W3C home > Mailing lists > Public > public-xml-processing-model-comments@w3.org > October 2008

Re: source on p:error

From: Norman Walsh <ndw@nwalsh.com>
Date: Tue, 14 Oct 2008 13:04:44 -0400
To: public-xml-processing-model-comments@w3.org
Message-ID: <m2wsgbw1jn.fsf@nwalsh.com>
ht@inf.ed.ac.uk (Henry S. Thompson) writes:
> Hmm.  What benefit follows?  If it has an inline, or an explicit
> connection, doesn't matter if it's primary or not.  In the (admittedly
> unlikely) case where you do in fact want to compute the error output,
> having it be non-primary is a pain.  I can't think of any situation
> where having it be primary will cause a problem. . .

It leads naturally to the confusing and very rarely useful situation
where the primary output of the preceding step becomes the error

If you're using p:error, you almost always want to provide a message.
Making the port non-primary will catch the error where you forget to
define it.

                                        Be seeing you,

Norman Walsh <ndw@nwalsh.com> | To enjoy yourself and make others enjoy
http://nwalsh.com/            | themselves, without harming yourself or
                              | any other; that, to my mind, is the
                              | whole of ethics.-- Chamfort

Received on Tuesday, 14 October 2008 17:05:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:28:25 UTC