Hi Joseph,
Thanks for all the details you've mentioned. I shall work on
various suggestions that you've suggested, in due course (and I
consider your suggestions important).
On Sun, Jun 23, 2024 at 1:10 AM Joseph Kesselman <[email protected]> wrote:
> To make Xalan a _real_ 3.0 processor we'd have to rework some of the core
> data model in addition to adding functionality. (The distinction between
> temporary trees and nodesets must be maintained in 2.0, and must be
> eliminated in 3.0, for example). What your work gives us, I believe, is a
> processor with 2.0 semantics to which some of the 3.0 features have been
> added. Which is an interesting thing, but it isn't 3.0, and I think the
> combination permits stylesheets that can't be reliably run on other 2.0 _or_
> 3.0 processors. I would ***much*** rather force the user to be aware of that
> when writing the stylesheet than get into arguments about it later.
>
> Hence my desire to give them a separate namespace. That's what I would have
> recommended if I'd seen the start of your work; I still think it's the best
> available solution.
About this point, IMHO I feel that giving separate namespace to XSLT
3.0 instruction implementations for now, on Xalan-J's xslt3 dev repos
branch, shall likely cause many users of Xalan-J's XSLT 3.0
implementation to become uninterested in this work, which I think is
not a desirable situation. I'm in midst of adding more XSLT 3
instruction implementations owithin Xalan-J's xslt3 dev repos branch
and IMHO I don't wish to change XSLT namespace within this
implementation.
I request you to please, grant your concurrence to my this suggestion,
and I shall be grateful for that. Of-course any Xalan-J release shall
require vote on this list to pass as per Apache Software Foundation
rules.
Many thanks.
--
Regards,
Mukul Gandhi
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
