On Wed, Jun 10, 2009 at 6:24 PM, Anne van Kesteren<[email protected]> wrote: > There is no paragraph preceding the list of steps making it all somewhat > confusing.
Editorial: I've broken the section up by using <h4>s. There are now two sub sections: * Terminology Used in Processing Algorithm * Processing Algorithm The Processing Algorithm section now begins with, "The algorithm to process a configuration document is as follows: " > It should probably be a requirement that the document is _namespace_ > well-formed. To prevent any confusion. > Editorial: right, now reads "If the document is not namespace well-formed [XML]..." > "If any of the following attributes are in error or invalid," is double and > also somewhat recursive. E.g. invalid depends on the definition of in error. > It is also unclear what "following attributes" refers to here. > Editorial: I've removed, "If any of the following attributes are in error or invalid..". I instead dealt with the attribute errors in place where they happen. This makes it consistent with the rest of the algorithm and hopefully gets rid of the confusion. > "If any other attribute, in any XML namespace, apart from xml:lang, is used, > then the attribute is in error and the user agent must ignore it." Is this > not again stating what has already been stated? Duplicating requirements > makes the specification very hard to read. > Editorial: Agreed. This is no longer duplicated as the first assertion has been removed. -- Marcos Caceres http://datadriven.com.au
