[xsl] Re: document not there ambiguity

[S Woodside <sbwoodside@xxxxxxxxx>]

On Wednesday, April 23, 2003, at 12:38  PM, Robin Berjon wrote:

> Since when have massive cross-posts become good netiquette?

I cross posted to three lists. Each one has a group of people who would 
be interested
axkit-users : some people use XSLT
xslt : because I'm using libxslt
xsl-list : because there are quite a few people on that list who either 
implement XSLT processors or are members of various WGs at the W3C that 
design XSLT/Xpath

Since this seems to me now to be an axkit specific problem that is no 
longer necessary. xsltproc does the IMO "right thing". I include 
xsl-list because there is a problem with the spec -- still in xslt 2.0

> S Woodside wrote:
> > On Wednesday, April 23, 2003, at 10:23  AM, Chris Leishman wrote:
> > > >   The spec states in section 12.1 Multiple Source Documents
> > > >  "If there is an error retrieving the resource, then the XSLT 
> > > > processor
> > > >   may signal an error; if it does not signal an error, it must 
> > > > recover
> > > >   by returning an empty node-set."
> > > >
> > > > So both behaviours are correct.
> > > Really...what where the W3C thinking?  Perhaps someone should start 
> > > a list of 'standard implementation choices for implementing the xslt 
> > > standard' (rolls eyes) and maybe that'll become YAWNS (Yet Another 
> > > W3C uNsuccessful Standard).
> > They're already on the way... the WD for xslt 2.0 says this about 
> > "unparsed-text" function
> > hooray :-\
> I wonder why people that complain about problems they find in specs 
> don't ask themselves "why is this here?" before bitching out loud. 
> Some specs, like all software, have misfeatures. In other cases there 
> is a real need and reason for the given feature, whether it is liked 
> or not.

I did ask myself why is this here and googled for it. What I found was 
(a) other people asking "why is this here" and (b) a variety of 
processors that implement it as a warning instead of an error, such as 
libxslt, 4suite, saxon (configurable), ...

there is a nice discussion starting here

   http://www.biglist.com/lists/xsl-list/archives/200101/msg00073.html

> There are cases in which non-failing document() is desired (eg 
> optional inclusion of data, often using the "relative to nodeset" 
> approach)

Yes.

>  and others where failing is better.

Can you explain a situation where that would be better?

> The spec addresses both.
>
> Which behaviour to choose is an application problem. That processors 
> do not give the option may be construed as a bug. That they don't 
> default to failing may be construed as a bug. It nevertheless remain 
> an application problem.

Since the spec is ambiguous, and people have complained about it on a 
number of occasions, without a satisfactory answer, therefore there is 
a problem with the spec.

Since at the time we have the spec as is, then yes, it is also an 
application problem. So it is an application problem. Are you saying 
that AxKit's behaviour is correct?

simon

> --
> Robin Berjon <robin.berjon@xxxxxxxxx>
> Research Engineer, Expway        http://expway.fr/
> 7FC0 6F5F D864 EFB8 08CE  8E74 58E6 D5DB 4889 2488

--
www.simonwoodside.com -- 99% Devil, 1% Angel


XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list