|No such thing as a small change|
How to know whether question in SOPW is entirely answered or not?by Chicken (Initiate)
|on Jun 17, 2002 at 20:43 UTC||Need Help??|
While perlmonks.org is a beautiful site where many perl questions are answered in a really short time (personaly I got a fruitful help several times, and several times provided answers to questions that was I happen to know).
This is great, but I see a small problem that do a situation slightly worse that it could be.
Sometimes answers do not entirely cover a question, due to either replier mistakes, or may be some questions coul be better covered by more than one perlmonk, because first adviser solves only a part of a problem and allows others to answer other aspects of a question.
And, as I know from my own experience, if I see that a question is replied twice or more, then I really do not bother reading deeper than reading a title, and go for a next question that may be of interest.
Example of such node is How to embed other GUI into Tk application which I consider answered to about 20%, but it has two replies and also is old. When a person who knows an answer will be here, he'll hardly notice it. I tend to think of such partly-answered questions (and wrongfully answered...) as "zombie" questions :) . Thay have almost no chance to be answered.
Zombie questions usually will fit into FAQ or Q&A sections, because it may be too specific and rarely answered.
Asking it once again (to "refresh" it) is obviously not good.
I think of "zombie" questions as of a problem, and I see it could be resolved by two ways.
1. (seems to me easier) May be it's worth to have a page, in addition to "newest nodes" as "unanswered nodes" where authors of questions could place link to node ids of questions they wish to be covered in more details, (or just to wait to be answered somehow), may be with "percentage" of current coverage that answers covers at the moment.
2. (more global, and hard to implement) Each question have a number 0-100 that is initially 0, and 100 when question is completely answered. I can't invent sane way of how this number could be edited. (may be following: replyer suggests it's coverage, and author of a question have a right to confirm this coverage?)
I hope best wishes to everyone,