These are pages which are usually derived from an original page and are usually the comments of people with regard to the original subject. They are usually thread mode pages and can be well defined comments with backup and support mixed with opinions having no more weight than that of the opinion-ator. As a result they are candidates for refactoring. They do however present a challenge if the refactorer does anything more that substituting a personal slant and interpretation, failing to present the many views which may be included. As a result they are rarely refactored, and if they are, they are rarely refactored well.
Discussion pages are usually fractured pages, derived from another page, or consisting of compilations from several pages.
Discussion pages need refactorers.
When such pages are refactored, they should be refactored with substance, not reshaped by the refactorer's opinions.
If we don't have formal studies, which is often the case with software tools, then opinions are often all we have to go on, such as developer behavior and thought processes when interacting with tools and techniques. To the best of my knowledge, WardCunningham never designated this wiki to be only for formal research. Formal research would be nice, but nobody wants to pay for it such that we have to use proxy evidence instead. I realize proxy evidence is messy to work with, but unless WikiWiki wins the lottery, we are stuck with difficult-to-work-with evidence sources. Some will argue for perfection-or-nothing, but that's a minority opinion, I believe, among WikiZens.
Surely refactoring in this context would mean making it clear for someone reading the discussion what is going on, without taking sides in a discussion. The phrase refactored with substance seems to imply a resolution of the discussion and that would mean changing the meaning. Adding new thoughts or attempting judgement would not I think be refactoring. Deleting opinions is also not refactoring in my opinion. -- JohnFletcher
I've tried about 3 times myself, and the other party was not satisfied with it. My re-interpreation apparently was too lossy, and I've had similar experiences with others re-stating my opinions in their own words. Without a 100% understanding of what was intended, translation-related meaning loss/alteration is quite likely. And if we had a 100% understanding, we probably wouldn't be arguing in the first place. (With the X-windows example, I cannot verify the reworked version accurately reflects the original authors' opinions because I cannot see inside their mind to compare.)
If two or more positions or classifications can be identified, they should be reflected in the refactoring. (perhaps resulting in a multiplication of pages).
This is a category classifying one such group of pages.