Re: Strict sequential identity rule?

On 22 Sep 2005 05:52:47 +0100, Colin Paul Adams
<colin@xxxxxxxxxxxxxxxxxx> wrote:
> >>>>> "Dimitre" == Dimitre Novatchev <dnovatchev@xxxxxxxxx> writes:
>
>    Dimitre> Pardon my bad English, but for me you are also a
>    Dimitre> "vendor", and a "free vendor" :o) of an XSLT processor.
>
> Well, it was really only tongue-in-cheek :-)
>
>    Dimitre> Probably vendors of XSLT processors could provide such
>    Dimitre> facilities?
>    >>  The former is easy to do (but should not be on by default, as
>    >> it would annoy too many people).  The latter is equally easy,
>    >> but is non-compliant behaviour. So it must be a non-default
>    >> option. In which case, I think the warning is better.
>
> The real point of my post was to find out which option you would
> prefer - a warning, or a suppression.

Both are useful  -- it would be nice to be able to selectively turn
them on/off using command-line switches.

In relation with my original post, I suggest that serious testing must
include test cases where the source xml documents have more than one
children of the document node. There will be cases, in which what we
thought was correctly working transformation would produce unexpected
results.

--
Cheers,
Dimitre Novatchev
---------------------------------------
Getting caught is the mother of invention.