[WikiEN-l] More efficient process suggestions

Sun Nov 9 02:28:51 UTC 2003

Alex Rosen wrote:

>>One problem is, which version are you voting for? If I vote 7 for an article, and then a few edits happen, what happens to my vote? We could weight the votes based on what percent of the article has changed since
the vote.<<

Current version, always. As we get closer to 1.0 we'll need to have monthly runs to find and tag articles which make the cut for 1.0 and searches to highlight the ones most in need of work. We'll also need 1.0MustInclude and 1.0MustExclude. At some point near the end of that process it may be necessary to protect articles if vandalism becomes a problem. We can see what happens and adjust accordingly. You can change your vote at any time if you like. It will be necessary to stop allowing new articles to qualify near the end as well, so a poor article with a few 9 votes can't make it in just because few people have seen it.

>> As far as using this to solve the inclusionism/ deletionism debate, we'd need two votes for that - quality and importance. I might want only
quality articles but of any importance, or vice versa.<<

Hit counters can provide an intital importance ranking. 1.0MustInclude and 1.0MustExclude go further. Might need an importance vote tagging scale. I hope that we'll have really good quality by then and it won't be required. WikiProjects can have a big role in this, organizing and rating the key topics for heir subject. We'll see how it develops after a year of 1.0Qualified monthly runs have passed and we've seen how well that gets things screened and improved.

>> Adding child-rating votes would require a lot more choices, since people have very different ideas about what's apprpriate for kids. We'd have to have ratings for sex, violence, language, etc. Fortunately the
vast majority of articles would be fine for kids, so nobody would need to bother with those controls most of the time.<<

Right. That's for the future beyond 1.0, though. Just illustrating how the tagging expands to include as many tags as we find useful for producing subsets. If we find a need for texasSchoolBoardYes and texasCSchoolBoardNo or catholicSchooYes and catholicSchoolNo, we can add them and let those with an interest use them.

>> Another problem I see is how to deal with vandalism. It's usually easy to determine if an article edit was made in bad faith, but there's no way to tell if someone's just being a punk with their vote. I guess we'd need to restrict voting to non-newbies, which is unfortunate...<<

Not so good to restrict it. We want to start with every person, known user or not, who views an article rating it so we get lots of quality votes and it becomes hard for small groups to change ranks. Then we can use the flags described earlier to fine tune the process. Or maybe we will need qualified votes at some point. We can wait see how it goes for a while and tweak as required...

Now you see why I find the deletion arguments frustrating. There are more useful ways to solve the problem, like sorting out how to make the junk vanish from sight.:) That's done by gradually changing the default quality scale people see from minQuality1.0 to minQuality5.0 and having the Wikipedia software just not show wikilinks for anything which doesn't make the cut. Yes, my deltionist leanings are showing in the long term views of how things might evolve.:)