Traditionally comments in Interpreted language files slows things down. It
there any practical experience with the impact of comment nodes in the XSLT
files, or best practice guidance?
I would like to find the balance point between too much and too little.
We have also considered an XSLT comment scrubber between the saved source
and the published source. It would be easy to build, but if the runtime
impact of leaving the comments in place is sufficiently small why introduce
all the other headaches?
I think the answer varies from one XSLT processor to another. My particular
interest is in .NET ( don't flame me please) but I have an interest in other
processors as well.
Greg Alvord