RANT: Whitespace is not an issue?

["David RR Webber" <david@drrw.info>]

OK - I just need to rant.  I watched hundreds of emails on

the W3C Schema WG on this whitespace.  I struggle when

these document-heads get religious about this stuff!

 

For me this is as simple as it gets from an eBusiness

stance - if you are designing eBusiness systems and

exchanges that can be broken by whitespace issues -

you need to go back and fix your design!

 

Also - when it comes down to it - if you are shipping

payloads where format specific information is critical

use CDATA [[ - that's what its there for!

 

Beyond that - there's a bigger principle that I'd like

to mention - that relates to interoperability and

implementablity of our specifications:

 

1) BPSS should only rely on W3C schema specific features

    as a last resort.  Basic XML mechanisms should

    be preferred whereever possible.

2) BPSS should strive to use only features of W3C schema

    that are pervasive, simple and widely adopted and

    consistently used.

 

My rationale is obvious - I want BPSS to not run foul of

any deployment and versioning issues WRT vendors schema

parser implementations.

 

I'm not sure this latest issue over whitespace

is something we need to spend any more time on than absolutely

necessary.  Maybe a simple design note - to the effect that we

expect people to build business processes where whitespace is

not an issue, and to use XML constructs to resolve ambiguous

information content accordingly.   If they are using BPSS to

control document workflows and whitespace within payloads

is important - then CDATA [[ should be used to remove

any possibility that payloads are manipulated by the transport

layer(s).

 

Thanks, DW.