Re: [xsl] [exsl] EXSLT 1.0 - Common, Sets and Math

Subject: Re: [xsl] [exsl] EXSLT 1.0 - Common, Sets and Math
From: Jeni Tennison <mail@xxxxxxxxxxxxxxxx>
Date: Wed, 7 Mar 2001 18:41:04 +0000
Hi David,

> yes, especially in the beginning, but a year down the line and you
> get an email from (say) Michael and a representative of the evil
> empire saying they've both implemented foobangwhizz(spodge) so can
> it go into exslt so users can use it portably on both their systems.

Mike collaborating with the evil empire?!? Is it only you that remains
unsullied by connections with the dark side? ;)

> At that point do you want to vet the semantics of the function, or
> do you just want to say. It's implemented <here>, documented <here>
> so others can implement, and no one else has used the name
> exslt:foobangwhizz -- so it goes in.

[This conjured a question: What is it that's currently stopping an
implementer from using another implementer's extension
element/function? I don't think there's anything proprietary about
extension elements/functions, is there? Is it simply that they all
thought up node-set() at the same time?]

I do agree that it would be beneficial to have a neutral repository of
extension element/function definitions and implementations with a
neutral namespace to match.

But what should happen if Miss Innovator comes along later and says
'Hey, that foobangwhizz function should really take three arguments,
like this'? It would be a real pain if the foobangwhizz that takes
three arguments has to be called foobangwhizz-3args or something. Of
course there's always a retreat to namespaces, but the whole point is
to avoid that.

So I think that there should be some kind of really minimal review
period (like a week or two) just so that when Miss Innovator turns
round three months later and suggests changes at least we can say "you
had your chance".



Jeni Tennison

 XSL-List info and archive:

Current Thread