Re: Initial whitespace in PI from XSLT, main body

> > Do we want the behavior of almost all current XML processing tools to
suddenly become "incompliant"?
> >
>
>   Wendell is proposing an application-level house style for use of
processing instructions, a rule for the next level up the protocol stack,
which is perfectly legitimate.

Sorry, this sounds completely not understandable:

  1. What is "application-level house style"?

  2. What is "a rule for the next level up the protocol stack" ?

And why are we dealing at such a level of detail of a relatively simple
feature, when there are much more serious issues with some specifications,
like being 1100 pages long and probably lacking a single reader who has
read them all and has grasped the intended (???) meaning?

Thanks,
Dimitre

On Sun, May 8, 2022 at 12:02 PM Michael Kay mike@xxxxxxxxxxxx <
xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:

> >
> > Do we want the behavior of almost all current XML processing tools to
> suddenly become "incompliant"?
> >
>
> Wendell is proposing an application-level house style for use of
> processing instructions, a rule for the next level up the protocol stack,
> which is perfectly legitimate.
>
> In fact I've heard others propose a stronger convention, namely that
> processing instructions should use the syntax of attribute="value" pairs,
> as in an element start tag: that's the design convention used for the
> xml-stylesheet processing instruction.
>
> Michael Kay
> Saxonica