at standards development, using XML instead of, or in addition to a common word processor.  The idea was introduced at the April 2000 NCITS meeting to more easily enable the electronic distribution (and sale) of standards. 

 

I joined a small ad-hoc group to investigate the possibility of using XML in standards development, other members are from DoD, Sun, and Microsoft.

 

I'd like to have your advise in some of my most pressing questions:

 

·     Is it possible to develop standards using XML instead of e.g. Microsoft Word ?

·     Is it feasible ?

·     What kind of training would be required ?

·     Are there XML editors that work WYSIWYG ?

·     How complicated would it be to develop a DTD - would we need more than one for TR and IS ?

·     Are there tools that allow the comparison of old and new versions of documents ?

·     Can a document be made "secure", meaning protected against changes by others than the editor ?

 

·     … and I am sure, you have a lot of other information with value for this idea.

 

Most importantly - do you think, it is a good idea ?

 

Especially in light of the experience with the 14651 table, I am hesitant to alert JTC1 or ITTF too early of our considerations.  

 

Please let me know, what you think of the idea.

 

Thanks

Arnold

=================================================================

 

First comments from Ken:

 

Arnold,

 

>

> NCITS is looking at standards development, using XML instead of, or in

> addition to a common word processor.  The idea was introduced at the April

> 2000 NCITS meeting to more easily enable the electronic distribution (and

> sale) of standards.

 

The premise seems twisted to me. It is clear right now that pdf is an

already existing format that makes it easy to enable the electronic

distribution and sale of standards.

 

When we are talking about *document* distribution, why not use (at least

for now) a very widespread, standard, and usable format? A format for

which anyone in the world can download and use a free viewer.

 

> I joined a small ad-hoc group to investigate the possibility of using XML

> in standards development, other members are from DoD, Sun, and Microsoft.

>

> You have experience and a passion for XML, I'd like to have your advise in

> some of my most pressing questions:

 

Well, Mark is the one with experience and a passion for XML. My company

is heavily involved, and I can get information from others to bring

to bear on these questions, but it is not my own forté.

 

>

> *   Is it possible to develop standards using XML instead of e.g.

> Microsoft Word ?

> *   Is it feasible ?

> *   What kind of training would be required ?

> *   Are there XML editors that work WYSIWYG ?

> *   How complicated would it be to develop a DTD - would we need more

> than one for TR and IS ?

 

I expect that this *would* be reasonably complex. And if NCITS already

has document templates that can be used with Microsoft Word, why reinvent

the wheel here for a completely new mode?

 

> *   Are there tools that allow the comparison of old and new versions of

> documents ?

> *   Can a document be made "secure", meaning protected against changes

> by others than the editor ?

 

PDF already has provisions for that.

>

> *   … and I am sure, you have a lot of other information with value for

> this idea.

>

> Most importantly - do you think, it is a good idea ?

 

No.

 

My sense is that the main reason for going to XML is to create documents

that have well-defined machine-readable content/formats. I want XML to

enable automation of invoice and other financial instrument interchange

between disparate, distributed, complex database applications, for example.

I want XML to have better control over content/format abstractions when

some document needs to be used across platforms of very different

rendering capabilities. But I don't want XML to arbitrarily replace existing

document formats for human-readable documents, just because in principle it

can be done.

 

--Ken

 

>

> If you think that a discussion on Unicore or any other list would be

> valuable, I have no problems with it - I wanted to have a "sanity-check"

> before going public.  Especially in light of the experience with the 14651

> table, I am hesitant to alert JTC1 or ITTF too early of our considerations.

>

> Please let me know, what you think of the idea.  The ad-hoc will possibly

> meet within 3 weeks, so some input would be nice before then.

>