Re: A serious detail point

At 12:12 19/4/97 CDT, Michael Sperberg-McQueen wrote:
>
>For any project which needs to freeze its choice of technology now,
>XML-link should not come into consideration for the simple reason
>that it is not finished and still volatile enough that the printed
>draft reads, on every page, DO NOT IMPLEMENT.  If you are recommending
>that the project DROP a plan to implement XML-LINK, you are doing
>the only correct thing.  It would be an irresponsibility bordering on
>malpractice to recommend that a project
>COMMIT itself, NOW, to using XML-link, in the form it WILL eventually
>take.

The point I was making was that XML-link as defined prior to April 6th was a
good fit for what a group of users wanted - once you killed locsrc it was no
longer suitable and we had to go to HyTime.

>On the substantive question of link management, I am still waiting
>for some coherent explanation of why you can't use general entities.

Because there is no way of distinguishing between calls to external entities
that must be recalled to make sense of the document, and those that are
"nice to have". How do I know which general entities contain the URLs of
documents that I must recall? How do I even know they contain a URL that
might be significant. (We can't lextype the contents of the entity
replacement string to force it to be a URL!)

>>>The update and maintenance problem is handled nicely by the general
>>>entity mechanism you illustrate.  The caching problem can be handled
>>>with affinity groups / BOS / whatchamacallits that say "Cache this
>>>one, I'm going to need it often".
>>
>>The problem is that XML has no concept of BOS, ...
>
>Then caching is an argument in favor of having one.  It is not, in
>my view, an argument for messing up XML addressing with locsrc.

The problem is how do you automatically generate a BOS on the fly. I could
see how to do this with locsrc - I can no longer see how to do it.
----
Martin Bryan, The SGML Centre, Churchdown, Glos. GL3 2PU, UK 
Phone/Fax: +44 1452 714029   WWW home page: http://www.sgml.u-net.com/

Received on Sunday, 20 April 1997 05:07:40 UTC