Re: [xsl] short and long empty xml elements

Subject: Re: [xsl] short and long empty xml elements
From: Abel Braaksma <>
Date: Fri, 13 Oct 2006 17:20:01 +0200
Michael Kay wrote:
I do completely aggree. The point is that the target system is not in our responsability but is used by a customer. I already tried to convince the customer to replace the system: no way. So we can try to find a nasty solution or we risk to loose the customer. We voted for the first way.

What would you do if you agreed to communicate with the customer using email, and then found that the customer's SMTP server didn't conform to any recognized standards? Would you go to an SMTP mailing list and ask how to bend the standard products to align them with your customer's buggy implementation?

I agree with your statement in general, but often we have to bend for the bigger companies having their own products with their own bugs...
What, for example, about HTML? For instance, XHTML is supposed to be valid XML, but it explicitly states that the following are not the same:

<script type="text/javascript" src="somfile.js"></script>
<script type="text/javascript" src="somfile.js" />

Also, it explicitly states that:


should be written as (note the space):

<tagname />

All this, because some vendors (notably MS here) decided not to follow the standards closely. As a result, there exists an XML format, names XHTML that is only one-way compliant: all XHTML is XML, but not all XML (even when it is literally to the same DTD), is not XHTML.

If you were to write XML (output method="xml") and it is supposed to be XHTML, you are out of luck. Luckily, the w3c people made an exception and added XHTML to the output methods (but I pity all these implementers that have to write serializers for all these exceptions).

-- Abel Braaksma

Current Thread