- From: Norman Walsh <Norman.Walsh@Sun.COM>
- Date: Thu, 21 Sep 2006 13:42:31 -0400
- To: public-xml-processing-model-wg@w3.org
- Message-ID: <878xkdjeuw.fsf@nwalsh.com>
/ Alex Milowski <alex@milowski.org> was heard to say:
| OK. Except now you are computing an optimization for something was
| easily known before. I could say the exact opposite in that you
| can easily recognize a here document and fabricate a step for it.
My argument against "here documents" *is about the syntax*, not the
semantics. So saying that I could fabricate the step automatically is
both true and not the point.
| Personally, I think this is far less usable than just putting the
| document where you need it.
|
| All this doesn't preclude the need for a "document" component. That
| component is necessary for inputs that are shared.
|
| The canonical use case here is the XSLT transform being embedded
| in the pipeline itself. What we have right now allows the XSLT step
| to be specified with one self-contained step (i.e. cut-n-paste is
| easy). This proposal would mean you'd have two steps to do the
| exact same thing.
Like I said, I'm not sure I'm going to get consensus for this
proposal, but I think the cost of here documents in terms of spec
complexity outweighs their benefits.
Be seeing you,
norm
--
Norman Walsh
XML Standards Architect
Sun Microsystems, Inc.
Received on Thursday, 21 September 2006 17:42:11 UTC