Subject: Re: [xsl] Long Namespaces|
From: "Karl Stubsjoen" <kstubs@xxxxxxxxx>
Date: Thu, 7 Jun 2007 16:24:02 -0700
As usual, the input has been fantastic; thanks. I have found that in my XSL development namespace use has become invaluable. Especially when the templates become nested through import and includes. In my case, there is a natural hierarchy which exists and the prefix/namespace tells the story. For example a variable which has been defined in the base template, the prefix should be xslb, be for base. I than have site level templates, application level templates, and page level templates, the prefixes (in same order) would be: xsls, xsla, xslp. So in your "page" template, top level template, if you were referencing the following variable: xsls:status, you'd know to seek the base level template for its declaration. This has become common practice for me. The complaint is that these short names are confusing, which I argue is not the case because of the availability and localization of the namespace declaration. Never the less, I've changed, the prefixes are now like this: xslSite, xslApplication, xslPage.
There's some kind of law about natural language that says short words are used more often than long words. Language has evolved that way because it improves communication. When you have a concept that is very frequently referenced, it's best to use a short name for it, because this reduces the time taken by the reader to recognize it. So for a namespace like xsl or xs which is used hundreds of times in a stylesheet, a short prefix works well. It's likely that the saxon namespace will be used less often, so a slightly longer name is appropriate.
I think a short prefix also helps the reader to focus their attention on the local part of the name, which is the part that carries more information.
There's a school of thought, of course, which tries to ban short names like "i" and "j" from programming entirely. This is of course a stupid over-reaction to programs that over-use such names. The fact of the matter is that bad programmers will produce unreadable code whatever disciplines you impose on them.
Michael Kay http://www.saxonica.com/
> -----Original Message----- > From: Karl Stubsjoen [mailto:kstubs@xxxxxxxxx] > Sent: 07 June 2007 18:58 > To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx > Subject: [xsl] Long Namespaces > > Is there any creed that suggests that namespace decelerations > are either bad or shunned upon? It seems that 3 letter > namespaces are the norm. Is anyone using verbose namespace > declarations regularly? > > Karl..