http://qs321.pair.com?node_id=18207

Thought #1 - It might be nice to see a breakdown of the reputation of a 'controversial' node, with both ++ and -- votes (say a node with more than one of both). It's quite likely that such a node merits further discussion :-)

Thought #2 - You can vote on a node, no matter how ancient it is. I suspect people very rarely vote on a node once it gets old. As the reputation of a node may be useful in helping someone searching the archives for an answer to a question, it would be good if the voting system timed out, and you could see the reputation for all nodes over, say, one month old, but no longer vote on them.

Andrew.

Replies are listed 'Best First'.
RE: Thoughts on voting and reputations (kudra: keep voting for old nodes)
by kudra (Vicar) on Jun 15, 2000 at 09:18 UTC
    I disagree with the second suggestion. The value of certain nodes doesn't erode over time. People who are discovering these nodes for the first time or who are using them for the first time should be able to vote on them. This is especially true because more people are voting now, so older nodes look undervalued in comparision to new ones.

    Haven't you ever run across a topic you knew nothing about, and so ignored, only to find yourself later dealing with it, going back to the related nodes, and finding a great deal of useful information? Haven't you ever searched for information on a topic, only to find just what you needed in a node that was created before you joined?

    I want the ability to vote for a good node regardless of its age. I don't know how many people do go through the archives looking for interesting old nodes, but I'm one person who does.

      The point of the second suggestion was not to devalue old nodes, but rather to help people searching the archives judge which answers to a question are the best - if you don't know about a subject, you will have difficulty telling the difference between a good answer and a great answer - though I imagine the really bad answers will tend to speak for themselves (or have helpful followups)

      It would just be a helpful cue to people who are new here or who are new to a subject as to which answers in the archives are the best or most helpful (or at least, those that other people think are the best or most helpful :-)

      Andrew.

        I must admit, I'm with kudra on this one.
        I often search for answers to problems I'm experiencing, and that search often leads me to old nodes. If I find something that I think is insightful and/or particularly good, then I will vote ++ for it.
        I see your point about being able to see the results of the voting and agree that it would be useful if you are trying to gauge the 'importance' of a node, but I think withdrawing the ability to vote on that node will just remove the ability for a monk to show his appreciation to the author.
        As you suggest I think that the 'greatness' of the node can be gauged by the comments attached to it and the reader, whether a newbie or not, should be able to judge for themselves.
RE: Thoughts on voting and reputations
by Ovid (Cardinal) on Jun 15, 2000 at 04:04 UTC
    #1 is an excellent idea. I have some nodes that have been voted on quite a bit (both ways) and I'd be curious to see the breakdown something like:

    Node name      ++   --   Total
    Perl vs. VB      +7   -1   +6

    Don't look at the HTML. It was a quick hack :)

RE: Thoughts on voting and reputations
by Ozymandias (Hermit) on Jun 15, 2000 at 02:42 UTC
    Excellent idea. I like number 2 better than 1, but they'd both be useful.

    - Ozymandias

RE: Thoughts on voting and reputations
by flyfishin (Monk) on Jun 15, 2000 at 21:27 UTC
    Comment on #2: I like being able to go vote on old nodes. Take for example SuperCruncher's Sexiest Women node. I want to be able to go back and vote it down hopefully pushing it onto the Worst Node list.

    Update: Poor example. What I believe about a certain node is irrelevant. My point is that I like being able to go back to old stuff. I often miss nodes and find myself leafing through old ones. If I find a great one I want to be able to add my ++ no matter how old. Or my -- if that is the case.

      I don't think there should be a contest about voting nodes onto the Worst Nodes. Of course, SuperCrunchers post was a possible offending post, as it did not talk that much about Cool Uses For Perl, but was more an advertisement of his site. But that does not justify ganging together just to put the node up onto the Worst Nodes. If you, for yourself, look at a node and think you want to waste a -- vote on it, that's OK. But you should never vote negative on a node, just because it already is on the Worst Nodes.

        I agree. I was simply voicing my opinion about being able to vote on old notes and used a poor example.
      Hmm, right idea, wrong reason, I feel. I think SuperCruncher's node was misguided but not really bad enough to justify a jihad against it. Worst nodes should get there of their own accord - this is the reason we don't get to see reputations before we vote on a node, and the reason for the caveat at Best Nodes.
RE: Thoughts on voting and reputations
by gregorovius (Friar) on Jun 16, 2000 at 00:43 UTC
    I like #1 as is.

    On #2 I think we could even out the 'goodness' of older versus newer nodes by making a weighted 'goodness' index. Ideally we would want this index to be the number of votes the post has received versus the number of people (of qualified voters) who have seen it. Since we probably don't have a record of the latter then we could divide by some other number akin to it. Example:
    goodness = votes / number of viewers goodness = votes / (number of registered monks / number of posts on th +e week of the posting) # maybe?
    It would be good to know what other statistics are kept of our use of the site in order to craft the best possible index.

    I also think that it is a good thing to not disclose the votes a given post has. We could make this or the proposed goodness index remain hidden and be used only as a sorting key for search results.