Re: Current Status of Discussion on RE/RS Handling

>In SGML, the records (which  is synonymous with your use of "lines")
>are those that the author saw when the document was formatted to his
>screen by his text editor.  Some editors preserve those record
>breaks; some don't. SGML makes sure that the data is the same in both
>cases by eliminating the REs that were caused solely by markup.

So how can you determine which is which? Unless the editor understands
all the rules of SGML, or the entity manager understands all the
rules of the editor, there seems to be a very large margin for
error in creation or interpretation. 

It seems to me that removing the rules for RE/RS entirely is certainly
no worse than this: in either case, there is a large body of software
(in one case applications, in the other, XML entity and storage
managers) that could get it wrong. The advantage of removong RE/RS
entirely is that you suddenly have a single, easily explained set of
rules for parsers, to which applications can cater.

I consider both this issue, and the characater set issues of far less
importance than getting XML *done*, so I'm willing to compromise my
position. However, it seems to me that in both cases, the decision to
go one way or the other is purely subjective: technical arguments
either way are available. I think the point of view I take generally
emphasises pragmatism.