Re: Query Languages for XML

Martin Bryan (mtbryan@sgml.u-net.com)
Mon, 17 Nov 1997 15:58:59 -0000


Joe Lapp wrote:

>I agree that under the appropriate circumstances you wouldn't have
>to lock the whole document. However, were you to do the trick with
>what is currently XSL, it seems to me that you would have to create
>a _replacement_ document and then replace the original document.

This presumes that the "document" is the thing you want to remove. What if:

a) the document was built from a set of entities?
b) only part of the document consisted of updatable data fields?

The key factor is "what proportion of the data needs to be modified?"

>If in the time between reading the original and generating the
>replacement another user reads the original, and if you the other
>user posts his replacement after you post your replacement, then
>your changes do not take.

Always a problem with databases, but fields that are "temporarily locked"
can always be assigned an attribute that the presentation software can use
to indicate that the data is in a state of flux to read-only users of the
data during the update period.

>Or maybe you are suggesting there is no need to replace the whole
>document using an XSL approach. XSL or some other XML standard would
>need to define a standard mechanism for identifying and modifying a
>portion of a document.

The XML/EDI crew will be looking into this problem as it is key to running
an electronic business using XML.

> I am aware of some sort of 'chunking'
>initiative, but I don't know exactly what the scope of the effort is.

Why not join the XML/EDI research teams (see http://www.xmledi.net for
details)