OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [xsl] ANNOUNCE: Petition to withdraw xsl:script from XSLT 1.1



At 14:29 01/03/2001 -0800, Steve Muench wrote:
>| At 13:31 01/03/2001 -0800, Steve Muench wrote:
>| >I (personally, not speaking for the WG here) don't happen to
>| >be a believer in some kind of look-it-up-and-magically-download-
>| >the-right-(trusted?)-implementation-from-some-central-website-
>| >for-the-language-my-processor-can-support-for-the-chip-architecture-
>| >and-OS-that-I'm-currently-running-(while-I'm-using-this-on-a-plane)
>| >scheme to download and run the "right" implementation of some function.
>| 
>| Nevertheless, xsl:script allows for precisely that through it's src
>| attribute. Given that, don't you think that relying on extension namespaces
>| pointing to an RDDL document containing 1) information on how a human being
>| can get the extension should he want to do it by hand and 2) a pointer to
>| implementations of the extension (possibly in several languages so that the
>| processors can choose the one they know about, or their favoured one)
>| should the processor be allowed to fetch extensions automatically (possibly
>| only within a trusted network) is far superior to what xsl:script offers ?
>
>The 'src' attribute purports no magic beyond the standard URI mechanism.

And, to paraphrase, how is that not look-it-up(DNS,locator
services,repositories,etc)-and-magically-download(HTTP,FTP,etc)-the-right(Co
ntent-negociation,browser/processor sniffing,etc)-(trusted?)-etc-etc-etc ?

>The implements-prefix="foo" points to a specific *namespace* URI
>for the function "library". A human-assisted process to download
>the right bits would only need this NS Uri as a key into whatever
>information you'd want to stage, but I don't envision the
>XSLT processor firing up modal dialogs or web browsers in the
>middle of stylesheet execution to allow the user to "Please pick
>which implementation of this function you'd like to use..."
>before continuing along its merry way with the current transformation.

Obviously, I don't envision modal dialogs and whatnot to ask the user
permission to get the extension. Simply, if the extension module is not
present and the processor hasn't been configured to download it
automatically, then it dies requesting it's dependency. If it has been
configured to automatically download extensions then it uses the RDDL
document to get the most appropriate module.

Embedding ties one to an implementation language for a given extension. I
don't understand the advantages this could have.

-- robin b.
The computer can't tell you the emotional story.  It can give you the exact
mathematical design, but what's missing is the eyebrows. -- Frank Zappa